HR: 1330h
AN: H12B-0976 [PDF]
TI: Statistical Development of Flood Frequency and Magnitude Equations for the Cosumnes and Mokelumne River
Drainage Basins, Sierra Nevada, California
AU: * Burns, R G
EM: rburns@water.ca.gov
AF: Dept. of Geology, California State University - Sacramento, 6000 J Street, Sacramento, CA 95819
AU: Meyer, R W
EM: rwmeyer@usgs.gov
AF: U.S. Geological Survey - Water Resources Division, Placer Hall,
6000 J Street, Sacramento, CA 95819-6129
AU: Cornwell, K
EM: cornwell@csus.edu
AF: Dept. of Geology, California State University - Sacramento, 6000 J Street, Sacramento, CA 95819
AB:
In-basin statistical relations allow for development of regional flood frequency and magnitude equations in the Cosumnes
River and Mokelumne River drainage basins. Current equations were derived from data collected through 1975, and do not
reflect newer data with some significant flooding. Physical basin characteristics (area, mean basin elevation, slope of
longest reach, and mean annual precipitation) were correlated against predicted flood discharges for each of the 5, 10, 25,
50, 100, 200, and 500-year recurrence intervals in a multivariate analysis. Predicted maximum instantaneous flood discharges
were determined using the PEAKFQ program with default settings, for 24 stream gages within the study area presumed not
affected by flow management practices. For numerical comparisons, GIS-based methods using Spatial Analyst and the Arc Hydro
Tools extension were applied to derive physical basin characteristics as predictor variables from a 30m digital elevation
model (DEM) and a mean annual precipitation raster (PRISM). In a bivariate analysis, examination of Pearson correlation
coefficients, F-statistic, and t \& p thresholds show good correlation between area and flood discharges. Similar analyses
show poor correlation for mean basin elevation, slope and precipitation, with flood discharge. Bivariate analysis suggests
slope may not be an appropriate predictor term for use in the multivariate analysis. Precipitation and elevation correlate
very well, demonstrating possible orographic effects. From the multivariate analysis, less than 6% of the variability in
the correlation is not explained for flood recurrences up to 25 years. Longer term predictions up to 500 years accrue
greater uncertainty with as much as 15% of the variability in the correlation left unexplained.
DE: 1821 Floods
SC: Hydrology [H]
MN: 2003 Fall Meeting