Implications of medium-range numerical weather model output in hydrologic applications: Assessment of skill and economic value

Integrating medium-range numerical weather model output into hydrologic models has been shown to yield useful information. The emphases in this study were to (i) develop effective mechanisms for integrating meteorological ensemble systems into hydrologic forecasts; and (ii) evaluate the skill and economic values of the subsequent hydrologic forecasts. The general framework was demonstrated under a number of conditioning paradigms for the Chute-du-Diable watershed located in Quebec, Canada. For this purpose, two downscaling models were employed to generate station total daily precipitation and average daily temperature. The performances of the downscaled outputs were compared against the raw model output using suites of diagnostic measures. The comparative results indicated that the downscaled outputs yielded more accurate and realistic forecasts than the raw model output. These outputs were then forced into an HBV hydrologic model in order to generate flow forecasts up to 14 days ahead. The approach effectively generated deterministic and probabilistic flows. The subsequent simulation results revealed that the downscale-based flows yielded greater skill values than the raw-based flows. The potential economic values of flow forecasts were assessed based on a simple optimal decision-making, cost–loss analysis technique. The principal outcomes emerging from the analyses include: (i) the economic benefits associated with probabilistic flow forecasts were more useful than their deterministic counterparts; and (ii) the downscale-based flow forecasts offered greater benefits, which are applicable to a much wider range of users, than the raw-based flow forecasts. The findings of the present study clearly illustrate the potential added value that may be obtained as a result of adequate downscaling, as opposed to using the raw model output, for hydrologic applications.