Potential value of operationally available and spatially distributed ensemble soil water estimates for agriculture

The focus of this paper is to develop a methodology to answer the question: do the spatially distributed soil water estimates produced by operational distributed hydrologic models provide potential benefits for agriculture? The formulation quantifies the potential value through a cost-loss analysis, whereby cost for the farmer is associated with the decision to irrigate the field and loss is associated with the decision not to irrigate while damaging soil water deficits occur. Farmer decisions are made in view of the likelihood of damaging events as estimated by the ensemble distributed model simulations of soil water deficit. The ensemble simulations account for parametric and radar rainfall uncertainty. The application area for the economic value analysis is the farmland of the Illinois River watershed in northwestern Arkansas (mainly) and eastern Oklahoma, for which operational-quality distributed model input is available. The land is used to produce hay for feed. The analysis indicates that there is substantial potential economic value in using the ensemble soil water estimates to make decisions regarding irrigation within the watershed for the months of July, August and September, when severe soil water deficits may occur. The benefits are higher for lower cost-loss ratios and for higher yield plants. They exhibit considerable spatial variability within the watershed in agreement with the spatial variability of the incidence of soil water deficits and with the spatial variability of the ability of the ensemble model simulations to reproduce this variability. The results of this study warrant additional analysis of the economic value of distributed model simulations in other regions, different distributed models and for other types of crops. Consideration of forecasts in addition to simulations is also an important next step. (c) 2005 Elsevier B.V. All rights reserved.