Accuracy of project-wide water uses from a water balance: a case study from Southern California

Abstract

A detailed water balance was conducted on the Imperial Valley in Southern California for the years 1987 to 1996. The area included all lands within the boundaries defined, including farms, towns, road, etc. This analysis included surface and subsurface inflows, rainfall, surface and subsurface outflows, evaporation, evapotranspiration, municipal and industrial uses, etc. Total water consumption was computed as the remainder in the water balance. The accuracy of this quantity was determined from standard statistical procedures based on the accuracy of the input data. For the Imperial Valley, total water consumption was estimated to within ±4.5%. The various sources of water and destinations were partitioned so that the accuracy of each component could be determined. The method also determines the source of uncertainty through the variance of each measurement. Subsystems were defined so that other quantities of interest could be determined. A canal subsystem water balance was used to determine more accurate estimates of water delivered to farms. A drainage/river subsystem water balance was used to determine the amount of irrigation return flows. Finally, the destinations of irrigation water were delineated and the Irrigation Consumptive Use Coefficient (ICUC) was calculated. The confidence interval, as a 10-year average, was 0.64 0.70. It was shown that the uncertainty of rainfall and how that water is partition has a significant influence on the accuracy of the estimated water consumption and ICUC.