CALIBRATION OF A WATER‐BALANCE MODEL FOR SMALL WATERSHEDS IN EASTERN OREGON1

Abstract

ABSTRACT: The BURP water‐balance model was calibrated for 13 small (0.46 to 7.00 mi2), forested watersheds in the Blue Mountains of eastern Oregon where snowmelt is the dominant source of runoff. BURP is the model name and is not an acronym. Six of the 16 parameters in BURP were calibrated. The subsurface recession coefficient and three subsurface water‐storage parameters were most sensitive for simulating monthly flow. Calibrated subsurface recession coefficients ranged from 0.988 to 0.998. The subsurface‐water storage parameters were calibrated at between 20 to 120 percent of their initial values obtained from a category III soil survey. That reconnaissance‐level survey was apparently too broad to accurately reflect subsurface‐water storage in small watersheds. Tests of model performance showed BURP is capable of producing accurate simulations of monthly flow for mountainous, snow‐dominated watersheds with shallow (< 4 ft) soils when calibrated with 2 to 4 years of streamflow data. A regression of observed versus simulated monthly flows with data from all watersheds combined showed that BURP accounted for 85 percent of the variability in observed flows, which ranged from 0.01 to 20.8 inches, but underpredicted high flow months, with a slope of 1.15 that is significantly different from 1.0 (p = 0.05). Without prior calibration, subsurface‐water storage parameters appeared to be the greatest source of potential error.