Isolating the Role of End-User Behavior in the Assessment of Seasonal Forecast Value

Recent improvements in model resolutions, initialization procedures, and representation of large scale hydro-meteorological processes contributed in advancing the accuracy of hydroclimatic forecasts, which are more and more skillful over the seasonal and longer timescales. These forecasts are potentially valuable for informing multisector strategic decisions, including irrigated agriculture, where they can improve crop choices and irrigation scheduling decisions. In this operational context, forecast accuracy is important but not necessarily proportional to the associated economic marginal benefit, which is also affected by how forecasts are employed by end-users. In this work, we contribute a novel framework to quantify the value of hydroclimatic forecasts by extending traditional quality assessments with estimates of the potential economic benefit of the forecasts to the end-user. We also explore the sensitivity of this benefit to both model set up and end-user behavioral factors. The approach is demonstrated on the Lake Como system (Italy), a regulated lake operated for flood protection and irrigation supply. Our framework relies on an integrated modeling chain composed of three building blocks: bias-adjusted seasonal meteorological forecasts are used as input to the continentally-calibrated E-HYPE hydrological model; predicted lake inflows are used for conditioning the daily lake operations; the resulting lake releases feed an agricultural model to estimate the net profit of the farmers in a downstream irrigation district. Results suggest that, on average, informing the Lake Como operations based on E-HYPE hydrological forecasts allows gaining about 1% of the farmers’ profit with respect to a baseline solution not informed by any forecast. This gain rises up to about 15% during intense drought episodes. Moreover, our analysis suggests that this value can be largely attributed to the hydrological model and its initial conditions, while the role of meteorological forcing emerges only during dry seasons. Lastly, our results show a high sensitivity to behavioral factors capturing different perception of risk and uncertainty, with the estimated forecast value being potentially undermined if end-users are not able to properly extract the most valuable information from the forecast ensemble.