OPERATIONAL MULTISITE FORECAST AND RESERVOIR MANAGEMENT IN NORTHERN

CALIFORNIA

KONSTANTINE P. GEORGAKAKOS, THERESA M. CARPENTER,

AND NICHOLAS E. GRAHAM

Hydrologic Research Center, San Diego, CA 92130

in collaboration with

ARIS P. GEORGAKAKOS, HUAMING YAO AND MARTIN KISTENMACHER

Georgia Water Resources Institute, Georgia Tech, Atlanta, GA 30334

Sponsored by National Oceanic and Atmospheric Administration

(Award No. NA07OAR4310457)

HRC LIMITED DISTRIBUTION REPORT NO. 34

Hydrologic Research Center

12780 High Bluff Drive, Suite 250, San Diego, CA 92130, USA

26 September 2010

Page left blank

i

ACKNOWLEDGEMENTS

The work described in this report was sponsored by the National Oceanic and Atmospheric

Administration (NOAA), Climate Program Office (CPO) and Office of Hydrologic Development

(OHD) under Award No. NA07OAR4310457. It is conceived and executed as a joint project

between the Hydrologic Research Center with expertise in hydrometeorological modeling and

forecasting and the Georgia Water Resources Institute with expertise in reservoir decision support

systems. We are grateful to Dr. Chet Koblinsky, CPO, and Dr. Gary Carter, OHD, for their

support of this Transition to Operations Project. The present research project is a complement to

the INFORM (Integrated Forecast and Reservoir Management) Demonstration Project and benefits

from the advice and support of the INFORM Oversight and Implementation Committee, which

consists of representatives from operational forecast and management Agencies in Northern

California. The authors thank the members of the INFORM Oversight and Implementation

Committee for their feedback and guidance during this project period. Information about the

INFORM Project and the OIC may be found in Georgakakos et al., HRC Technical Report No. 5, August

2006 (available in print from the Hydrologic Research Center and on line at the California Energy

Commission site http://www.energy.ca.gov/pier/project_reports/CEC-500-2006-109.html.

The authors also wish to thank Robert Hartman and Pete Fickenscher of the California Nevada

River Forecast Center for their continuing advice on Northern California operational hydrologic

modeling, and associated data and software support. The ideas and opinions in this report are of the

authors and need not reflect those of the funding or collaborating Agencies.

.

This report should be cited as follows:

Georgakakos, et al. 2010: Operational Multisite Forecast and Reservoir Management in Northern

California. HRC Limited Distribution Report No. 34. Hydrologic Research Center, San Diego, CA, 26

September 2010(NA07OAR4310457), 104 pp.

ii

Page left blank

iii

EXECUTIVE SUMMARY

This report documents activities pertaining to the three-year research project entitled: Multiscale

Operational Forecasting and Reservoir Management for Northern California. The project activities

facilitate the transition to operations of the INFORM (Integrated Forecast and Reservoir

Management) system forecast and decision components for Northern California.

Project activities consisted of: (a) meetings and collaboration with operational forecast and

management Agencies in Northern California aiming to plan the transition to operations of the

INFORM systems, (b) streamlining of the INFORM operational forecast component for land

surface hydrology to emulate with fidelity the current operational hydrologic forecast system of the

California Nevada River Forecast Center (CNRFC), (c) analysis of multi-month water-supply

forecasts and water resources impacts, and (d) generation and assessment of ensemble forecasts and

risk-based water use trade-offs for 2008 through 2010 in support of operational water resources

management by the California Department of Water Resources and the Central Valley Operations of

the Bureau of Reclamation. All assessments were carried out with a start forecast date in the interval

March 1 to 15, which corresponds to the start of critical water use period (growing season). In each

year, the forecast and assessment horizons were nine months.

The meetings with the operational forecast and management Agencies of Northern California

resulted in plans for the transition process that includes alignment of the INFORM and CNRFC

operational hydrologic components. Alignment was necessary because the definition of the

hydrologic segments was updated by the CNRFC since the time of the first definition of INFORM

(2005) concerning the Folsom, Oroville, New Bullards Bar, and Shasta drainages (the Trinity

configuration has not changed). The INFORM multi-month ensemble reservoir-inflow forecasts,

driven by the National Centers of Environmental Prediction (NCEP) CFS (Climate Forecast

System) ensemble forecasts, were used by the INFORM decision component together with the

Sierra Nevada snowpack and large Northern California reservoir initial conditions for March 1 of

the years 2008, 2009 and 2010. INFORM results suggested that compared to the previous two years

(2006 and 2007), in 2008 the Northern California system of reservoirs would have significantly

reduced ability to accommodate the multiple objectives and to meet water demands in Northern

California. The analysis of INFORM forecasts and management results with a March 1 initial date

iv

using observed data indicates that the INFORM system provides reliable predictions and

management trade-offs for a range of climatic conditions for Northern California.

The transition to operations of the INFORM system is on-going (not complete) because: (a) the US

National Weather Service operational forecast system configuration is currently in transition (from

the current NWSRFS to the FEWS/CHPS system) and alignment of the INFORM system with the

CNRFC operations requires additional changes to conform to the FEWS/CHPS software

architecture; (b) NCEP CFS products are currently in transition (making available complete three-

dimensional fields for lead times out to 45 days rather than selected fields and surface variables only)

and the INFORM system forecast component should be changed to take advantage of the new

information for more reliable ensemble predictions in the first forecast month; and (c) upstream

regulation significantly biases operational ensemble reservoir inflow predictions, which at the present

state of INFORM development mirror the operational predictions pertaining to unimpaired

reservoir inflow predictions.

A newly initiated project with the California Energy Commission aims to address (a) and (b) above

(INFORM II project, 2009-2012), and a project with the NWS Office of Hydrologic Development

provides the theoretical basis and formulation for (c) (Project on Upstream Regulation Adjustment

to Ensemble Streamflow Prediction, 2009-2010). Transitioning the tested formulations for upstream

regulation effects into INFORM and River Forecast Center operations is a necessary future

development.

v

TABLE OF CONTENTS

List of Tables vii

List of Figures ix

Introduction 1

First Year Activities (2007-2008) 11

Second Year Activities (2008-2009) 27

Third Year Activities (2009-2010) 32

Discussion and Concluding Remarks 51

References 57

Appendix A: Preliminary Assessment of the INFORM Phase I Forecast-Decision System Results vs. Observed Data for the 2006, 2007, and 2008 Seasons 59

Appendix B: Operational Multiscale Forecast and Reservoir Management in Northern California, Assessments 2009 75

vi

Page left blank

vii

LIST OF TABLES

Table 2-1: New Snow and Soil Parameters for the Folsom Drainage 13

Table 2-2: Channel Routing Reaches and Model Parameters for Folsom Drainage 14

Table 3-1: Snow water equivalent estimated for each INFORM hydrology model

subcatchment. 29

Table B-1: Monthly Average Inflows for Selected Locations (TAF) 95

Table B-2: Reservoir Monthly Parameters 96

Table B-3: Monthly Minimum and Target River Flows 96

Table B-4: Monthly Demands 103

Table B-5: Initial Reservoir Storages on March 15, 2009 104

viii

Page left blank

ix

LIST OF FIGURES

Figure 1-1: Schematic diagram of the distributed INFORM system configuration with data links

indicated. Black arrows signify real-time data links while grey arrows signify off-line

data links. 4

Figure 1-2: A schematic of the INFORM Reservoir and River System 5

Figure 1-3: INFORM DSS Modeling Framework 7

Figure 2-1: Folsom Lake drainage and tributary basins as configured in 2005. NFDC1 – North

Fork American drainage; MFAC1 – Middle Fork American drainage; CBDC1 –

South Fork American drainage; FOLC1 – Local drainage of Folsom Lake. 12

Figure 2-2: Folsom Lake drainage and tributary basins configuration as updated in 2008 (red

outline). The additional segments provide information upstream of smaller

reservoirs in the Forks of the American River. 13

Figure 2-3: Schematic representation of INFORM processing pathways that utilize CFS and

GFS data for the generation of ensemble streamflow predictions with lead time up to

9 months. 15

Figure 2-4: Blended ensemble streamflow predictions of Folsom Reservoir inflow produced by

the INFORM forecast component with initial conditions on 1 March 2008 and

based on GFS (short term up to 16 days) and CFS (longer term up to 9 months).

The first 30 days are shown to highlight the blending period. 17

Figure 2-5: Mean inflow forecasts for 2006, 2007, and 2008 and historical mean inflows for all

major reservoirs. 17

Figure 2-6: Northern California system storage volumes on 1 March 2008 for all major Northern

California reservoirs. 18

Figure 2-7: Water Supply vs. Carryover Storage vs. Energy Tradeoffs for 2008. 20

x

Figure 2-8: Mean Carry Over Storage and Energy Generation Comparisons for 2006, 2007, 2008

for 50% water 21

Figure 2-9: Delta outflow and X2 location Forecast Ensembles. 22

Figure 4-1: Long Range Inflow Forecasts 35

Figure 4-2: Mid Range Inflow Forecasts 36

Figure 4-3: Forecasted Inflow Mean Comparison; Trinity 37

Figure 4-4: Forecasted Inflow Mean Comparison; Shasta 37

Figure 4-5: Forecasted Inflow Mean Comparison; Oroville 38

Figure 4-6: Forecasted Inflow Mean Comparison; Folsom 38

Figure 4-7: Basin average inflow comparisons 39

Figure 4-8: Reservoir Initial Storages 39

Figure 4-9: Sample Tradeoff Plot 1; 40

Figure 4-10: Sample Tradeoff Plot 2; 40

Figure 4-11: Reservoir Elevation Sequences 41

Figure 4-12: Reservoir Release Sequences 42

Figure 4-13: Reservoir Energy Generation Sequences 43

Figure 4-14: X2 Location Sequences 44

Figure 4-15: Delta Outflow Sequences 45

Figure 4-16: Mean Water Delivery Comparisons 46

Figure 4-17: System Energy Generation Comparisons 46

Figure 4-18: Mid Range Reservoir Elevation Sequences 47

Figure 4-19: Mid Range Reservoir Release Sequences 48

xi

Figure 4-20: Mid Range Energy Generation Sequences 49

Figure A-2.1: Historical Average Inflows (Solid Bars Correspond to Modeled Nodes) 61

Figure A-2.2: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow

Sequences, 2006 63

Figure A-2.3: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow

Sequences, 2007 64

Figure A-2.4: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow

Sequences; 2008 65

Figure A-2.5: System Monthly Water Diversion Sequences 67

Figure A-2.6: Observed Total Diversions 67

Figure A-2.7: Observed Total System Initial and Terminal Storage 68

Figure A-2.7: Observed System Energy Generation 69

Figure A-2.8: System Carry-over Storage Comparison 71

Figure A-2.9: System Energy Generation Comparison 71

Figure A-2.10: INFORM Tradeoff: Total Water Delivery vs. System Carryover Storage; 2006 72

Figure A-2.11: INFORM Tradeoff: Total Water Delivery vs. System Carry-over Storage; 2007 72

Figure A-2.12: INFORM Tradeoff: Total Water Delivery vs. System Carry-over Storage; 2008 73

Figure B-1: Long Range Inflow Forecasts 80

Figure B-2: Mid Range Inflow Forecasts 81

Figure B-3: Forecasted Inflow Mean Comparison; Trinity 82

Figure B-4: Forecasted Inflow Mean Comparison; Shasta 82

Figure B-5: Forecasted Inflow Mean Comparison; Oroville 83

xii

Figure B-6: Forecasted Inflow Mean Comparison; Folsom 83

Figure B-7: Basin average inflow comparisons 84

Figure B-8: Reservoir Initial Storages 84

Figure B-9: Sample Tradeoff Plot 1; 85

Figure B-10: Sample Tradeoff Plot 2; 85

Figure B-11: Reservoir Elevation Sequences 86

Figure B-12: Reservoir Release Sequences 87

Figure B-13: Reservoir Energy Generation Sequences 88

Figure B-14: X2 Location Sequences 89

Figure B-15: Delta Outflow Sequences 90

Figure B-16: Mean Water Delivery Comparisons 91

Figure B-17: System Energy Generation Comparisons 91

Figure B-18: Mid Range Reservoir Elevation Sequences 92

Figure B-19: Mid Range Reservoir Release Sequences 93

Figure B-20: Mid Range Energy Generation Sequences 94

1

CHAPTER 1: INTRODUCTION

1.1 OBJECTIVES

Current operations in the water sector typically do not use climate forecasts for real time hydrologic

predictions and do not use climate-driven hydrologic predictions for reservoir management. The

investigators have demonstrated the value of climate forecasts in operations through feasibility

studies and through the development of a demonstration project in close collaboration with

operational forecast and management agencies (e.g., Georgakakos et al. 2005). A near real time system

(INFORM) has been developed and implemented for the five largest Northern California reservoirs

(Folsom, New Bullards Bar, Oroville, Shasta and Trinity) producing ensemble inflow predictions

and risk-based planning and management reservoir regulation policies on the basis of NOAA NCEP

Global Forecast System (GFS) and Climate Forecast System (CFS) products (HRC-GWRI 2006).

The GFS products are used for producing ensemble reservoir inflow forecasts out to 16 days, while

the CFS forecasts are used to produce ensemble inflow forecasts out to nine months. To maintain

synchronization with NOAA hydrologic forecast models, the hydrologic components of the

INFORM system are linked with NWS operational databases at the California Nevada River

Forecast Center. A series of technology transfer seminars has paved the way for the transfer of the

decision component to the California Department of Water Resources (DWR) and the Bureau of

Reclamation (Central Valley Operations – CVO). The present project provides support for

collaboration with operational forecast and management agencies in Northern California to continue

validation of the system output in an operational environment and to develop protocols for using

system products in operational forecast and management of Northern California water supply.

The specific scientific objectives of the present project work are to:

a) Complete the validation of integrated system operational products (ensemble precipitation,

streamflow simulations and forecasts, and management scenario assessments and benefits)

through the three additional California wet seasons covered by the project activities.

b) Making necessary adjustments to the component formulations and implementations.

c) Collaborate with CNRFC, DWR and CVO staff pertaining to reciprocal technology transfer,

and training of their staff to ensure a successful transition of the system (both physical

2

hydrology and decision support components) to operations and appropriate use by

operational agencies.

d) Assess and report on the manner and the extent to which the INFORM products are used

by the operational forecast and management agencies.

To provide context for the present research, in the following sections we provide an overview of the

salient features of the INFORM system, followed by a summary of the report contents.

1.2 OVERVIEW OF INFORM SYSTEM

The INFORM software system consists of a number of diverse components for data handling,

model runs, and output archiving and presentation. At its current state of development and input

data availability, it is a distributed system with on-line and off-line components. The system routinely

captures real-time National Center for Environmental Predictions (NCEP) ensemble forecasts. It

uses both ensemble synoptic forecasts from NCEP’s Global Forecast System (GFS) and ensemble

climate forecasts from NCEP’s Climate Forecast System (CFS). The former are used for producing

real-time short-term forecasts, and the latter are used off-line for producing longer-term forecasts as

needed. The reason for the difference between the GFS and CFS processing is the data type

available in real time from each source (three dimensional 6-hourly fields from GFS, and monthly

average surface precipitation and temperature from CFS).

The INFORM ensemble forecast output feeds an off-line decision system for producing risk-based

short- and long-term management alternatives for a nine-month decision horizon. The INFORM

forecast component is implemented at the Hydrologic Research Center (HRC) for real time use and

with data links to the California Nevada River Forecast Center (CNRFC) databases. In addition, the

ensemble reservoir inflow forecasts and maps of the ensemble surface precipitation forecasts of

INFORM out to several days are posted on a secure internet site for INFORM developing

institutions and collaborating forecast and management agencies. The INFORM decision

component is implemented at the Georgia Water Resources Institute (GWRI), the U. S. Bureau of

Reclamation (USBR) and the California Department of Water Resources (DWR) for off-line use.

Figure 1-1 shows a schematic of the system distributed configuration, indicating the data links. The

3

arrows point to the site of the database from which the organization initiating the link receives and

deposits data.

GFS ensemble forecasts of three-dimensional atmospheric fields are captured, archived, ingested

and quality controlled in real time for further use. Downscaling components that use the ingested

ensemble fields produce corresponding ensemble gridded forecasts of surface precipitation and

temperature over the INFORM application area of Northern California. A Geographic Information

System (GIS) locates the gridded forecasts over the terrain of Northern California in geodetic

coordinates and estimates mean areal precipitation and surface air temperature for all ensembles and

forecast lead times, and for the hydrologic catchments that comprise the drainage areas of interest.

Hydrologic models use the downscaled ensemble forecast mean areal quantities as input to produce

ensemble forecasts of snow depth and snow melt during the cold season, and of surface and

subsurface runoff and streamflow, including reservoir-site inflow, throughout the year.

CFS ensemble forecasts of surface air temperature and precipitation with monthly resolution and

with a nine-month maximum forecast lead time are also captured in real time by the INFORM data

ingest system at HRC. At a user-specified time, a probabilistic downscaling component uses the

ensemble CFS forecasts and produces high spatial and temporal resolution surface precipitation and

temperature estimates for each hydrologic catchment in the INFORM region. The hydrologic

component of INFORM is then engaged to produce ensemble reservoir inflow estimates for the

primary reservoir sites of interest. Downscaling and hydrologic forecasting is done off-line (typically

once per month) in this case of CFS processing. The short-term (GFS-based) and long-term (CFS-

based) ensemble reservoir inflow forecasts of INFORM are blended to produce a consistent series

of input to the decision component.

The INFORM DSS is designed to support the decision making process, which is characterized by

multiple decision makers, multiple objectives, and multiple temporal scales. Toward this goal, the

INFORM DSS includes a suite of interlinked models that address reservoir planning and

management at hourly, daily, seasonal, and over-year time scales. The DSS includes models for each

major reservoir in the INFORM region, simulation components for downstream river reaches as

necessary to incorporate downstream decision objectives, optimization components suitable for use

with ensemble forecasts, and a versatile user interface. The decision software runs off-line, as

forecasts become available, to derive and assess planning and management strategies for all key

4

system reservoirs. The DSS is embedded within a user-friendly, graphical interface that links the

models with the database and helps visualize and manage results. A policy assessment model has

also been developed and is part of the DSS.

Training and collaboration with staff of CNRFC, USBR and DWR has produced an efficient

distributed INFORM system for risk-based management and planning.

Figure 1-1: Schematic diagram of the distributed INFORM system configuration with data links indicated. Black

arrows signify real-time data links while grey arrows signify off-line data links.

1.2.1 RESERVOIR SYSTEM

The scope of the originally proposed INFORM Decision Support System (INFORM DSS) included

four reservoirs: Trinity, Shasta, Oroville, and Folsom. However, the operational planning and

5

management of these reservoirs is depended upon the downstream facilities and water uses

including the Sacramento-San Joaquin Delta, and the export system to Southern California. Thus,

based on extensive discussions with the INFORM Oversight Committee, the California Department

of Water Resources, the US Bureau of Reclamation, and the US Corps of Engineers, it was decided

to expand the scope of the original four reservoir system to include most downstream elements that

have a bearing on planning decisions. More specifically, the original project scope was expanded to

include the elements shown on Figure 1-2.

Figure 1-2: A schematic of the INFORM Reservoir and River System

This system encompasses the Trinity River system, the Sacramento River system, the Feather River

system, the American River system, the San Joaquin River system, and the Sacramento-San Joaquin

Delta. Major regulation and hydropower projects on this system include the Clair Eagle Lake

San Joaquin River

San Luis

Clair Engle Lake

Trinity Power Plant

Lewiston

Lewiston

JF Carr

Whiskeytown

Shasta

Keswick

ShastaSpring Cr

Keswick

Oroville

Thermalito

Folsom

Natoma

New Melones

Tulloch

Goodwin

Oroville

Folsom

Nimbus

Melones

Tracy Pumping

Banks Pumping

San Joaquin River

Amer

ican

Riv

er

Feat

her R

iver

Sacramento River

Trinity River Clear Creek

Yuba River

Bear River

Delta-Mendota Canal

California Aqueduct

O’Neill Forebay

To Dos Amigos PP

To Mendota Pool

Sacramento San Joaquin River DeltaReservoir/

Lake

Power Plant

Pumping Plant

River Node

Reservoir/Lake

Power Plant

Pumping Plant

Reservoir/Lake

Power Plant

Pumping Plant

River Node

ISV

IFT

IES,IMC,IYB,ITI

DDLT,DBS,DCCWD,DNBA

DDM

DFDM

DDA

DSF

DSB

Black Butte

New Bullards Bar

6

(Trinity Dam) and the Whiskeytown Lake on the Trinity River, the Shasta-Keswick Lake complex

on the upper Sacramento River, the Oroville-Thermalito complex on the Feather River, the Folsom-

Nimbus complex on the American River, and several storage projects along the tributaries of the

San Joaquin River including New Melones. The Sacramento River and the San Joaquin River join to

form an extensive Delta region and eventually flow out into the Pacific Ocean. The Oroville-

Thermalito complex comprises the State Water Project (SWP), while the rest of the system facilities

are federal and comprise the Central Valley Project (CVP).

The Northern California river and reservoir system serves many vital water uses, including providing

two-thirds of the state’s drinking water, irrigating 7 million acres of the world’s most productive

farmland, and being home to hundreds of species of fish, birds, and plants. In addition, the system

protects Sacramento and other major cities from flood disasters and contributes significantly to the

production of hydroelectric energy. The Sacramento-San Joaquin Delta provides a unique

environment and is California’s most important fishery habitat. Water from the Delta is pumped

and transported through canals and aqueducts south and west serving the water needs of many more

urban, agricultural, and industrial users.

An agreement between the US Department of the Interior, Bureau of Reclamation, and the

California Department of Water Resources (1986) provides for the coordinated operation of the

SWP and CVP facilities (Agreement of Coordinated Operation-COA). The agreement aims to

ensure that each project obtains its share of water from the Delta and protects other beneficial uses

in the Delta and the Sacramento Valley. The coordination is structured around the necessity to meet

the in-basin use requirements in the Sacramento Valley and the Delta, including Delta outflow and

water quality requirements.

The expanded INFORM system is intended to “drive” the decision making process at the long

range (planning) level. An overview of the INFORM DSS is provided next to better clarify the role

of the expanded system.

At present, a number of tools are being used by the federal and state agencies responsible for the

management of the northern California water resources system. Such tools include spreadsheet

models (USBR), hydropower scheduling models (USBR), simulation models (DWR, USACE), and

forecasting models (CNRFC). However, these tools are neither vertically (planning to management

to operations) nor horizontally (agency wise) fully integrated. Perhaps, the most significant

7

contribution of the INFORM project is that it provides an integration framework and a common set

of tools for all operational agencies involved.

1.2.2 INFORM DSS OVERVIEW

The INFORM DSS modeling framework is illustrated in Figure 1-3. The DSS includes multiple

modeling layers designed to support decisions pertaining to various temporal scales and objectives.

The three modeling layers shown in the figure include (1) short range and near real time operations

decision support (which has hourly resolution and a horizon of one day), (2) mid range reservoir

management (which has an daily resolution and a horizon of several months), and (3) long range

planning (which has a monthly resolution and a horizon of one or two years). The INFORM DSS

also includes an assessment model which replicates the system response under various inflow

scenarios, system configurations, and policy options.

Figure 1-3: INFORM DSS Modeling Framework

Response Functions• Energy• Flood Damage• Spillage

Water DistributionFlow RegulationHydro Plant OperationEmergency Response

Monthly Decisions• Releases/EnergyTarget Conditions• State Variables

Planning Tradeoffs

• Water Supply/Allocation• Energy Generation• Carry-over Storage• Env.-Ecosystem Management

Development Tradeoffs

• Urban/Industrial • Agriculture• Power System• Socio-economic & Ecological

Sustainability

Operational Tradeoffs

• Flood Management• Water Distribution• Energy Generation• Env.-Ecosystem Management

Response Functions• Energy• Flood Damage• Spillage

Scenario/Policy Assessment

Monthly / Several Decades

Actual Hydrologic Conditions

Actual Demands

Climate-Hydrologic Forecasts

Demand Forecasts• Water• Food • Energy• Env.-Ecosystem

Climate-Hydrologic Forecasts

Demand Forecasts• Water Supply• Power Load/Tariffs• Flood Damage• Env.-Ecosystem Targets

Inflow Scenarios

Development/Demand Scenarios• Water/Energy Projects• Water/Benefit Sharing Agreements

Daily Decisions• Releases/EnergyTarget Conditions• State Variables

Short Range / Near Real Time Decision Support

Hourly / 1 Day

Mid Range Decision Support

Daily / several Months

Long Range Decision Support

Monthly / 1-2 Years

Infrastructure Develpmnt.Water Sharing CompactsSustainability Targets

Management Policy

Man

agem

ent

Agen

cies

/Dec

isio

ns

Plan

nin

g A

gen

cies

/Dec

isio

ns

Oper

atio

nal

Pla

nnin

g a

nd M

anag

emen

tO

ff-lin

eA

sses

smen

ts

8

The INFORM DSS is designed to operate sequentially. In a typical application, the long range

planning model is activated first to consider long range issues such as water conservation strategies

for the upcoming year in view of the climate and hydrologic forecasts. As part of these

considerations, the DSS quantifies several tradeoffs of possible interest to the planning and

management agencies and system stakeholders. These include, among others, assessments regarding

relative water allocations to water users throughout the system (including ecosystem demands),

reservoir carry over storage, reservoir coordination strategies and target levels, water quality

constraints, and energy generation targets. This information is provided to the planning and

management agencies to use as part of their decision process together with other information. After

completing these deliberations, key decisions are made on monthly water supply contracts, reservoir

releases, energy generation, and reservoir coordination strategies. The INFORM DSS planning level

is linked to the INFORM forecast component through the use of the long range forecasts (9-month

forecast ensemble). The mid range management model is activated next to consider system

operation at finer time scales. The objectives addressed here are more operational than planning

and include flood management, water supply, and power plant scheduling. This model uses mid

range forecast ensembles with a daily resolution and is intended to quantify the relative importance

of, say, upstream versus downstream flooding risks, energy generation versus flood control, and

other applicable tradeoffs. Such information is again provided to the management agencies (the

operational departments) to use it within their decision processes to select the most preferable

operational policy. Such policies are revised as new information on reservoir levels and flow

forecasts is acquired. The model is constrained by the long range decisions, unless current

conditions indicate that a departure is warranted. Lastly, the short range and near real time

operations models are activated to determine the turbine and spillway operation that realize the

hourly release decisions made by the mid range decision process. The results of this model can be

used for near real time operations.

In developing the INFORM DSS, particular attention has been placed on ensuring consistency

across modeling layers, both with respect to physical system approximations as well as with respect

to the flow of decisions. For example, the mid range management model utilizes aggregate power

plant functions that determine power generation based on reservoir level and total plant discharge.

These functions are derived by the short range and turbine load dispatching models which

determine the optimal turbine loads for each plant corresponding to the particular reservoir level

9

and total discharge. Thus, the mid range model “knows” how much power generation will actually

result from a particular daily release decision. Furthermore, the mid range model generates similar

energy functions to be used by the long range planning model. In this manner, each model has a

consistent representation of the benefits and implications of its decisions.

The three modeling layers discussed earlier address planning and management decisions. The

scenario/policy assessment model addresses longer term planning issues such as the implications of

increasing demands, inflow changes, storage re-allocation, basin development options, and

mitigation measures. The approach taken here is to simulate and compare the system response

under various inflow, demand, development, and management conditions.

Altogether, the purpose of the INFORM DSS is to provide a modeling framework responsive to the

information needs of the decision making process at all relevant time scales and water uses. The

INFORM DSS has been provided to the INFORM participating agencies. Training and

demonstration workshops have been conducted to ensure that agency personnel have the necessary

knowledge and experience to correctly use and interpret the results of the software.

1.3 REPORT ORGANIZATION

The report is organized by year of activities performed. The next three chapters summarize the

activities for the first (2007-2008), second (2008-2009) and third (2009-2010) year of project tenure.

Chapter 5 presents a discussion of lessons learned and challenges remaining, as well as conclusions

drawn from the activities of this project. Chapter 6 contains the list of references of the report.

10

Page Left Blank

11

CHAPTER 2: FIRST YEAR ACTIVITIES (2007-2008)

2.1 INTRODUCTION

First year project activities focused on the alignment of the INFORM forecast component

hydrologic models with the operational hydrologic models used at CNRFC for real-time flow

prediction, on the development of INFORM assessments for the spring and summer of 2008 based

on initial conditions on 1 March 2008 and on CFS operational forecasts, and on meetings with

operational forecast and management agencies for strategic planning, coordination and information

exchange. This Chapter summarizes these first-year activities (and associated findings) of this

transition-to-operations project.

2.2 PROJECT ACTIVITIES

2.2.1 HYDROLOGIC MODEL UPDATES

The INFORM hydrologic model component is designed to emulate the hydrologic forecast models

used in real time operations by the CNRFC. The design of the models is based on the snow, soil

models used as part of the National Weather Service River Forecast System (NWSRFS) with

hydrologic segments defined for each large watershed of interest as defined for operational

forecasting. The alignment was first done in 2005. After discussions with staff of the CNRFC it

was found necessary to re align the segments defined to conform to the updates of the operational

hydrologic forecast system. These updates involved the configuration of hydrologic segments

(basins) within large watersheds that provide inflow to the INFORM main reservoirs, and the

estimation of snow and soil model parameters. They also involved changes in the unit hydrograph

estimates for several of the segments that necessitated changes in the channel routing models of

INFORM. It is noted that the INFORM routing models use more detail to route the water in the

stream network than do the operational models, but changes in the configuration of the hydrologic

segments in some cases necessities changes in the channel routing model configuration as well.

Work focused on the large watersheds that provide inflows to Folsom (American River), New

Bullards Bar (Yuba River), Oroville (Feather River), and Shasta (Pit-McCloud-Sacramento Rivers).

The first step was to produce, test and implement software that decodes the segment definition files

12

for the NWSRFS and generates parametric files for the INFORM snow and soil components for

user specified basins. The second step was to generate software to read the unit hydrograph

coordinates from the NWSRFS segment definition files, and to estimate the parameters of a cascade

of conceptual reservoirs (INFORM routing model) that provide the least-square fit to the unit

hydrograph coordinates. The third step was to change the channel segment topological

configuration in the INFORM system parametric files to correspond to the updates in NWSRFS

segment definitions.

As an example, we show the updates done to the INFORM parametric input for the American River

that provides inflow to the Folsom Reservoir. Figure 2-1 shows the original segment

(subcatchment) configuration for the Folsom drainage. It consisted of three tributary basins

(drainages of the North, Middle and South Forks of the American River) and the local drainage

basin of the Folsom Lake. The updated configuration involves several more segments as shown in

Figure 2-2. Updating was necessary to produce inflows to upstream reservoirs in the watershed and

to correct basin boundaries. The definition of the new hydrologic segments necessitated changes in

the NWSRFS segment definition files that contain parameters for the snow, soil and unit

hydrograph models. New generic software written during the first year of the project reads the new

parameters of the segment definition files and creates the parametric tables necessary for the

INFORM models.

Figure 2-1: Folsom Lake drainage and tributary basins as configured in 2005. NFDC1 – North Fork American

drainage; MFAC1 – Middle Fork American drainage; CBDC1 – South Fork American drainage;

FOLC1 – Local drainage of Folsom Lake.

CBDC1

MFAC1

FOLC1

NFDC1

13

Figure 2-2: Folsom Lake drainage and tributary basins configuration as updated in 2008 (red outline). The

additional segments provide information upstream of smaller reservoirs in the Forks of the American

River.

Table 2-1 shows the new snow and soil parameters for all the segments now defined for the Folsom

drainage. Note that in some cases an upper and a lower subcatchment is defined to capture the

snow line separation of upper and lower elevation areas for snow and soil water accounting. These

values are used by the INFORM hydrologic forecast component models to produce ensemble

inflow forecasts for the Folsom Reservoir.

Table 2-1: New Snow and Soil Parameters for the Folsom Drainage SNOW PARAMETERS American % NFDC1UP NFDC1LW MFAC1UP MFAC1LW CBDC1UP CBDC1LW FOLC1L FMDC1 HLLC1 RRGC1 UNVC1 AKYC1 1.2000 1.1000 1.2800 1.1000 1.2000 1.1000 1.1000 1.1500 1.0900 1.1600 1.2500 1.2500 0.9000 0.9500 0.9500 1.0000 0.9000 0.9500 0.9000 1.0500 0.9000 1.0000 1.1000 1.2000 0.0700 0.0800 0.0500 0.0500 0.0500 0.0500 0.0500 0.0300 0.0300 0.0300 0.0500 0.0600 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.1500 0.0200 0.1200 0.0800 0.0400 0.0800 0.0400 0.0400 0.0800 0.0800 0.0300 0.0200 0.0100 0.2500 0.2500 0.2500 0.2500 0.2500 0.2500 0.2500 0.1000 0.1000 0.2500 0.2500 0.2500 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 2.0000 2.0000 0.1200 0.0800 0.0800 0.0800 0.0800 0.0800 0.0400 0.0400 0.0400 0.0800 0.0600 0.0200 0.3000 0.3000 0.3000 0.3000 0.3000 0.3000 0.1000 0.1500 0.1500 0.3000 0.3000 0.3000 1.9000 1.5000 1.9000 1.5000 1.8000 1.5000 1.5000 1.4000 0.5000 1.5000 1.5000 1.0000 1600.0000 409.0000 1600.0000 409.0000 1600.0000 409.0000 200.0000 1600.0000 1600.0000 1600.0000 1600.0000 1600.0000 19.0000 11.0000 19.0000 11.6000 19.5000 10.4000 6.2500 19.2000 20.7000 19.0000 19.4000 21.7500 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.6000 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 0.4500 19.8100 10.6700 19.8100 10.6700 19.8100 10.6700 4.8800 19.8100 19.8100 19.8100 19.8100 19.8100 1.2500 1.0200 0.9500 0.9500 1.2000 0.9500 0.9600 1.2000 1.4500 1.3000 1.1400 1.0300 SAC PARAMETERS American 145.0000 155.0000 155.0000 155.0000 155.0000 155.0000 155.0000 115.0000 95.0000 115.0000 120.0000 115.0000 50.0000 50.0000 50.0000 40.0000 40.0000 40.0000 30.0000 40.0000 30.0000 40.0000 40.0000 30.0000 250.0000 310.0000 300.0000 310.0000 350.0000 380.0000 140.0000 150.0000 150.0000 90.0000 170.0000 150.0000 210.0000 70.0000 110.0000 180.0000 60.0000 140.0000 180.0000 50.0000 70.0000 20.0000 70.0000 60.0000 130.0000 90.0000 60.0000 120.0000 110.0000 120.0000 45.0000 80.0000 70.0000 80.0000 100.0000 170.0000 0.2500 0.1500 0.2000 0.2500 0.2000 0.2500 0.4500 0.2000 0.4000 0.2500 0.2500 0.5000 0.0020 0.0040 0.0020 0.0020 0.0020 0.0050 0.0030 0.0020 0.0020 0.0020 0.0020 0.0020 0.0800 0.1000 0.0600 0.1000 0.0600 0.1000 0.0900 0.0600 0.0600 0.0600 0.0600 0.0600 12.0000 9.0000 14.0000 16.0000 10.0000 16.0000 12.0000 15.0000 14.0000 12.0000 12.0000 15.0000 1.0500 1.1000 1.2000 1.1000 1.2000 1.1000 1.3000 1.3000 1.2000 1.2000 1.2000 1.2000 0.3000 0.2000 0.2000 0.2000 0.2000 0.2000 0.3000 0.2000 0.1500 0.0600 0.3000 0.3000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.2500 0.0000 0.0000 0.0000 0.0000 0.0000 0.1000 0.1000 0.0400 0.1500 0.1300 0.1200 0.0050 0.0080 0.0050 0.0080 0.0050 0.0080 0.0400 0.0050 0.0050 0.0200 0.0050 0.0050 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000 1.0000

FOLC1

NFDC1

MFAC1

CBDC1

AKYC1

HLLC1

UNVC1

RRGC1

FMDC1

14

Changes in the hydrologic segment definition may necessitate changes in the routing configuration

of the channel routing model of the INFORM forecast component. Kinematic routing is used for

these mountainous watersheds with the routing equations written in the form of a cascade of linear

reservoirs for each routing channel reach. This model can emulate the unit hydrograph model used

in operations by CNRFC by forcing the parameters of the cascade to produce unit pulse response

equivalent to the unit hydrograph response. In this case “equivalent” is in terms of least square

error (see Sperfslage and Georgakakos 1996 for details). During this first year of the project software

was written to produce the parameters of the cascade of linear reservoirs for each segment for which

there is a unit hydrograph defined by CNRFC. In addition, changes in the association between

routing channel reaches and newly defined hydrologic segments were made to allocate correctly the

lateral local inflow to the channel reaches. Table 2-2 shows the channel routing parameters for the

Folsom drainage as obtained after the updating process.

Table 2-2: Channel Routing Reaches and Model Parameters for Folsom Drainage KINEMATIC CHANNEL ROUTING MODEL PARAMETERS Reach n alpha Area Segment Description 1 2 5.69 327.6 1 %NF-UP 2 2 3.36 557.8 2 %NF-LO 3 2 6.66 277.6 3 %MF-UP 4 2 3.43 538.9 4 %MF-LO 5 2 8.01 191.1 5 %SF-UP 6 2 2.39 639.9 6 %SF-LO 7 2 5.63 293.7 7 %J1-LO 8 2 5.63 146.8 7 %LOCAL-NF+MF 9 2 3.38 587.3 7 %LOCAL-SF 10 3 5.73 148.9 8 %FMDC1 (ALL UP) 11 2 3.46 295.1 9 %HLLC1 (ALL UP) 12 3 3.99 123.3 10 %RRGC1 (ALL UP) 13 2 2.11 217.5 11 %UNVC1 (ALL UP) 14 2 2.11 499.7 12 %AKYC1 (ALL UP)

Analogous procedures were followed for all the watersheds in the INFORM Northern California

area and new parametric files were obtained for INFORM. This process is a vitally important

process to permit transition to operations of the INFORM system in a way that will be easily

ingested and sustainable (without the need for extensive changes in model structure and parameters

for use by field personnel). The first year work also produced a set of utility codes that may be used

in the future when new updates are made to the hydrologic segment definitions by the CNRFC or

other operational forecast agencies.

15

2.2.2 INFORM ASSESSMENTS FOR 2008 OPERATIONS

Prior assessments made using INFORM system ensemble forecast results produced for a date of 1

March in each year and for a 9-month period may be found in HRC-GWRI (2006, 2007). This

section documents the activities pertaining to the assessments made for 2008 using the INFORM

system and initial conditions on 1 March 2008. The ensemble forecasts produced by the INFORM

system component that served as input to the INFORM system decision component were driven by

Climate Forecast System (CFS) ensemble predictions. Figure 2-3 shows the processing pathway

followed for the CFS-driven ensemble predictions used herein. It is noted that, as opposed to the

GFS-driven real time INFORM ensemble predictions that utilize the 3-D ensemble predictions of

Global Forecast System (GFS - right pathway in Figure 3), the CFS-driven longer-term predictions

utilize the predicted monthly surface variables of precipitation and temperature.

Figure 2-3: Schematic representation of INFORM processing pathways that utilize CFS and GFS data for the

generation of ensemble streamflow predictions with lead time up to 9 months.

16

These ensemble CFS predictions of the surface variables on a monthly scale are used to condition

the operational Ensemble Streamflow Prediction (ESP) procedure that employs historical

precipitation and temperature time series information. A description of the probabilistic approach

for conditioning and downscaling is given in Carpenter and Georgakakos (2001). Discussion of the

blending of the shorter-term (0 – 16 days) GFS-driven ensemble inflow predictions with the longer-

term (1 – 9 months) CFS-driven ensemble inflow predictions is given in HRC-GWRI (2006). We

only note here that blending is accomplished through the persistence of the snow pack and soil

moisture states in INFORM. We also note that the on-going second phase of the INFORM project

will produce CFS-driven ensemble predictions using available 3-D information from CFS output,

much like is done for GFS output.

Comparison of the 1 March 2008 CFS ensemble predictions of surface precipitation with the

climatology of such predictions for the last 20 years indicated that the nine month period is expected

to be near the average. Therefore conditioning on CFS this year produced the same results as the

unconditioned ESP procedure (Carpenter and Georgakakos 2001). As an example of the results, Figure

2-4 shows the Folsom Reservoir ensemble inflow predictions, conditioned on operational CFS

ensemble predictions of surface precipitation, for the blending period of the first month. Analogous

results were produced for all the other reservoir inflows for the same prediction horizon of 9

months. Comparison of the 2008 ensemble forecast predictions to those of the last two years and

to the historical averages may be made by reference to the results of Figure 2-5. This Figure displays

the inflow means for each case of year and major reservoir. Oroville and Shasta show significantly

lower mean forecast inflows in 2008 than in the previous two years.

These ensemble inflow predictions were used by the INFORM decision component to assess the

impacts on reservoir operations for the prediction horizon. The components of the INFORM

decision system used for the assessments are shown in Figure 1-2. The components are described in

detail in HRC-GWRI (2006). They include: Trinity River System (Clair Engle Lake, Trinity Power

Plant, Lewiston Lake, Lewiston Plant, JF Carr Plant, Whiskeytown, Clear Creek, and Spring Creek

Plant); Shasta Lake System (Shasta Lake, Shasta Power Plant, Keswick Lake, Keswick Plant, and the

river reach from Keswick to Wilkins); Feather River System (Oroville Lake, Oroville Power Plants,

Thermalito Diversion Pond, Yuba River, and Bear River); American River System (Folsom Lake,

Folsom Plant, Natoma Lake, Nimbus Plant, Natoma Plant, and Natoma Diversions); San Joaquin

River System (New Melones Lake, New Melones Power Plant, Tulloch Lake, Demands from

17

Goodwin, and Inflows from the main San Joaquin River); and Bay Delta (Delta Inflows, Delta

Exports, Coordinated Operation Agreement--COA, and Delta Environmental Requirements).

Figure 2-4: Blended ensemble streamflow predictions of Folsom Reservoir inflow produced by the INFORM

forecast component with initial conditions on 1 March 2008 and based on GFS (short term up to 16

days) and CFS (longer term up to 9 months). The first 30 days are shown to highlight the blending

period.

Figure 2-5: Mean inflow forecasts for 2006, 2007, and 2008 and historical mean inflows for all major reservoirs.

18

Forecasted inflows were provided to the decision component with start date 1 March (112 traces, 9 -

month horizon, and five locations: Clair Engle Lake (Trinity), Shasta, Oroville, Folsom, and Yuba).

Historical monthly average values were used for locations where forecasted inflows were not

available. The decision model was run using monthly reservoir parameters and constraints (max,

min, and target storage levels; evaporation rates); minimum river flow and Bay Delta requirements;

and base monthly demands at all locations. The decision objective was to develop the tradeoff

between water supply deliveries and carry-over storage that meets all other stated system

requirements.

The initial water volumes in the major reservoirs in Northern California on the 1st of March play a

significant role in management operations. Figure 2-6 shows that in 2008, reservoir storage at the

beginning of March was lower than that of the years 2006 and 2007, especially for the Oroville and

Shasta reservoirs. This together with the lower mean inflow volumes predicted (as mentioned earlier

for the results of Figure 2-5), is expected to substantially influence system operations in 2008.

Figure 2-6: Northern California system storage volumes on 1 March 2008 for all major Northern California

reservoirs.

Reservoir Initial Storages On March 1st

2013

3872

2992

463

2020 20191902

3786

2997

594

20021895

1490

2660

1456

375

1532

1774

0

500

1000

1500

2000

2500

3000

3500

4000

4500

Trinity Shasta Oroville Folsom New Melones San Lius

Stor

age

(100

0 AF

)

Y2006Y2007Y2008

19

More specifically, the cumulative initial system storage on March 1st, 2008, was approximately 50%

of the total storage capacity, in contrast to the initial storage situation on March 1st, 2007, which was

approximately 80% of the total storage capacity, and to the total storage on March 1st, 2006, which

was 81% of the total storage capacity. The low initial storage and the below normal inflow forecasts

combine to create stressful conditions for the Northern California river system in 2008.

These stresses are depicted on the tradeoffs of Figure 2-7 which quantify the expected relationship

between water deliveries, carry over storage, and energy generation for 2008 (red line) and the

average hydrology (black line). The figure shows that for the same water delivery fraction, carry

over storage and energy generation are expected to be below average in 2008. For water deliveries

corresponding to 50% of the base line demands, carry over storage and energy generation are

expected to be 17.6% and 9.5% less than those that would have materialized under normal

hydrologic conditions.

Figure 2-8 compares the end of November storages (carry over storages) and energy generation for

2006, 2007, and 2008 for water deliveries equal to 50% of the baseline demand. The figure illustrates

that meeting water deliveries at the 50% baseline level in 2008 has the potential to deplete carry over

storage level down to 20% of the system capacity. Furthermore, hydro-generation in 2008 is

expected to be 32% less than in 2006 and 24% less than in 2007.

Alternatively, the previous results imply that to maintain the same system carry over storage, water

deliveries in 2008 must be reduced almost by 50% of those in 2006 or 45% of those in 2007, a very

significant reduction.

20

Figure 2-7: Water Supply vs. Carryover Storage vs. Energy Tradeoffs for 2008.

Total Demand Fraction vs. Terminal Storage Tradeoff

8871

8040

7210

6373

5533

88808372

8078

7478

6718

4000

5000

6000

7000

8000

9000

10000

0 0.1 0.2 0.3 0.4 0.5 0.6

Demand Fraction

Term

inal

Sto

rage

(100

0 A

F)

Y2008

Historical Mean

Total Demand Fraction vs. System Energy Tradeoff

3736

4164

4556

4923

5254

47554932

5233

5515

5802

3000

3500

4000

4500

5000

5500

6000

0 0.1 0.2 0.3 0.4 0.5 0.6

Demand Fraction

Ener

gy (G

WH

)

Y2008

Historical Mean

21

Figure 2-8: Mean Carry Over Storage and Energy Generation Comparisons for 2006, 2007, 2008 for 50% water

delivery fraction.

The previous discussion is focused on mean quantities. The INFORM DSS provides the means to

assess the entire frequency distrbution as illustrated on Figure 2-9. This figure shows the delta

outflow and X2 location forecast ensembles for the 10% water deliveries case. All model runs

ensure that the delta X2 constraint (maximum 80% distance from the Golden Gate Bridge) is met

for a user-specified reliability, set to 90% in the above described experiments.

Simulated Total Terminal Mean Storage Comparison

95378995

5533

0

2000

4000

6000

8000

10000

12000

Y2006 Y2007 Y2008

1000

AF

Simulated System Mean Energy Comparison

7841

6946

5254

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

Y2006 Y2007 Y2008

GW

H

22

Figure 2-9: Delta outflow and X2 location Forecast Ensembles.

2.2.3 MEETINGS WITH OPERATIONAL FORECAST AND MANAGEMENT AGENCIES

During the first year of the project two meetings were held in Sacramento at the CNRFC offices.

The first on 20 March involved the INFORM Oversight and Implementation Committee (OIC) and

the second on 3 April involved representatives from the Department of Water Resources (DWR)

and CNRFC. The first meeting was to provide strategic advice to the PIs of the project by

representatives of operational forecast and management Agencies that are interested in the use of

INFORM forecasts and decision support and to present to such representatives the assessments for

2008 (see previous section). The second meeting was to discuss the INFORM management focus

for the California Department of Water Resources, the updating of the INFORM hydrologic

modeling components (see section 2.1 above), the possible expansion of the OIC to include

representatives from additional Agencies and expertise, and to develop a plan for treating upstream

regulation in an ESP framework. The Summaries of meeting proceedings for both meetings follow.

23

2.2.3.1 Meeting of 20 March 2008

PARTICIPANTS

Agency Representatives

Michael Anderson California Department of Water Resources

Paul Fujitani Central Valley Operations, U.S. Bureau of Reclamation

Robert Hartman California Nevada River Forecast Center, National Weather Service, NOAA

Joe O’Hagan California Energy Commission (through a conference call)

Tom Morstein-Marx Central Valley Operations, U.S. Bureau of Reclamation

Eric Strem California Nevada River Forecast Center, National Weather Service, NOAA

Dingchen Hou National Centers of Environmental Prediction, NOAA (through a

(for Zoltan Toth) conference call)

INFORM Co-PIs and INFORM Project Scientists

Aris Georgakakos (Co-PI) Georgia Water Resources Institute

Kosta Georgakakos (PI) Hydrologic Research Center

Nick Graham (Co-PI) Hydrologic Research Center

Martin Kistenmacher Georgia Water Resources Institute

Huaming Yao Georgia Water Resources Institute

LOCATION AND TIME

The meeting was held at the Joint Operations Center (3310 El Camino Ave.) in Sacramento on the

20th of March 2008. The meeting started at 1:15PM and ended at 3:15PM.

PURPOSE AND INFORMATIONAL MATERIAL

The meeting served as the first critical review meeting for INFORM (Integrated Forecast and

Reservoir Management), Phase II. The meeting presentations in PDF form were made available to

all the OIC members and participants through the following link at the HRC web site:

http://www.hrc-lab.org/INFORM/. In addition, the published reports of the previous phases of

INFORM were also made available via links to the INFORM Core Program Office web site and

through transmission via email. The INFORM Core Program Office web site is:

24

http://www.hrc-lab.org/projects/dsp_projectSubPage.php?subpage=inform.

INFORM STATUS PRESENTATION AND DEMONSTRATION RESULTS FOR LAST

THREE YEARS

The Co-PIs summarized the technical activities of (a) the three-year first phase of INFORM (funded

by CALFED, the California Energy Commission, and the Climate Program Office of NOAA), (b)

the Interim Phase of INFORM (funded by the Climate Program Office and the Office of

Hydrologic Development of NOAA), and (c) the activities of Phase II of INFORM so far (funded

by NOAA Climate Program Office and with contracting in progress with the California Energy

Commission and the California Department of Water Resources). The Co-PIs discussed the design

and implementation of the forecast and decision support components of INFORM and the focus of

the demonstration effort so far, including a presentation of the principle results for the

demonstration winter seasons: ’05 - ’06, ’06 – ’07, and ’07 – ’08. It was noted that with the current

long term (up to 9 months) and large scale forecasts by the National Centers of Environmental

Prediction (NCEP) of the National Oceanic and Atmospheric Administration (NOAA), the current

conditions of snow pack and soil moisture, and the current levels of reservoir storage in Northern

California, the INFORM forecast and decision components indicate that this year (’07 – ’08) has a

good chance of being significantly below average in terms of possible reservoir releases available to

meet demands. These assessments were made with information at the beginning of the month of

March 2008.

DISCUSSION

The participants discussed the role of INFORM for real time operations and the possible expansion

of the Oversight and Implementation Committee to include other key participants for the second

phase of INFORM (e.g., possibly with knowledge and experience in forecasting and management

associated with flood control, fisheries, Bay Delta issues, energy industry concerns). The OIC

members also emphasized the importance of generating INFORM products that are easily ingested

within routine operations of the forecast and management agencies at the end of Phase II. The

INFORM developers emphasized their desire to produce INFORM products and software that can

be used by management agencies to determine the range of the most profitable release policies at

various risk levels and with due consideration of short and long term forecasts prior to the

25

engagement of their existing detailed simulation models that are used to finalize the most

appropriate release policy. The suggestion was made to link some of the INFORM demonstration

activities with Folsom reoperation, as a possible demonstration effort. It was also mentioned that a

significant component of the Phase II project will be to continue the effort to make forecast and

management agencies familiar with the basis and use of the INFORM methodology, which has been

designed and tested over a period of two decades to address decision problems in operations given

uncertain short and long term forecast information.

As a result of the discussions held, it was decided to have a follow-on visit of the INFORM PI in

two weeks time to coordinate with Northern California forecast and management agencies the

details of the expansion of the OIC, and the development of a phased plan for agency science

cooperation and technology transfer in the context of INFORM Phase II.

2.2.3.2 Meetings of 3 April 2008

The meeting was with the Department of Water Resources representatives at first and then with

CNRFC representatives in Sacramento at the Operations Center. The participants were Gary

Bardini and Mike Anderson from DWR and Robert Hartman from CNRFC. A number of ideas

were discussed with respect to the phase II of INFORM and items of particular interest to these

forecast and management Agencies were identified. Among other things, the real time production

of ensemble forecasts driven by GFS and CFS operational ensemble predictions for the Northern

California domain with resolution of 10 km was identified as a useful complement to operations at

CNRFC, and the uncertainty management ability of the decision component of INFORM was

identified as particularly useful for DWR operations. Several potential additional members of the

OIC committee were identified and the PI of the project was tasked to approach such individuals to

discuss their potential participation once the second phase of INFORM is fully contracted (in

progress). With respect to more technical matters, the issue of upstream regulation was the focus of

much of the technical discussion with CNRFC.

Current Ensemble Streamflow Prediction (ESP) procedures provide risk based information through

the production of ensemble members for full natural flows for regulated points or actual flows for

unregulated points. However, users of this information are typically concerned with actual flows

even at points with upstream regulation. It is desired to design and test a procedure to incorporate

26

upstream regulation to the production of an ESP. Instead of a site specific implementation, which

appears to be infeasible for general application as part of the NWS ESP procedures, it appears more

appropriate to develop a structured parameterized approach to simulate upstream regulation so that

it may be used for various sites with the requirement that historical data is available to estimate the

parameters and possibly the structure functions of the methodology.

The American River was discussed as a potential case study for this analysis, with emphasis on the

Middle and South Forks with significant upstream regulation. To obtain data and information for

the structure function approach to be followed, participation of Agencies responsible for upstream

regulation is desirable. A workshop for bringing such agencies together was discussed for May 2008.

2.3 CONCLUDING REMARKS

The first year of the NOAA transition to Operations Project titled: Operational Multiscale Forecast

and Reservoir Management in Northern California was devoted to close collaboration with

operational forecast and management agencies, alignment of the forecast and management systems

developed by the PIs with those of the operational Agencies, and production of assessments for the

ability of the system to meet requirements and satisfy objectives of operation for 2008. Based on

Agency recommendations a number of new utility software components were developed to facilitate

the transition of the INFORM forecast and management components to operations. The 2008

assessment for Northern California water supply concluded that 2008 will be a low year for Oroville

and Shasta both in terms of expected inflows but also in terms of initial storage volume, and thus

will lead to lower terminal storage volumes, lower energy production and less flexible trade-offs. It

appears likely that 2008 water deliveries might be 40 to 50% less than those in 2007 and 2006.

27

CHAPTER 3: SECOND YEAR OF ACTIVITIES (2008-2009)

3.1 INTRODUCTION

The second year activities focused on establishing the validity of the system quantitative results for

the Northern California water management agencies through an independent evaluation of the

forecasts and assessments for the first three years using the INFORM system against observations

of hydrologic conditions and against actual reservoir management outcomes. The results were

presented in a meeting of the agencies in Northern California. These results together with the

continuing collaboration with the Northern California forecast and management agencies

contributed to the funding of the second 3-year phase of INFORM (INFORM II) that will

complement as matching funding the transition to operations that is partially funded by the present

project. In addition during the second year, the project team worked on the development of

INFORM assessments for the spring, summer and fall of 2009 based on initial conditions on 15

March 2009 and on GFS/CFS operational forecasts. Chapter 3 summarizes these second-year

activities (and associated findings) of this transition-to-operations project.

3.2 PROJECT ACTIVITIES

3.2.1 EVALUATION OF INFORM REAL TIME FORECAST AND MANAGEMENT

ASSESSMENTS (2006, 2007 AND 2008)

The California Energy Commission undertook the independent evaluation of the first three years of

the INFORM project real time assessments (spring-fall of 2006, 2007, and 2008) to assess the utility

of the project from the perspective of the management agencies against actual observations of

hydrologic conditions and actual reservoir release decisions. As part of this process, during the

period September – December 2008, HRC and GWRI cooperated with the Energy Commission and

with their consultants to develop an initial evaluation of the project results and effectiveness. The

HRC-GWRI resultant report is included herein in its entirety as Appendix A.

The main conclusions are:

28

(a) This initial comparison indicates that the information provided by the INFORM system in

operational planning and management of the northern California river and reservoir system is

relevant, reliable, and decision worthy. This is corroborated by the favorable comparison of the

predicted and actually observed system outputs (water deliveries, carry-over storages, and energy

generation amounts) over the March to November 2006, 2007, and 2008 seasons.

(b) The value of the INFORM forecast-decision system ultimately depends on whether the

information it generates provides management agencies with good appreciation of the benefits and

risks associated with different decisions and helps them adopt those that serve well the interests of

the system stakeholders. The Northern California management Agencies were very supportive of

the system in interviews associated with this evaluation.

The results of the evaluation were discussed in several conference calls and a presentation for the

management agencies of Northern California. This evaluation and the continuing collaboration of

the HRC-GWRI team with operational forecast and management agencies in Northern California

contributed to the decision of the California Energy Commission to continue the support of the

INFORM project for another 3 years (phase II, 2009-2012) as matching funding for needed

INFORM upgrades (requested by the management agencies), the complete transition to operations

and continuing evaluation of the real time forecasts and decision assessments.

3.2.2 INFORM ASSESSMENTS FOR 2009 OPERATIONS

Prior assessments made using INFORM system ensemble forecast results produced for a date of 1

March in each year and for a 9-month period may be found in HRC-GWRI (2006, 2007) and

Georgakakos et al. (2008). This section documents the activities pertaining to the assessments made

for 2009 using the INFORM system and initial conditions on 15 March 2009. The ensemble

forecasts produced by the INFORM system component that served as input to the INFORM

system decision component were driven by Climate Forecast System (CFS) ensemble predictions.

Figure 2-1 shows the processing pathway followed for the CFS-driven ensemble predictions used

herein. It is noted that, as opposed to the GFS-driven real time INFORM ensemble predictions that

utilize the 3-D ensemble predictions of Global Forecast System (GFS - right pathway in Figure 2-1),

the CFS-driven longer-term predictions utilize the predicted monthly surface variables of

precipitation and temperature.

29

These ensemble CFS predictions of the surface variables on a monthly scale are used to condition

the operational Ensemble Streamflow Prediction (ESP) procedure that employs historical

precipitation and temperature time series information. A description of the probabilistic approach

for conditioning and downscaling is given in Carpenter and Georgakakos (2001). Discussion of the

blending of the shorter-term (0 – 16 days) GFS-driven ensemble inflow predictions with the longer-

term (1 – 9 months) CFS-driven ensemble inflow predictions is given in HRC-GWRI (2006). We

only note here that blending is accomplished through the persistence of the snow pack and soil

moisture states in INFORM.

To generate the ensemble reservoir inflow forecasts for this year, we introduced uncertainty in the

initial snow water equivalent (uncertainty range of 15% of initial value) for all the subcatchments

modeled in INFORM to account for repeated failures of the transmission of the full ensemble of

GFS forcing from NCEP. These failures introduced uncertainty to the estimation of the snow water

equivalent of the snow pack by the INFORM hydrology component in March 2009. Table 3-1

shows the snow water equivalent of all the model subcatchments for 2009 and compares it to that of

earlier years.

Table 3-1: Snow water equivalent estimated for each INFORM hydrology model subcatchment.

(The imposed uncertainty range is shown in parenthesis)

FOLSOM DRAINAGE (Number of subcatchments=7) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 455 565 982 856 (730-980) 2 2 48 90 37 ( 32- 43) 3 225 358 630 521 (445-597) 4 35 168 179 141 (121-162) 5 340 355 676 757 (647-868) 6 0 77 42 49 ( 42- 56) 7 0 3 0 0 ( 0- 0) OROVILLE DRAINAGE (Number of subcatchments=12) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 210 245 516 451 (391-504) 2 75 102 327 159 (138-178) 3 142 186 289 188 (164-211) 4 3 91 47 0 ( 0- 0) 5 353 500 909 605 (525-677) 6 0 47 65 0 ( 0- 0)

30

7 10 130 71 92 ( 79-102) 8 0 55 6 4 ( 3- 5) 9 350 488 753 613 (532-685) 10 6 121 276 97 ( 84-108) 11 335 575 1133 775 (673-867) 12 0 27 0 0 ( 0- 0) SHASTA DRAINAGE (Number of subcatchments=10) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 m 65 67 67 ( 59- 77) 2 i 33 37 0 ( 0- 0) 3 s 128 185 113 ( 98-129) 4 s 30 40 2 ( 2- 3) 5 i 707 1044 946 (826-1080) 6 n 32 96 66 ( 58- 76) 7 g 422 798 697 (608-795) 8 ! 65 108 77 ( 67- 88) 9 ! 37 0 0 ( 0- 0) 10 ! 34 72 55 ( 48- 63) TRINITY DRAINAGE (Number of subcatchments=2) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 680 440 705 735 (642-839) 2 5 73 111 35 ( 30- 40) YUBA DRAINAGE (Number of subcatchments=4) 3/1/2006 3/1/2007 3/1/2008 3/13/2009 1 407 506 507 395 (337-452) 2 0 66 73 0 ( 0- 0) 3 457 450 787 650 (556-745) 4 0 21 0 0 ( 0- 0)

The results show that the snow water equivalent in 2009 for all the reservoir drainages but Shasta

and Trinity is estimated to be lower than in 2008, especially for the southernmost drainages of the

American and Yuba Rivers included in the INFORM system. The ensemble reservoir inflow

forecasts were generated and were used as input to the INFORM decision support (DSS)

component.

A detailed analysis of the decision model assessment runs is included in Appendix B. A summary of

the findings is provided below.

The 2009 assessment portends a dry year similar to 2008. In fact, system performance is expected to

be somewhat worse than 2008 due to the already depleted reservoir storages. Depending on the

31

amount of carry-over storage target (at the end of 2009), average water deliveries are estimated to be

in the range of 2 to 2.4 million acre feet (MAF), which represents only 35% of the average 2007

amount an 30% of the average 2006. Beyond this water deliveries commitment, the likelihood that

Shasta and Oroville may fully deplete becomes appreciable. Average energy generation over the

March 15 to November 15 period ranges from 4,152 to 4,397 GWH, or approximately 55% of the

2007 generation and 45% of the 2006 generation.

In light of the above assessment results, we recommend that 2009 water management policies and

commitments be conservative and revised adaptively as inflow information becomes available.

3.3 CONCLUDING REMARKS

The second year of the NOAA transition to operations Project titled: Operational Multiscale

Forecast and Reservoir Management in Northern California was devoted to evaluating the

effectiveness of the INFORM real time forecasts and management assessments in collaboration with

operational forecast and management agencies, and production of assessments for the ability of the

system to meet requirements and satisfy objectives of operation for 2009. The system evaluation

indicated that INFORM is a useful tool for integrated forecast and management in Northern

California. The 2009 assessment for Northern California water supply management concluded that

2009 will be a difficult year in terms of meeting demand, with higher demand deficits possible than

in 2008. Adaptive management is recommended to maximize benefits for the system under these

conditions.

32

CHAPTER 4: THIRD YEAR ACTIVITIES (2009-2010)

4.1 INTRODUCTION

The third year of the NOAA TRACS project was devoted to the assessments of the 2010 period and

collaboration with the Northern California operational forecast and management agencies for the

profitable incorporation of the INFORM ensemble forecasts and management planning assessments

in operational decision making. Several meetings took place in Sacramento with the California

Nevada River Forecast Center (CNRFC) of the US National Weather Service (NWS), and with the

California Energy Commission, the California Department of Water Resources (DWR) and the U.S.

Bureau of Reclamation, Central Valley Operations (CVO), to coordinate the interpretation and use

of the INFORM outlooks and assessments in a beneficial manner for the agencies. Also, an external

evaluation of the INFORM system was commissioned by the Energy Commission with input by

end-user agencies and developers with positive results for the system. As a result of the activities of

this project and of the positive external review, additional funding by the agencies was generated to

produce enhancements to the existing INFORM system to accommodate agency requirements.

The present chapter focuses on the INFORM assessments for 2010, which were communicated to

the agencies in April 2010 and which provide one of the input pieces of information for decision

making regarding water allocation in California. The next section describes these assessments in

detail.

4.2 ASSESSMENTS FOR YEAR 2010 – DRY SEASON IN CALIFORNIA

The application described here utilizes the following input data:

· Forecasted inflows start from March 15, 2010 (88 traces, 9 month horizon, and five

locations: Clair Engle Lake, Shasta, Oroville, Folsom, and Yuba);

· Historical monthly average values are used for locations where forecasted inflows are not

available;

33

· Monthly reservoir parameters and constraints (max, min, and target storage, evaporation

rates);

· Minimum river flow requirements;

· Base monthly demands at all locations;

· Reservoir initial storages are set to their actual values on March 15, 2010.

The management system configuration is shown in Figure 1-3. As in past years, assessments used

the INFORM ensemble reservoir inflow forecasts generated using Climate Forecast System (CFS)

forcing from the National Centers of Environmental Prediction (NCEP) of NOAA.

Inflows: The forecasted monthly inflow ensembles are shown in Figures 4-1 and 4-2. The

comparisons between the forecasted inflow mean and the corresponding historical means for four

major reservoirs are plotted in Figures 4-3 through 4-6. As shown, the forecasted inflow means at

both Folsom and Trinity are higher than the historical values during high flow months and lower

during low flow months. The forecasted means at both Shasta and Oroville are lower than their

historical means. Figure 4-7 shows the forecasted basin total inflow means for 2006 to 2010. The

result for 2010 is close to that of 2007, slightly higher than those of 2008 and 2009. Overall, the

forecasts indicate that 2010 will be approximately 5% drier than the average year. Figure 4-8 shows

the initial reservoir storages on March 1 for years from 2006 and 2010, which indicates a slight

recovery from the previous year at the same time for all major reservoirs.

Water Deliveries and Energy Generation: Using the forecasted inflows, tradeoffs are generated

by changing the base demands for all locations with fractions 60% to 100%. The tradeoffs between

the total reservoir carryover storages and the system energy versus the system water deliveries are

depicted in Figures 4-9 and 4-10. As demands increase, the reservoir carryover storages decrease.

Energy generation increases as downstream demands increase because of higher reservoir releases.

The results show that the system can meet water deliveries up to 4,194 TAF, at the 92% tradeoff

point. Meeting demands beyond this level would result in significant reservoir drawdown (especially

at Shasta and Oroville) and diminished carryover storage.

Figure 4-16 compares water deliveries over the 2006-2010 five-year period assuming the actual initial

storages and approximately the same carryover storage of 7,500 TAF. As shown, the expected water

34

deliveries are reduced by 40% compared to 2006 and 33% compared to 2007, but increased by 80%

compared to 2008, 100% to 2009. Similarly, average system energy generation for 2010 is shown in

Figure 4-17, showing a decrease of 42% relative to 2006 and 35% relative to 2007, but an increase of

12% to 2008, 20% to 2009. In summary, the forecast for 2010 shows a wetter year than the two

previous years 2008 and 2009, but still drier than the historical average.

Selected reservoir elevation, release, and energy generation sequences corresponding to 3,728

thousand acre feet (TAF), 84% of base demands, are shown in Figures 4-11 through 4-13.

X2 and Delta Outflow: The X2 location sequences are shown in Figure 4-14, indicating all traces

below 80 km, the maximum constraint set in the study. The X2 location stays within this constraint

for all tradeoff points. The Delta outflow sequences are plotted in Figure 4-15.

The first month releases determined in the long range planning model are passed on to the mid

range model for generation of daily operation decisions. The daily sequences of elevation, release,

and energy generation of four major reservoirs Trinity, Shasta, Oroville, and Folsom are generated

and plotted in Figures 4-18 to 4-20. The average daily release values can be used to guide day-to-day

operations.

4.2.1 SUMMARY

The 2010 assessment portends a wetter year than the last two years 2008 and 2009, but still drier

than the historical average year. System performance is expected to improve comparing to 2009.

Depending on the amount of carry-over storage target (at the end of 2010), average water deliveries

are estimated to be in the range of 3 to 4.5 million acre feet (MAF), which represents an 80%

increase over 2008, and 100% increase over 2009. Beyond this water deliveries commitment, there

is a small likelihood that Shasta and Oroville may experience significant drawdown (below 50% of

their total capacity). Average energy generation over the March 15 to November 15 period ranges

from 4,437 to 5,520 GWH, approximately 12% higher than 2008 and 20% higher than 2009.

In light of the above assessment results, we recommend that 2010 water management policies and

commitments follow closely those of an average water year, but be adaptively revised as future

inflow information becomes available.

35

Figure 4-1: Long Range Inflow Forecasts

36

Figure 4-2: Mid Range Inflow Forecasts

37

Figure 4-3: Forecasted Inflow Mean Comparison; Trinity

Figure 4-4: Forecasted Inflow Mean Comparison; Shasta

Forecasted Inflow Means - Trinity

0

500

1000

1500

2000

2500

3000

3500

4000

4500

Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10

cfs

F2010

His.

Forecasted Inflow Means - Shasta

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10

cfs

F2010

His.

38

Figure 4-5: Forecasted Inflow Mean Comparison; Oroville

Figure 4-6: Forecasted Inflow Mean Comparison; Folsom

Forecasted Inflow Means - Oroville

0

2000

4000

6000

8000

10000

12000

14000

Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10

cfs

F2010

His.

Forecasted Inflow Means - Folsom

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Mar-10 Apr-10 May-10 Jun-10 Jul-10 Aug-10 Sep-10 Oct-10 Nov-10

cfs

F2010

His.

39

Figure 4-7: Basin average inflow comparisons

Figure 4-8: Reservoir Initial Storages

Forecasted Basin Total Average Inflows

16962

20448

16754

13528 13981

16159

0

5000

10000

15000

20000

25000

Historical F2006 F2007 F2008 F2009 F2010

cfs

Reservoir Initial Storages On March 1st

2013

3872

2992

463

2020 20191902

3786

2997

594

20021895

1490

2660

1456

375

1532

1774

1045

2014

1385

432

1211

831

1178

2014.5

1397

426

1278

1449

0

500

1000

1500

2000

2500

3000

3500

4000

4500

Trinity Shasta Oroville Folsom New Melones San Lius

Stor

age

(100

0 AF

)

Y2006Y2007Y2008Y2009F2010

40

Figure 4-9: Sample Tradeoff Plot 1;

Figure 4-10: Sample Tradeoff Plot 2;

Total Water Delivery vs. Carryover Storage Tradeoff

85638059

7493

6881

6263

89918460

7968

73416786

4000

5000

6000

7000

8000

9000

10000

2500 3000 3500 4000 4500 5000

Water Delivery (TAF)

Term

inal

Sto

rage

(100

0 A

F)

Y2010

Historical Mean

Total Water Delivery vs. System Energy Generation Tradeoff

4438

4689

4957

5262

5521

4555

4809

5079

5349

5620

3000

3500

4000

4500

5000

5500

6000

2500 3000 3500 4000 4500 5000

Water Delivery (TAF)

Ener

gy (G

WH

)

Y2010

Historical Mean

41

Figure 4-11: Reservoir Elevation Sequences

42

Figure 4-12: Reservoir Release Sequences

43

Figure 4-13: Reservoir Energy Generation Sequences

44

Figure 4-14: X2 Location Sequences

45

Figure 4-15: Delta Outflow Sequences

46

Figure 4-16: Mean Water Delivery Comparisons

Figure 4-17: System Energy Generation Comparisons

Simulated System Mean Water Delivery

6196

5541

20781864

3728

0

1000

2000

3000

4000

5000

6000

7000

Y2006 Y2007 Y2008 Y2009 Y2010

TAF

Simulated System Mean Energy Generation

8670

7573

4400 4152

4956

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Y2006 Y2007 Y2008 Y2009 Y2010

GW

H

47

Figure 4-18: Mid Range Reservoir Elevation Sequences

48

Figure 4-19: Mid Range Reservoir Release Sequences

49

Figure 4-20: Mid Range Energy Generation Sequences

50

Page Left Blank

51

CHAPTER 5: DISCUSSION AND CONCLUDING REMARKS

5.1 INTRODUCTION

This chapter describes the lessons learned from this research effort that aims to transition the end-

to-end INFORM (Integrated Forecast and Reservoir Management) system from the demonstration

phase to operations. It also presents recommendations that appear beneficial for the near future

along the lines of this work. The assessments are based solely on the authors experience during the

activities of the project documented in this report and they do not necessarily reflect the beliefs of

the participant agencies.

It is important to mention at the outset that INFORM provides science-based support for reservoir

managers of large multi-objective reservoir systems. As such, it targets skilled technical (engineering

and science) as well as management (planning) personnel, rather than the general public.

Furthermore, it is designed to support decisions that involve several stakeholders and decision

makers (some with conflicting objectives). The following agencies participate in the INFORM

Oversight and Implementation Committee: California Department of Water resources, California

Energy Commission, NOAA National Weather Service, NOAA Climate Program Office, U.S.

Bureau of Reclamation Central Valley Operations, and U.S. Army Corps of Engineers. The

reservoir system targeted by INFORM provides the majority of water to the California Bay Delta

region and as such it provides two-thirds of the drinking water and 7 million acres worth of

cropland irrigation for California, contributes to the hydroelectric generation of 15% of the State’s

electricity, and regulates the flow, temperature and salinity of aquatic environments to safeguard the

habitats of numerous fish, birds, mammals, and plant species.

5.2 SIX PROJECT LESSONS

Transition to operations of the end-to-end INFORM (Integrated Forecast and Reservoir

Management) system involved steps that begin with the design of the system to conform to

operational norms in close collaboration with relevant agencies, include workshops with operational

forecast and management agencies for reciprocal technology transfer with the INFORM system

52

developers, and provide relevant operational system assessments on key dates in the decision cycle

associated with water management in Northern California.

The INFORM design uses modular forms for data ingest and models so that operational models

and procedures may be used and easily compared to more advanced models for the benefit of

operational forecast and management agencies. The project experience is that this is a fundamental

requirement of any system that aims to contribute information for forecast and management

operations in a sustainable manner. This approach to design allows an incremental transition to

more advanced models starting from models with which the operational agency personnel is familiar

and with products that they have been trained to use. In addition, it helps validate synthesis

products that use operational models as components together with other more advanced component

models. Lastly, after the transition to operations is complete, it generates systems that are

sustainable by the operational agency personnel without significant continuing involvement of the

developers. For the INFORM system, operational data ingested are the ensemble predictions of the

National Center of Environmental Predictions (NCEP) Global Forecast System (GFS) and Climate

Forecast System (CFS), the operational hydrologic models and update procedures of the California

Nevada River Forecast Center (CNRFC) for snow and soil water continuous accounting, and the

operational spreadsheets for reservoir management of the Bureau of Reclamation and the California

Department of Water Resources.

A second important lesson is that, prior to beginning the transition process, operational agencies

wish to see non-trivial comparisons of the current operational system output to corresponding

output from the new system to consider supporting transitions to operations. For the INFORM

system Carpenter and Georgakakos 2001, Yao and Georgakakos 2001, HRC-GWRI 2007 provide such

comparisons with existing procedures and show in a conclusive manner that it has significant skill in

most cases. As a result California Agencies financially supported the transition with a ratio of

funding of approximately 3 (California Agencies) to 1 (NOAA).

A third lesson relates to the role that the output of models, systems and procedures has in decision

making pertaining to reservoir management. That is, this output provides advisory information for

decision makers, while decisions may also rely on additional information and subjective judgment, as

well as non-quantifiable objectives. Thus, the more the systems to be transitioned facilitate this

advisory role for their products and are amenable to use with additional information, the more

53

suited they are perceived to be for transition to operations. The INFORM system produces risk-

based non-inferior trade-offs for use by decision makers with additional information to arrive at

decisions regarding reservoir releases.

A fourth lesson is that decision makers assign importance to the existence of uncertainty measures

associated with the products of the system that is to be transitioned to operations. Both operational

forecast and management agencies are concerned with uncertainty in recent years when gradual

changes are seen in climatic variables (e.g., earlier snowmelt in the Sierra Nevada and thus more

variable flows to be accommodated by the California reservoirs). Ensemble predictions of inflows

and other hydrologic variables are important as are commensurate risk-based decision component

outputs. Reliability of such uncertainty measures is important and validations of such reliability are

important prerequisites of a transition to operations. For instance, management agencies required

that prior to using them in operations the ensemble reservoir inflows were based and were

sanctioned by the operational forecast agencies. The INFORM system design was modified to allow

alignment of the INFORM hydrologic model states to those of the California Nevada River

Forecast Center operational model states. Also, California Agencies sought and obtained external

independent reviews of the INFORM system effectiveness and utility for California operations.

Such reviews were favorable for the INFORM project. Lastly, the effects of upstream regulation to

downstream ensemble streamflow predictions were identified as important by the operational

forecast agencies. Changes in the INFORM structure are necessary (not yet made) to enhance the

reliability of the ensemble reservoir inflow forecasts during the dry California summer and fall.

The fifth lesson has to do with the variable of the response of the operational forecast and

management agencies to this transition to operations effort. Although INFORM is an integrated

system with both forecast and reservoir management components, individual agencies were

interested in specific components of the system because they believe that the other components

have existing operational counterparts that are producing satisfactory results. It is not clear if this is

an individual’s preference or agency policy. Nevertheless, it suggests that large integrated systems

will meet favorable agency response if they are modular not only in design but also in the products

generated, and furthermore even in the demonstration phase if they make these products available to

the interested agencies in formats that are easy to digest and conform to the agency product types.

After substantial agency discussion that involved the Oversight and Implementation Committee, the

INFORM system was modified to generate both mean areal precipitation and reservoir inflow

54

ensemble products, and several products from the decision component pertaining to agency defined

management objectives.

The sixth lesson reinforces the value of frequent workshops to producing usable and sustainable

systems for operations. Throughout the project such meetings led to several changes in the initial

system and system-product configuration and lent credence to the initial feasibility studies and

system skill and utility. They also led to significant science cooperation and technology transfer

pertaining to several topics of interest both for the agencies participating and for the developers.

For example, this cooperative effort led to developing a common basis for system assessments,

identifying the current basis of multi-agency decision making in California, isolating a significant part

of a very large and complex California water management system which could be modeled by

INFORM for operational support, considering the process from ensemble reservoir inflow

predictions to risk-based trade-offs, identifying uncertainty pathways within INFORM, defining

system retrospective studies for validation, contributing to real time product validation, and others.

5.3 TRANSITION TO OPERATIONS PROJECT CHALLENGES

There were several challenges that the authors faced during the efforts to transition INFORM to

operations. Accommodating the suggestions (presented in the previous section in the form of

lessons) was a significant but necessary challenge for successful progress toward the operational use

of INFORM products. However, there was a significant challenge that the project faced which was

only partly accommodated, mainly with external resources.

The main challenge pertains to on-going changes to operational systems and products by both

operational forecast and management agencies, which made the incorporation of operational

components within the INFORM system to be transitioned (see previous section for the need) a

moving target. This is particularly so for the changes in the National Centers of Environmental

Prediction CFS products and the changes of the California Nevada River Forecast Center model

configuration (number and configuration of modeled basins and upstream regulation effects to

downstream ensemble streamflow predictions) and, most importantly, forecast system architecture

(from the National Weather Service River Forecast System to the Common Hydrologic Prediction

System (CHPS)). Such changes, although necessary for improving the products produced by these

55

agencies, were not foreseen at the time of the development of the project and required substantial

changes in INFORM system design and software architecture (as well as training of the developers

in the new systems), accruing significant additional cost. In addition, they prolonged the transition

process substantially beyond the developers control as external agency timetables are not aligned

with the transition timetable set for this project. The option of freezing the system to a particular

input product or software architecture effectively stops the transition to operations process, and it is

thus undesirable. INFORM developers sought additional funding to accommodate these changes

and were able to cover within the project timetable a part of the needed changes as documented in

this report (Chapter 2). However, additional changes are necessary and these are presented in the

recommendations section below.

5.4 RECOMMENDATIONS

Perhaps the most important recommendation to NOAA Climate Program Office is to continue

supporting activities of agency communication and training pertaining to the profitable use of the

INFORM ensemble forecasts and water management assessments for Northern California. The

time and resources invested are bearing fruit (this is based on third party independent reviews

solicited by the California Energy Commission) and they will help improve management in a

changing climate and demand environment in California. The authors believe that the lessons

learned in the INFORM Project transition-to-operations process highlight the fact that some of

impediments to the application of probabilistic weather and climate forecast information to end-user

systems are quite general. Thus the larger community dealing with these issues would benefit from

INFORM's near-unique experience of having produced an end-to-end system suitable for

operations and the many layers of detail (too often ignored) underlying the idea of using weather and

climate forecast information in real, large and managed end-user systems. These benefits could be

realized by having NOAA recognize and publicize the achievements and lessons learned in the

INFORM project and its transition-to-operations.

The second recommendation regards the needed changes to INFORM to complete the transition to

operations phase. This pertains to the ingest of the new CFS ensemble predictions (complete

variable fields of three-dimensional ensemble prediction suitable for dynamic downscaling) so that

56

the INFORM forecast component can take advantage of these fields for more reliable predictions in

the 0 to 30 day range for Northern California. It also pertains to the changes required to

accommodate the new CHPS architecture of the National Weather Service operational forecast

system. This work is now being supported by Phase II of the INFORM project sponsored by the

California Energy Commission (2009-2012), while requisite training is supported by the Technology

Transfer Program of the Hydrologic Research Center.

A third recommendation concerns the transition to operations of a sustainable and useful procedure

for accounting for upstream regulation effects on ensemble reservoir inflow predictions. The Office

of Hydrologic Development of the National Weather Service has supported the necessary research

and demonstration phase, and the transition to operations of the resultant methodologies remains to

be implemented as a new transition to operations project. It is recommended that this be supported

as needed to enhance credibility of NWS predictions mainly during the dry periods of the year

(summer and fall in Northern California).

57

CHAPTER 6: REFERENCES

Carpenter, T.M., and K.P. Georgakakos, 2001: Assessment of Folsom Lake Response to Historical

and Potential Future Climate Scenarios, 1, Forecasting. Journal of Hydrology, 249,148-175.

Georgakakos, K.P., Graham, N.E., Carpenter, T.M., Georgakakos, A.P., and Yao, H., 2005:

Forecasts and multiobjective reservoir management in Northern California. EOS 86(12), 122, 127.

Georgakakos, K.P., Graham, N.E., Carpenter, T.M., Georgakakos, A.P., Yao, H., and Kistenmacher,

M., 2008. Operational Multiscale Forecast and Reservoir Management in Northern California, First Year Progress

Report. HRC TN 34, Hydrologic Research Center, San Diego, CA, 18pp.

HRC-GWRI, July 2006. Integrated Forecast and Reservoir Management for Northern California:

System Development and Initial Demonstration. California Energy Commission, PIER Energy-

Related Environmental Research (Aquatic Resources). CEC-500-02-008, 244pp. (Available on

line: http://www.energy.ca.gov/pier/final_project_reports/CEC-500-2006-109.html)

HRC-GWRI, 2007. Integrated Forecast and Reservoir Management (INFORM) Demonstration Project, Winter

’06 – ’07 Operations. HRC TN 29, Hydrologic Research Center, San Diego, CA, 37 pp.

Sperfslage, J. A. and K. P. Georgakakos, 1996: Operational Implementation of the Hydrologic Forecast System

(HFS) Operation as part of the National Weather Service River Forecast System (NWSRFS). HRC Technical

Report No. 12 Hydrologic Research Center, San Diego, California 213pp.

Yao, H, and A. Georgakakos, "Assessment of Folsom Lake Response to Historical and Potential

Future Climate Scenarios," Journal of Hydrology, 249, 176-196, 2001

58

Page Left Blank

59

APPENDIX A: PRELIMINARY ASSESSMENT OF THE INFORM

PHASE I FORECAST-DECISION SYSTEM RESULTS VS. OBSERVED DATA FOR THE 2006, 2007, AND 2008 SEASONS

A.1 INTRODUCTION AND OVERVIEW

The purpose of this analysis is to assess the value of the INFORM DSS forecast-decision system

(http://www.energy.ca.gov/pier/project_reports/CEC-500-2006-109.html) in the planning and

management of the northern California reservoir system (Figure 1-2).

This should only be viewed as a preliminary and tentative assessment because (a) the INFORM

project is at its first of two planned development and demonstration phases and not all system

components have been completed, (b) a good part of the necessary actual data are unavailable at the

time of this writing, and (c) the time and resources for this effort are inadequate for a

comprehensive evaluation. It is stressed that a major objective of the second INFORM phase

(scheduled to begin in January, 2009) is to perform such a comparison and assess the INFORM

system benefits in pragmatic detail. To be meaningful, this comparison should be done after careful

planning of realistic INFORM applications in support of existing decision processes, with evaluation

of results jointly performed by stakeholder agency experts and INFORM developers.

The actual system data used in this assessment has been obtained from the USBR and DWD

websites. USBR data relate to water deliveries, reservoir storages, and energy generation for projects

managed by USBR. DWR data pertain to inflow data for all reservoirs. Energy generation data for

Oroville (state project) is unavailable at the time of this writing. The comparison is based on data

from March to November of 2006 and 2007, and March to October of 2008.

Based on this data availability, the comparison is carried out with respect to inflows at main system

nodes, water deliveries, carry-over storages of the major reservoirs, and energy generation.

60

A.2 OBSERVED DATA

A.2.1 INFLOW DATA AND FORECAST ASSESSMENT

The INFORM DSS requires inflows at the following nodes (Figure 1-2):

• Trinity

• Whiskeytown

• Shasta

• Keswick-Wilkins

• Oroville

• Folsom

• Yuba

• Sacramento Miscellaneous

• Eastside Streams

• Delta Miscellaneous Creeks

• New Melones

• San Joaquin River (SJR)

However, inflow forecasting models have been developed during the first INFORM phase for the

following locations only:

• Trinity

• Shasta

• Oroville

61

• Folsom

• Yuba

Forecasting models for all other nodes are scheduled for development during the second project

phase. Thus, in the INFORM forecast-decision runs to be presented, at the locations where models

are not yet available, the forecasts comprise historical seasonal means. This is expected to introduce

discrepancies which can be significant as the modeled tributary inflows on average comprise about

58% of the total. Figure A-2.1 displays the average historical natural inflows at different system

nodes, highlighting the modeled nodes with solid bars.

Figure A-2.1: Historical Average Inflows (Solid Bars Correspond to Modeled Nodes)

The observed inflows for the assessment seasons are posted on the California Data Exchange

Center website at http://www.usbr.gov/mp/cvo/deliv.html. The original data are in daily time

step. Monthly average sequences and other statistics are derived using daily data.

Historical Average Inflows (1968-1996)

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Trin

ity

Whi

skey

tow

n

Sha

sta

Kes

wic

k-W

ilkin

s

Oro

ville

Fols

om

Yub

a

Sac

ram

ent M

isc

Eas

tsid

eS

tream

s

Del

ta M

isc

Cre

eks

New

Mel

ones

SJR

cfs

62

As part of this assessment, nine month forecasts are issued for the five above-mentioned nodes

(Trinity, Shasta, Oroville, Folsom, and Yuba) with a starting date of March 1, 2006, 2007, and 2008.

These forecasts are not updated in the subsequent months as would be done in practice, highlighting

the preliminary nature of this assessment. Furthermore, at the present stage of INFORM system

development (end of Phase I), probabilistic rather than dynamic downscaling is performed to

produce basin precipitation and temperature fields from the large scale operational Climate Forecast

Model (CFS) predictions to drive the hydrologic components of INFORM for the basins modeled.

Thus, the generated precipitation and temperature fields reflect the characteristics of the historical

climatology of above and below average events more than the actual large scale forecast field

characteristics from the CFS.

Figures A-2.2., A-2.3, and A-2.4 present the monthly sequences of historical means, forecasted

mean/maximum/minimum, and observed sequences at the four nodes corresponding to the major

system reservoirs (Trinity, Shasta, Oroville, and Folsom).

The March 1, 2006, forecasts indicated an above average water year for Trinity, Shasta, and Folsom,

and an average water year for Oroville (with the exception of the March mean forecast which was

predicted above average). The observed data showed that all locations registered above average

flows. The forecast results are overall reliable in that they contain the actual sequences throughout

the forecast horizon. Forecast reliability is expected to increase in a real time application where

forecasts are routinely updated as the year evolves.

The March 1, 2007, forecasts indicated an above average water year for Trinity and Folsom, average

for Shasta, and below average for Oroville. The observed data showed that all locations received

below average flows. However, again, all observed sequences fell within the forecast ensemble.

Lastly, the March 1, 2008, forecasts indicated an average water year for Trinity, above average for

Folsom, and below average for Shasta and Oroville. The observed data showed that all locations

received below average flows.

System wide, the mean forecasts suggested a wet year for 2006, average year for 2007, and a dry year

for 2008. With the exception of 2007, the actual observations were consistent with model

predictions. Furthermore, all observed inflows were within the forecasted ranges.

63

Figure A-2.2: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow Sequences, 2006

2006 Forecasts; Trinity

0

2000

4000

6000

8000

10000

12000

Mar-06 Apr-06 May-06

Jun-06 Jul-06 Aug-06

Sep-06 Oct-06 Nov-06

cfs

Max.

Avg

Min

Obs.

His.

2006 Forecasts; Shasta

0

5000

10000

15000

20000

25000

30000

35000

40000

Mar-06 Apr-06 May-06

Jun-06 Jul-06 Aug-06

Sep-06 Oct-06 Nov-06

cfs

Max.

Avg

Min

Obs.

His.

2006 Forecasts; Oroville

0

5000

10000

15000

20000

25000

30000

35000

Mar-06 Apr-06 May-06 Jun-06 Jul-06 Aug-06 Sep-06 Oct-06 Nov-06

cfs

Max.

Avg

Min

Obs.

His.

2006 Forecasts; Folsom

0

5000

10000

15000

20000

25000

Mar-06 Apr-06 May-06 Jun-06 Jul-06 Aug-06 Sep-06 Oct-06 Nov-06

cfs

Max.

Avg

Min

Obs.

His.

64

Figure A-2.3: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow Sequences, 2007

2007 Forecasts; Trinity

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07

cfs

Max.

Avg

Min

Obs.

His.

2007 Forecasts; Shasta

0

5000

10000

15000

20000

25000

30000

35000

40000

Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07

cfs

Max.

Avg

Min

Obs.

His.

2007 Forecsats; Oroville

0

5000

10000

15000

20000

25000

Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07

cfs

Max.

Avg

Min

Obs.

His.

2007 Folsom

0

5000

10000

15000

20000

25000

Mar-07 Apr-07 May-07 Jun-07 Jul-07 Aug-07 Sep-07 Oct-07 Nov-07cf

s

Max.

Avg

Min

Obs.

His.

65

Figure A-2.4: Monthly Forecasted (Mean, Maximum, and Minimum) and Observed Inflow Sequences; 2008

A.2.2 ACTUAL WATER DELIVERIES

System water diversions represent the total amount water taken out of the system. The observed

monthly data were taken from the USBR website ( http://www.usbr.gov/mp/cvo/deliv.html).

Table 21 of the website (Central Valley Project Diversions) contains the monthly diversions in acre-

feet for all projects operated by USBR. The following nodes are related to the INFORM DSS

system:

• Contra Costa

• Delta-Mendota

• Actual Fed Dos Amigos

66

• Madera

• Friant-Kern

• Corning

• Folsom-South

• Tehama-Colusa

• Thermalito Forebay

Historical water withdrawals for Thermalito Afterbay are provided by the Department of Water

Resources (DWR) of California.

The total system diversions are taken to be equal to the sum of diversions from all above nodes.

Thus, the actual monthly system water diversions (taken to represent deliveries) from March to

November for 2006 and 2007, and March to October for 2008 are shown in Figure A-2.5. The

corresponding total volumes are plotted in Figure A-2.6.

The water delivery shows a clear seasonal pattern. It starts to increase in May, peaks in July, and

gradually reduces to the lowest level in November. The total actual deliveries in 2006 were 5810

TAF (thousand acre-feet). In 2007, this amount was reduced to 4695 TAF, and in 2008 even further

to 3368 TAF. Thus, actual deliveries were reduced by almost 50% from 2006 to 2008.

67

Figure A-2.5: System Monthly Water Diversion Sequences

Figure A-2.6: Observed Total Diversions

Observed Monthly System Diversions

0

200

400

600

800

1000

1200

Mar Apr May Jun Jul Aug Sep Oct Nov

AF200620072008

Total Diversions from 2006 to 2008

0

1000

2000

3000

4000

5000

6000

7000

2006 2007 2008

AF

68

2.3 ACTUAL CARRY-OVER STORAGES

Actual reservoir storage sequences are obtained from the USBR website at

http://www.usbr.gov/mp/cvo/deliv.html. For comparison purposes, this analysis pertains to the

five major reservoirs: Trinity, Shasta, Oroville, Folsom, and New Melones. The initial time refers to

the beginning of March, and the terminal time refers to the end of November for 2006 and 2007,

and the end of October for 2008. The total system storage is the sum of the storages of all major

reservoirs. The observed data are plotted in Figure A-2.7. A clear downward trend is observed for

the terminal storages with a 50% reduction from 2006 to 2008.

Figure A-2.7: Observed Total System Initial and Terminal Storage

A.2.4 ACTUAL ENERGY GENERATION

Actual monthly energy generation are posted on the USBR website at

http://www.usbr.gov/mp/cvo/deliv.html For comparison purposes, this analysis includes

generation amounts only from five major reservoirs: Trinity, Oroville, Shasta, Folsom, and New

Melones. Data for Oroville is provided by DRW. The total generation values from March to

System Carryover Storages of Major Reservoirs

0

2000

4000

6000

8000

10000

12000

2006 2007 2008

TAF

Initial

Carryover

69

November for 2006 and 2007, and March to October for 2008 are shown in Figure A-2.8. This

figure shows a 33% reduction in actual energy generation from 2006 (6759 GWH) to 2007 (4479

GWH), and a more than 50% reduction from 2006 to 2008 (2975 GWH).

Figure A-2.7: Observed System Energy Generation

A.2.5 INFORM DSS RESULTS

The INFORM DSS uses the forecast ensembles from March to November to generate tradeoffs

among the system objectives such as water deliveries, energy generation, and carryover storage,

given applicable operational constraints. Three system runs are carried out for the years 2006, 2007,

and 2008. The tradeoffs illustrate the many feasible combinations of the above objectives and form

the basis for the comparison presented herein. More specifically, in each run, the tradeoff point that

matches the actual deliveries is identified and the carry-over storage and energy generation amounts

are compared with the actual corresponding quantities. This is a biased comparison against the INFORM

results, which in this analysis are generated by an initial nine-month run without subsequent monthly

Energy Generation from Major Plants

0

1000

2000

3000

4000

5000

6000

7000

8000

2006 2007 2008

GW

H

70

updating as would be its intended practical use. The comparison is depicted in Figures A-2.8 (carry-over

storage) and A-2.9 (energy generation) and is discussed below:

· An above average forecast was issued for 2006. Based on the forecasts and initial reservoir

storage, the INFORM DSS estimated that the system could support up to 7500 TAF of

water deliveries. The tradeoff curve was generated for a water demand range from 3850 to

7700 TAF (Figure A-2.10). The actually met demand of 5810 TAF was close to the first

tradeoff point 5887 TAF. Figures A-2.8 and A-2.9 show that the actual system carry-over

storage and energy generation are within the predicted ranges, with the later practically equal

to the forecasted mean.

· A relatively average to dry forecast ensemble was issued for 2007. Based on the forecasts

and high initial reservoir storage, the INFORM DSS estimated that the system can support

up to 5700 TAF water supply. The actual demand met was 4695 TAF which was close to

the first tradeoff point of 4694 TAF (Figure A-2.11). The actually observed inflows were in

the low forecast range, which resulted in a lower than the mean forecasted carryover storage

(Figure A-2.8). However, this value is still within the predicted range. The same comment

applies for energy generation (Figure A-2.9).

· A dry forecast was issued for 2008. With already reduced initial storage, the INFOM DSS

estimates that the system can only support up to 3000 TAF demand. The tradeoff curve

ranged from 700 to 3800 TAF. The actual water delivery (to October) of 3367 TAF is close

to the fifth tradeoff point (3315 TAF; Figure A-2.12). Figures A-2.8 and A-2.9 show very

good correspondence between average predicted and observed carry-over storage and energy

generation.

71

Figure A-2.8: System Carry-over Storage Comparison

Figure A-2.9: System Energy Generation Comparison

Carryover Storage Comparisons

0

2000

4000

6000

8000

10000

12000

2006 2007 2008

TAF

ObservedDSS AVGDSS MinDSS MAX

Energy Generation from Major Plants

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

2006 2007 2008

GW

H

ObservedDSS AVGDSS MINDSS MAX

72

Figure A-2.10: INFORM Tradeoff: Total Water Delivery vs. System Carryover Storage; 2006

Figure A-2.11: INFORM Tradeoff: Total Water Delivery vs. System Carry-over Storage; 2007

Tradeoff Curve for 2006

0

1000

2000

3000

4000

5000

6000

7000

8000

3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000

Water Delivery (TAF)

Syst

em C

arry

over

Sto

rage

(TA

F)

Tradeoff Curve for 2007

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000

Water Delivery (TAF)

Syst

em C

arry

over

Sto

rage

(TA

F)

73

Figure A-2.12: INFORM Tradeoff: Total Water Delivery vs. System Carry-over Storage; 2008

A.3 CONCLUSIONS

The previous preliminary comparison indicates that the information provided by the INFORM

system in operational planning and management of the northern California river and reservoir

system is relevant, reliable, and decision worthy. This is corroborated by the favorable comparison

of the predicted and actually observed system outputs (water deliveries, carry-over storages, and

energy generation amounts) over the March to November 2006, 2007, and 2008 seasons.

The value of the INFORM forecast-decision system ultimately depends on whether the information

it generates provides management agencies with good appreciation of the benefits and risks

associated with different decisions and helps them adopt those that serve well the interests of the

system stakeholders. In the second project phase, more detailed assessments are planned to quantify

the system response with and without this information.

It is expected that the value of the INFORM system will increase after the second project phase

which aims at completing its development and implementing it operationally. More specifically,

forecast improvements are anticipated as a result of the following planned activities:

Tradeoff Curve for 2008

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

0 500 1000 1500 2000 2500 3000 3500

Water Delivery (TAF)

Syst

em C

arry

over

Sto

rage

(TA

F)

74

· Enlarging the domain of hydro-meteorological modeling to include the entire application

area and Delta tributary flows;

· Enhancing the downscaling of the Climate Forecast System (CFS) model forecasts with a

downscaling procedure as used for the shorter term ensemble operational forecasts from the

Global Forecast System (GFS). Both GFS and CFS are run by NOAA at the National

Centers of Environmental Prediction (NCEP) and are the primary operational

meteorological forecast input to the INFORM forecast-management system; and

· Implementing the forecast component on a cluster computer so that more ensembles from

GFS and CFS may be processed in real time for more accurate evaluation of forecast

uncertainty for short and long lead times.

Likewise, decision system improvements are expected due to the following planned activities:

· Developing and incorporating a system-wide daily resolution model;

· Incorporating a river temperature model for assessing fishery needs;

· Managing system uncertainties in a spatially optimal manner; and

· Working with the stakeholder agencies to integrate the INFORM forecast-decision system

within the existing decision processes and to assess/demonstrate its value for planning and

management.

75

APPENDIX B: OPERATIONAL MULTISCALE FORECAST AND

RESERVOIR MANAGEMENT IN NORTHERN CALIFORNIA, ASSESSMENTS 2009

The contribution of Georgia Water Resources Institute to the NOAA TRACS project is to develop

and demonstrate the utility of Decision Support Systems (DSS) for reservoir management. The DSS

consists of databases, interfaces, and various application programs interlinked to provide meaningful

and comprehensive information to decision makers. During the current NOAA funding period,

GWRI efforts focused on assessing the system performance over the period from March 15, 2009,

to November 15, 2009. Inflow forecasts were provided by HRC starting from March 15, 2009. The

findings are discussed below.

B.1 INTEGRATED INFORM DSS AND FORECASTING MODELS - AN OVERVIEW

The INFORM DSS includes three modeling layers (Figure 1-3) designed to support decisions

pertaining to various temporal scales and objectives. The three modeling layers include (1) turbine

load dispatching (which models each turbine and hydraulic outlet and has hourly resolution over a

horizon of one day), (2) short/mid range reservoir control (which has a daily resolution and a

horizon of one month), and (3) long range reservoir control (which has a monthly resolution and a

horizon of up to one year).

Both the long range control model and the mid/short range control model use inflow forecasts as

inputs. The integration of the decision models and inflow forecasting models are done through data

exchange. The forecasted inflows are saved in a pre-formatted Excel file. The DSS provides easy

tools to read the data in the Excel file and save it into the database. The DSS also provides tools to

plot and validate the forecasting results.

The long range control model is designed to consider long range issues such as whether water

conservation strategies are appropriate for the upcoming year using the provided hydrologic

forecasts. As part of these considerations, the DSS would quantify several tradeoffs of possible

interest to the management agencies and system stakeholders. These include, among others, relative

76

water allocations to water users throughout the system (including ecosystem demands), reservoir

coordination strategies and target levels, water quality constraints, and energy generation targets.

This information would be provided to the forum of management agencies (the planning

departments) to use it as part of their decision process together with other information. After

completing these deliberations, key decisions would be made on monthly water supply contracts,

reservoir releases, energy generation, and reservoir coordination strategies.

The short/mid range control model considers the system operation at finer time scales. The

objectives addressed are more operational than planning and include flood management, water

supply, and power plant scheduling. This model uses hydrologic forecasts with a daily resolution and

can quantify the relative importance of, say, upstream versus downstream flooding risks, energy

generation versus flood control, and other applicable tradeoffs. Such information is again provided

to the forum of management agencies (the operational departments) to use it within their decision

processes to select the most preferable operational policy. Such policies are revised as new

information on reservoir levels and flow forecasts comes in. The model is constrained by the long

range decisions, unless current conditions indicate that a departure is warranted.

The three modeling layers address planning and management decisions. The scenario/policy

assessment model addresses longer term planning issues such as increasing demands, infrastructure

change (water transfers options), potential hydro-climatic changes, and mitigation measures. The

approach taken in this DSS layer is to simulate and inter-compare the system response under various

inflow, demand, development, and management conditions.

Altogether, the INFORM DSS provides a comprehensive modeling framework responsive to the

information needs of the decision making process at all relevant time scales.

B.2 INFORM DSS RESULTS FOR 2009 FORECASTS

The application described here utilizes the following input data:

· Forecasted inflows start from March 15, 2009 (88 traces, 9 month horizon, and five

locations: Clair Engle Lake, Shasta, Oroville, Folsom, and Yuba;

77

· Historical monthly average values are used for locations where forecasted inflows are not

available (Table B-1);

· Monthly reservoir parameters and constraints (max, min, and target storage, evaporation

rates; Table B-2);

· Minimum river flow requirements (Table B-3);

· Base monthly demands at all locations (Table B-4);

· Reservoir initial storages are set to their actual values on March 15, 2009.

The forecasted monthly inflow ensembles are shown in Figures B-1 and B-2. The comparisons

between the forecasted inflow mean and the corresponding historical means for four major

reservoirs are plotted in Figures B-3 through B-6. As shown, the forecasted inflow means at both

Folsom and Trinity are higher than the historical values during high flow months and lower during

low flow months. The forecasted means at both Shasta and Oroville are lower than their historical

means. Figure B-7 shows the forecasted basin total inflow means for 2006 to 2009. The results for

2009 are similar to those for 2008. Overall, the forecasts indicate that 2009 will be approximately

18% drier than the average year. Figure B-10 shows the initial reservoir storages on March 1 for

years from 2006 and 2009. The current year has the lowest initial storage values for all major

reservoirs, which adds to the system stress.

Using the forecasted inflows, tradeoffs are generated by changing the base demands for all locations

with fractions 10% to 50%. The tradeoffs between the total reservoir carryover storages and the

system energy versus the system water deliveries are depicted in Figures B-9 and B-10. As demands

increase, the reservoir carryover storages decrease. Energy generation increases as downstream

demands increase because of higher reservoir releases. The results show that the system can meet

water deliveries up to 2,330 TAF, at the 50% tradeoff point. Meeting demands beyond this level

would result in significant reservoir drawdown (especially at Shasta and Oroville) and diminished

carryover storage.

78

The reservoir sequences and other system outputs corresponding to all tradeoff points are saved in

the INFORM DSS database. Selected reservoir elevation, release, and energy generation sequences

corresponding to 1,860 thousand acre feet (TAF), 40% of base demands are shown in Figures B-11

through B-13. The X2 location sequences are shown in Figure B-14, indicating all traces below 80

km, the maximum constraint set in the study. The X2 location stays within this constraint for all

tradeoff points. The Delta outflow sequences are plotted in Figure B-15.

Figure B-16 compares water deliveries over the 2006-2009 four-year period, assuming the actual

initial storages and the same carryover storage of 7,500 TAF. As shown, the expected water

deliveries are reduced by 70% compared to 2006 and 65% compared to 2007. Similarly, average

system energy generation for 2009 is shown in Figure B-17, showing a decrease of 53% relative to

2006 and 45% relative to 2007. Generally, the impact of reduced inflows on water deliveries and

energy generation is not equally proportional because of the Bay Delta water requirements.

The first month releases determined in the long range planning model are passed on to the mid

range model for generation of daily operation decisions. The daily sequences of elevation, release,

and energy generation of four major reservoirs Trinity, Shasta, Oroville, and Folsom are generated

and plotted in Figures B-18 to B-20. The average daily release values can be used to guide day-to-

day operations.

B.3 SUMMARY

The 2009 assessment portends a dry year similar to 2008. In fact, system performance is expected to

be somewhat worse than 2008 due to the already depleted reservoir storages. Depending on the

amount of carry-over storage target (at the end of 2009), average water deliveries are estimated to be

in the range of 2 to 2.4 million acre feet (MAF), which represents only 35% of the average 2007

amount an 30% of the average 2006. Beyond this water deliveries commitment, the likelihood that

Shasta and Oroville may fully deplete becomes appreciable. Average energy generation over the

March 15 to November 15 period ranges from 4,152 to 4,397 GWH, or approximately 55% of the

2007 generation and 45% of the 2006 generation.

79

In light of the above assessment results, we recommend that 2009 water management policies and

commitments be conservative and revised adaptively as inflow information becomes available.

80

Figure B-1: Long Range Inflow Forecasts

81

Figure B-2: Mid Range Inflow Forecasts

82

Figure B-3: Forecasted Inflow Mean Comparison; Trinity

Figure B-4: Forecasted Inflow Mean Comparison; Shasta

Forecasted Inflow Means - Trinity

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Nov-09

cfs

F2009

His.

Forecasted Inflow Means - Shasta

0

2000

4000

6000

8000

10000

12000

14000

16000

18000

Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Nov-09

cfs

F2009

His.

83

Figure B-5: Forecasted Inflow Mean Comparison; Oroville

Figure B-6: Forecasted Inflow Mean Comparison; Folsom

Forecasted Inflow Means - Oroville

0

2000

4000

6000

8000

10000

12000

14000

Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Nov-09

cfs

F2009

His.

Forecasted Inflow Means - Folsom

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

Mar-09 Apr-09 May-09 Jun-09 Jul-09 Aug-09 Sep-09 Oct-09 Nov-09

cfs

F2009

His.

84

Figure B-7: Basin average inflow comparisons

Figure B-8: Reservoir Initial Storages

Forecasted Basin Total Average Inflows

16962

20448

16754

13528 13981

0

5000

10000

15000

20000

25000

Historical F2006 F2007 F2008 F2009

cfs

Reservoir Initial Storages On March 1st

2013

3872

2992

463

2020 20191902

3786

2997

594

20021895

1490

2660

1456

375

1532

1774

1045

2014

1385

432

1211

831

0

500

1000

1500

2000

2500

3000

3500

4000

4500

Trinity Shasta Oroville Folsom New Melones San Lius

Stor

age

(100

0 AF

)

Y2006Y2007Y2008Y2009

85

Figure B-9: Sample Tradeoff Plot 1;

Figure B-10: Sample Tradeoff Plot 2;

Total Water Delivery vs. Carryover Storage Tradeoff

89938500

79947488

6982

101499643

91368629

8122

4000

5000

6000

7000

8000

9000

10000

11000

0 500 1000 1500 2000 2500

Water Delivery (TAF)

Term

inal

Sto

rage

(100

0 A

F)

Y2009

Historical Mean

Total Water Delivery vs. System Energy Generation Tradeoff

3439

3675

3917

4152

4397

3424

3681

3952

4218

4486

3000

3200

3400

3600

3800

4000

4200

4400

4600

0 500 1000 1500 2000 2500

Water Delivery (TAF)

Ener

gy (G

WH

)

Y2009

Historical Mean

86

Figure B-11: Reservoir Elevation Sequences

87

Figure B-12: Reservoir Release Sequences

88

Figure B-13: Reservoir Energy Generation Sequences

89

Figure B-14: X2 Location Sequences

90

Figure B-15: Delta Outflow Sequences

91

Figure B-16: Mean Water Delivery Comparisons

Figure B-17: System Energy Generation Comparisons

Simulated System Mean Water Delivery

6196

5541

20781864

0

1000

2000

3000

4000

5000

6000

7000

Y2006 Y2007 Y2008 Y2009

TAF

Simulated System Mean Energy Generation

8670

7573

4400 4152

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

Y2006 Y2007 Y2008 Y2009

GW

H

92

Figure B-18: Mid Range Reservoir Elevation Sequences

93

Figure B-19: Mid Range Reservoir Release Sequences

94

Figure B-20: Mid Range Energy Generation Sequences

95

Table B-1: Monthly Average Inflows for Selected Locations (TAF)

Month Whisktown Keswick-Wilkens

Sacrament Misc

Eastside Streams

Delta Misc Creeks New Melones SJR

Jan 8. -211.27 -100. 80.67 25.5 76. 133.Feb 4. -299.69 -220. 60.44 25.5 43. 31.Mar 2. -370.28 -330. 20.72 29. 34. 33.Apr 1. -267.47 -175. 21.89 19. 33. 28.May 1. -117.56 45. 28.71 11.1 31. 33.Jun 2. -125. -15. 33.2 0.8 30. 71.Jul 2. -31.24 121. 30.74 0.9 30. 62.Aug 4. 564.46 981. 21.52 1.2 30. 63.Sep 8. 841.7 1465. 21.52 1.8 30. 78.Oct 12. 1767.58 2482. 40.03 32.3 40. 94.Nov 45. 1021. 1763. 67.33 17.4 70. 103.Dec 16. 74.65 328. 146.34 15.4 110. 126.

96

Table B-2: Reservoir Monthly Parameters

Name Month Smax (TAF) Smin (TAF) Starget (TAF) Evap Rate (feet)

Clair Engle Jan 2287.00 312.63 2287.00 0.17

Clair Engle Feb 2287.00 312.63 2287.00 0.13

Clair Engle Mar 2287.00 312.63 2287.00 0.20

Clair Engle Apr 2287.00 312.63 2287.00 0.39

Clair Engle May 2287.00 312.63 2287.00 0.51

Clair Engle Jun 2287.00 312.63 2287.00 0.58

Clair Engle Jul 2287.00 312.63 2287.00 0.76

Clair Engle Aug 2287.00 312.63 2287.00 0.71

Clair Engle Sep 2287.00 312.63 2287.00 0.60

Clair Engle Oct 2287.00 312.63 2287.00 0.30

Clair Engle Nov 2287.00 312.63 2287.00 0.15

Clair Engle Dec 2287.00 312.63 2287.00 0.09

WhiskeyTown Jan 237.90 200.00 205.70 0.17

WhiskeyTown Feb 237.90 200.00 205.70 0.13

WhiskeyTown Mar 237.90 200.00 205.70 0.20

WhiskeyTown Apr 237.90 200.00 237.90 0.39

WhiskeyTown May 237.90 200.00 237.90 0.51

WhiskeyTown Jun 237.90 200.00 237.90 0.58

WhiskeyTown Jul 237.90 200.00 237.90 0.76

WhiskeyTown Aug 237.90 200.00 237.90 0.71

WhiskeyTown Sep 237.90 200.00 238.00 0.60

WhiskeyTown Oct 237.90 200.00 230.00 0.30

WhiskeyTown Nov 237.90 200.00 205.70 0.15

WhiskeyTown Dec 237.90 200.00 205.70 0.09

Shasta Jan 4552 1168 4552 0.17

Shasta Feb 4552 1168 4552 0.13

97

Shasta Mar 4552 1168 4552 0.20

Shasta Apr 4552 1168 4552 0.39

Shasta May 4552 1168 4552 0.51

Shasta Jun 4552 1168 4552 0.58

Shasta Jul 4552 1168 3882 0.76

Shasta Aug 4552 1168 3252 0.71

Shasta Sep 4552 1168 3252 0.60

Shasta Oct 4552 1168 3872 0.30

Shasta Nov 4552 1168 4252 0.15

Shasta Dec 4552 1168 4552 0.09

Oroville Jan 3538 855 3458 0.17

Oroville Feb 3538 855 3538 0.13

Oroville Mar 3538 855 3538 0.20

Oroville Apr 3538 855 3538 0.39

Oroville May 3538 855 3538 0.51

Oroville Jun 3538 855 3343 0.58

Oroville Jul 3538 855 3163 0.76

Oroville Aug 3538 855 3163 0.71

Oroville Sep 3538 855 3163 0.60

Oroville Oct 3538 855 3163 0.30

Oroville Nov 3538 855 3163 0.15

Oroville Dec 3538 855 3163 0.09

Folsom Jan 975 83 805 0.17

Folsom Feb 975 83 975 0.13

Folsom Mar 975 83 975 0.20

Folsom Apr 975 83 975 0.39

Folsom May 975 83 975 0.51

Folsom Jun 975 83 975 0.58

Folsom Jul 975 83 700 0.76

98

Folsom Aug 975 83 575 0.71

Folsom Sep 975 83 575 0.60

Folsom Oct 975 83 575 0.30

Folsom Nov 975 83 575 0.15

Folsom Dec 975 83 675 0.09

New Melones Jan 2420 273 2230 0.17

New Melones Feb 2420 273 2420 0.13

New Melones Mar 2420 273 2420 0.20

New Melones Apr 2420 273 2420 0.39

New Melones May 2420 273 2420 0.51

New Melones Jun 2420 273 2270 0.58

New Melones Jul 2420 273 1970 0.76

New Melones Aug 2420 273 1970 0.71

New Melones Sep 2420 273 1970 0.60

New Melones Oct 2420 273 1970 0.30

New Melones Nov 2420 273 1970 0.15

New Melones Dec 2420 273 2040 0.09

Tulloch Jan 67 57 57 0.00

Tulloch Feb 67 57 57 0.00

Tulloch Mar 67 57 58 0.00

Tulloch Apr 67 57 60 0.00

Tulloch May 67 57 67 0.00

Tulloch Jun 67 57 67 0.00

Tulloch Jul 67 57 67 0.00

Tulloch Aug 67 57 67 0.00

Tulloch Sep 67 57 62 0.00

Tulloch Oct 67 57 57 0.00

Tulloch Nov 67 57 57 0.00

Tulloch Dec 67 57 57 0.00

99

San Luis Jan 2027 450.00 1000.00 0.17

San Luis Feb 2027 631.60 1464.02 0.13

San Luis Mar 2027 748.10 1806.84 0.20

San Luis Apr 2027 835.60 1975.02 0.39

San Luis May 2027 879.92 1976.43 0.51

San Luis Jun 2027 694.72 1546.00 0.58

San Luis Jul 2027 442.12 1062.95 0.76

San Luis Aug 2027 181.12 642.62 0.71

San Luis Sep 2027 9.72 352.64 0.60

San Luis Oct 2027 8.32 312.90 0.30

San Luis Nov 2027 115.02 354.13 0.15

San Luis Dec 2027 286.72 514.21 0.09

Table B-3: Monthly Minimum and Target River Flows

Name Month Rmin (cfs) Rtarget (cfs)

Lewiston Jan 300 300

Lewiston Feb 300 300

Lewiston Mar 300 300

Lewiston Apr 300 300

Lewiston May 3939 300

Lewiston Jun 2507 783

Lewiston Jul 1102 450

Lewiston Aug 450 450

Lewiston Sep 450 450

Lewiston Oct 373 0

Lewiston Nov 300 300

Lewiston Dec 300 300

100

Clear Creek Jan 150 150

Clear Creek Feb 200 200

Clear Creek Mar 200 200

Clear Creek Apr 200 200

Clear Creek May 200 200

Clear Creek Jun 200 200

Clear Creek Jul 200 200

Clear Creek Aug 200 200

Clear Creek Sep 200 200

Clear Creek Oct 200 200

Clear Creek Nov 90 90

Clear Creek Dec 90 90

Spring Creek Jan 325 325

Spring Creek Feb 306 306

Spring Creek Mar 2749 2749

Spring Creek Apr 252 252

Spring Creek May 813 813

Spring Creek Jun 1681 1681

Spring Creek Jul 2602 2602

Spring Creek Aug 2114 2114

Spring Creek Sep 2017 2017

Spring Creek Oct 1138 1138

Spring Creek Nov 504 504

Spring Creek Dec 244 244

Keswick Jan 3250 3250

Keswick Feb 3250 3250

Keswick Mar 3250 3250

Keswick Apr 8000 8000

Keswick May 9600 9600

101

Keswick Jun 11000 11000

Keswick Jul 14500 14500

Keswick Aug 12000 12000

Keswick Sep 5500 5500

Keswick Oct 7200 7200

Keswick Nov 5700 5700

Keswick Dec 3250 3250

Wilkins Jan 0 0

Wilkins Feb 0 0

Wilkins Mar 0 0

Wilkins Apr 5000 5000

Wilkins May 5000 5000

Wilkins Jun 5000 5000

Wilkins Jul 5000 5000

Wilkins Aug 5000 5000

Wilkins Sep 5000 5000

Wilkins Oct 5000 5000

Wilkins Nov 0 0

Wilkins Dec 0 0

FeatherBelowThermalito Jan 1250 0

FeatherBelowThermalito Feb 1250 0

FeatherBelowThermalito Mar 1250 0

FeatherBelowThermalito Apr 1250 0

FeatherBelowThermalito May 2030 0

FeatherBelowThermalito Jun 0 2706

FeatherBelowThermalito Jul 0 5692

FeatherBelowThermalito Aug 5040 5156

FeatherBelowThermalito Sep 0 4386

FeatherBelowThermalito Oct 1980 2683

102

FeatherBelowThermalito Nov 1750 1815

FeatherBelowThermalito Dec 1250 0

AmericanRiverbelowNimbus Jan 800 0

AmericanRiverbelowNimbus Feb 800 0

AmericanRiverbelowNimbus Mar 1000 0

AmericanRiverbelowNimbus Apr 1500 0

AmericanRiverbelowNimbus May 2300 0

AmericanRiverbelowNimbus Jun 1800 0

AmericanRiverbelowNimbus Jul 0 0

AmericanRiverbelowNimbus Aug 0 0

AmericanRiverbelowNimbus Sep 0 0

AmericanRiverbelowNimbus Oct 0 0

AmericanRiverbelowNimbus Nov 1000 0

AmericanRiverbelowNimbus Dec 800 0

Goodwin Jan 175 175

Goodwin Feb 150 150

Goodwin Mar 268 268

Goodwin Apr 760 760

Goodwin May 800 800

Goodwin Jun 561 561

Goodwin Jul 396 396

Goodwin Aug 352 352

Goodwin Sep 240 240

Goodwin Oct 200 200

Goodwin Nov 200 200

Goodwin Dec 200 200

DeltaExit Jan 6001 6001

DeltaExit Feb 11398 11398

DeltaExit Mar 11401 11401

103

DeltaExit Apr 7848 7848

DeltaExit May 9319 9319

DeltaExit Jun 7092 7092

DeltaExit Jul 6505 6505

DeltaExit Aug 4261 4261

DeltaExit Sep 3008 3008

DeltaExit Oct 4001 4001

DeltaExit Nov 4655 4655

DeltaExit Dec 4505 4505

Table B-4: Monthly Demands

Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov DecThermolito 35 0 11 67 189 178 200 178 78 95 104 71

Folsom Pumping 4 4 4 7 8 12 13 12 10 7 5 4Folsom South Canal 1 1 1 1 2 3 4 4 3 2 1 1

OID/SSJID 0 0 14 60 90 90 95 95 74 14 0 0CVP Contractors 0 0 0 0 0 0 0 0 0 0 0 0

CCWD 14 17 18 18 14 14 13 13 13 10 11 13Barker Slough 2 2 1 2 4 5 7 7 6 5 3 3

Federal Tracy PP 258 233 258 250 135 169 270 268 260 258 250 258Federal Banks On-Peak 0 0 0 0 0 0 28 28 28 0 0 0

State Banks PP 390 355 241 68 108 125 271 278 238 175 193 390State Tracy PP 0 0 0 0 0 0 0 0 0 0 0 0

Delta Mendota Canal 30 60 100 120 190 220 270 240 180 110 40 30Federal Dos Amigos 40 50 60 70 110 180 238 178 68 30 30 30Federal O'Neil to Dos

Amigos 0 1 1 1 1 2 2 1 0 0 0 0San Felipe 6 6 10 15 19 20 21 20 13 11 8 8

South Bay/San Jose 2 2 2 5 5 7 7 8 7 12 8 6State Dos Amigos 105 127 158 105 348 348 423 388 269 229 196 61

Delta Consumptive Use -56 -37 -10 63 121 191 268 252 174 118 55 2

Freeport Treatment Plant 14 13 14 12 12 12 12 13 12 12 12 13

104

Table B-5: Initial Reservoir Storages on March 15, 2009

Reservoirs Max. Storage Min. Storage Initial Storage Ini Act. Sto. Fraction (%)Clair Engle Lake 2287 312.63 1123.2 41.05

Whiskeytown 237.9 200 212.07 31.85Shasta 4552 1168 2529 40.22Oroville 3538 855 1739.5 32.97Folsom 975 83 647.3 63.26

New Melones 2420 273 1278 46.81Tulloch 67 57 55 -20.00

San Luis 2027 0 931 45.93