[House Hearing, 105 Congress]
[From the U.S. Government Publishing Office]



 
 OVERSIGHT HEARING ON THE WATER MANAGEMENT IMPLICATIONS OF THE 1997/98 
                                EL NINO

=======================================================================

                           OVERSIGHT HEARING

                               before the

                    SUBCOMMITTEE ON WATER AND POWER

                                 of the

                         COMMITTEE ON RESOURCES
                        HOUSE OF REPRESENTATIVES

                       ONE HUNDRED FIFTH CONGRESS

                             FIRST SESSION

                               __________

                    OCTOBER 30, 1997, WASHINGTON, DC

                               __________

                           Serial No. 105-58

                               __________

           Printed for the use of the Committee on Resources


                                


                      U.S. GOVERNMENT PRINTING OFFICE
 46-056 CC                   WASHINGTON : 1998
------------------------------------------------------------------------------
                   For sale by the U.S. Government Printing Office
 Superintendent of Documents, Congressional Sales Office, Washington, DC 20402



                         COMMITTEE ON RESOURCES

                      DON YOUNG, Alaska, Chairman
W.J. (BILLY) TAUZIN, Louisiana       GEORGE MILLER, California
JAMES V. HANSEN, Utah                EDWARD J. MARKEY, Massachusetts
JIM SAXTON, New Jersey               NICK J. RAHALL II, West Virginia
ELTON GALLEGLY, California           BRUCE F. VENTO, Minnesota
JOHN J. DUNCAN, Jr., Tennessee       DALE E. KILDEE, Michigan
JOEL HEFLEY, Colorado                PETER A. DeFAZIO, Oregon
JOHN T. DOOLITTLE, California        ENI F.H. FALEOMAVAEGA, American 
WAYNE T. GILCHREST, Maryland             Samoa
KEN CALVERT, California              NEIL ABERCROMBIE, Hawaii
RICHARD W. POMBO, California         SOLOMON P. ORTIZ, Texas
BARBARA CUBIN, Wyoming               OWEN B. PICKETT, Virginia
HELEN CHENOWETH, Idaho               FRANK PALLONE, Jr., New Jersey
LINDA SMITH, Washington              CALVIN M. DOOLEY, California
GEORGE P. RADANOVICH, California     CARLOS A. ROMERO-BARCELO, Puerto 
WALTER B. JONES, Jr., North              Rico
    Carolina                         MAURICE D. HINCHEY, New York
WILLIAM M. (MAC) THORNBERRY, Texas   ROBERT A. UNDERWOOD, Guam
JOHN SHADEGG, Arizona                SAM FARR, California
JOHN E. ENSIGN, Nevada               PATRICK J. KENNEDY, Rhode Island
ROBERT F. SMITH, Oregon              ADAM SMITH, Washington
CHRIS CANNON, Utah                   WILLIAM D. DELAHUNT, Massachusetts
KEVIN BRADY, Texas                   CHRIS JOHN, Louisiana
JOHN PETERSON, Pennsylvania          DONNA CHRISTIAN-GREEN, Virgin 
RICK HILL, Montana                       Islands
BOB SCHAFFER, Colorado               RON KIND, Wisconsin
JIM GIBBONS, Nevada                  LLOYD DOGGETT, Texas
MICHAEL D. CRAPO, Idaho

                     Lloyd A. Jones, Chief of Staff
                   Elizabeth Megginson, Chief Counsel
              Christine Kennedy, Chief Clerk/Administrator
                John Lawrence, Democratic Staff Director

                                 ------                                

               Subcommittee on Water and Power Resources

                JOHN T. DOOLITTLE, California, Chairman
KEN CALVERT, California              PETER A. DeFAZIO, Oregon
RICHARD W. POMBO, California         GEORGE MILLER, California
HELEN CHENOWETH, Idaho               OWEN B. PICKETT, Virginia
LINDA SMITH, Washington              CALVIN M. DOOLEY, California
GEORGE P. RADANOVICH, California     SAM FARR, California
WILLIAM M. (MAC) THORNBERRY, Texas   ADAM SMITH, Washington
JOHN B. SHADEGG, Arizona             RON KIND, Wisconsin
JOHN E. ENSIGN, Nevada               LLOYD DOGGETT, Texas
ROBERT F. SMITH, Oregon              ---------- ----------
CHRIS CANNON, Utah                   ---------- ----------
MICHAEL D. CRAPO, Idaho
                  Robert Faber, Staff Director/Counsel
                    Valerie West, Professional Staff
                    Steve Lanich, Democratic Counsel



                            C O N T E N T S

                              ----------                              
                                                                   Page

Hearing held October 30, 1997....................................     1

Statement of Members:
    Doolittle, Hon. John T., a Representative in Congress from 
      the State of California....................................     1

Statement of Witnesses:
    Andrews, Richard, Director, Governor's Office, Emergency 
      Services, California.......................................    27
        Prepared statement of....................................    74
    Friday, Elbert W., Director, Office of Oceanic and 
      Atmospheric Research, National Oceanic and Atmospheric 
      Administration, Department of Commerce.....................     3
        Prepared statement of....................................    34
    Georgakakos, Aris, Director, Georgia Water Resources 
      Institute, Georgia Institute of Technology.................    22
        Prepared statement of....................................    60
    Georgakakos, Konstantine, President, Hydrologic Research 
      Center, Scripps Institution of Oceanography, University of 
      California, San Diego......................................    20
        Prepared statement of....................................    58
    Hall, Stephen K., Executive Director, Association of 
      California Water Agencies..................................    29
        Prepared statement of....................................    78
    Leetmaa, Ants, Director, Climate Prediction Center, National 
      Oceanic and Atmospheric Administration, Department of 
      Commerce...................................................     5
        Prepared statement of....................................    39
    Schaefer, Mark, Deputy Assistant Secretary for Water and 
      Science, Acting Director of U.S. Geological Survey, 
      Department of the Interior.................................     8
        Prepared statement of....................................    44
    Sorooshian, Soroosh, Professor, Department of Hydrology, 
      University of Arizona......................................    25
        Prepared statement of....................................    73



 OVERSIGHT HEARING ON THE WATER MANAGEMENT IMPLICATIONS OF THE 1997/98 
                                EL NINO

                              ----------                              


                       THURSDAY, OCTOBER 30, 1997

        House of Representatives, Subcommittee on Water and 
            Power, Committee on Resources, Washington, DC.
    The Subcommittee met, pursuant to notice, at 2:10 p.m., in 
room 1334, Longworth House Office Building, Hon. John T. 
Doolittle (chairman of the subcommittee) presiding.

STATEMENT OF THE HONORABLE JOHN T. DOOLITTLE, A REPRESENTATIVE 
            IN CONGRESS FROM THE STATE OF CALIFORNIA

    Mr. Doolittle. [presiding] The Subcommittee on Water and 
Power will come to order.
    The Subcommittee is meeting today to hear testimony 
concerning the water management implications of the 1997/98 El 
Nino. The current El Nino event is expected to be the most 
severe El Nino since 1982/83. Indeed, many scientists believe 
that it is the most severe event of this kind since records 
have been kept, about 150 years ago. But what does that mean 
for our weather? Will every quirk in the weather this season be 
seen as a manifestation of El Nino?
    There are some pretty good indications that this El Nino 
will mean serious flooding in portions of the United States, 
probably on the West Coast. It is equally probable that we are 
going to have drought conditions somewhere as well. The real 
challenge is to know when and where these conditions are going 
to develop. And will we know that information early enough to 
make a difference?
    Knowing that we have a significant El Nino event that is 
likely to affect our weather, it is much like knowing we are 
going to have a severe winter along the eastern seaboard. We 
still don't know whether there are going to be bitter cold 
temperatures or snowstorms which will shut down New York City, 
Boston, or Washington, DC on any given day. That information 
will only be known a few days in advance, and of course, like 
all predictions, they can be off, as we know from watching the 
nightly weather forecast.
    Much of the press coverage and most of the speeches in the 
political arena might lead one to believe that flooding is the 
only concern. However, as important as flooding is, it is only 
one of the critical issues associated with the El Nino 
phenomenon. The impact on water supply could be almost as 
significant. Millions of people in this country rely on water 
stored in reservoirs to meet their munic-

ipal, industrial, and agricultural water supplies. The western 
half of the United States is particularly dependent on these 
reservoirs for water supplies about 6 months out of every year.
    Initially, you might think that more rainfall and snowfall 
would translate into fuller reservoirs and expanded water 
supplies. Unfortunately, the reverse can happen. Every year, 
prior to the months when most of the precipitation occurs, the 
water level in reservoirs is lowered to accommodate the 
anticipated flood flows. However, significant water is retained 
in the reservoirs for water supplies in case the rain never 
materializes. The challenge throughout the rainy season is to 
respond to the storms, trying to pass the flood flows as they 
occur.
    If good information is available, reservoir managers can 
provide extra protection by releasing water even before a storm 
begins to affect a reservoir, but there is the dilemma. If a 
reservoir does not refill, water shortages are a very real 
probability. In some cases the available advanced warning will 
allow time to avert flooding; in others it will not. The 
critical question is: How far in advance will water managers 
and flood control experts know that a crisis is pending in a 
given watershed?
    With all of the general storm information available, the 
data to make a decision in any particular watershed will only 
be known 48 hours to 72 hours in advance. Many factors affect 
the ability to predict accurately whether a given watershed or 
reservoir will receive enough precipitation to justify 
releasing the water before the storm begins. If a reservoir 
manager fails to make releases, there is a greater chance there 
will be flooding on the river system. If the reservoir manager 
decides to begin releases and the water is not recovered that 
season, hundreds of thousands of acre feet of water can be lost 
that will be needed for deliveries in the summer months.
    If the wrong decisions are made often enough, much of the 
water stored in Federal, State, and private reservoirs will be 
lost. Such a scenario is not only highly possible, it in fact 
occurred in California on a smaller scale during this past 
1996/97 water season. It is possible that the El Nino brewing 
in the Pacific will dissipate and cause no significant impact. 
It is also possible that we may have a long, wet winter with 
storms evenly spaced over the season, which will be easier to 
manage. It may be that all of the storms pass over the extreme 
southwestern and southern part of the country, but, based on 
past El Nino events, it is also highly possible that we could 
have a long, wet rainy season with erratic and unpredictable 
impacts in many different parts of the country. Certainly, any 
portion of the western United States could be affected.
    The strongest El Nino ever recorded so far occurred in 1982 
through 1983. It was responsible for 2,000 deaths and $8 to $12 
billion damages worldwide. The current El Nino event appears to 
be even stronger than that one. If we are to avoid or mitigate 
some of those impacts, we must build on what we know, but also 
realize--and I think this is key--what we don't know.
    In river systems without reservoirs or with inadequate 
reservoir capacity, we will not be able to blunt the effects of 
these storms. Even in river systems with adequate reservoirs, 
we will not have perfect knowledge, and we run the risk of 
either flooding and/or re-

leasing water that will be lost for delivery when needed next 
summer.
    We, clearly, have better data today than we had when the 
last big El Nino hit us in 1982/83. It gives us an opportunity 
to prepare much more than we had then. But while witnesses will 
talk about what we know now, I am asking them in their oral 
testimony to focus on what we still do not know. I am asking 
them to also focus on the difficulties facing water managers in 
determining when and how much water to release and what are the 
weak links in communications and in placing reliance on the new 
information.
    Finally, I ask the witnesses to focus on how we can learn 
from this El Nino to improve our response to similar events 
that will come upon us in the future. I look forward to hearing 
our witnesses.
    We have our first panel seated already, and let me ask you, 
please, if you will rise and raise your right hands.
    [Witnesses sworn.]
    Mr. Doolittle. Let the record reflect that each answered in 
the affirmative.
    And, gentlemen, thank you. You may be seated. We appreciate 
your coming here today.
    The custom of the Committee is to use those lights before 
you as a rough guide to the length of your testimony, which I 
think with one exception is going to be 5 minutes. If you have 
further things to say, just because the red light goes on, 
don't feel obliged to stop in mid-sentence, and feel free to 
complete the sentence or the paragraph, and if we don't have 
too many Members, the time limitation is not going to be 
particularly relevant. But they're a guide for you, and I think 
the yellow light will go on at the beginning of the fifth 
minute.
    And with that, our first witness is Dr. Elbert W. Friday, 
who is Director of the Office of Oceanic and Atmospheric 
Research for the National Oceanic and Atmospheric 
Administration within the Department of Commerce. Dr. Friday.

STATEMENT OF ELBERT W. FRIDAY, DIRECTOR, OFFICE OF OCEANIC AND 
    ATMOSPHERIC RESEARCH, NATIONAL OCEANIC AND ATMOSPHERIC 
             ADMINISTRATION, DEPARTMENT OF COMMERCE

    Mr. Friday. Thank you very much, Mr. Chairman. I request 
that my written testimony be entered into the record and I 
would like to give a few brief comments in response to your 
charge to the witnesses today.
    Mr. Doolittle. I appreciate that, and let me say that your 
statement and all others will be entered into the record, as 
you have submitted them. So please feel free to share with us 
your views during the oral testimony.
    Mr. Friday. Thank you, sir. It is, indeed, a pleasure to be 
here to address what you have identified, and I think we all 
agree, is a very, very important issue.
    The El Nino situation that exists today is, indeed, being 
blamed for everything, and yet we have to be very careful that 
we don't overattribute meteorological and hydrological events 
to this particular pattern. You were very correct in pointing 
out that we know a lot more today than we did in 1982 and 1983. 
Indeed, in the 1982/83 event, we didn't even understand that El 
Nino was occurring until we were about halfway through the 
event. We recognized during that event that we needed to learn 
a great deal more about it in order to be able to better 
respond to the next major event that occurred.
    In a 10-year period, beginning in 1984, we did a remarkable 
number of things. We instrumented the Pacific Ocean and the 
tropical regions, so that we could identify changes in the 
ocean long before they began to affect the atmosphere, this 
coupled system that we have. We were able then this year, using 
the information from that research data base, to begin to 
predict the beginning of the El Nino in January, and Ants 
Leetmaa, who will be testifying next, was then able to go out 
with an actual forecast for a very strong El Nino in the 
springtime of this year, long before we started to see these 
major effects starting.
    We made a conscious decision in NOAA and in the Federal 
Government in general that it was important, even though we 
were basing this on research results, that we try to provide to 
the user community the information that we had, recognizing 
that it was not complete, but recognizing it was better than 
not having any information at all, and that's a critical point 
that we need to make.
    It is important also to understand--you asked for the 
uncertainties that we have in the process--that when we talk 
about the effects of El Nino, we are talking about not 
absolutes, but relatively shifting a probability distribution 
of wet or dry, of warm or cold. We're talking about loading the 
dice, if you would, so that they may come up slightly more 
frequently a 7, but they will still come up an occasional snake 
eyes. Our job in the forecast process is to try to understand 
ahead of time how they've been loaded in the various sections 
of the country.
    The ability that we have now to integrate all of these data 
are based on the fact that we have been archiving data for a 
very long period of time. You indicated records going back 150 
years. Detailed records for every meteorological and 
oceanographic parameter are archived in various aspects of the 
government, and those data are now being mined in order to be 
able to understand more completely how these effects occur, 
what kind of information that we can say about the future 
events.
    The observation system that we have out in the Pacific 
Ocean has clearly demonstrated its advantage to being able to 
predict the onset of an El Nino, and this array of instruments 
out in the tropical ocean, it is absolutely critical that we 
continue to maintain and monitor that. And, indeed, we in NOAA 
are certainly looking forward to Congress supporting the 
operational funding of that, which we have in the President's 
budget this year.
    NOAA has not operated by itself in this. The effort of 
research has been a cooperative effort across many aspects of 
the Federal Government. It includes National Oceanic and 
Atmospheric Administration, the National Science Foundation, 
National Air and Space Administration, and Office of Naval 
Research from the Department of Defense doing the initial 
research into this activity.
    This year virtually every aspect of the Federal Government 
is working together, as you see from the witnesses that are 
here at this table, to be able to pass the information that we 
have developed onto the user community, to allow them the best 
possible information in making their decisions.
    Again, I want to stress that these are not absolutes. It is 
difficult to attribute any individual weather event to the El 
Nino. What we do know about the patterns that exist is the fact 
that it tends to affect the overall interactions between the 
ocean and the atmosphere. So it tends to affect the storm 
tracks. It tends to affect the jetstreams. And, yet, we still 
don't know the precise manner in which this is going to occur.
    You were correct in pointing out, sir, that the important 
information for the day-to-day management of all of our 
activities, including reservoir management, is going to be much 
more short-term in the nature, on the order of two or 4 days. 
We are beginning to be able to extend, through the National 
Centers for Environmental Prediction, that forecasting 
capability out to 10 days, and we hope to be able to improve 
that, but that's an area that we really have to work on.
    We have a capability now that we've demonstrated in at 
least one location in the country of being able to provide 
probability distributions of river stages out in the future 
going out to several weeks, and, again, that's a development 
effort that we need to continue to pursue and work with the 
user community to make sure that that can be effectively 
applied to all of the various users of the information.
    The forecast process that we're undergoing now represents a 
lot of research that has gone on. We have just taken action 
within NOAA to announce that we are going out over the next few 
weeks to identify things additionally that we need to observe 
during this event, so that we will have the data base, we will 
have the information base necessary to truly evaluate this 
entire situation, so that we can do a better job on the next 
one.
    Thank you, sir.
    [The prepared statement of Mr. Friday may be found at end 
of hearing.]
    Mr. Doolittle. Our next witness will be Dr. Ants Leetmaa.

STATEMENT OF ANTS LEETMAA, DIRECTOR, CLIMATE PREDICTION CENTER, 
NATIONAL OCEANIC AND ATMOSPHERIC ADMINISTRATION, DEPARTMENT OF 
                            COMMERCE

    Mr. Leetmaa. Yes.
    Mr. Doolittle. He is the Director of the Climate Prediction 
Center within NOAA and the Department of Commerce. Dr. Leetmaa.
    Mr. Leetmaa. Thank you, Mr. Chairman. I welcome the 
opportunity to come here and talk to you. I would like to 
discuss, first, the nature of what the El Nino forecasts are 
for coming seasons, then outline some of the expected impacts, 
especially over the western region of the U.S., and some of the 
forecast issues as you alluded to, because there are many of 
those.
    Both Joe Friday and I have with us here Dr. Danny Fread of 
the Office of Hydrology, who is directly involved in the 
hydrological forecasting. So if some questions come up 
regarding that, he is here to help us.
    Let me start with the forecast for El Nino itself. The 
Climate Prediction Center is forecasting the current event to 
peak around the end of this year or early next year. Present 
observations show that the event is still slowly growing. 
Hence, the forecast for the next 3 months are pretty much on 
the mark.
    At the end of the year, this event in amplitude and spatial 
extent will be comparable to the 1982/83 one, which until now 
had been considered the event of the century. Most of the model 
forecasts--and there are many model forecasts besides those 
done at the National Centers for Environmental Prediction--
forecast that this event will decay toward normal or in some 
cases even cold conditions, La Nina conditions, by next summer. 
How fast it decays--and there is some disagreement here--is 
critical to the U.S. forecast during late winter and spring.
    In 1982/83, many of the heaviest impacts were in the 
February-to-April period. So although the event is decaying at 
that time, because of the tropics are warming up, because of 
the annual cycle, the sea surface temperatures in fact will 
remain anomalously warm, and so the Climate Prediction Center 
is forecasting that we will basically be under the influence of 
strong El Nino conditions through the January-to-March period.
    As Dr. Friday indicated, the onset of this event and its 
evolution have been well-predicted three to 6 months in 
advance, and this is letting us do something which we haven't 
been able to do before in the past, which is to anticipate what 
might take place and then try to mitigate potential harmful 
impacts.
    The forecasts for El Nino are reasonably well-established. 
The forecast for El Nino impacts over the United States are 
less so. Currently, we use two primary sources of information. 
The historical record is stratified according to when moderate 
or large El Ninos have taken place. This gives us an indication 
of what we think the average El Nino response would look like. 
Equally importantly, it lets us know what the probabilities are 
that such an event will take place or such impacts will take 
place. Clearly, each El Nino does not have the same impacts.
    We also routinely run atmospheric model forecasts out to 
the next several seasons. Both of these give us very much the 
same picture of what might happen, but there are differences in 
detail. So, basically, at the present time we still base our 
forecasts, the seasonal forecasts, primarily on the statistics, 
but modify those in regions where we think the models are 
adding something.
    One of the major issues that we face is that this event is 
obviously not an average El Nino, and one might ask the 
relevance of what using the moderate El Nino criteria, and so 
we anticipate and we already know that the impacts will depend 
on the size of El Nino. So in discussing what might happen in 
this particular event, we make the comparison to 1982/83, which 
was an event of the same size.
    There was one other event comparable to both of these 
events in recent history, and that occurred in 1877/78, but the 
data for that at the present time over much of the United 
States is limited, and so we're limited in terms of what we can 
say.
    Although much of the U.S. is forecast to be impacted, let 
me focus for this hearing on California. California, during 
winters with moderate El Ninos, receives about 130 to 140 
percent of normal rainfall during its rainy season from 
November to March. This converts to about 5 extra inches of 
rain averaged over much of the State. During 1982/83, which is 
a plausible scenario, California, as much of the Southwest, 
received 150 to 200 percent normal rainfall. This translates 
over coastal California and central California and the southern 
Sierras to approximately 10 extra inches of rain spread over 
that area, and for much of northern California the amount was 
greater than about 16 inches. So that's a lot of water.
    These two estimates, one for the average impact for 
moderate El Ninos and what happened in 1982/83, give one 
estimate of the possible spread of impacts. Ranges in 
forecasts, however, result from both the chaotic component of 
Nature and the uncertainty and the limitations of our current 
forecast capabilities. Although research will improve the 
forecasts, this range in what we forecast will always remain.
    If, for example, Nature could repeat the next two seasons 
several times over, keeping in place the same El Nino 
anomalies, the impacts for each of those winters, in fact, 
would be slightly different. For example, during 1877/78, 
another event of the century, central and northern California 
received more rain than in 1982/83, while southern California 
received less.
    Many serious shortcomings of the current forecasts are in 
the nature of the forecasts themselves. We, basically, at 
present forecast seasonal averages. As you indicated, El Nino 
rains come as storms. Series of storms are what cause problems. 
Research is indicating that the number of storms and their 
intensity is a function of the strength and type of El Nino. 
Depending also on the type of storm, we can either get rain or 
snow. It's the rain events that cause a lot of flooding, 
although ultimately the snow melts and there can be problems 
then also.
    So, basically, we have issues that we haven't addressed 
yet, which is exactly how El Nino modifies the nature of the 
storms, the number of the storms, the intensity, and also 
what's going to happen to the snow pack.
    As we look toward the future, ultimately, what we want for 
effective water management practices is a suite of forecast 
products of different lead times from hours, to days, to a 
week, to 2 weeks, to a month, to a season, and even decadal. 
Currently, the Weather Service produces short-range forecasts 
which are the most accurate and have the most local detail. For 
this event, the Weather Service is also developing a weekly 
update, a threat assessment, trying to extend the forecast into 
that week two area, so we can give emergency managers a heads-
up as to what potentially might be happening in terms of the 
intense flooding events.
    The Weather Service also provides the link between 
meteorological forecasts and reservoir operations through the 
River Forecast Centers. These work side by side with the Bureau 
of Reclamation, the Corps of Engineers, USGS, and State 
personnel. As part of the Weather Service modernization, they 
will start using the seasonal forecasts to ultimately extend 
hydrologic forecasts out to several seasons also.
    Let me conclude by saying that it is unlikely that in the 
foreseeable future or ever we will have meteorological 
forecasts of the ac-

curacy that water managers ideally would like to see. Nature 
inherently has an unpredictable component. The key is for the 
providers of the forecasts and the users to work together on an 
interagency basis, so each understands the other's capabilities 
and limitations and needs, and works toward enabling the 
transfer of this information into operations. Although we have 
started working in that direction, we need to accelerate the 
efforts.
    Also, I hope that the newness and the uncertainty of the 
seasonal forecasts associated with this El Nino do not lead to 
an underestimation of the potentially harmful impacts that 
might take place. Thank you.
    [The prepared statement of Mr. Leetmaa may be found at end 
of hearing.]
    Mr. Doolittle. Thank you.
    Our next witness, who will be recognized for 10 minutes, is 
Dr. Mark Schaefer, Deputy Assistant Secretary for Water and 
Science and Acting Director of the U.S. Geological Survey 
within the Department of the Interior. He will be accompanied 
by Dr. Thomas J. Casadevall, Regional Director of the Western 
Region, U.S. Geologic Survey, Water Resources, and by Dr. David 
A. Matthews, Manager of River Systems and Meteorology within 
the Bureau of Reclamation.
    Dr. Schaefer, you're recognized.

  STATEMENT OF MARK SCHAEFER, DEPUTY ASSISTANT SECRETARY FOR 
 WATER AND SCIENCE, ACTING DIRECTOR OF U.S. GEOLOGICAL SURVEY, 
                   DEPARTMENT OF THE INTERIOR

    Mr. Schaefer. Thank you, Mr. Chairman, and good afternoon. 
We appreciate the opportunity to be here to discuss the 
Department's activities related to El Nino. This episode is of 
particular interest and concern to the Department because 
changes in precipitation and temperature, stream flow, et 
cetera, all impact our facilities and the responsibilities that 
we have for managing land and water resources.
    Across the Nation, the U.S. Geological Survey is monitoring 
stream flow at more than 7,000 locations. Nearly half of these 
transmit information in real time that is made available to 
individuals through the Internet and through other means. This 
network will provide first indications of the effects of El 
Nino on rivers and streams and reservoirs, and it forms the 
basis for the flood forecasting and warning activities that my 
colleagues have just described.
    The Department has been extensively involved in evaluating 
current El Nino forecasts. Daily stream flow data from 50 years 
of USGS records reveals marked increase in short-term flood 
during El Nino episodes, even when total seasonal stream flow 
is not profoundly affected. Studies also show that previous El 
Ninos are linked with many of the most severe floods in 
southern California and other southwestern States. The USGS has 
been active in communicating these lessons to others and in 
incorporating this information in our management activities.
    You asked about what we don't know. The predicted strength 
of this event raises unusual problems for us because we do not 
know whether the most intense El Ninos simply yield more 
intense ver-

sions of the regional water resources and hazard effects that 
we typically see with El Ninos, large or small, or whether they 
result in whole new patterns of effects. For example, during 
the strong El Nino of 1982 and 1983, most of the West Coast 
experienced substantial increases in rainfall.
    The Bureau of Reclamation has been closely tracking the 
long-range forecasts, with support of the National Weather 
Service's River Forecast Centers and their Climate Prediction 
Center, and their Technical Service Center in Denver. 
Reclamation is developing a procedure to integrate all of this 
information into its operations at its various facilities 
around the country. Water operations managers in the five 
regional offices and in the 60 area and project offices are 
being kept informed of any changes in forecast. They're getting 
information on a weekly and daily basis, and they are going to 
incorporate that into any management decisions that they make.
    Reclamation has also begun the process of informing all 
water managers of the potential impacts of this El Nino, as 
well as the current state of our climate predictions. The 
Bureau has opened dialog with its water managers and climate 
researchers from the National Weather Service and from other 
agencies. We're working together on this, and as my colleagues 
pointed out, we have a lot more information available to us 
today than we did in the early eighties, and we think we're in 
a good position to respond to whatever may occur.
    As you know, the Bureau of Reclamation has many dams and 
reservoirs, and I just wanted to underscore this through one of 
our charts here. We have 596 dams and reservoirs, and as you 
pointed out, Mr. Chairman, in the western United States, in 
particular, individuals and businesses, farmers, and so on, 
depend on the decisions that the Bureau of Reclamation makes. 
It's a tricky balancing act to try to respond to weather 
predictions and ensure that we're preserving public health and 
safety, while at the same time maintaining the water levels we 
need for agriculture and for other uses.
    Long-range forecasts are too general for us to target the 
specific watersheds and the precise locations that may be 
affected by storm events related to El Nino. However, past 
events and current information suggests that we should prepare 
for frequent heavy precipitation this winter in the mountains 
of California, Arizona, New Mexico, and other areas that affect 
the operations of reclamation in the mid-Pacific, upper and 
lower Colorado, and southern Great Plains regions. This 
translates into potentially high water levels in reservoirs of 
the Central Valley Project, Colorado River Storage Project, 
Lake Powell, Lake Mead, and the Rio Grande Basin Projects, as 
well as smaller projects in these areas.
    And as you know, Mr. Chairman, the Central Valley Project 
is Reclamation's largest and most complex project, comprised of 
storage dams, reservoirs, pumping plants, canals, and so on, 
and delivers water throughout about a 40,000-square-mile area. 
The 1982/83 water year was the wettest on record throughout the 
Central Valley Basin, and record precipitation throughout that 
winter led to very large spring snowmelt runoff that lasted 
until June.
    Reclamation is tracking weather forecasts. It's 
incorporating the forecasts into ongoing modeling activities, 
and it's planning possible management actions to respond to 
whatever may be on the horizon. Significant operational actions 
to accommodate potential high inflows next spring are already 
underway. Reclamation's water operations and management 
activities in each project area are already in a delicate 
balance of supply and demand.
    USGS is using El Nino forecast information in conjunction 
with historical El Nino flood discharge relations to evaluate 
the status and condition of its stream-gauging equipment in 
regions where floods are likely, and this will help ensure that 
our gauges are fully operational when storms do occur.
    Our coastal and marine program, in cooperation with NOAA/
NASA, is evaluating the potential land loss in coastal areas 
from severe storms, and USGS is developing a variety of 
geospatial data. It's preparing maps and other products, and 
these will be available on an emergency basis. We have teams 
that can work around the clock in order to develop whatever 
products are needed.
    USGS is collaborating with Reclamation and various NOAA 
operational and research groups and State agencies to test and 
improve the hydrologic value of current weather and climate 
forecasts.
    The USGS is also contributing valuable information 
regarding the potential for landslide occurrences associated 
with El Nino weather effects, and I'd like to give you an idea 
of what some of this information looks like. These are two 
maps. This shows the landslide hazard outlook for October 
through December this year, and this shows the outlook for 
January through March. Now this is a product that was developed 
jointly by USGS and by NOAA. In fact, I think this is the way 
of the future, where the agencies will work more closely to 
develop integrated data products that will be useful to the 
public and to resource managers.
    But what you can see here, these are the areas of high 
landslide potential in the U.S. So this is the map that was 
developed by USGS. We overlaid that on NOAA's information 
related to the El Nino episode, and what you want to look for 
are the purple or dark blue areas, where we're expecting high 
precipitation and where there is already a high landslide 
potential. So you can see we took two fairly simple datasets 
and have developed some information that we think is very 
powerful and will be useful in helping to ensure that people 
are prepared in regions like this, some areas along the 
California coast, for this type of event.
    In addition, the floods and droughts associated with El 
Nino can seriously affect water quality. Increased loads of 
nutrients and toxic chemicals may be washed into rivers during 
flood conditions. The USGS has a national water quality 
assessment program that does water quality monitoring. It does 
that generally over the long term, but these individuals are 
available on a short-term or emergency basis to evaluate water 
quality, if that becomes necessary in a particular area.
    These planned actions are expected to meet the needs of 
water managers and mitigate the effects of extreme weather 
events that may arise from this El Nino. However, unanticipated 
needs may arise, some of which may require the attention of 
Congress. We'll work closely with you, should such needs arise.
    Long-term future needs include enhanced stream-gauging 
operations in areas where flood probabilities are increased. 
I'd like to show you another visual here. This is the USGS 
stream-gauging network nationwide. The network is an extensive 
network, obviously, but, in addition, it has decreased by 5 
percent in the last 7 years, from about 7,400 stations in 1990 
to about 7,000 stations in 1997.
    Now that's a bit of the bad news. The good news is that the 
number of stations that are telemetered, that have the link to 
satellites, so that we can virtually instantly get that 
information, has gone up significantly. And let me demonstrate 
that in this figure.
    Here you can see our overall number of stream gauges has 
been constant or slightly decreasing, but the proportion of 
stations that have the telemetry, which is indicated here by 
these dark bars, has actually increased, and this is what's 
particularly important to the Weather Service and to our 
managers. We get this information out on the World Wide Web 
typically with a 4-hour delay, and under emergency conditions 
we can get it out much more quickly.
    So we're increasing our telemetered sites. We're decreasing 
slightly on our total number of sites, and we're trying to 
address total issue, and we're trying to make sure that we can 
keep adding to the telemetered sites. Right now we're at about 
43 percent with no telemetry; 57 percent that have telemetry.
    We look forward to working with Congress and with this 
Subcommittee as this unprecedented event unfolds. And, again, I 
appreciate the opportunity to appear here this afternoon, and 
my colleagues and I would be happy to try to answer any 
questions you may have.
    [The prepared statement of Mr. Schaefer may be found at end 
of hearing.]
    Mr. Doolittle. At this point we'll recess for the vote and 
then reconvene for questions.
    [Recess.]
    Mr. Doolittle. Well, thank you for waiting for us, for me.
    Dr. Leetmaa, was it you that referred to the storm of--was 
it 1878?
    Mr. Leetmaa. Yes.
    Mr. Doolittle. Yes.
    Mr. Leetmaa. 1887.
    Mr. Doolittle. Thank you, 1887----
    Mr. Leetmaa. 1877/78.
    Mr. Doolittle. 1877/78?
    Mr. Leetmaa. Right.
    Mr. Doolittle. And I guess we don't have too much data on 
that, but we have some, and that was a storm, if I understood 
you correctly, where it was wetter in northern California than 
in southern California.
    Mr. Leetmaa. That's right. We had several stations. We had 
Red Bluff, which I think is California still.
    Mr. Doolittle. Yes.
    Mr. Leetmaa. San Francisco, Sacramento, and San Diego. And 
if you look at the comparison of that one to 1982/83, what you 
had in San Francisco for the January, February, March period, 
280 percent normal rainfall versus 215 for 1982/83; Sacramento, 
you had 240 percent versus 200 percent in 1982/83; Red Bluff 
was 400 percent versus 280 percent.
    Obviously, after this event, we'll have some more 
confidence in knowing how these big events behave. So one of 
the ways we obviously try to bracket what's going to happen is 
to try to look at comparable kinds of things. So this is one 
more piece of evidence that, in fact, it probably will be quite 
wet in not only just southern California, but central 
California and northern California also.
    Mr. Doolittle. I apologize. This speaker went off up here, 
and I only sort of half-heard your last two or three sentences. 
Could you just restate that again?
    Mr. Leetmaa. As I indicated, for any given El Nino event, 
there is an uncertainty in how much, what the range of 
precipitation might be. There's an unpredictability, in 
essence, and one of the ways we can study that unpredictability 
is to look events that look similar to the ones that we've 
observed or we can use numerical models, and both of those 
would then give you some spread on what happens. So this 
particular event is indicating that it probably is appropriate 
to think about both central and northern California also being 
heavily impacted this year.
    Mr. Doolittle. Was the 1877/78 event an unusually large El 
Nino event?
    Mr. Leetmaa. Yes, scientists look back, and when they 
compare sort of the characteristics, the global 
characteristics, it certainly was comparable to 1982/83.
    Mr. Doolittle. It's interesting to hear that testimony 
because so much focus has been given to Southern California and 
the extraordinary amounts of rainfall they're supposed to have, 
which of course may, in fact, happen, but it could also happen 
that Northern California instead might get that impact.
    Mr. Leetmaa. Yes.
    Mr. Doolittle. And for cities like Sacramento, that could 
be particularly serious--well, for the whole State's water 
supply, if a reaction to this prediction or estimate is to 
release too much water in anticipation of the 200 or 400 
percent rainfall that they had at Red Bluff over 100 years ago. 
And really it's going to be tough. And what I'm getting out of 
this is it's just a rough guess. Yes, we know a lot more than 
we did, but it's still a rough guess that could be right or 
might not be right.
    Mr. Leetmaa. That's right. I think the climate forecasts 
are always going to have a degree of uncertainty. I mean, 
they're basically pushing the mean rainfalls in a certain 
direction, but around those mean rainfalls there will be a 
spread; there will be a range. And so even 10 years from now or 
20 years from now, it would be unlikely if we can basically 
narrow that spread down. So one has to look at the historical 
record and get some sense of what that spread looks like.
    Mr. Doolittle. Was it you who said that your short-range 
forecasts tend to be the most accurate?
    Mr. Leetmaa. Yes, that was probably in my statement.
    Mr. Doolittle. But, even with that, isn't there a high 
degree of inaccuracy, even with the short-range forecasts?
    Mr. Leetmaa. Perhaps I should let Dr. Friday, who used to 
be the head of the Weather Service, address that.
    [Laughter.]
    Mr. Doolittle. All right. I'd be curious to know--it seems 
like somewhere I heard or read that there's about 50 percent 
accuracy on your short-range weather forecasts.
    Dr. Friday. Oh, actually, I believe the 50 percent accuracy 
number, it's the low end of the range on some of the seasonal 
outlooks. The short-range weather forecast, as far as storm 
events are concerned, heavy rain, thunderstorms, the accuracy 
on those forecasts are now up in the 80 percent range. That did 
not used to--that was not the case five or 6 years ago, before 
we put in the next-generation weather radars around the 
country.
    Mr. Doolittle. Oh, OK. So we've improved dramatically.
    Dr. Friday. We have dramatically improved the accuracy of 
storm and rain event forecasting since the introduction of the 
modernized National Weather Service.
    Mr. Doolittle. Dr. Friday, I was interested--and I don't 
think I knew about the weather buoys, that they're floating out 
there, I guess in the eastern Pacific, that were installed a 
few years ago in order to track El Nino. I am curious to know 
more about that. It seems that would have been quite a feat. 
How many of those were installed and how far apart are they? 
I'd like to know a little bit about that.
    Dr. Friday. It is a spectacular engineering and scientific 
success. After we looked back after the 1982/83 El Nino and we 
said, ``Wow, what was that that hit us?'', we recognized the 
fact of what it was and identified the fact that we really 
didn't know it was coming because of the limitations of the 
data-observing systems of the time was so great that we didn't 
have anything to base understand how it evolved.
    An international research program was put together with 
cooperation from many countries around the globe, and in the 
United States, as I indicated, it was supported by NOAA, NASA, 
National Science Foundation, and the Office of Naval Research 
in DOD. Technology was developed, because people had indicated 
that you couldn't put data buoys out in the middle of the 
Pacific Ocean and have them stay in one place--it was just an 
engineering nightmare--technology was developed at NOAA's 
laboratory in Seattle, the Pacific Marine Environmental 
Laboratory, which allowed us to install in water that was 3, 
3.5 miles deep data buoys right on the equator, and within a 
few degrees north and south of the equator. And over a 10-year 
period, we have installed 70 of those data buoys, all the way 
from the coast of South America, all the way over to Indonesia.
    Now one of our NOAA ships operates full time out there just 
simply maintaining these data buoys, going out and repairing 
components and replacing components that fail. It is this 
information that not only measures the surface temperature, but 
it also measures the temperature below the surface of the earth 
down to 500 meters all the way across this area, and it was the 
subsurface information that gave us the hint back in January 
that we starting to see the development of an El Nino. And it 
was that subsurface information, coupled with the beginning of 
the surface impacts, that allowed Ants Leetmaa to be able to 
forecast the significance of this back in the springtime of 
this year. So we have maintained those buoys under a research 
component for many years, and in this year's Presidential 
budget we are asking for those to be funded as a fully 
operational system, so that we maintain that throughout--in 
perpetuity, to keep track of this phenomena.
    Mr. Doolittle. Could you explain how they keep these buoys 
in place?
    Dr. Friday. I was afraid you were going to ask that.
    [Laughter.]
    Dr. Friday. The buoys themselves are relatively small, and 
they are connected to an anchor on a very long nylon cable, and 
the design of the equipment on the buoy is so that it wasn't 
literally carried along with the equatorial current. The 
specific design to dampen the vibrations that would occur and 
all of that--the way they handled the moorings, the way they 
handled the coupling, that was the engineering feat, and it was 
a very complex design of those systems, so that it would not be 
literally carried away by the current.
    But, in addition to the 500 meters of instrumented cable, 
with temperature sensors and in some cases salinity sensors 
that are directly under the buoys, there is another three miles 
of heavy-duty nylon cable that goes all the way to the surface 
of the bottom of the ocean, and there are large weights 
attached there to hold it in place.
    Mr. Doolittle. So each of these buoys is actually anchored 
to the bottom of the ocean?
    Dr. Friday. Yes, sir.
    Mr. Doolittle. That's amazing.
    Dr. Friday. What I said, this was an amazing engineering 
feat, and it has resulted in a tremendous success story in 
being able to understand how the ocean and the atmosphere links 
together to cause the change in water resources capacity for 
California.
    Mr. Doolittle. Did somebody have to go down to the bottom 
of the ocean?
    Dr. Friday. No, sir, they're deployed from--they have been 
deployed from the NOAA vessels, operating research vessels, and 
we now have, as I indicated, one vessel, the KAIM'MIMOANA, 
which operates out of Hawaii, that its full-time job is to go 
out and maintain those buoys.
    Mr. Doolittle. So this is a heavy weight resting on the 
bottom of the ocean?
    Dr. Friday. Actually, it's usually three or four railroad 
wheels.
    Mr. Doolittle. Railroad weights?
    Dr. Friday. The wheels from a railroad----
    Mr. Doolittle. Oh, railroad wheels?
    Dr. Friday. They're very heavy.
    [Laughter.]
    Mr. Doolittle. That's very interesting.
    Dr. Friday. I would invite you, sir, if you're interested 
in this technology, to visit the laboratory in Seattle. All the 
capabilities, all the engineering work has been done in the 
NOAA laboratories there, and it is a spectacular engineering 
site.
    Mr. Doolittle. So if it's nylon cable, I guess that has an 
indefinite lifespan?
    Dr. Friday. No, sir, it does not. It does eventually decay. 
Also, we have not eliminated all the problems with that. These 
cables ac-

tually hum a little because of the currents passing them, and 
sharks don't like that. So when you pull up this equipment, you 
find a lot of shark teeth and shark prints on the instrument 
chain and on the upper reaches of the moorings.
    Mr. Doolittle. And how many years have these been there?
    Dr. Friday. The network was completed in 1994, and we have 
been operating it as a complete network ever since. We are now 
involved in trying to put some additional buoys in the northern 
Pacific because we know that although the tropical Pacific El 
Nino signal accounts for a large amount of the variability, 
there is also some changes in the northern Pacific Ocean that 
we are beginning to recognize that may affect the exact way 
that the El Nino storm system plays out. But the difficulty is 
the northern Pacific is much more hostile than the tropical 
Pacific is. The type of storms that affect the tropical Pacific 
are relatively mild compared to the tremendous storms up coming 
off the Gulf of Mexico and in that arena, that may have waves 
that just are incredibly large. And so we have another major 
engineering feat to be able to do that.
    Our laboratory in Seattle says that they have solved that 
problem. They have now installed two of the buoys out there, 
and this winter is going to tell us whether they're really 
solved that problem with the winter storms there or not.
    Mr. Doolittle. Can you describe briefly what's different 
about those than the ones you've got?
    Dr. Friday. No, sir, I can't, but I can provide that 
information to you.
    Mr. Doolittle. OK. Maybe they're just heavier weights and 
thicker cable or something.
    [The information referred to follows:]
----------
    The ATLAS buoys in the TAO array moored along the 
Equatorial Pacific are 2.3m diameter fiberglass torroids. They 
are inexpensive to fabricate, structurally rigid, and very 
stable instrument platforms. However, they have high drag due 
to the hull shape and limited weight capability. A new buoy has 
been designed for the North Pacific with lower drag and 
increased weight capability. An effort has been made to keep 
the cost down by using proven design and fabrication methods 
employed in the production of the torroid buoys. The new buoys 
are approximately 30 percent larger and require larger mooring 
components including wire rope, nylon line, hardware, and 
anchors. Consequently a high latitude mooring costs more that 
an equatorial mooring--probably an additional $10,000 each. The 
top of the instrument tower is typically 4 to 5 meters above 
the water surface.

    The initial effort to move into the North Pacific was 
driven by a NOAA funded project to develop a real-time Tsunami 
Reporting system using open ocean observations and satellite 
communications. (PMEL received additional funding from DARPA.) 
Two moorings were deployed in FY 97 and additional deployments 
of these moorings are scheduled in the future (see http://
www.pmel.noaa.gov/tsunami-hazard/). We are also supported by 
the National Ocean Partnership Program to place two air-sea/
upper ocean climate reference station moorings in the North 
Pacific during FY98-99. The first will be at Ocean Weather 
Station P. and the second will be NW of Hawaii near 165W. These 
buoys will carry a very different sensor package than the 
Tsunami buoys, but the buoy is the same. You correctly stated 
in your testimony that our goal has been to develop a mooring 
system that can carry out a variety of tasks in the ice-free 
but still very demanding North Pacific environment.

    Dr. Friday. There is also something to do with the exact 
design of the instrument chain itself, so that it's better able 
to ride the waves, and the fact that the mooring line itself is 
not taut here, but it's flexible, so that the buoy can ride up 
and down on that. But I don't know--I'm sorry, I just simply 
don't know the precise details of that.
    Mr. Doolittle. Just out of curiosity, have you had problems 
with ships running into those or something?
    Dr. Friday. Every time I have an opportunity of addressing 
any of the marine community, I try to point out the fact that 
buoys are their friends. They should not be used to tie up 
ships and they should not be used for target practice, because 
that happens to frequently the buoys, not only the deep ocean 
buoys along the coast.
    Mr. Doolittle. And these buoys are 6 or 8 feet high?
    Dr. Friday. They're probably on the order of 20-feet tall. 
They're on the order of around 10-foot across, and as I 
indicated, the instrumentation chain goes down for about 500 
meters below the surface.
    Mr. Doolittle. Well, that's very impressive. What did each 
of those cost?
    Dr. Friday. They cost approximately $50,000 to $60,000 
apiece, but it does take the operation of one entire 
oceanographic research vessel to maintain them on an annual 
basis.
    Mr. Doolittle. So that just plies the waters back and 
forth.
    Dr. Friday. Back and forth. The total annual operating cost 
to maintain this network is approximately $5 million a year. 
Now that, in one respect, seems like a lot of money, but when 
you understand, as you read off the impacts of the 1982/83 El 
Nino, with the economic loss and the loss of life associated 
with that, that's a very small cost to pay to be able to have 
the information to provide advanced forecasts of the situation 
as it evolves.
    Mr. Doolittle. Well, it certainly seems to me that it's 
money well-spent. I commend you and the organization for being 
able to cause that to occur and hope you can get the next phase 
taken care of.
    Dr. Schaefer, you indicated that a number of stream gauges 
have been taken offline in recent years. Could you explain why 
that's happening when the need for these gauges and additional 
data seems ever more critical?
    Mr. Schaefer. I'll give you the beginning of the answer, 
and then I'd like to turn, if I may, to Dr. Casadevall to 
expand on that.
    We've been, of course, under funding pressure in general at 
the USGS, and we've had to make some difficult choices to fund 
our projects. I think, in general, what we're seeing right now 
are indications that we're beyond cutting into fat, and we're 
beginning to cut into the meat of some parts of our program. We 
recognize that this is a problem, and we want to address it, 
and I hope to be able to address it in the coming fiscal year, 
and to some extent in this fiscal year.
    Also, what's happening is this system is really a system 
that comes about through the work of many cooperators on the 
State and local level, and our cooperators are also under 
financial pressure. Most of our gauging sites actually exist 
because of a 50/50 cost share. So when one of the cooperators 
pulls out, we're not in the position to pick up the other 50 
percent. Now in some cases we have been successful in finding 
others that are able to do that, and we've been able to retain 
some stations that otherwise would have been lost.
    But let me, if I may, ask Dr. Casadevall to expand on that.
    Mr. Casadevall. Good afternoon, Mr. Chairman.
    With regard to the gauge system, as Dr. Schaefer showed you 
in the graphic, we have about 7,000 gauges nationally. The 
image he showed you was the coterminous U.S. We also have 
gauges in the States of Hawaii and in Alaska.
    Over the last few years, we've decreased the number of 
gauges from about 7,400 down to about 7,000, but what we've 
increased are the number of gauges that send data in real time 
through radio telemetry. And for the kinds of issues that we're 
addressing here in relation to El Nino, it's really essential 
that managers, water managers, dam and reservoir managers, city 
planners, and others have data in real time. So, on the one 
hand, we've decreased the number of total gauges, but some of 
those gauges, remember, were only visited on a monthly basis, 
and data is retrieved or was retrieved from those static sites 
on a monthly basis. On the real-time telemetry, we can get data 
as frequently as every 15 minutes. This is really the essential 
data that planners need and that folks in the Bureau of 
Reclamation and the folks in the Flood Forecast Office of NOAA 
need to be able to make decisions that affect people's lives 
and safety, which really is the first priority for us in this 
activity.
    Mr. Doolittle. Could you describe a gauge? In the 
Sacramento River, how big would one be? How costly is it? Where 
is it typically located?
    Mr. Casadevall. Well, if you've ever crossed a bridge, for 
example, going across a river--and on the Sacramento River we 
can use an example--you've probably noticed a small house that, 
for better comparison, it looks like an outhouse; it looks like 
a portapotty. It's about 8 feet tall; it's about 4 feet on a 
side. Often on the top you'll see an inclined panel. That's a 
solar panel that is used to power the electrical components 
that are part of the gauge. What you don't see is that 
extending out from the gauge house down into the river is a 
hydraulic sensor or a pressure sensor that passes under to the 
base of the river. And it's information from that pressure 
sensor that gets recorded in the gauge house. That information 
then gets--if it's a telemeter site, radio-telemetered or 
satellite-telemetered back to a recording station in 
Sacramento, for example, at our USGS Water Resources District 
Office.
    On the other hand, if it's a site that's only visited once 
a month, a hydro-tech, a hydro-technician will go out and 
retrieve the record. The record will be recorded right there 
onsite.
    So you'll often see these, and I invite you the next time 
you're on a bridge driving across the river, keep an eye out 
for the outhouse, and if we're smart, you'll see the new USGS 
logo in bright green letters. We plan to put these on our more 
than 7,000 stream gauges around the country.
    Mr. Doolittle. And that sits in the bottom of the river and 
measures the velocity of the water?
    Mr. Schaefer. Well, it measures the pressure of the water 
on the bed or the river, and as you add more water to the 
stream or to the river, of course, the pressure or the 
hydrostatic----
    Mr. Doolittle. OK.
    Mr. Schaefer. [continuing] head increases on that. In 
addition, we've made very precise cross-sectional measurements, 
and we do what we call calibrating that gauge, so that when 
there is a pressure change, we know, because of calibration, we 
know how much additional water is now being carried by the 
channel.
    Mr. Doolittle. Does all of that data feed into, I guess in 
the case of California, or at least the Sacramento area, into 
that place on El Camino there, the State and Federal 
Governments joint operation center?
    Mr. Schaefer. In El Camino--in Menlo Park, California or--
--
    Mr. Doolittle. No, no.
    Mr. Schaefer. Oh, up in Sacramento?
    Mr. Doolittle. El Camino and Watt in Sacramento.
    Mr. Schaefer. The data from our continuous telemetered 
sites, the majority of those in the State of California are 
available on the Internet, on the USGS California State 
website. If we had a computer here right now, we could find the 
river closest to your home near Sacramento and we could find 
what gauge was there----
    Mr. Doolittle. Oh, OK.
    Mr. Schaefer. [continuing] and we could look at the stage 
of the river right now in real time with data that's up to 4 
hours old.
    Mr. Doolittle. Is there a standard cost on a stream gauge?
    Mr. Casadevall. Well, let's suppose there's a gauge already 
there and we wanted to add telemetry. To add telemetry, we're 
talking about a telemeter package costing between about $3,500 
and about $5,000. To install a gauge itself, to put the 
outhouse-type structure in, to put the piping in, to put the 
pressure transducer in, and the associated equipment, I believe 
there's a range in prices depending on where the site it and 
how much hardening of the site you have to do, but my 
recollection from years ago, that it's in the range of $35,000 
to $60,000 per site. Once again, it depends on how hard we want 
the site to be, how hardy the site should be, to withstand 
flood conditions.
    Mr. Doolittle. Well, I think the Committee analysis 
mentioned that the middle fork of the American River is not 
very well-monitored, and I don't know what it takes to have it 
be well-monitored, but, obviously, we're spending the money out 
in the Pacific, which is money well-spent. Are there plans to, 
especially in a State like California with high mountains and 
the short distance from the mountains to sea level, et cetera, 
and right there on the coast, the first to receive the brunt of 
El Nino or some bizarre storm--are there some plans to increase 
monitoring in some of those types of areas?
    Mr. Schaefer. Well, right now in the State of California we 
have more than 700 stream gauges. That's about 10 percent of 
our network nationally. I would have to check with my 
colleagues to find out what the plans are for this winter, and 
we'd be happy to get that information for you, and----
    Mr. Doolittle. Well, I'd be interested in that, and 
interested in supporting your efforts to increase your 
stations, both in the ocean and on land.
    [The information referred to follows:]
    Mr. Doolittle. I guess I'm going to have to run out for a 
vote again. If we ask a couple more questions, we'll excuse you 
folks, because it's a half hour's worth of votes, apparently.
    For Dr. Matthews, how do Bureau of Reclamation managers 
factor snow melt and runoff into reservoir management 
decisions?
    Mr. Matthews. The Bureau of Reclamation works very closely 
with the National Weather Service in their forecasting systems, 
and we take the snow water equivalent analyses throughout the 
mountainous West from the National Operational Hydrologic 
Remote Sensing Center. Their data has up to one kilometer 
resolution. We incorporate that through the National Weather 
Service River Forecast System to forecast runoff, where that's 
available.
    We also look at that data directly and evaluate what the 
snowpack snowwater equivalent is and use that. We have some 
research tools that are currently being used to physically 
estimate the runoff from both accumulated snow and temperature 
forecasts. So that snowpack information is used on a routine 
basis throughout the spring runoff season.
    Mr. Doolittle. We had instances in California where they've 
had a lot of snow, and then a tropical storm comes in and melts 
it. I think 1986 was one of those events. Is that likely to be 
the case or is it a small possibility? El Nino is a warmer 
phenomenon, right? So are we more vulnerable to receiving 
storms that might melt our snowpack?
    Mr. Matthews. We could be. In many of the El Ninos, we have 
a subtropical jet stream which brings in warm air out of the 
subtropics, and in that case there could be an earlier snowpack 
meltdown, which is what happened to a certain extent in the 
California flood January, 1997, I believe.
    Chet Bowling is the expert here from our Central Valley 
Operations Office. I think he would be in a better position to 
describe the California events than I would.
    Mr. Doolittle. OK, we'll come back to him.
    Dr. Schaefer, Dr. Friday, in his testimony, stated that 
some regions of the U.S. are relatively more vulnerable to the 
effects of El Nino. Does the USGS concentrate its stream 
gauging in the areas that are more sensitive to climate 
variability?
    Mr. Schaefer. In part it does, and in part it doesn't. I'm 
sorry to give you a two-handed answer there. The stream gauge 
network really originally advanced for other than flood control 
and monitoring reasons. It was really put in place to help 
people monitor water conditions for use in irrigation, for 
example. And so that means that it does make sense for us to 
step back periodically and say, all right, now that we have 
this dual-use for this system and we're beginning to apply it 
to flood-related questions, where should we look strategically 
to put additional sites? And that's the reason why we're not 
only concerned about the overall downward trend in sites, but, 
frankly, we'd like to have it be an overall upward trend, so 
that we can begin to fill those gaps.
    Mr. Doolittle. I'd like to encourage you, as the chairman 
of this Subcommittee, to get that trend going back the other 
direction upward, because I think it's clear we're going to 
need to know as much as we can. Even knowing more now than we 
did before, we still are in the dark to a large extent.
    I will have further questions, and I'll just tender them in 
writing and ask you to supply the answers as expeditiously as 
possible, and hold the record open for that point.
    [The information referred to follows:]
    Mr. Doolittle. And then I'll excuse the first panel, and 
thank you all very much for your testimony.
    Yes, sir?
    Mr. Schaefer. Sir, I'd just like to thank you. This is a 
very technical area, and people don't often ask questions about 
how our monitoring systems operate. We appreciate your 
interest, and we would like to work with you in making sure 
that we've got the most robust monitoring system that we can 
get.
    Mr. Doolittle. Thank you. We'll excuse this panel and we'll 
recess. When we come back, we'll begin the final panel.
    [Recess.]
    Mr. Doolittle. I thank you, ladies and gentlemen, for 
persevering. We had an unusual series of votes, and they're 
going to have an even more unusual series shortly. So we'll try 
and do this in the little period between that.
    Let's call up the second panel, and please come forward and 
raise your right hands, please.
    [Witnesses sworn.]
    Mr. Doolittle. Let the record reflect that each answered in 
the affirmative.
    Thank you, gentlemen. Please be seated.
    Our first witness will be Dr. Konstantine Georgakakos, 
president of the Hydrologic Research Center, Scripps 
Institution of Oceanography, the University of California at 
San Diego. And Dr. Georgakakos, you're recognized.

  STATEMENT OF KONSTANTINE GEORGAKAKOS, PRESIDENT, HYDROLOGIC 
     RESEARCH CENTER, SCRIPPS INSTITUTION OF OCEANOGRAPHY, 
              UNIVERSITY OF CALIFORNIA, SAN DIEGO

    Mr. Konstantine Georgakakos. Thank you very much, Mr. 
Chairman, for the opportunity to testify concerning the 
involvement of the Hydrologic Research Center in the 
forecasting and in forecasting research associated with the 
predicted El Nino event. The Hydrologic Research Center is a 
nonprofit research corporation in San Diego, California, and in 
this testimony I will focus in particular on Center research 
activities pertaining to operational stream flow forecasting 
(as opposed to climate forecasting) for flood warning and water 
resources management.
    Two Center research activities are directly relevant. The 
first one, assisting the California-Nevada River Forecast 
Center of the U.S. National Weather Service in improving the 
short-and long-term forecasts of Folsom Lake inflow in east 
central California. These forecasts will be used in the 
operation and management of the Lake waters and flow release.
    And the second one, assessing the utility of integrated 
forecast-control methodologies for the operation and management 
of reservoir systems. Case studies in Iowa and California are 
now being conducted for this research.
    Our research relies on mathematical hydrologic models of 
the watershed processes and of the flow forecast uncertainty. 
Watershed process considered for the case studies are snow 
accumulation and melt, and surface and subsurface flow. The 
flow forecast uncer-

tainty exists for three reasons. No. 1, incomplete coverage of 
watershed area by sensors--and I mean by precipitation sensors, 
temperature, wind, and so on--No. 2, large errors in 
meteorological forecasts used to drive the hydrological models, 
and that's a very large source of uncertainty for the smaller-
scale hydrological models; and then mathematical model 
approximations of complex, natural processes.
    Our results so far support the following conclusions:
    No. 1, Snowmelt volume estimates in hydrologic forecasts 
are strongly affected by the density of the recording 
precipitation sensors in mountainous areas such as California. 
The accuracy of the computed snowmelt volumes within the 1,800, 
roughly, square miles of Folsom Lake Watershed decreases 
substantially with decreased watershed coverage by 
precipitation measurement sensors. For example, there are 
currently large areas drained by the Middle Fork of the 
American River with poor precipitation sensor coverage.
    No. 2, substantial improvement of operational flow 
forecasts is attained when the current flow forecast systems 
are upgraded to include models for uncertainty and updating 
from flow measurements in real time. The improvement is mainly 
in the reduction of forecast errors for unusually high or low 
flow rates. Thus, in many cases meteorological forecasts with 
large uncertainty, as obtained in mountainous California during 
a storm event, can be used to derive useful hydrologic 
forecasts.
    No. 3, the third conclusion, in a case study involving data 
since 1904 from the Iowa River at the Iowa City gauging station 
in Iowa, statistically significant seasonal stream flow 
associations to ENSO were found. Analogous associations were 
not found for the American River at Folsom Lake. These studies 
are ongoing, and although strong associations between ENSO and 
seasonal stream flow volumes in extremes in the Southwestern 
U.S. have been found, no such statistical associations have 
been identified for central and northern California. It doesn't 
mean they don't exist. It is possible that the three-to-seven-
year ENSO signal is concealed by the extreme year-to-year 
variability of stream flow in the Folsom Lake Watershed.
    The fourth conclusion, substantial benefits for operational 
reservoir water management were obtained for the Saylorville 
Reservoir in Iowa, one of our case studies, when flow forecasts 
were used as input to the decision process with due account for 
forecast uncertainty, and that is important. On the basis of 
extensive computer simulations of the watershed-reservoir 
system, it was found that using coupled forecast-control 
methodologies reduces reservoir management sensitivities to 
climatic variability and to the large uncertainties associated 
with the forecast of such variability by current climate 
models. Analogous simulations are in progress for Lake Folsom 
in California, and my brother, Professor Aris Georgakakos of 
Georgia Tech, will testify to that effect.
    I wish to make the following topical recommendations:
    One, conduct detailed studies to quantify the uncertainty 
associated with the estimation of precipitation and snowmelt 
over the Sierra Nevada in California.
    Two, advanced stream flow forecast procedures should be 
implemented and utilized in parallel to current operational 
ones to evaluate increased benefits to operations. Such 
procedures should include models for uncertainty of 
meteorological forecasts and should utilize stream flow 
observations to improve the forecasts continuously in real 
time.
    Third, in parallel to No. 2 above, coupled forecast-control 
methodologies with due account for forecast uncertainty should 
be implemented in prototype watersheds. In this context, the 
utility of climate forecasts for increasing the benefits of 
reservoir management should be quantified.
    Real-time flow forecasting and reservoir water management 
are important operational functions for mitigating natural 
disasters. These functions, vital for present-day communities, 
have their bases on hydrologic science and engineering and 
water resources systems analysis. As the requirements of the 
public for safety and reduction of damage losses from natural 
disasters increase, it is important to formulate a national 
plan for the increased effectiveness of these operational 
functions.
    I have argued elsewhere in the scientific literature that 
to achieve these goals and in analogy to the establishment of 
our National Center for Atmospheric Research in the atmospheric 
sciences, it appears necessary to establish a National Center 
for Hydrology and Water Resources. I firmly believe that, with 
the establishment of such a national center, much progress will 
be made almost immediately on a national level by concerted 
efforts to enhance the flow of information from research to 
operations. Such a center is envisioned as a collaborative 
effort among universities, the Federal Government, and the 
private sector.
    With this last important recommendation, I now conclude my 
testimony. I will be pleased to answer your questions.
    [The prepared statement of Mr. Konstantine Georgakakos may 
be found at end of hearing.]
    Mr. Doolittle. Thank you, sir. And our next witness is Dr. 
Aris Georgakakos, director of the Georgia Water Resources 
Institute, Georgia Institute of Technology. Dr. Georgakakos.

    STATEMENT OF ARIS GEORGAKAKOS, DIRECTOR, GEORGIA WATER 
      RESOURCES INSTITUTE, GEORGIA INSTITUTE OF TECHNOLOGY

    Mr. Aris Georgakakos. Thank you, Mr. Chairman. It is, 
indeed, a pleasure and an honor to testify before this 
Committee. I'm an expert in the operational management of 
reservoir systems, and my testimony particularly addresses the 
value of climate and hydrologic forecasting in reservoir 
operation.
    With proper management, reservoirs can provide vital 
services to human communities, including the mitigation of 
severe floods and droughts, the generation of hydroelectric 
energy, the provision of water supply to urban, industrial, and 
agricultural areas, recreation, navigation, and the sustainable 
management of riverine ecosystems.
    However, the extent to which reservoirs succeed in 
providing these services depends critically upon the manner in 
which they are operated. Consider, for example, the Folsom 
Reservoir in east central California, shown on figure 1 of my 
written testimony, which is expected to provide flood control, 
generate electricity, provide water supply for irrigation, and 
maintain a certain downstream flow rate in the American River 
for water quality and ecosystem preservation. In the interest 
of hydropower, the reservoir should always be full to create 
the highest possible hydraulic head for the turbines and 
maximize their power output. However, from a flood control 
standpoint, there is a need to draw the reservoir level down in 
anticipation of floods, free up storage space, and accommodate 
flood volumes without causing downstream damage.
    In a world without uncertainty, the reservoir managers 
would know the magnitude of the flood precisely and could run 
the turbines at full power prior to the flood to lower the 
reservoir just enough to receive and contain the flood volume. 
This scenario would be ideal because it would avoid downstream 
damage; it would pass the entire flood through the turbines 
without wasting power to spillage, and it would maintain the 
reservoir as high as possible, maximizing its value for 
hydropower, flood prevention, and the other water uses.
    Unfortunately, in the real world in which we live, 
reservoir managers can only guess the magnitude of the upcoming 
floods through imprecise climate, weather, and stream flow 
forecasts, and the challenge is to balance the risk for flood 
damage against the adverse impacts on hydropower and other 
reservoir uses. Whether their decisions are successful or not 
depends critically on two factors: First, the quality of stream 
flow forecasts and, second, the ability to fully utilize them 
through an integrated and flexible decision system and process.
    In principle, good, quality stream flow forecasts are 
expected to benefit reservoir management. However, the actual 
benefits depend on many system-specific factors, such as the 
lead time and reliability of forecasts, reservoir size related 
to inflow volume, hydrologic characteristics of the outlet 
structures, turbine discharge capacities, flood damage 
thresholds, and the levels and timing of other water demands.
    Thus, to assess the value of stream flow forecasts in the 
management of Folsom, I developed a computer model which 
includes a forecasting, a decision, and a simulation component. 
A brief description of this model and its underlying 
assumptions appears in my written testimony and will not be 
elaborated here, other than to say that it represents most 
Folsom features, which were kindly provided by the Bureau of 
Reclamation and the Folsom operators, and is designed to assess 
the relative differences in reservoir performance under 
different forecast scenarios of low, intermediate, and perfect 
skill. These assessments are made by recreating the Folsom 
response over the historical period from 1964 to 1995, assuming 
that the reservoir was operated with the guidance of the 
decision support system.
    Folsom's performance is measured in accordance with three 
criteria: flood damage in millions of dollars, annual energy 
generation revenue in millions of dollars, and annual spillage 
in million cubic feet. The results indicate that Folsom would 
benefit substantially from improved forecast skill. Most 
notably, flood damage would be mitigated from approximately 
$5.3 billion in the case of low-skill forecast to about $220 
million in the case of intermediate-skill forecast. Relative to 
energy generation, the value of intermediate over low skill 
forecasts is approximately $1 million per year. And in the 
extreme scenario of perfect forecast skill, flood damages would 
be fully mitigated, and energy revenues would be increased by 
another $2 million per year.
    While the previous results are annualized, the actual year-
by-year benefits would be much higher. For example, figures 3 
and 4 of my written testimony illustrate this comparison for 
the high-flow year of October 1, 1996 to September 30, 1965. 
For this particular period, the low-skill forecast scheme would 
cause heavy flood damages, on the order of $4.3 billion, 
whereas the other two would completely avoid flooding. 
Similarly, energy revenues would increase by approximately $4 
million from the low to the intermediate-skill forecast 
scenario, and almost $8 million from the intermediate to the 
perfect case.
    One last point to emphasize is that the above-referenced 
benefits can only be realized if forecasts are used in 
connection with dynamic decision schemes that fully account for 
forecast uncertainty. By contrast, static reservoir rule 
curves, which are traditionally used in the operation of 
reservoir systems, would fall short of realizing the value of 
improved forecasts.
    I have so far argued that good quality, long-lead stream 
flow forecasts, coupled with appropriate decision support 
systems, improve reservoir management. An important and 
relevant question is: By how much can stream flow forecasts 
actually be improved?
    A reliable answer to this question can only come from the 
continuing research on coupled climate, weather, and hydrologic 
prediction systems. My experience with such integrated 
approaches in the midwestern U.S., east-central Africa, and 
South America is promising, albeit at a preliminary stage.
    The most concrete improvements in short-range--that is, up 
to one month--stream flow forecasting can be realized from the 
use of hydrologic watershed models. To assess the value of such 
models, I also conducted an experiment using an adaptation of 
the National Weather Service River Forecast System, coupled 
with the Folsom decision system, in collaboration with Dr. 
Konstantine Georgakakos of the Hydrologic Research Center in 
San Diego, California. The results were comparable to those of 
the intermediate-skill forecast experiment mentioned earlier, 
indicating that such forecast decision systems could accrue 
significant operational benefits.
    Lastly, as a first step toward assessing the value of El 
Nino information in the management of Folsom, I investigated 
the correlation between sea surface temperatures at the 
equatorial Pacific Ocean and monthly Folsom inflows. I was 
unable to find any significant relationship between these two 
variables, which led me to conclude that a strong El Nino does 
not necessarily imply a predictable change in the weather 
patterns over the Folsom drainage basin. This conclusion, 
however, may not apply to other regions of the western United 
States. In fact, similar case studies have shown that, if 
strong enough, this correlation between El Nino and stream 
flows does improve forecast skill and reservoir operations.
    In conclusion, I'd like to reiterate that integrated 
forecast decision systems can significantly mitigate flood 
damage, increase the value of energy generation, and 
potentially benefit all other water uses. However, the 
magnitude of these benefits are system-specific and can only be 
assessed on a case-by-case basis.
    In addition, better forecasting procedures do not by 
themselves imply operational improvements. Such improvements 
can only be realized through coupled forecast decision systems 
and institutional processes. In this regard, there is a 
pressing need to make water resources professionals fully aware 
of the capabilities and benefits of integrated decision systems 
and processes relative to traditional operational practices. An 
effective means to stimulate this transfer of technology from 
researchers to the user community is through the support of 
prototype demonstration projects throughout different regions 
of the U.S. with the involvement of both groups. I would like 
to urge the Committee to support such demonstration projects 
through existing or new funding programs, an example of which 
is the GENEX Continental Scale International Project (GCIP) of 
the National Oceanic and Atmospheric Administration (NOAA).
    Thank you for the opportunity to present this testimony on 
the implications of recent advances for the management of water 
resources.
    [The prepared statement of Mr. Aris Georgakakos may be 
found at end of hearing.]
    Mr. Doolittle. Thank you, sir.
    Our next witness will be Dr. Soroosh Sorooshian, professor 
of the Department of Hydrology, University of Arizona. Dr. 
Sorooshian.

   STATEMENT OF SOROOSH SOROOSHIAN, PROFESSOR, DEPARTMENT OF 
                HYDROLOGY, UNIVERSITY OF ARIZONA

    Mr. Sorooshian. Thank you, Mr. Chairman. As we've heard 
today and read in many places, we do know that this year's El 
Nino is the strongest yet recorded. We also know that 
statistical evidence points to above-average winter 
precipitation for the Southwest United States. However, these 
facts do not constitute sufficient information for water 
resources and emergency managers to initiate major changes in 
operating practices. Such decisions require reliable 
information about the expected arrival time of significant 
storm systems, their expected duration, and intensities.
    A storm system that arrives during warm weather may not 
result in snow accumulations at high elevations, but may 
instead produce large amounts of runoff and potential flooding. 
Without accurate information about the timing and quantities of 
precipitation and stream flow, it is very difficult to plan 
timely evacuation from flood-prone areas or to pre-release 
large quantities of stored reservoir water to help mitigate the 
flooding.
    On the other hand, precipitation which arrives during a 
cooler period may accumulate as snow at high elevations. In 
this case, water resources managers have greater flexibility to 
evaluate options and to decide on an appropriate operational 
strategy. Under this scenario, it would be a lot less risky to 
commit to other reservoir releases, knowing that melting of the 
above-average snowpack will provide the water necessary to fill 
the reservoirs later in the season.
    However, rapid warming could cause sudden large releases of 
melt water leading to late winter and/or spring flooding. It is 
critically important, therefore, that accurate estimates of the 
volume of water in the snowpack and timely and accurate 
temperature forecasts during the melt season be available.
    I wish to strongly emphasize that sound water resources 
management decisions in the Southwest will require far more 
information than merely the knowledge of a strong El Nino 
signal. Water resources managers are rightly reluctant to order 
early reservoir releases without further information. In the 
few instances that I know of where early decisions have been 
made regarding reservoir releases and other water management 
issues, the knowledge that this will be a strong El Nino year 
has been only one of several useful pieces of information, but 
not the sole decision factor.
    For instance, the Salt River Project, which supplies water 
to the greater Phoenix area, has incorporated the information 
about the strong El Nino signal into its decision to reduce 
ground water pumping by some 40,000 acre feet this year. This 
requirement will instead be satisfied from the reservoir system 
based on expectations that an above-average spring snowmelt 
runoff will fill the reservoirs to their normal level.
    Water resources and emergency managers are accustomed to 
making decisions based on probabilistic information. While the 
knowledge of a strong El Nino year has enhanced the probability 
for a wetter-than-average year in the Southwest, it has not 
reduced uncertainty of many other factors critical in making 
decisions regarding major deviations from normal operating 
practices. In order to enhance the quality and usefulness of 
both short-term, meaning hours to days, and extended, weeks to 
months, forecasts, the reliability of hydrologic prediction 
systems for the western U.S. must be improved. The primary 
components of this system are the quantitative precipitation 
forecasts, extended stream flow prediction, and more accurate 
methods for estimating snow accumulation, particularly in the 
mountainous regions, and high resolution accurate rainfall 
measurements are critical for forecasting rapidly developing 
flood events.
    The strength of the current El Nino signal has attracted a 
lot of media attention and has generated much-needed public 
attention to this climatic phenomenon. The climate research 
community is to be commended for developing the capability of 
predicting it with such a high degree of accuracy. Perhaps the 
greatest benefit of this prediction to the water resources 
management community has been in encouraging very close 
cooperation among the Federal and state agencies responsible 
for various aspects of water resources management and 
hydrologic services.
    As an example, the cooperation over the past several months 
between the U.S. Bureau of Reclamation, the National Weather 
Service's Colorado Basin River Forecast Center, and the USGS 
has resulted in close coordination for sharing modeling and 
observational information required for improved management of 
the reservoir systems on the Colorado River. Continued 
cooperation among these agencies will be critical to the 
development of an operational hydro-

logic prediction system for the western U.S. to be used for 
water resources management in both El Nino and non-El Nino 
years.
    It's worth noting that while statistical evidence points to 
a wetter-than-average year in the Southwest during a strong El 
Nino year, the wettest winter on record, which was 1993 in the 
White Mountains and the surrounding areas of southern Arizona, 
was not an El Nino year. A reliable hydrologic prediction 
system is crucial for the efficient management of western water 
resources, irrespective of whether we are experiencing an El 
Nino weather pattern or not.
    Finally, most of the Southwest at the current time is below 
average in terms of its precipitation. In the Tucson area, we 
are about three inches below normal, and I hope you and other 
folks here will pray for us to receive the additional 
precipitation expected from a strong El Nino year. Thank you 
very much.
    [The prepared statement of Mr. Sorooshian may be found at 
end of hearing.]
    Mr. Doolittle. Thank you, sir. Our next witness is Mr. 
Richard Andrews, director of the Governor's Office of Emergency 
Services, the State of California. Mr. Andrews.

  STATEMENT OF RICHARD ANDREWS, DIRECTOR, GOVERNOR'S OFFICE, 
                 EMERGENCY SERVICES, CALIFORNIA

    Mr. Andrews. Thank you, and thanks for the opportunity to 
speak today on the many activities underway in California in 
preparation for El Nino, as well as issues related to Federal 
policies and practices that influence our preparedness and 
disaster recovery efforts.
    Despite the considerable uncertainties about what the 
impacts of El Nino will be, we are confident that State and 
local governments will be well-prepared to meet the challenges 
that severe winter weather may bring. As is well-known, 
California has in this decade had repeated and varied 
experience in coping with the consequences of natural 
disasters, including large fires, earthquakes, and floods. We 
have improved our local and State response systems after each 
event. California has, I believe, the most effective emergency 
response system in the Nation.
    The El Nino forecasts have helped accelerate ongoing 
preparedness efforts that followed this January's historic 
floods in central and northern California. In January 1997, the 
State experienced serious flooding in 48 counties, with damages 
totaling nearly $2 billion. These floods came only two years 
after a series of winter storms in the first quarter of 1995, 
where losses also totaled $2 billion, led Governor Wilson to 
declare for the first time in the State's history all 58 
counties as disaster areas.
    At the height of the 1997 winter storms, Governor Wilson 
established the Flood Emergency Action Team to review the 
lessons and to establish long-term strategies to protect 
Californians from future flood disasters. Following a series of 
hearings throughout the impacted region, the Flood Emergency 
Action Team made more than 50 recommendations that had been 
implemented to improve the State's flood-fighting systems, 
including improving emergency response coordination between 
public safety agencies at the local and State level, the Army 
Corps of Engineers, and local flood mainte-

nance organizations, improving also and expanding existing 
flood data. To expand the State's existing flood data, the 
California Department of Water Resources has installed 
telemetry linking nearly 50 stream-gauging sites in areas that 
have high flood probability.
    On October 6, Governor Wilson convened an El Nino Summit in 
Sacramento. In this Executive Order, signed on that date, the 
Governor directed that OES and the Department of Water 
Resources establish technical assistance teams and conduct a 
series of regional workshops throughout the State to review 
specific State and local preparedness action. Governor Wilson 
also signed legislation allocating $7.4 million for El Nino 
preparedness measures, including the prepositioning of flood-
fighting resources as forecasts become more specific.
    In addition to the actions being undertaken by the State, 
local governments, community groups, and businesses are taking 
unprecedented preparedness measures. Supplies have been 
stockpiled, storm drains cleared, evacuation procedures 
reviewed, and strategies for the care and shelter of 
individuals updated. Many cities and counties have held special 
flood preparedness drills and developed specific emergency 
plans for possible El Nino impacts. California will continue to 
provide the public with the best available information through 
the Internet, briefings and workshops, training sessions, 
technical assistance teams, and the stockpiling of equipment.
    The Federal Government is an important partner in the 
State's overall preparedness effort and plays an essential role 
in helping communities recover from the impacts of natural 
disaster. In a letter to President Clinton on October 6, 
Governor Wilson urged action on several concerns about the 
current Federal policies and practices that impacted the pace 
of recovery from past floods and which also affect current and 
future preparedness and recovery efforts, including, first, 
urging the Army Corps of Engineers to accelerate the timetable 
to make repairs to levies damaged in central California in 
January 1997. Policy disputes between Corps officials in 
California and Washington slowed recovery during the spring and 
summer months. Progress is now being made, and if we're 
fortunate to have winter storms hold-off until the end of 
November, we have received indication from the Clinton 
Administration that all but one of the critical central 
California flood control repairs should be in place to handle 
the flows. We would hope that this year's experiences of 
unnecessary delays would not be repeated following future 
flooding.
    In the event levies are not fully repaired, direct the 
Corps to undertake response preparations to improve their 
emergency response under Public Law 84-99. Again, we have 
received general assurances from the administration that these 
suggestions will be acted upon.
    Second, direct all Federal regulatory agencies to 
consolidate the needed approvals for flood channel clearance. 
Governor Wilson has directed all State agencies to place the 
highest priority on expediting approvals for this work. A large 
number of Federal agency approvals are also needed. Local 
agencies need clear and consistent permit requirements and 
procedures from Federal agencies.
    On October 24, the Corps of Engineers issued a nationwide 
31 permit for channel-clearing and sediment removal in Los 
Angeles County. Over this past weekend, the county began some 
of this important work. Governor Wilson remains concerned, 
however, that the Environmental Protection Agency has indicated 
that it will seek a policy review of the cumulative impacts of 
channel-clearing activities with the intention of requiring 
further mitigation work.
    Fourth, direct FEMA and the Corps of Engineers to modify 
policies for local agencies that conduct flood fights on flood 
control works. As a result of Federal policies arising from the 
1993 Midwest floods, there now exists in some important 
instances a disincentive for local agencies to assist in flood 
fights.
    Finally, direct FEMA to implement the recommendations 
relating to public assistance processes and policies made by 
the California congressional delegation in a September 15th 
letter to the FEMA Director.
    All Californians are grateful for the assistance we 
received from Congress and the administration as we join 
together on the levies to battle the flood waters of last 
year's storms. We are also grateful for the assistance we 
continued to receive in the flood's aftermath as communities 
and individuals worked to repair the damage and clean up homes, 
farms, and businesses. State agencies, local governments, 
community groups, and individuals are currently engaged in an 
unprecedented preparedness effort in advance of this year's 
winter. Our common goal is to reverse Mark Twain's famous 
aphorism: ``Everybody talks about the weather, but nobody does 
anything about it.''
    Thank you very much.
    [The prepared statement of Mr. Andrews may be found at end 
of hearing.]
    Mr. Doolittle. The next witness and final witness is Mr. 
Stephen K. Hall, executive director of the Association of 
California Water Agencies. Mr. Hall.

 STATEMENT OF STEPHEN K. HALL, EXECUTIVE DIRECTOR, ASSOCIATION 
                  OF CALIFORNIA WATER AGENCIES

    Mr. Hall. Thank you, Mr. Chairman.
    As you may know, the agencies that we represent are the 
local agencies that are the closest to the flooding when it 
occurs. So we have a lot at stake in this discussion, and we 
thank the Committee for its interest.
    I'd like to make three quick points regarding the weather 
in California and what we'll do about it. First, as you've 
heard this afternoon, and as we all well know, the weather in 
California is highly variable and unpredictable. In the last 20 
years, we've had 2 years that have fallen into the normal range 
in rainfall totals. Every other year has either been dry or 
wet.
    And there's probably no better example of the variability 
of weather than 1997, when we were both dry and wet. In January 
of this year, we received 30 inches of rainfall in the northern 
California watershed. Our largest flood control reservoir 
nearly filled in a week's time. The flooding that occurred 
caused $2 billion in damage; 120,000 people were put out of 
their homes, 9 people lost their lives. The spring that 
followed was the driest that we've had in 104 years, and so at 
the end of the irrigation season we faced the in-

credible situation of having record rainfall and flooding 
followed by delivery cutbacks to irrigators late in the season.
    That was due in part to the extreme weather variability, 
but also due to the state of the system that we use to operate 
for flood and water supply. We've done a pretty good job--in 
fact, I would say a very good job--in recent years of improving 
our capabilities to manage the system for both flood control 
and water supply, but the system itself is outdated and 
undersized. The heart of the system is in the Central Valley, 
our largest watershed. It's an impressive system, has 23 
reservoirs, over 1,800 miles of levies and channels, but most 
of that was built before modern construction techniques were 
available, and as the 1997 floods showed, it is very 
vulnerable. We will not finish all of the repairs necessary 
from those floods before the next flood season begins.
    It also lacks storage. As an example, on the Colorado River 
they average 15 million acre feet of annual runoff. They have 
60 million acre feet of storage, 4 years' worth of annual 
runoff.
    On the Sacramento River system, we have 16 million acre 
feet of storage to take care of 22 million acre feet of runoff, 
less than one year. That's why we floo so often. That lack of 
capacity means water managers have to make a lot of tough 
choices. The same system that protects lives and property also 
serves the Nation's largest economy. Water managers will always 
err on the side of public safety, as they should, but that 
means there is even more stress to operate for water supply.
    We've made great strides in stretching supplies, but the 
problem is getting worse because in the last 20 years our 
population has essentially doubled. We have done nothing to add 
to our flood control or water supply infrastructure.
    That leads to my third point: Not only have we not grown 
the system, there have been substantial new demands placed on 
it. The Bay-Delta Estuary, which is the heart of our water 
supply and flood control system, is also an estuary that 
contains over 120 species, some of which are endangered. We 
have dedicated several million acre feet of water to protection 
of those species in the last several years. That has put a 
substantial additional demand on the system which is straining 
our water supplies even further.
    I might note that the environment also suffers from floods 
as well as water shortages. This year biologists are very 
concerned about the impacts to salmon populations from the 
floods early in the year that wiped out substantial habitat. 
New storage on the system would help regulate flows to protect 
lives, the economy, and the environment.
    From these observations that we will continue to have 
variable weather, we do have a system that is too small and out 
of date, and that we have put substantial new demands on the 
system in recent years, one has to conclude that we need to 
improve and add to our existing system. We are in support of a 
recently announced study by the Corps of Engineers to undertake 
a comprehensive reassessment of the flood control system in the 
Central Valley. We, likewise, appreciate what the Wilson 
administration is doing to prepare for the El Nino year that 
we're facing, and we agree with them about the need to take 
action at the Federal level.
    Perhaps most importantly, we are actively involved in the 
CALFED process, a comprehensive look at the Bay-Delta to focus 
on ecosystem restoration, improved flood control, and improved 
water supply. We believe that it is our best hope in the 
foreseeable future to bring about the kinds of improvements 
necessary in our system to meet the needs of the people of the 
State.
    We appreciate the fact that the Federal Government has been 
involved in the CALFED process because there is clearly a 
Federal interest in doing so. We look forward to working with 
you, Mr. Chairman, and with the Congress in future years to 
complete the planning process in CALFED, and then move on to 
implementing the plan that it produces.
    Thank you.
    [The prepared statement of Mr. Hall may be found at end of 
hearing.]
    Mr. Doolittle. Thank you. Well, we got through that, and I 
only missed one vote, but I can't miss another one. So let's 
take advantage of what time we have for questions.
    Mr. Hall, your testimony talks about the need for 
additional water storage in California. Are you referring to 
on-stream storage, off-stream, or both?
    Mr. Hall. We're certainly open to both. Certainly, there is 
a need for improved flood control on-stream, and there are 
great potential advantages to off-stream storage, even for 
flood control, as well as water supply. So we would support 
certainly a lot of off-stream storage, and in selected cases 
on-stream.
    Mr. Doolittle. Thank you.
    Mr. Andrews, do you agree with Mr. Hall's general 
assessment about the strains on California's water supply 
system?
    Mr. Andrews. Yes, absolutely, and as part of the second 
phase of the Governor's Flood Emergency Action Team work, we're 
addressing issues of the long-term strategy that's needed 
because, clearly, as Mr. Hall indicated, the system itself is 
simply limited, and we can operate it as effectively as we 
possibly can, using the best forecasting information available, 
but there are simply inherent limitations to the system that 
can only be addressed by a more long-term comprehensive 
solution.
    Mr. Doolittle. To what extent has the State water project 
been evaluated for the potential impacts of the El Nino event?
    Mr. Andrews. Well, the Department of Water Resources is 
continually working with NOAA and with the National Weather 
Service, with the Reclamation groups, to evaluate it. David 
Kennedy, the Director of Water Resources, is confident that 
they're taking every measure that they can, given the 
uncertainty of the forecast right now, to try to make sure that 
the system can handle, to the extent it can given its overall 
capacity, what is now being forecast.
    Mr. Doolittle. Thank you.
    Dr. Sorooshian, in your opinion, how well are other water 
agencies and utilities which operate dams, which may or may not 
have the flood control component, integrated in the decisions 
taking place between state and Federal agencies--excuse me--in 
the discussion taking place between state and Federal agencies?
    Mr. Sorooshian. Mr. Chairman, as I pointed out, it's really 
a wonderful thing out of this El Nino year that much closer 
coopera-

tion is already taking place. Certainly, once this period is 
over, I think the agencies should be encouraged to evaluate to 
see how their cooperation has improved the amount of exchange 
of information to minimize--for instance, in the late '80s 
there was an episode of warming in the Colorado which resulted 
in major flooding and damage to the Grand Canyon Dam. When one 
in hindsight looks at it, if better information exchanges had 
taken place, perhaps better operating policies could have been 
implemented to avoid situations like that.
    So, in general, I am very optimistic, and we see a lot more 
interaction between agencies.
    Mr. Doolittle. Thank you.
    Dr. Konstantine Georgakakos, how important is it in your 
opinion to be able to get real-time data from all the stream 
gauges, if we are to make substantial improvement of operation 
flow forecasts?
    Mr. Konstantine Georgakakos. I think it's very important, 
especially in view of the uncertainty in the meteorological 
forecasts, and probably for short hydrologic forecast lead 
time, say between 6 and 18 hours.
    Mr. Doolittle. Mr. Andrews, you said that through the 
Department of Water Resources, California has installed nearly 
50 stream-gauging sites in areas that have high flood 
probability. To what extent is that data collection activity 
coordinated with USGS?
    Mr. Andrews. The initial intent was to have the USGS join 
in that effort, and my understanding is that, because of 
funding limitations, they were not able to deploy additional 
instruments or to provide real-time telemetry on additional 
instruments. So that the state went ahead and both installed 
additional instruments, but the major step was to link the 50 
states with real-time telemetry.
    Mr. Doolittle. So, do you forward that information along to 
USGS anyway?
    Mr. Andrews. Oh, yes, the information is forwarded onto 
USGS.
    Mr. Doolittle. OK.
    Mr. Andrews. It's just that we had hoped for a more robust 
array than we were able to find----
    Mr. Doolittle. I think we're going to have to find them 
some more money to do their part of the bargain here.
    Dr. Aris Georgakakos, you state that the benefits to 
reservoir management that you describe in your testimony can be 
realized only if forecasts are used in connection with dynamic 
decision schemes, but that static reservoir rule curves, which 
are traditionally used in the operation of reservoir systems, 
would fall short of realizing the value of improved forecasts. 
Could you elaborate some on what you mean by that, dynamic 
decision schemes versus static reservoir rule curves?
    Mr. Aris Georgakakos. The main difference is really the 
ability to incorporate the uncertainty of the hydrologic 
forecast. The usual, the traditional operating rules of 
reservoirs do not really handle that very well. While dynamic 
systems follow the hydrologic forecast, they incorporate the 
uncertainty within them and have a better sense of the 
operation of the system, and that's basically what I meant.
    Mr. Doolittle. Gentlemen, I sincerely regret that I haven't 
had more time to ask you and the other panel more questions. 
I'll ask you, if you would, please, to respond to some of our 
written questions.
    [The information referred to may be found at end of 
hearing.]
    Mr. Doolittle. I appreciate the work that you are doing and 
the efforts that you have made to come here this afternoon, and 
now into this evening. I hope we aren't making you miss your 
plane flights.
    We will hold the record open and would ask you to respond 
as quickly as possible to those supplementary questions.
    And with that, the hearing is concluded.
    [Whereupon, at 6:04 p.m., the Subcommittee adjourned 
subject to the call of the Chair.]
    [Additional material submitted for the record follows.]
    Statement of Elbert W. Friday, Director, Office of Oceanic and 
Atmospheric Research, National Oceanic and Atmospheric Administration, 
                      U.S. Department of Commerce

    Mr. Chairman and Members of the Subcommittee: The past 
fifteen years have witnessed remarkable advances in the 
observations, understanding and predictions of climate 
variability, especially as related to El Nino. This is an 
outstanding scientific success story in which NOAA research has 
played a leading role. We have entered a new era where we can 
now observe and closely monitor El Nino conditions as they 
develop, and can also provide measurably skillful forecasts of 
future conditions for at least a few seasons in advance. We 
have further established connections between El Nino and 
regional and global climate variations which can begin to be 
incorporated into decision making in climate-sensitive sectors, 
such as energy and water resources.
    These are extraordinary steps forward. At the same time, it 
is important to keep in mind that the science of climate 
prediction is still in its infancy, and is developing rapidly. 
Therefore, much of what I describe will be products of ongoing 
research by NOAA and its partners and, clearly, much more 
remains to be done in this area. I must also emphasize that 
climate predictions are by their nature probability forecasts. 
That is, El Nino alone will not determine what happens in our 
Nation's weather over the next year, but rather it will shift 
probabilities in such a way as to make certain climate events 
more likely, and others less likely. Further, El Nino does not 
influence all parts of our Nation's weather in the same way, 
and some regions are relatively more vulnerable to its effects. 
My testimony will outline principal features of El Nino, 
describe the status of the current event, and summarize recent 
research results on climate risks associated with El Nino over 
the United States.

BACKGROUND

    What is El Nino, and how is it related to larger scale 
patterns of climate variability? In brief, El Nino refers to a 
naturally occurring phenomenon in the equatorial Pacific Ocean 
which is characterized by an unusual warming of the sea surface 
temperatures extending from the South American coast to near 
the dateline. The flip side of El Nino, often called La Nina, 
is characterized by abnormally cold sea surface temperatures 
over the same region. Both El Nino and La Nina have major 
effects on global and regional climate. El Nino (and La Nina) 
events happen on average every two to seven years and, once 
established, persist for six to twelve months (Exhibit 1). The 
persistence of such conditions provides one key as to why El 
Nino is potentially useful for climate predictions.
    A second key is that El Nino is not an isolated oceanic 
phenomenon but, rather, involves coupled ocean-atmospheric 
interactions in the tropical Pacific. The coupled phenomenon is 
called the El Nino--Southern Oscillation, or ENSO, where 
Southern Oscillation refers to systematic changes in tropical 
atmospheric pressure patterns. El Nino conditions, such as we 
are experiencing now, are associated with a pronounced 
weakening of the trade winds which ordinarily blow from east to 
west across the tropical Pacific. The tropical rainfall 
patterns also shift, with the heaviest rainfall moving eastward 
following the warmer water. The changes in tropical rainfall 
alter the global wind patterns and, in particular, the jet 
streams, which in turn affect our Nation's weather. La Nina 
also exerts important effects on global wind patterns through 
essentially the same processes, although in many cases the 
climate changes are reversed. For North America, the largest 
and most systematic effects on climate with either El Nino or 
La Nina events are usually experienced in winter and spring.
    By the early 1980s, there was growing recognition in the 
climate research community that ENSO was a major source of 
climate variability, and that potentially useful seasonal 
predictability of this phenomenon was possible. However, in 
1982, the onset of the largest El Nino in this century was not 
even recognized, let alone predicted. The failure to identify 
the initiation of this major event was due to inadequacies in 
data, to deficiencies in our conceptual understanding of 
fundamental causal mechanisms, and to the rudimentary nature of 
prediction models at that time.
    To address important challenges in ENSO observations and 
prediction, a new research program on the Tropical Oceans and 
Global Atmosphere, or TOGA, was initiated in 1985. TOGA was 
sustained by cooperative efforts of four agencies: NOAA, the 
National Science Foundation, NASA, and the Department of 
Defense's Office of Naval Research, with NOAA leading the 
interagency research program on interannual climate variability 
since 1982. As noted in a 1996 report of the National Research 
Council, the TOGA program, which formally ended in 1994, has 
left a series of important legacies, including:
    (1) an observational system supported by NOAA, called the 
TOGA TAO (Tropical Atmosphere--Ocean) Array, that provides a 
vital set of in situ observations for monitoring El Nino 
conditions and for initializing computer models for El Nino 
forecasts;
    (2) the ability to project El Nino conditions in the 
tropics a few seasons in advance with measurable skill;
    (3) an improved ability to estimate the global atmospheric 
response to projected oceanic conditions in the tropical 
Pacific;
    (4) the increased use of climate information and forecasts 
as a factor in decision-making in climate-sensitive areas.
    These major advances are all critical to today's 
discussion. As one example, unlike 1982, the current ENSO 
observing system has allowed us to identify and describe the 
evolution of this year's major El Nino event as it is 
occurring, rather than after the fact. The in situ observations 
also provide fundamental input data for the computer models 
that are now being used to forecast the evolution of this 
event.

The 1997 El Nino Event

    NOAA is closely monitoring the evolution of the 1997 El 
Nino event, and is making its data and analyses available in 
real-time through various media, including the World Wide Web, 
through sites at the Climate Diagnostics Center (CDC), Pacific 
Marine Environmental Laboratory (PMEL), Climate Prediction 
Center (CPC), and Office of Global Programs (OGP). The latest 
TAO array data indicate that very strong El Nino conditions 
continue in the tropical Pacific, with sea surface temperature 
anomalies exceeding 4.5C (9F) in the eastern Pacific. Sea 
surface anomalies through much of the central and eastern 
tropical Pacific are at the highest observed values in at least 
the last 50 years, exceeding even the 1982-83 event at this 
time of year. It is important to note that by early last 
winter, the National Centers for Environmental Prediction 
(NCEP) coupled model was forecasting a transition to warm 
conditions over the summer, and by May, NCEP was confident in 
predicting that this would be a major El Nino event.
    To get a further picture of the magnitude of the current 
event, it is useful to compare its evolution with that of the 
six prior strongest El Nino events over the last 50 years 
(Exhibit 2). The measure used for this purpose is an index that 
includes tropical winds, pressures, cloudiness and sea surface 
temperatures, to more accurately reflect the coupled nature of 
the ENSO phenomenon. By this measure, this event is by far the 
strongest we have seen for this time of year, and second only 
to the 1982-83 event in absolute magnitude, with the latter 
event reaching a peak in late winter to early spring of 1983. 
Another important aspect of the current El Nino is that it 
experienced the most rapid sustained growth of any event of the 
record, from its initiation in spring through this past summer. 
Although the latest data point (for August-September) suggests 
a leveling off, note that such behavior has also occurred in 
earlier events, including 1982-83, and therefore should not be 
taken as a clear indication that the event has peaked. It is 
both the rapid growth and absolute magnitude of this event 
which have raised legitimate concerns about adverse climate 
impacts, both in the U.S. and worldwide.
    So far, extreme climate events associated with this event 
have been mainly outside the North American sector, although 
the strong suppression of hurricane activity in the Atlantic 
and Gulf of Mexico is quite likely connected to this event, as 
are the unusual northward paths of tropical cyclones along the 
west coast of Mexico into the southwestern U.S. As indicated 
earlier, however, the strongest and most systematic effects of 
El Nino on North American climate typically occur in winter and 
spring. I will now discuss some research results on potential 
future climate risks associated with this event.

Potential Climate Risks

    Climate predictions are inherently probability forecasts, 
with the fundamental goal being to estimate how the probability 
distributions of various quantities, such as temperature and 
rainfall, will change subject to particular conditions; in this 
case, El Nino. An important new direction of NOAA's recent 
research efforts in this area is toward the development of 
extreme event predictions at extended range (beyond several 
days). A basic goal of this research is to identify regions and 
time periods where the risk of large-scale extreme events, such 
as droughts or floods, is significantly increased (or 
decreased).
    So how do we do this? There are two basic approaches: 
first, through analysis of past behavior (an empirical 
approach) and, second, through application of numerical 
forecast models through a relatively new technique called 
ensemble predictions. The approaches are complementary, and 
both are being actively pursued in NOAA, although here results 
derived from the former approach will be emphasized.
    An example is shown in Exhibit 3, and is derived from one 
hundred years of data. The figure shows the regions at highest 
risk for seasonal precipitation extremes in spring following El 
Nino winters and, for comparison, also following La Nina 
winters.
    Based on past El Nino events, the areas at highest risk for 
much above normal precipitation in winter and spring include 
the Southern California coast, much of the Southwest extending 
into the southern Plains, and portions of the Gulf Coast and 
Southeast. Areas at increased risk of much below normal 
precipitation include portions of the far northern Rockies and 
northern Plains in winter and the northern Ohio Valley in late 
winter to early spring. Other analyses also strongly support 
the possibility that portions of California and the Southwest 
are at significantly increased risk of much above normal 
precipitation, and that high stream flow values and floods are 
much more frequent in these areas in El Nino years. For 
example, more detailed analyses for the lower Colorado basin 
suggest approximately a 60 percent chance of much above normal 
precipitation (defined as wettest 20 percent of all seasons) in 
springs following El Nino winters.
    In contrast, the Southwest is particularly vulnerable to 
drought in springs following La Nina winters. This was the case 
in 1995-96, when a severe drought affected the Southwest, Texas 
and portions of the southern and central Plains, resulting in 
several billion dollars in losses. A potential positive aspect 
of this year's El Nino is that there is a significantly reduced 
risk of drought in these areas, particularly as compared to La 
Nina years. The strong ENSO signal in the Southwest, together 
with the high sensitivity of this region to hydrologic 
variations, suggest that this is a critical region to consider 
for issues concerning climate variability and water management.
    It is important to briefly mention other aspects of weather 
and climate pertinent to U.S. water management. First, although 
ENSO is the most widely-recognized mode of climate variability, 
it is not the only one, and scientists are now studying other 
modes of variability, such as the North Atlantic Oscillation 
(NAG) and Pacific Decadal Oscillation (PDO), to gain additional 
predictive ability. Second, critical water management issues 
extend over a broad array of time scales, both shorter and 
longer than ENSO variability. For example, although El Nino may 
increase the likelihood of flooding in a given region, in 
general, it will not determine the timing of a specific flood 
event. Short range prediction of floods is the operational 
responsibility of NOAA, and research efforts are also underway 
to extend the maximum lead times for identifying risks of 
specific flood events. Some preliminary work between NOAA and 
USGS scientists for the Merced River suggests that use of 
ensemble model forecasts from the National Centers for 
Environmental Prediction (NCEP) provides useful lead time 
information on flow variability out to approximately ten days. 
An additional issue is longer-term climate variations, such as 
multi-year or decadal droughts, that have episodically affected 
large portions of the U.S., as in the 1930s. Such events would 
have potentially devastating consequences were they to recur. 
Understanding and predicting longer-term climate variability is 
also an area of active research by NOAA.
    New climate research products discussed above and others 
are being developed by NOAA and its research partners for a 
broad array of potential applications. At present, there is 
ongoing coordination between NOAA and other agencies including 
the National Weather Service River Forecast Centers, the Bureau 
of Reclamation and the Natural Resources Conservation Service 
in developing usable information for assessing flood and 
drought risks in areas showing empirical ENSO-related signals.
    Focused research on regional climate variations and their 
hydrologic impacts should rapidly increase our ability to 
predict extremes of high and low water availability. This new 
information can be put to use in water management systems, with 
the potential for substantial economic and other benefits. The 
present need is for continuing intensive research on prediction 
of extreme weather and climate events, the regional hydrologic 
response to climatic extremes, and effects of hydrologic 
extremes on water demand, water management, and aquatic 
resources generally.
    Mr. Chairman, that concludes my testimony. I thank you for 
this opportunity to discuss El Nino and potential U.S. climate 
risks which may affect water management decisions. Recent 
substantial advances have been made in understanding and 
predicting climate variability, and these research advances 
hold great potential to benefit society. I believe the future 
holds even greater promise for return on our research 
investment.
    I would be pleased to answer any questions that you or 
other members of the Committee may have.

[GRAPHIC] [TIFF OMITTED] T6056.001

[GRAPHIC] [TIFF OMITTED] T6056.002

    Statement of Ants Leetmea, Director, Climate Prediction Center, 
  National Centers for Environmental Prediction, National Oceanic and 
        Atmospheric Administration, U.S. Department of Commerce

    Mr. Chairman and Members of the Subcommittee:
    Fifteen years ago a small group of scientists and research 
managers based the development of a major research effort in 
climate on the premise that patterns of extreme weather were 
not totally random events and that the key to predictive 
insight lay in the interaction between the ocean and the 
atmosphere. That international multidisciplinary research 
effort today in the foundation underlying the National Oceanic 
and Atmospheric Administration's (NOAA) world-recognized 
advances in forecasting climate based on dynamical predictions 
of the states of El Nino. These advances are leading towards an 
understanding of the connections between El Ninos and U.S. 
temperature and rainfall variability, and the possible 
utilization of this predictive information into decision-making 
in water and energy sectors. We currently are in the maturing 
stages of the 1997/98 El Nino which is forecast to be 
comparable in strength the 1982/83 event, the previous ``event 
of the century.'' During 1982/83 the United States suffered 
roughly $3 billion in losses related to storm damages and 
flooding. It is appropriate for the Chairman and the Members to 
ask how much better prepared we are for this event and the 
potential eventualities associated with it than the 1982/83 one 
and what we might do in the future to be better prepared.

El Nino and Forecasts for the 1997/98 Event

    In simple terms, the distribution of sea surface 
temperatures in the tropical Pacific determines the overlying 
atmospheric circulation. However, the overlying tropical 
circulation through current conditions and its past history 
determines the sea surface temperature. This coupled 
interaction between the atmosphere and the ocean results in a 
quasi-periodic oscillation termed the E1 Nino/Southern 
Oscillation (ENSO) which has a period of 4 to 7 years. When the 
tropical Pacific is warmer than normal, this is the El Nino 
phase; when it is colder than normal, this is the La Nina 
phase. Both phases have strong impacts on precipitation and 
temperature patterns over the U.S.
    Forecasts for E1 Nino at NOAA's Climate Prediction Center 
(CPC) are made using coupled numerical models of the ocean and 
atmosphere that require supercomputers for their execution.
    The initial conditions for the forecasts utilize ocean 
measurements from the ENSO observing system, measurements from 
satellites, and reports from volunteer observing ships. 
Forecasts out to a year in advance are made once a week. The 
first forecasts that indicated a likelihood of an El Nino 
episode this year were made in November of 1996. El Nino 
conditions began to visibly develop in March of this year; 
subsequently positive departures from normal sea surface 
temperatures rapidly increased, reaching record levels for the 
period August-September. The current El Nino is forecast to 
peak at the end of the year or early next year. At that time 
its magnitude and spatial extent will be comparable to the 
1982/83 event, previously considered the event of the century. 
Conditions are forecast to return to about normal during the 
summer of 1998.
    El Nino causes large scale shifts in the distribution of 
tropical rainfall. This is expected to result in a stronger 
than normal jet stream over the eastern North Pacific and 
southern United States during the 1997-98 winter and continuing 
through the early spring of 1998. This pattern is expected to 
result in wetter than normal conditions across much of the 
southern United States from California eastward to the 
Carolinas. Drier than normal conditions are forecast over the 
northern High Plains and Ohio valley. During 1982/83 many of 
the areas expected to be wetter than normal, e.g. states such 
as California, Utah, Illinois, Missouri, Arkansas, and 
Louisiana, experienced extensive flooding and storm damage.

Forecasts for United States Rainfall and Temperature

    Currently the forecasts for El Nino impacts are made using 
both statistical techniques and dynamical forecasts. If 
rainfall and temperature records are ranked for the past 102 
years according to when moderate or strong El Nino were 
present, the tendency for much of the southern third of the 
U.S. to be wetter than normal (and drier than normal during a 
La Nina) is clear (figure 1) while regions of below normal 
rainfall include the areas surrounding Montana and the Ohio 
valley. The magnitude of the impacts and the regions that are 
impacted vary with season. Strongest impacts are experienced 
during our winter. These average rankings, conditioned on the 
presence of a moderate or strong El Nino can be converted to 
percent of normal rainfall. The seasonal maps for the 
temperature and rainfall impacts for the United States, as well 
as more detailed state by state summaries for states that will 
be likely impacted can be found on the web site for the CPC 
(http://nic.fb4.noaa.gov). The most likely scenario for given 
regions is that given by what has been observed for moderate El 
Ninos, e.g. figure 1. However, since the 1997/98 event is not a 
moderate event, but another event of the century, what happened 
in 1982/83 should be considered as a plausible scenario. The 
dynamical model forecasts, which are just coming on line as 
forecast tools, suggest some truth to this plausible scenario. 
The impacts for California and the major river basins are 
discussed in the following:

    California: California during winters with moderate El 
Ninos receives about 130 to 140 percent of normal rainfall 
(figure 2). This converts to about five inches extra rain for 
most of the state. During 1982/83, an event comparable to the 
current one, California, as well as much of the southeast, 
received 150 to 200 percent of normal rainfall. Over coastal 
southern and central California and the southern Sierra this 
resulted in about ten inches extra rain for the January through 
March period. For much of northern California the amount was 
even greater at about 16 inches. This was the result of more 
storms during this period that on average carried more rainfall 
per storm. However, we don't anticipate that these storms will 
reach the intensity of those during last year, December 1996-
January 1997, although this cannot be precluded. These drew 
their moisture and temperature from the deep tropics, whereas 
the moderate El Nino and 1982/83 storms picked up their 
moisture and temperature from the mid-latitude Pacific.
    Colorado Basin: During moderate El Ninos the Basin receives 
from normal rainfall in western Colorado to over 190 percent of 
nominal in southern Arizona. Generally the northern part of the 
basin has a weaker rainfall signal associated with El Nino than 
the southern part. Southern Arizona has the largest signal. 
Since the overall region receives little rainfall, the large 
percent of normals for a season only amount to a few inches. 
During 1982/83 most of the region, except for western Colorado, 
received greater than nominal rainfall with 125 percent of 
normal in the northern part to over 200 percent or so in 
southwest Arizona. Despite the modest rainfall amounts, the 
observed Colorado river water year runoff reached its largest 
values for the past 45 years in 1982/83 of roughly 23 Million 
Acre Feet or about 180 percent of normal during the spring of 
1983.
    Columbia Basin: For moderate El Ninos in fall and winter 
there are weak probabilities for below normal rainfall in 
western Washington, British Columbia, Idaho, and western 
Montana. The probabilities are also increased for warmer than 
normal temperatures with the largest warm anomalies on average 
being present in March and April. These conditions lead to a 
lower than average snowpack in late winter and early spring 
with higher early spring runoffs and lower ones later. The 
potential exists for water allocation problems in summer. 
However, during the strong 1982/83 event, much of the region 
received above normal rainfall; hence, there is a possibility 
these conditions might not arise.
    Upper Missouri Basin: Deviations from normal rainfall and 
temperature conditions over the upper Missouri regions, 
Montana, Wyoming, and northwestern North Dakota during moderate 
El Ninos are much the same as over the Columbia Basin except 
that the likelihood of below normal rainfall is enhanced. 
However, over Nebraska and Kansas wetter than normal conditions 
are to be expected. Without feeding these impacts into river 
and reservoir flow models, it is not clear what the integral 
effects on Missouri River flows will be. During 1982/83 the 
pattern of above and below rainfall deviations described 
earlier was observed.

Other Factors Influencing Western U.S. Rainfall: Decadal 
Variability

    The history of rainfall variations for coastal southern 
California, Arizona, and New Mexico since 1930 indicates that, 
in addition to El Nino and La Nina impacts, there have been 
strong decadal variations in the amount of rain received. The 
signal in the Pacific northwest is the opposite of that in the 
southwest, e.g. when the southwest tends to have above normal 
rainfall, the northwest tends to have below normal rainfall. 
The period from the mid-1940s to the mid-1970s was one where 
the southwest received below normal rainfall. Since then 
rainfall amounts have been greater than normal, e.g. a relative 
climatic optimum. When a La Nina has been present, the 
southwest almost always experiences below normal rainfall. The 
most recent case of this was during 1995-96 when the region 
experienced a severe drought which cost the nation roughly $4 
billion. When an El Nino is present, conditions tend to be 
wetter than normal, and the probability of being wetter than 
normal roughly doubles.
    This decadal variability is referred to by researchers as 
the Pacific Decadal Oscillation (PDO). The spatial pattern of 
rainfall variations over the United States associated with it 
is quite similar to that of El Nino (figure 1). This suggests 
that the same kind of ocean-atmosphere interactions are 
responsible for both. Although understanding the origins of the 
PDO is still a research issue, its impacts are already included 
in the CPC seasonal climate forecasts. The forecasts include a 
component of features that persist for 10 to 15 years. A more 
common appreciation of this decadal variability could be of use 
for water managers.

    The Future

    NOAA's Strategic Plan has a Goal, Implement Seasonal to 
Interannual Climate Forecasts, that has a focus in the 
implementation of a capability to forecast water resources up 
to several seasons in advance. This has to be a cooperative 
program with other Federal and State Agencies, Universities, 
and the private sector. Floods and droughts, especially the 
Great Flood of 1993, have emphasized the need for improved 
short term and seasonal predictions to support flood/drought 
and water management and damage mitigation. In addition to 
these improvements in hydrometeorological predictions, NOAA is 
modernizing its hydrologic predictions with an Advanced 
Hydrologic Prediction System (AHPS). AHPS builds on the current 
capability of the National Weather Service's River Forecast 
Centers who currently issue stage forecasts for only one, two, 
and three days into the future at most forecast points and 
crest forecasts out to about one week for a few selected 
forecast points. AHPS will build upon the skill of NOAA's 
seasonal forecasts to provide new hydrologic forecast products 
out to seasons in the future. The seasonal forecasts are 
required to predict the likelihood of extreme events such as 
droughts and floods happening, whereas, the short term 
forecasts are needed to predict the details and magnitude of 
such events. The allocation of water among competing demands 
(e.g. fisheries, irrigation, hydropower and municipalities) 
looms as a national problem. AHPS and the NWS River Forecast 
Centers are the logical links to the operations; practices of 
the Bureau of Reclamation and the Corps of Engineers. AHPS has 
been implemented in the Des Moines River Basin and in that 
location has successfully demonstrated the coupling between 
hydrology, meteorology, and climatology on daily, weekly, 
monthly, and seasonal time-scales. Demonstrable products for 
the Des Moines AHPS project can be seen at http://
www.crh.noaa.gov/dmx/ahps.
    Significant challenges still remain to be overcome in order 
to fully integrate the effects of El Nino and decadal 
variability into the operating practices of agencies that 
manage water resources. Many of these are technical and no 
doubt will be more fully described by the other experts at this 
hearing. Nevertheless, one can anticipate that the general tone 
of their testimonies is one of optimism that the time is here 
to start this process. Indeed the current event has initiated a 
number of studies into seeing what the limitations currently 
are and how far the current technology can be pushed in 
bringing the climate forecasts down to the river basin scales. 
These studies over the next six months will significantly 
accelerate the progress towards utilization of climate forecast 
in water management.
    One of the lessons from this event will be that an enhanced 
focus resulting from urgency gets people's attention and 
cooperation. The challenge after this event will be to maintain 
this momentum. No one Agency has all the required expertise. 
Getting the necessary Agencies to work towards a common focus, 
especially under current budgetary constraints will not be 
easy. What can be done, what needs to be done, the directions 
to be taken, and the potential partnerships will be much more 
evident as the 1997/98 El Nino proceeds.
    Mr. Chairman, that concludes my testimony. I want to thank 
you and the Committee for this opportunity to discuss the 
current El Nino and the possible impacts on water resources in 
the west. Let me finish by giving a partial answer to the 
question I posed at the beginning of my testimony--``How much 
better prepared we are for this event and the potential 
eventualities?'' In 1982/83 we did not know until September of 
1982 that we were in an event, let alone the ``event of the 
century,'' nor what the potential regional impacts would be 
over the United States. For this event the forecasts started 
indicating a likelihood of an event about a year ago, and we 
knew from the forecasts and the observations that this would be 
a major event by late May of this year. We also knew by that 
time the potential for heavy rainfall in California and much of 
the southwest this coming winter for this coming winter--a full 
six months in advance!
    I would be pleased to answer any questions that you or 
other members might have.

[GRAPHIC] [TIFF OMITTED] T6056.003

[GRAPHIC] [TIFF OMITTED] T6056.004

[GRAPHIC] [TIFF OMITTED] T6056.005

[GRAPHIC] [TIFF OMITTED] T6056.006

[GRAPHIC] [TIFF OMITTED] T6056.007

[GRAPHIC] [TIFF OMITTED] T6056.008

[GRAPHIC] [TIFF OMITTED] T6056.009

[GRAPHIC] [TIFF OMITTED] T6056.010

[GRAPHIC] [TIFF OMITTED] T6056.011

[GRAPHIC] [TIFF OMITTED] T6056.012

[GRAPHIC] [TIFF OMITTED] T6056.013

[GRAPHIC] [TIFF OMITTED] T6056.014

[GRAPHIC] [TIFF OMITTED] T6056.015

[GRAPHIC] [TIFF OMITTED] T6056.016

[GRAPHIC] [TIFF OMITTED] T6056.017

[GRAPHIC] [TIFF OMITTED] T6056.018

 Statement of Konstantine P. Georgakakos, Sc.D., President, Hydrology 
                 Research Center, San Diego, California

    Mr. Chairman and members of the Subcommittee, my name is 
Konstantine Georgakakos. I am the President and Founding 
Director of the Hydrologic Research Center, a nonprofit 
research corporation in San Diego, California. I am also a 
Research Scientist IV with the Scripps Institution of 
Oceanography of the University of California in San Diego. I 
consider it an honor to be here and to be invited to testify 
concerning the involvement of the Hydrologic Research Center in 
the forecasting, and in forecasting research, associated with 
the predicted El Nino event. In this testimony I will focus in 
particular on Center research activities pertaining to 
operational streamflow forecasting for flood warning and water 
resources management.
    Two Center research activities are directly relevant:

          (a) Assisting the California-Nevada River Forecast Center of 
        the U.S. National Weather Service, in improving the short- and 
        long- term forecasts of Lake Folsom inflow in east central 
        California. These forecasts will be used in the operational 
        management of the Lake waters and flow releases.
          (b) Assessing the utility of integrated forecast-control 
        methodologies for the operational management of reservoir 
        systems. Case studies in Iowa and California are now being 
        conducted for this research.
    Both activities were initiated earlier this year and in 
both cases, the El Nino Southern Oscillation (or ENSO) 
phenomenon is an important consideration. These activities are 
funded by the Federal Government, the former by the Bureau of 
Reclamation and the latter by the National Oceanic and 
Atmospheric Administration.
    Our research on the aforementioned issues builds on earlier 
research results obtained for the midwestern and south-central 
U.S. It relies on proven mathematical hydrologic models of the 
watershed processes and the flow forecast uncertainty. 
Watershed processes considered are snow accumulation and melt, 
and surface and subsurface flow. The flow forecast uncertainty 
exists for three reasons:

        --incomplete coverage of watershed area by sensors;
        --large errors in the long lead-time meteorological forecasts 
        used to drive the hydrologic models;
        --mathematical model approximations of complex natural 
        processes.
    Our results so far support the following conclusions:

    (1) The accuracy of the computed snowmelt volumes within 
the Folsom Lake watershed decreases substantially with 
decreased watershed coverage by precipitation measurement 
sensors. For example, there are currently large areas drained 
by the Middle Fork of the American River with poor sensor 
coverage. The Middle Fork drains approximately 28 percent of 
the 1861-square-mile Lake Folsom watershed. In the record flood 
of 1964 it contributed more than 40 percent of the peak flow 
rate of 148,000 cubic feet per second.
    (2) Substantial improvement of operational flow forecasts 
is attained when the current flow forecast systems are upgraded 
to include models for uncertainty and updating from flow 
measurements in real time. The improvement is mainly in the 
reduction of forecast errors for unusually high or low flow 
rates.\1\
    (3) In a case study involving data since 1904 from the Iowa 
River at the Iowa City gauging station in Iowa, statistically 
significant seasonal streamflow associations to ENSO phenomenon 
were found.\2\ For example, for El Nino conditions, there is a 
70 percent chance to have above-normal streamflow, lagged three 
to five seasons from the reference time of El Nino. Such 
associations may be used in long-lead forecasting operations. 
Studies, in progress using streamflow data from the American 
River, have not revealed a strong statistical association with 
ENSO. It is possible that the 3-7 year ENSO signal is concealed 
by the extreme year-to-year variability of streamflow in the 
Folsom Lake watershed.
    (4) Substantial benefits for operational reservoir water 
management were obtained for the Saylorville reservoir in Iowa 
when flow forecasts were used as input to the decision process 
with due account for forecast uncertainty. On the basis of 
extensive computer simulations of the watershed-reservoir 
system it was found that using coupled forecast-control 
methodologies reduces reservoir-management sensitivities to 
climate variability and to the large uncertainties associated 
with the forecast of such variability by current climate 
models. Analogous simulations are in progress for Lake Folsom 
in California collaboration with Professor Aris Georgakakos of 
the Georgia Institute of Technology. Preliminary results show 
again the effectiveness of such coupled forecast-control 
methodologies.
    I wish to make the following topical recommendations:

    (1) Conduct detailed studies to quantify the uncertainty 
associated with the estimation of precipitation and snowmelt 
over the Sierra Nevada in California. Such studies should be 
supported by newly established special dense sensor networks, 
which should be left in place for 5-10 years. The effects of 
such uncertainty on the streamflow forecasts for Rivers 
draining the Sierra Nevada may be then quantified, and steps 
may be taken to improve the accuracy and reliability of the 
streamflow forecasts.
    (2) Advanced streamflow forecast procedures should be 
implemented and utilized in parallel to current operational 
ones to evaluate increased benefits to operations. Such 
procedures include models for uncertainty of meteorological 
forecasts and utilize streamflow observations to improve the 
forecasts continuously. Several watersheds within the U.S. 
should be identified prototypes for this effort.
    (3) In parallel to (2) above, coupled forecast-control 
methodologies with due account for forecast uncertainty should 
be implemented in the suggested prototype watersheds to 
evaluate increased benefits to operational reservoir 
management. In this context, the utility of climate forecasts 
for increasing the benefits of reservoir management should be 
quantified. Our studies to date make me confident that it is 
through the research and implementation of such methodologies 
we will be able to utilize the uncertain long-range climate 
forecasts.
    Mr. Chairman and members of the Subcommittee, real-time 
flow forecasting and reservoir water management are important 
operational functions for mitigating natural disasters. These 
functions, vital for present-day communities, have their basis 
on Hydrologic Science and Engineering, and Water Resources 
Systems Analysis. As the requirements of the public for safety 
and reduction of damage losses from natural disasters increase, 
it is important to formulate a national plan for the increased 
effectiveness of these operational functions. I have argued 
elsewhere in the scientific literature that to achieve this 
goal and in analogy to the establishment of our National Center 
for Atmospheric Research in the Atmospheric Sciences, it 
appears necessary to establish a National Center for Hydrology 
and Water Resourcest.\4\ I firmly believe that with the 
establishment of such a national center much progress will be 
made almost immediately on a national level by concerted 
efforts to enhance the flow of information from research to 
operations. Such a center is envisioned as a collaborative 
effort among universities, the Federal Government and the 
private sector. With this last important recommendation I now 
conclude my testimony. I will be pleased to answer your 
questions.

    References

\1\  Hydrologic Research Center Technical Report 1, June 1996: 
Implementation and Testing of the HFS Operation as Part of the 
National Weather Service River Forecast System (NWSRFS); 
Hydrologic Research Center Technical Note 3, May 1996: 
Improvement of Hydrologic Model Forecasts Using a Real Time 
Updating Technique.
\2\ Guetter, A.K., and K.P. Georgakakos, 1996: Are the El Nino 
and La Nina Predictors of the Iowa River Seasonal Flow? Journal 
of Applied Meteorology, 35(5), 34-35.
\3\ Georgakakos, K.P., Bae, D.-H., Mullusky, M.G., and A.P. 
Georgakakos, 1995: Hydrologic Variability in Midwestern 
Drainage Basins: Diagnosis, Prediction and Control. Chapter II-
2 in Preparing for Global Change. A Midwestern Perspective, SPB 
Academic Publ. bv, 61-90.
\4\ Georgakakos, K.P., 1995: On the Establishment of a U.S. 
National Center for Hydrologic Research and Technology 
Transfer. Journal of Hydrology, 172, 15-21.

[GRAPHIC] [TIFF OMITTED] T6056.019

[GRAPHIC] [TIFF OMITTED] T6056.020

[GRAPHIC] [TIFF OMITTED] T6056.021

[GRAPHIC] [TIFF OMITTED] T6056.022

[GRAPHIC] [TIFF OMITTED] T6056.023

[GRAPHIC] [TIFF OMITTED] T6056.024

[GRAPHIC] [TIFF OMITTED] T6056.025

[GRAPHIC] [TIFF OMITTED] T6056.026

[GRAPHIC] [TIFF OMITTED] T6056.027

[GRAPHIC] [TIFF OMITTED] T6056.028

[GRAPHIC] [TIFF OMITTED] T6056.029

[GRAPHIC] [TIFF OMITTED] T6056.030

[GRAPHIC] [TIFF OMITTED] T6056.031

  Statement of Soroosh Sorooshian, Professor, Dept. of Hydrology and 
      Water Resources, The University of Arizona, Tucson, Arizona

    Honorable Chairman and Members of the Subcommittee:
    As early as last winter, the scientific evidence based on 
observations and model simulation results clearly indicated 
that a strong El Nino weather pattern has been developing in 
the western equatorial Pacific Ocean. With further 
observations, it has become clear that this El Nino is going to 
be one of the strongest weather patterns on record. Several 
coastal countries located near the Equator have already been 
impacted. The west coast of Mexico experienced disastrous 
consequences as a result of Hurricane Pauline. Chile is 
experiencing record amounts of rainfall, resulting in major 
flooding.
    The effects on the United States (so far) have been 
relatively small. The exception is the precipitation resulting 
from Hurricane Nora, which was mostly confined to southeastern 
California, western Arizona, extreme southeastern Nevada, and 
the southwest portion of Utah. However, the current predictions 
are that higher than-average precipitation will occur in 
California and the Southwest United States between November and 
February.
    The impact of a strong El Nino on water resources is 
expected to be significant. Statistical analyses correlating 
historical precipitation amounts over the Southwest U.S. with 
El Nino years indicate a strong likelihood of above-average 
precipitation (and associated runoff). Water resources managers 
are, therefore, evaluating the potential impact of El Nino on 
their decisions and strategies for operation of water resources 
systems. Reservoir releases for the purposes of water supply 
and power generation, etc., must be scheduled, and emergency 
management plans must be drawn up to minimize the adverse 
impact of potential flooding. These decisions pose a major 
challenge, due to the type of information which is required.
    We do know that this year's El Nino is the strongest yet 
recorded, and that the statistical evidence points to above-
average winter precipitation for the Southwest United States. 
However, these facts do not constitute sufficient information 
for water resources and emergency managers to initiate major 
changes in operating practice. Such decisions require reliable 
information about the expected arrival times of significant 
storm systems, their expected durations, and intensities, etc. 
A storm system that arrives during warm weather may not result 
in snow accumulation at higher elevations, but may, instead, 
produce large amounts of runoff and potential flooding. Without 
accurate information about the timing and quantities of 
precipitation and streamflow, it is very difficult to plan 
timely evacuations from flood-prone areas or to pre-release 
large quantities of stored reservoir water to help mitigate the 
flooding. On the other hand, precipitation which arrives during 
a cooler period may accumulate as snow at higher elevations. 
Water resources managers then have greater flexibility to 
evaluate options and to decide on an appropriate operational 
strategy. Under this scenario, it would be a lot less risky to 
commit to early reservoir releases, knowing that melting of the 
above-average snow pack will provide the water necessary to 
fill the reservoir(s) later in the season. However, rapid 
warming could cause sudden large releases of meltwater, leading 
to late-winter and/or spring flooding. It is critically 
important, therefore, that accurate estimates of the volume of 
water in the snow pack and timely and accurate temperature 
forecasts during the melt season be available.
    I wish to strongly emphasize that sound water resources 
management decisions in the Southwest will require far more 
information than merely the knowledge of a strong El Nino 
signal. Water resources managers are, rightly, reluctant to 
order early reservoir releases without further information. In 
the few instances that I know of, where early decisions have 
been made regarding reservoir releases and other water 
management issues, the knowledge that this will be a strong El 
Nino year has been only one of several useful pieces of 
information, but not the sole deciding factor. For instance, 
the Salt River Project--which supplies water to the greater 
Phoenix area--has incorporated the information about the strong 
El Nino signal into its decision to reduce ground-water pumping 
by some 40,000 acre-feet for this past year. This requirement 
will instead be satisfied from their reservoir system, based on 
expectations that an above-average spring snowmelt runoff will 
fill the reservoirs to their normal level.
    Water resources and emergency managers are accustomed to 
making decisions based on probabilistic information. While the 
knowledge of a strong El Nino year has enhanced the probability 
for a wetter-than-average year in the Southwest, it has not 
reduced the uncertainty of many other factors critical in 
making decisions regarding major deviations from normal 
operating practice. In order to enhance the quality and 
usefulness of both short-term (hours to days) and extended 
(weeks to a month) forecasts, the reliability of the hydrologic 
prediction system for the West-

ern United States must be improved. The primary components of 
this system are: the Quantitative Precipitation Forecasts 
(QPFs), Extended Streamflow Prediction (ESP), more accurate 
methods for estimating snow accumulation, particularly in the 
mountainous regions, and high-resolution accurate rainfall 
measurements critical for forecasting rapidly developing flood 
events.
    The strength of the current El Nino signal has attracted a 
lot of media attention and has generated much-needed public 
attention to this climatic phenomenon. The climate research 
community is to be commended for developing the capability of 
predicting it with such a high degree of accuracy. Perhaps the 
greatest benefit of this prediction to the water resources 
management community has been in encouraging very close 
cooperation among the Federal and state agencies responsible 
for various aspects of water resources management and 
hydrologic services. As an example, the cooperation over the 
past several months between the U.S. Bureau of Reclamation, the 
National Weather Service's (NWS) Colorado Basin River Forecast 
Center (CBRFC), and the U.S. Geological Survey (USGS) has 
resulted in close coordination for sharing modeling and 
observational information required for improved management of 
the reservoir system on the Colorado River. Continued 
cooperation among these agencies will critical to the 
development of an operational hydrologic prediction system for 
the Western U.S. to be used for water resources management in 
both El Nino and non-El Nino years. It is worth noting that, 
while statistical evidence points to a wetter-than-average year 
in the Southwest during a strong El Nino year, the wettest 
winter on record (1993) in the White Mountains and the 
surrounding areas of southern Arizona was not an El Nino year. 
A reliable hydrologic prediction system is crucial for the 
efficient management of western water resources irrespective of 
whether we are experiencing an El Nino weather pattern or not.

                           Supplemental Sheet

    Sooroosh Sorooshian
    Dept. of Hydrology and Water Resources
    Harshbarger Buiding 11; Room 122
    Tucson, Arizona 85721
    Phone: (520) 621-1661
    FAX: (520) 626-2488
                                ------                                


      Statement of Richard Andrews, Director of Emergency Services

    Mr. Chairman and Members of the Subcommittee. I am Richard 
Andrews, Director of Emergency Services for Governor Pete 
Wilson.
    Thank you for the opportunity to speak today on the many 
activities in California in preparation for El Nino and issues 
related to Federal policies and practices and their influence 
on our preparedness and disaster recovery efforts.

Background

    As has been widely publicized, a large El Nino event is 
underway and is forecast to continue through the winter months. 
Though the specific forecasts continue to be updated and, at 
this time, remain uncertain, we have been told by various 
meteorologists and climatologists to expect substantially more 
rainfall than normal, especially in the central and southern 
coastal areas of California. Some forecasts suggest that the 
storms may come as early as November and, if the initial rains 
are sufficient to soak the ground, we could have an early and 
long flood season. Based on the records from past El Nino 
episodes, the flood risk would be higher in the coastal and 
mountain areas of southern California and not as great on major 
Central Valley rivers arising in the Sierra.
    Higher ocean levels are expected, one-third to one-half 
foot above normal. Coastal damage and erosion are most severe 
when storm waves coincide with peak high tides. Potential 
coastal impacts include: structural damage from wave 
overtopping, onshore flooding and wave impact, and land loss 
due to cliff failure and beach erosion.
    Public lands and infrastructure, as well as private 
property, are at risk. State beaches are some of the most 
vulnerable properties, with bulkheads, seawalls, parking lots 
and restrooms in danger. Boating facilities and boats are also 
at risk. During the winter of 1982-83, also a strong El Nino 
period, over $100 million of damage occurred to the public and 
private coastal infrastructure. One third of the damage 
(approximately $35 million) occurred to public recreational 
facilities along the coastline in Central and Southern 
California. Farther north, while the Central Valley flood 
control system performed well, ten Delta islands were flooded 
due to levee failures.
    Impacts to inland watersheds could be highly variable. The 
Department of Fish and Game notes that the impact to 
California's inland rivers and streams will depend upon the 
condition of the watersheds. For watersheds that are stable, 
the high volume of rainfall and runoff that frequently 
accompanies an El Nino may wash sediment out of the gravel, 
thereby improving spawning conditions for fish. For unstable 
watersheds, El Nino may pose a more severe threat, as it can 
cause hillsides to collapse into rivers or streams.

State Preparedness

    Despite the considerable uncertainties about what the 
impacts of El Nino will be, we are confident that the state and 
local governments will be well prepared to meet the challenges 
that a severe winter weather season may bring. As is well 
known, California has, in this decade, had repeated and varied 
experience in coping with the consequences of natural 
disasters, including large fires, earthquakes and floods. We 
have improved our local and state response systems after each 
event; California has, I believe, the most effective emergency 
response system in the nation.
    The El Nino forecasts have helped accelerate on-going 
preparedness efforts that followed this January's historic 
floods in central and northern California. In January 1997 the 
state experienced serious flooding in forty-eight of the 
state's fifty-eight counties, with damages totaling nearly $2 
billion. These floods came only two years after a series of 
winter storms in the first quarter of 1995--where losses also 
totaled $2 billion--led Governor Wilson to declare all fifty-
eight counties as disaster areas, the first time in the state's 
history that the entire state was impacted by a natural 
disaster.
    At the height of the l997 winter storms, Governor Wilson 
established the Flood Emergency Action Team (FEAT) to 
coordinate the review of the lessons learned from the historic 
flooding and to establish long-term strategies to protect 
Californians from future flood disasters. Following a series of 
hearings throughout the impacted region, the FEAT's final 
report made more than fifty recommendations that have been 
implemented to improve the state's flood fighting systems, 
including:

         Improving emergency response coordination between 
        public safety agencies at the local and state level with the 
        Army Corps of Engineers and local flood maintenance 
        organizations.
                  The Office of Emergency Services (OES) has conducted 
                a series of workshops on evacuation procedures, 
                emergency alerts, the requirements for care and shelter 
                of large numbers of evacuees, procedures for handling 
                evacuations of livestock and pets, and the emergency 
                closure of the Sacramento-San Joaquin Delta waterways 
                during flood conditions.
                  Together with the Department of Water Resources, OES 
                has worked with reclamation officials on procedures for 
                requesting resources and coordinating with the state's 
                Standardized Emergency Management System. OES's 
                training facility at San Luis Obispo is conducting 
                training courses specifically focused on flood fight 
                management.
         Improving and expanding existing flood data. To expand 
        the state's existing flood data, the California Department of 
        Water Resources has installed nearly 50 stream gauging sites in 
        areas that have high flood probability. Earlier this year the 
        state called on the U.S. Geological Survey to expand its 
        surface water data collection program so that we might obtain 
        better and earlier data about flood threats. Unfortunately, due 
        to budget limitations, the Federal Government has indicated 
        that such expansion is unlikely.
    On October 6, Governor Wilson convened an El Nino summit in 
Sacramento for all state agencies and local leaders. In his 
Executive Order signed on that date, Governor Wilson directed 
the Office of Emergency Services and the Department of Water 
Resources to establish technical assistance teams and conduct a 
series of regional workshops throughout the state to review 
specific state and local preparedness actions. The first 
workshop was held in Long Beach on October 24, a second will be 
held in San Diego on November 3, with six additional workshops 
scheduled across the state in the coming weeks.
    Governor Wilson also signed legislation allocating $7.4 
million for El Nino preparedness measures, including the 
propositioning of flood fighting resources as forecasts become 
more specific. Moreover, Governor Wilson declared this week, 
October 27-31 as Winter Storm and Flood Preparedness Week 
throughout the state, with a series of events being held to 
provide continuing focus on the need for all residents, 
businesses and local governments to take basic emergency 
preparedness measures.
    In addition, all state agencies have been carrying out El 
Nino preparedness measures, including:
         California was one of the first entities to provide 
        comprehensive information on El Nino on the Internet. The 
        Office of Emergency Services' home page contains comprehensive 
        information on preparedness measures, current weather 
        forecasts, and situation reports. CERES, the California 
        Environmental Resources Evaluation System, one of the world's 
        largest environmental web sites, provides comprehensive 
        information on the El Nino phenomenon, and links to other home 
        pages containing pertinent information.
         The California Conservation Corps, one of the state's 
        most valuable emergency response resources, is offering 
        sandbagging workshops throughout the state.
         The Department of Parks and Recreation, who's 
        facilities sustained significant damage in the 1982-83 El Nino, 
        is cleaning drainage channels and taking general preparedness 
        measures at its facilities.
         The state Department of Transportation, as part of its 
        general winter preparedness efforts, is stockpiling additional 
        emergency response and road clearing resources at its 
        facilities throughout the state.
    In addition to the actions being undertaken by state 
agencies at the direction of Governor Wilson, local 
governments, community groups and businesses throughout the 
state are taking unprecedented preparedness measures. Supplies 
have been stockpiled, storm drains cleared, evacuation 
procedures reviewed and strategies for the care and shelter of 
individuals who may be displaced by the winter weather 
developed. Many cities and counties have held special flood 
preparedness drills and developed specific emergency plans for 
possible El Nino impacts.
    California will continue to provide the public with the 
best available information through the Internet, briefings and 
workshops, training sessions, technical assistance teams and 
the stockpiling of equipment. Local and state agency 
preparedness will continue throughout the coming weeks.

Coordination with Federal Water Management Agencies

    There are several areas of coordination between state and 
Federal agencies that need to be enhanced.
    The state's Department of Water Resources has the primary 
responsibility for working with Federal water management 
agencies. Flood operations and reservoir management can be 
divided into five major geographical regions in California 
where there are significant flood control reservoirs on the 
major river and stream systems. Most flood control reservoirs 
are multi-purpose in that flood control is one of several 
purposes for which the reservoirs were authorized and 
constructed.
    In California, most reservoirs that have Federal flood 
control reservations are operated by State, Federal or local 
water management agencies. The flood control reservation and 
seasonal operation rules are predetermined by the U.S. Corps of 
Engineers. Any deviation from the Corps' reservoir operation 
rules must be coordinated with the Corps. Last January, local, 
state and Federal coordination was exceptional.
    The regional scenarios for this winter include:

          San Francisco Bay Area--Significant damage occurred 
        in 1982-83 from high winds and high tides, and localized 
        flooding, particularly in the Alviso district of San Jose. The 
        potential for damage this winter depends on the timing and 
        intensity of storm systems, and whether storms occur during 
        high tides.
          Sacramento River Watershed--Forecasts indicate that 
        rainfall and runoff are expected to be slightly above normal. A 
        review of historical records indicates that extreme events, 
        like those experienced in February 1986 and January of this 
        year, do not correlate with strong El Nino events. In 1982-83 
        rainfall and runoff was well distributed through the flood 
        season and the flood control system--with the exception of the 
        Sacramento-San Joaquin Delta--was able to safely regulate 
        runoff through the flood control reservoirs and the levee 
        system. Peak releases from flood control reservoirs were much 
        less than experienced this past January. For example, Oroville 
        Dam released a maximum of 60,000 cubic feet per second (cfs) in 
        1982-83 versus 160,000 cfs in January 1997; Folsom Dam released 
        35,000 cfs in 1982-83 vs.115,000 cfs this January.
          Sacramento-San Joaquin Delta--While the levied 
        Sacramento Flood Control System performed well during the last 
        strong El Nino, in 1982-83 the Sacramento-San Joaquin Delta 
        experienced significant problems. High tides combined with the 
        nearly half-foot rise in sea level as a consequence of El Nino, 
        causing levee failures at five Delta islands. The state has 
        expended nearly $100 million to strengthen Delta levees.
          San Joaquin River Watershed--In 1983 an extremely 
        large snowpack and spring runoff impacted the San Joaquin. This 
        year, runoff from a similar event would likely be regulated 
        safely through the flood control reservoirs; but, the levied 
        flood control system in the San Joaquin valley would likely 
        experience high water for a long period. Considering the 
        vulnerable condition of the newly repaired flood control levees 
        to erosion and under seepage, a heightened state of alert will 
        need to be in place for this critical watershed.
          Tulare Lake Basin Watershed--As in the San Joaquin, 
        snowmelt runoff in the spring will likely be safely regulated 
        through the flood control reservoirs. However, excessive 
        runoff--similar to that of 1983--will likely flood the Tulare 
        Lakebed and potentially threaten surrounding communities.
          South Coast Basin Watershed--This region was hardest 
        hit by the 1982-83 El Nino. Coastal wave damage was severe as 
        were flash floods in the interior counties. Flood control 
        reservoirs and debris basins, primarily operated by local flood 
        control agencies, performed well in the past and are expected 
        to perform effectively this winter. Flood control reservoirs 
        and debris basins, primarily operated by local agencies to 
        Corps regulations, performed well in 1982-83. Since that time, 
        reservoir operational plans have been updated as a result of 
        changes in channel capacities and conditions downstream of 
        reservoirs. Earlier notification procedures have been 
        implemented in recent years for warnings based on reservoir 
        operations.

Issues Requiring Additional Attention

    The Federal Government is an important partner in the 
state's overall preparedness efforts and plays an essential 
role in helping communities recover from the impacts of natural 
disaster.
    In a letter to President Clinton on October 6, Governor 
Wilson urged action on several concerns about current Federal 
policies and practices that impact the pace of recovery from 
past floods and winter storms and will also affect current and 
future preparedness and recovery efforts, including:

         Urge the Army Corps of Engineers to accelerate the 
        timetable to make repairs to levees damaged in central 
        California during the January 1997 floods. Policy disputes 
        between Corps officials in California and Washington slowed 
        recovery during the spring and summer months. Progress is now 
        being made, and if we're fortunate to have winter storms hold 
        off until near the end of November, we have been assured by the 
        Administration in an October 24 response to Governor Wilson's 
        letter that all but one of the critical central California 
        flood control repairs should be in place to handle the flows. 
        We would hope that this year's experiences of unnecessary 
        delays would not be repeated following future flooding.
         In the event the levees are not fully repaired, direct 
        the Corps to undertake necessary response preparations to 
        improve the Corps emergency response under Public Law 84-99. 
        Some steps in implementing this strategy have already been 
        taken; several others are needed. In advance of high water, the 
        Corps should used their authority under Public Law 84-99 to do 
        the following:

                 Direct emergency repairs at the most critical 
                sites if rainfall interferes with the construction 
                schedule. This could be accomplished through 
                appropriate contract provisions change orders.
                 Redesign the repairs at unrepaired critical 
                sites to employ materials and equipment appropriate for 
                wet condition if rainfall impacts construction.
                 Award contracts with construction firms and 
                aggregate suppliers in the central valley in advance of 
                high water. These time and materials contracts would 
                allow the Corps to simply direct the contractor to make 
                repairs even as the repair is being redesigned.
                 Secure equipment and materials that are 
                difficult or time consuming to obtain during an 
                emergency. Examples include expandable vinyl bladders 
                for raising effective levee heights, and barges for 
                working in the waterway. Items could be mobilized or 
                stockpiled upon forecasts for high water, provided they 
                are secured in advance.
         Direct all Federal regulatory agencies to consolidate 
        the needed approvals for flood repair and preparation work. 
        Governor Wilson has directed all state agencies to place the 
        highest priority on expediting approvals for this work, a large 
        number of Federal agency approvals are also needed. Local 
        agencies need clear and consistent permit requirements and 
        procedures from Federal agencies. Governor Wilson has 
        designated the Department of Water Resources the principal 
        state agency to coordinate needed state agency approvals. Such 
        a ``one stop shop'' strategy should be considered for future 
        permitting processes for essential public safety related 
        facilities.
        On October 24 the Corps issued a Nationwide 31 permit for 
        channel clearing and sediment removal. Over this past weekend 
        Los Angeles County began some of this important work. Governor 
        Wilson remains concerned, however, that the Environmental 
        Protection Agency has indicated that it will seek a policy 
        review of the cumulative impacts of channel clearing activities 
        with the intention of requiring further mitigation work.
         Direct the Federal Emergency Management Agency (FEMA) 
        and the Corps of Engineers to modify policies for local 
        agencies that conduct flood fights on flood control works. As a 
        result of Federal policies arising from the 1993 Midwest 
        floods, there now exists a disincentive for local agencies to 
        assist in flood fights. The Federal Levee Policy asserts that 
        the Corps is the primary Federal agency for flood fighting. 
        Since the adoption of this policy, FEMA has established a 
        virtual ``hands off'' policy for funding flood fighting on any 
        flood control work. FEMA states that the Corps is responsible 
        for flood fighting on all flood control works, and that FEMA is 
        therefore prohibited by the Stafford Act from funding such 
        efforts. At the same time, the Corps has responsibility for 
        conducting flood fighting operations, but finds itself, in 
        extreme events like the January storms in California, 
        overextended and unable to respond in a timely manner to every 
        request. The Corps is forced to prioritize its limited 
        resources, omitting or deferring some essential flood fighting 
        projects. Because of the requirements for prompt action, local 
        agencies cannot wait for the often cumbersome Corps contracting 
        processes to be implemented. However, if they take action to 
        meet immediate flood fight needs without going through the 
        Corps, the local agencies find themselves unable to receive 
        reimbursement from either FEMA or the Corps. This illogical 
        Federal policy creates a perverse disincentive that results in 
        more damage than necessary.
         Direct FEMA to implement the recommendations relating 
        to Public Assistance processes and policies made by the 
        California Congressional delegation in their September 15, 1997 
        letter to the FEMA Director.

Conclusion

    All Californians are grateful for the assistance we 
received from Congress and the Administration as we joined 
together on the levees to battle the floodwaters of last year's 
storms. We are also grateful for the assistance we continue to 
receive in the flood's aftermath as communities and individuals 
work to repair the damage and clean up homes, farms, and 
businesses.
    State agencies, local governments, community groups and 
individuals are currently engaged in an unprecedented 
preparedness effort in advance of this winter's weather. While 
the exact outcome of El Nino remains to be seen, we still have 
time to undertake needed common-sense actions to prepare for 
the worst, and prevent the loss of life and property damage we 
suffered earlier this year and in the last El Nino.
                                ------                                


   Statement of Stephen K. Hall, Executive Director, Association of 
                    California Water Agencies (ACWA)

I. Introduction

    Mr. Chairman and Members of the Subcommittee, thank you for 
providing me an opportunity to submit this statement regarding 
the water management implications of the 1997/98 El Nino on 
behalf of the Association of California Water Agencies (ACWA). 
ACWA is a statewide, non-profit association which represents 
more than 440 public water agencies who collectively manage and 
deliver 90 percent of the state's urban and agricultural water.
    It is important to understand the role our members play in 
managing water resources, the uniqueness and fragility of 
California's water system and how a weather event like El Nino 
can play havoc on a water system already stressed to the 
breaking point.
    Local water agencies serve several functions in 
California's water management system, but the responsibilities 
most relevant to the discussion of El Nino are flood control 
and water supply. While there has been a deluge of attention on 
the disastrous flooding that can be caused by an El Nino, much 
less attention has been given to the water supply issue. There 
are several points that need to be made about California's 
flood control and water supply system in such a discussion.
    First, California's weather is hard to predict and 
extremely variable. The norm is arrived at by averaging the 
extremes of drought and flood. So-called ``normal years'' 
rarely occur.
    Second, California's water managers at the Federal, state 
and local level have done an excellent job of managing flood 
control and water supply given the constraints under which they 
operate. However our constraints are enormous. Much of 
California's water system was built decades ago, before modern 
construction techniques were available. Not only is 
California's system outdated, it is inadequate. Our storage 
capability is far below what we see on other river systems. 
This outdated, undersized system has created tension between 
flood control and water supply for those charged with managing 
this system.
    Third, added to the tension between managing our system for 
flood control and water supply, we have placed substantial new 
demands on the system by releasing large quantities of water 
for fish and other environmental purposes.
    These observations lead to the inescapable conclusion that 
we must act now to shore up our existing system and expand it 
in order to meet California's legitimate flood control, 
environmental and water supply needs.
    In California, water agencies routinely balance flood 
control and water supply needs. Specifically, water managers 
must keep enough reservoir storage space available to manage 
floods during heavy precipitation, while ensuring that adequate 
water is stored to meet water supply needs for cities and farms 
and to protect against drought.

II. California's Weather Picture: Feast or Famine

    Even without an El Nino in the mix, California's weather is 
highly unpredictable. Planning for a new water year is often 
akin to reading tea leaves.
    According to the California Department of Water Resources, 
weather in our state has become increasingly unpredictable over 
the last 50 years. Whether that variability is due to global 
warming, cyclical changes or some other explanation, erratic 
swings from too much to not enough make it very difficult to 
map out management plans for the state's water systems.
    Adding insult to injury, we have the variability of El 
Nino, which could bring either floods or drought to the state. 
During California's protracted seven-year drought, an El Nino 
in 1991-92 helped cause a critically dry year. Conversely, a 
strong El Nino in 1982 caused devastating floods similar to 
conditions predicted for this current El Nino.
    Since 1950, California has experienced 12 El Nino events; 
eight of which have been similar in strength to the current El 
Nino. Of those 12, five have led to drier than normal 
conditions, while seven have resulted in above normal 
precipitation.
    In the past 20 years, California has experienced only two 
years that experts can call ``normal''--where water supplies 
for the year have ended up close to the average. And even in 
those years--1989 and 1993--large swings in the weather pattern 
surprised forecasters and planners. Large amounts of rain can 
come in the first few months of fall, and that could be the 
last rain we see all year. Such was the case in 1997--a year of 
devastating floods. (See Exhibit 1.)

III. 1997: Floods and Drought

    Water officials recently closed the book on a 1997 water 
year that included some of the worst flooding we've seen this 
century. While not an ``El Nino year,'' the 1997 floods are 
being viewed as a model, due to their warm, El Nino-like 
characteristics.
    Over the three-day period centered on New Year's Day, more 
than 30 inches of rain poured onto Sierra Nevada watersheds 
already saturated by one of the wettest Decembers on record.
    The deluge overwhelmed many of the water systems in 
Northern California. The sheer volume of runoff exceeded the 
flood control capacity of Don Pedro Dam on the Tuolumne River 
and Millerton Reservoir on the upper San Joaquin River with 
large, uncontrolled spills of excess water. In just a week's 
time Shasta, the largest reservoir in our system, filled to 
near capacity 1.4 million acre-feet--taking in nearly two 
million acre-feet. At its peak, inflow was measured at a record 
236,700 cubic-feet-per-second. Most of the other large dams in 
Northern California were full or nearly full at the end of the 
storms.
    The effects of last January's storm have been far-reaching. 
Total flood damages reached nearly $2 billion, including $300 
million in damage to flood control facilities and $206 million 
in damages to various public facilities. Nearly 300 square 
miles of agricultural land were flooded, causing close to $300 
million in damage to agriculture. In addition, 120,000 people 
were forced from their homes and nine lives were lost.
    The damage could have been even more catastrophic. State 
and Federal water projects stemmed the worst of the floods, due 
in large part to the changes to the system implemented after 
the 1986 flood. Millions of homes and businesses were spared as 
a result.
    1997 was the third wet year in a row for Northern 
California. State water officials say total runoff statewide 
was 145 percent of average. But, perhaps even more dramatic 
than the deluge of rain in January was the lack of rain that 
followed. The spring of 1997 was the driest in 104 years. The 
wet pattern began in November, culminated in January, then shut 
off like a switch for the rest of the year.
    Water shortages forced the Bureau of Reclamation to cut 
some deliveries by 10 percent. Statewide, we begin the next 
water year with only average supplies in storage.

IV. California's Flood Control and Water Supply System

    To deal with California's weather extremes, flood control 
plans were established for the Sacramento Valley beginning in 
the late 1880s to improve navigability and protect population 
centers. Once called the ``Nile of the West,'' the Sacramento 
River yields about 35 percent of the state's water supply.
    The flood plan evolved into one of the most complex flood 
management and water distribution systems. Today's integrated 
Federal, state and local flood management network includes 23 
reservoirs with flood detention space and 1,760 miles of 
Federal levees, channels, and overflow bypasses and weirs in 
the Central Valley.
    This network, in concert with other Federal, state and 
local facilities, also supplies fresh water for urban, 
agricultural and commercial demands--the majority of which are 
in the drier regions of central and southern California--and 
flood protection for the Sacramento and San Joaquin valleys. 
Water supply provided by these projects also helps to fuel 
California's $800 billion economy.
    Key projects include the Federal Central Valley Project 
(CVP) and California's State Water Project (SWP). The CVP has a 
storage capacity of 11 million acre-feet and delivers about 7 
million acre-feet of water to agricultural and urban uses. The 
SWP delivers about 2 million acre-feet annually to farms and 
cities. The single most important aspect of California's 
complicated water system is the Sacramento-San Joaquin River 
Delta. Its channels through the state and Federal projects 
provide drinking water for two-thirds of the state, in addition 
to irrigation water for more than 4.5 million acres of the 
nation's most productive farmland.
    This is an impressive system, but it is far less than what 
we see on other river systems. According to the California 
Department of Water Resources, total storage on the Sacramento 
River system with average annual runoff of 22 million acre-feet 
is less than one-year or 16 million acre-feet. In comparison, 
the Colorado River system--with an average annual runoff of 
only 15 million acre-feet--boasts a storage capacity of 60 
million acre-feet or enough for a four-year supply.
    The lack of storage capacity has led to the tension between 
operating the system for flood control, the protection of life 
and property, and operating the system for water supply to meet 
the needs of the nation's largest economy. And the problem is 
growing worse. Since the last major element of our water 
management system was added in the early 1970s, the state's 
population has essentially doubled. Local water managers have 
done a good job in balancing this tension. Urban water managers 
have managed to meet the needs of the rapidly growing 
population through conservation, reclamation and innovative 
water transfers and exchanges. Meanwhile, California 
agriculture is today producing 50 percent more in food and 
fiber with the same amount of water that it was using 20 years 
ago. We're also doing a better job of protecting lives and 
property. It is widely acknowledged that the floods that have 
occurred in recent years could have been far more devastating 
had it not been for strong efforts to coordinate the local, 
state and Federal flood control operations.
    However, the experience of 1997 has shown the deficiencies 
in our system that not even innovative management can overcome. 
The devastating floods of January 1997, followed by water 
delivery cutbacks later in the year, point out that our 
existing system must be irked and expanded in order to protect 
California from floods while maintaining a healthy environment 
and a strong economy.
    Governor Pete Wilson reinforced this theme in a recent 
letter sent to President Bill Clinton, In the letter, Wilson 
expresses concern about the state of the Central Valley's levee 
system severly damaged by last year's floods. Wilson stated, 
``. . . I am deeply concerned that the Corps' work will not be 
completed in time . . . There is a substantial risk that not 
all repairs will be completed prior to this November when heavy 
rains are expected throughout California.''

V. The Environmental Factor

    The experience of recent years also points out that we have 
new demands on the system unrelated to growth in our population 
or economy. The Bay-Delta system--the hub of our water 
management system--also forms the largest estuary on the west 
coast and serves as a unique habitat for the more than 120 
species of wildlife, some of which are protected under the 
Federal Endangered Species Act (ESA), such as the winter run 
Chinook salmon and delta smelt.
    Fisheries are also affected by weather extremes. Many 
experts are concerned that this year's El Nino may be bad news 
for already-depleted salmon and trout populations. Many fish 
populations were decimated due to damage from last year's high 
waters. With another year of heavy rainfall predicted, concern 
is mounting for salmon populations especially--as one expert 
put it, ``For salmon, there's a strong apprehension bordering 
on mild panic.''
    While fresh water flows are by no means the only 
consideration in producing healthy fish populations, those 
flows are certainly one of the factors viewed as essential to 
produce healthy habitat throughout the system. To the extent 
our water management system falls behind in meeting our 
environmental water supply and flood control needs, conflict 
over those competing priorities will continue to increase.

VI. How The System Is Being Operated Today

    As we review the statistics from the extraordinary storms 
of 1997 and others which preceded it, we are reminded that 
tremendous precipitation and runoff do not necessarily equate 
to a bonanza in terms of water supply benefits. Several times 
over the past two decades, vast quantities of fresh water ran 
out to the Pacific Ocean because we don't have enough 
reservoirs to store the water. Capturing storm runoff which is 
excess to Delta environmental needs could provide sorely-needed 
water for urban, agricultural and environmental needs.
    As water managers struggle to maintain the balance between 
flood control, environmental and water supply needs working 
within existing storage constraints, water supply tends to 
suffer the most.
    Again, the 1996-97 water year is a case in point. With a 
tremendously wet January, water delivery commitments were made 
from both the state and Federal projects including voluntary 
operations in the Bay-Delta to benefit fisheries. As the wet 
months turned to dry, the state and Federal projects strained 
to make all delivery commitments to urban and agricultural 
water users. In the end, Federal customers did not receive full 
deliveries.
    Initiated in 1994 as part of the historic Bay-Delta Accord, 
the CALFED Bay-Delta Program, a joint state-Federal 
partnership, has been charged with developing a long-term plan 
to address the environmental decline and water supply 
reliability issues in the Bay Delta system. Included within the 
scope of work is the need to better address Bay-Delta conflicts 
such as those brought about by balancing flood control with 
water supply and environmental needs.
    As dictated by the accord, further constraints have been 
applied to both the CVP and SWP to protect fisheries. 
Approximately 850,000 acre-feet of project yield in critical 
periods is now being dedicated to the environment--over and 
above the original 3.65 million acre feet required under prior 
water rights decisions.
    In addition to efforts by CALFED, the U.S. Corps of 
Engineers is preparing a comprehensive reassessment of the CVP 
flood supply system, which is sorely needed. We must rethink 
our overall approach to flood control and water supply as we 
prepare to meet the needs of California into the next century.

VII. Conclusions

    Through the benefit of our collective experiences, water 
agency managers and environmental regulators are beginning to 
reach common ground on the need and potential strategies to 
improve our approach to flood control and water supply issues.
    First, there is general agreement that we need to improve 
our overall water system by addressing the need for additional 
water storage and by improving our water delivery system. This 
will also necessitate a need for even tighter operating 
regimens.
    Second, there is general recognition that both have a 
tremendous stake in finding workable solutions to these 
challenges, including the need to expand the ever-shrinking 
water pie.
    As this testimony has sought to highlight, the time to act 
on the consensus to improve our water management system is now. 
Those of us in the water management community believe that 
CALFED provides the best opportunity to take such action. That 
is why we are strongly supporting CALFED and are participating 
actively in the CALFED process. We believe it is imperative 
that CALFED produces a plan that will restore the ecosystem 
comprised of the Bay-Delta and its watershed as well as produce 
an implementable plan to improve our water supply and flood 
control capabilities into the next century. There is consensus 
in California along those lines, and we hope and trust the U.S. 
Congress will agree that there is not only a state but a 
Federal interest, and that the Federal Government will actively 
participate in preparing and implementing the CALFED plan.
    I appreciate the opportunity to speak before you today to 
underscore the importance of focusing on the water supply 
implications of weather events like El Nino and opportunities 
that exist to deal with those impacts.
