11 - Stochastic precipitation-runoff modeling for water yield from a semi-arid forested watershed

Summary

ABSTRACT

A stochastic precipitation-runoff modeling approach is used to estimate water yield from a particular forested watershed in North Central Arizona. The procedure uses selected theoretical probability distribution functions and a random number generator to describe and simulate various precipitation characteristics, such as storm depth, duration, and time between storm events. The spatial characteristics of precipitation events are described in terms of their orographic and areal distribution patterns while temporal distributions are expressed in terms of daily events in the watershed. The generated precipitation events are used as input into a precipitation-runoff model to estimate water yield from a particular forested watershed. The method uses geographic information systems (GIS) to subdivide the study watershed into cells assumed to be homogenous with respect to watershed characteristics, such as elevation, aspect, slope, overstory density, and soil type. The total water yield is the accumulated surface runoff generated at the watershed outlet. The outcome is the development of an improved model for estimating water yield which takes into consideration uncertainty, as well as temporal and spatial watershed characteristics. This method is useful not only for providing water resources managers with a good estimate of the amount of water yield, but also for determining the reliability or failure of a source to meet desired downstream water demands.

INTRODUCTION

This chapter is concerned with the development of an appropriate precipitation-runoff model for estimating water yield from a semi-arid forested watershed. This involves combining a stochastic precipitation model and a deterministic runoff model. The first one is selected to capture the inherently uncertain characteristics of precipitation, while the latter is chosen to simplify an otherwise complex surface runoff estimation method.