Precipitation-Runoff Processes in the Feather River Basin, Northeastern California, and Streamflow Predictability, Water Years 1971-97

By Kathryn M. Koczot, Anne E. Jeton, Bruce McGurk, and Michael D. Dettinger



Scientific Investigations Report 2004-5202

Sacramento, California 2005

In cooperation with the
California Department of Water Resources

Complete accessible text of report (5.35 MB PDF)

Complete accessible text of report cover (179 KB PDF)

To view PDF documents, you must have the Adobe Acrobat Reader (free from Adobe Systems) installed on your computer.
(download free copy of Acrobat Reader).


     Precipitation-runoff processes in the Feather River Basin of northern California determine short- and long-term streamflow variations that are of considerable local, State, and Federal concern. The river is an important source of water and power for the region. The basin forms the headwaters of the California State Water Project. Lake Oroville, at the outlet of the basin, plays an important role in flood management, water quality, and the health of fisheries as far downstream as the Sacramento-San Joaquin Delta. Existing models of the river simulate streamflow in hourly, daily, weekly, and seasonal time steps, but cannot adequately describe responses to climate and land-use variations in the basin. New spatially detailed precipitation-runoff models of the basin have been developed to simulate responses to climate and land-use variations at a higher spatial resolution than was available previously. This report characterizes daily rainfall, snowpack evolution, runoff, water and energy balances, and streamflow variations from, and within, the basin above Lake Oroville. The new model's ability to predict streamflow is assessed.

    The Feather River Basin sits astride geologic, topographic, and climatic divides that establish a hydrologic character that is relatively unusual among the basins of the Sierra Nevada. It straddles a north-south geologic transition in the Sierra Nevada between the granitic bedrock that underlies and forms most of the central and southern Sierra Nevada and volcanic bedrock that underlies the northernmost parts of the range (and basin). Because volcanic bedrock generally is more permeable than granitic, the northern, volcanic parts of the basin contribute larger fractions of ground-water flow to streams than do the southern, granitic parts of the basin. The Sierra Nevada topographic divide forms a high altitude ridgeline running northwest to southeast through the middle of the basin. The topography east of this ridgeline is more like the rain-shadowed basins of the northeastern Sierra Nevada than the uplands of most western Sierra Nevada river basins. The climate is mediterranean, with most of the annual precipitation occurring in winter. Because the basin includes large areas that are near the average snowline, rainfall and rain-snow mixtures are common during winter storms. Consequently, the overall timing and rates of runoff from the basin are highly sensitive to winter temperature fluctuations.

    The models were developed to simulate runoff-generating processes in eight drainages of the Feather River Basin. Together, these models simulate streamflow from 98 percent of the basin above Lake Oroville. The models simulate daily water and heat balances, snowpack evolution and snowmelt, evaporation and transpiration, subsurface water storage and outflows, and streamflow to key streamflow gage sites. The drainages are modeled as 324 hydrologic-response units, each of which is assumed homogeneous in physical characteristics and response to precipitation and runoff. The models were calibrated with emphasis on reproducing monthly streamflow rates, and model simulations were compared to the total natural inflows into Lake Oroville as reconstructed by the California Department of Water Resources for April-July snowmelt seasons from 1971 to 1997. The models are most sensitive to input values and patterns of precipitation and soil characteristics. The input precipitation values were allowed to vary on a daily basis to reflect available observations by making daily transformations to an existing map of long-term mean monthly precipitation rates that account for altitude and rain-shadow effects.

    The models effectively simulate streamflow into Lake Oroville during water years (October through September) 1971-97, which is demonstrated in hydrographs and statistical results presented in this report. The Butt Creek model yields the most accurate historical April-July simulations, whereas the West Branch model yields the least accurate simulations. Accuracy may reflect the quality of the streamflow measurements (or reconstructions) used in the calibration process. The overall simulated inflows to Lake Oroville reproduce reconstructed inflows with relative errors of -10 and -5 percent on monthly and annual time scales, respectively. The root-mean-squared errors of the simulated Lake Oroville inflows are 133,000 and 478,000 acre-feet for monthly and annual time scales, respectively. The accuracy of simulations appears to deteriorate for the period 1998-2000. Signatures of North Pacific decadal climate variations were observed in the Feather River Basin as a shift in the month of maximum streamflow (from April during the cooler Pacific decadal phase to March during the warmer decadal phase). The calibration period was dominated by the warmer (1977-98) phase. Since 1998, the simulations represent years in the newly re-established cool decadal phase. The response of the models to this subtle climatic fluctuation requires more evaluation.

    Streamflow predictions for the April-July snowmelt season were made with the Feather River model using a standard "ensemble streamflow prediction" (ESP) methodology. In the ESP methodology, April-July weather records from past years were used to drive the model through its plausible range of April-July streamflow totals for the current year, yielding a probabilistic forecast. Retrospective "predictions" using the ESP method were compared to the actual flows for each year from 1971 to 2000 to evaluate the reliability of the ESP results. These comparisons indicate that ESP-estimated flow probabilities are more accurate for the largest and smallest flows and tend to underestimate the likelihood of intermediate flow rates. Presumably, these comparisons can provide a guide for adjusting the confidence levels for any given ESP forecast in the future.





Purpose and Scope

Previous Studies


Physical Characteristics of the Feather River Basin

Location and Land Cover

Geology and Soils







Watershed Modeling

Spatial Representation

Watershed Processes

Model Areas

North Fork Tributary of the Feather River

Butt Creek and Almanor

East Branch

Lower North Fork

Middle Fork

South Fork

West Branch



Model Development

Model-Area Delineations

Precipitation Estimates for Hydrologic Response Units (HRUs)

Model Calibration and Error Analysis

Simulated and Remotely Sensed Snow Cover Comparison

Applications of the Models

Water-Balance Assessment

Seasonal Forecast Modeling using Ensemble Streamflow Prediction (ESP)

Model Limitations

Summary and Conclusions

References Cited


Document Accessibility: Adobe Systems Incorporated has information about PDFs and the visually impaired. This information provides tools to help make PDF files accessible. These tools convert Adobe PDF documents into HTML or ASCII text, which then can be read by a number of common screen-reading programs that synthesize text as audible speech. In addition, an accessible version of Acrobat Reader 5.0 for Windows (English only), which contains support for screen readers, is available. These tools and the accessible reader may be obtained free from Adobe at Adobe Access.


Water Resources of California

U.S. Department of the Interior, U.S. Geological Survey
Persistent URL:
Page Contact Information: Contact USGS
Last modified: Thursday, January 10 2013, 06:40:16 PM
FirstGov button  Take Pride in America button