Statistical Models and Methods for Rivers in the Southwest

Abstract

Riverflow modeling is believed useful for purposes of decision making with respect to reservoir control, irrigation planning, and flood forecasting and design of structures to contain floods. This author holds the view that present riverflow models in vogue are unsatisfactory because, for one thing, sample simulations according to these models do not resemble observed southwestern river records. The purpose of this paper is to outline a general Markov model which assumes only that rivers have a finite memory. We show how to calibrate the model from river records and then present evidence to support our contention that some success has been realized in mimicking typical flows by our simulation procedure.