Do Forecasts with Significant Error Have Value in Water Leasing Decisions?

Abstract:

Uncertainty about weather conditions over the growing season creates significant risk about important aspects of irrigated agricultural production (e.g., crop phenology and productivity, irrigation demand, etc.). Emerging technologies such as improved forecasts of seasonal water availability and crop productivity can reduce uncertainty for farmers in numerous decisions (e.g., crop mix, planting date, contracting, and water use vs. lease decisions). The usefulness of forecasts in the decision process is a function of multiple factors: their availability with sufficient accuracy and lead times, the costs (vs. benefits) of making an incorrect (or correct) decision, and the relative probabilities of outcomes. Using a case study of hay production within the Yakima River Basin of the U.S. Pacific Northwest, our objective is to (a) quantify the value of various levels of improvements in a seasonal forecast of irrigation water availability that informs hay grower’s yearly decision to use water to irrigate hay versus forgo irrigation and lease water to someone else, (b) quantify how this value changes with forecast lead times, and (c) quantify the value of a perfect forecast. For this, we take a hindcasting approach using downscaled North American Multi Model Ensemble (NMME) seasonal forecast data for the period 1979 to 2019, and compare outcomes of the hay grower’s use vs. lease decision made with and without a seasonal forecast where the default is historical climate average. The coupled crop-hydrology model VIC-CropSyst is used to determine crop productivity. Decisions made using forecasts available on February 1st, March 1st, April 1st, and May 1st are used to quantify forecast value as a function of lead time. Results indicate that forecast value does increase with lead time by improving water leasing decisions, but only up to a point. Our results show that forecasts even with significant error have a positive economic value. The results can inform the type of water market contracts that can be considered to improve regional water allocation efficiencies, and mitigate risk in agricultural production due to hydroclimatic variability.