On the reliability of spatially disaggregated global ensemble rainfall forecasts

• DOI: 10.1002/hyp.9509
• Publication Date: 2012-08-03T23:42:44Z
• Authors: Étienne Gaborit; François Anctil; Vincent Fortin; Geneviève Pelletier
• Keywords: Interpolation

Abstract Ensemble rainfall forecasts are of high interest for decision making, as they provide an explicit and dynamic assessment the uncertainty in forecast. However, hydrological forecasting, their low resolution currently limits use to large watersheds. To bridge this gap, various implementations a spatial statistical downscaling method were compared, bringing Environment Canada's global ensemble from 100 × 70‐km down 6 4‐km while increasing each pixel's variance preserving its original mean. This was applied nine consecutive days summer 2009 with strong rain events over Quebec City, Canada. For comparison purposes, simpler methods also implemented such bilinear interpolation, which disaggregates without modifying variance. The meteorological products evaluated, using different scores diagrams, against observed values taken City gauge network. most important conclusions work that overall quality preserved during disaggregation procedure disaggregated variance‐enhancing similar than interpolation products. dispersion members were, course, much improved variance‐enhanced products, compared is decisive advantage. Therefore, there implementing disaggregate forecasts. Copyright © 2012 Her Majesty Queen right Published by John Wiley & Sons, Ltd.