Extraction of information content from stochastic disaggregation and bias corrected downscaled precipitation variables for crop simulation

• Published in: Stochastic environmental research and risk assessment Vol. 27; no. 2; pp. 449 - 457
• Main Authors: Mishra, Ashok K, Ines, Amor V. M, Singh, Vijay P, Hansen, James W
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer-Verlag 01.02.2013; Springer Nature B.V
• Subjects: Agricultural production; Aquatic Pollution; basins; Bias; Chemistry and Earth Sciences; Climate models; Climate science; Computational Intelligence; Computer Science; crop models; Crop yield; Crops; cumulative distribution; Earth and Environmental Science; Earth Sciences; Effective precipitation; Entropy; Environment; Food security; Global climate; Math. Appl. in Environmental Science; Mathematical models; Original Paper; Physics; Precipitation; probability; Probability Theory and Stochastic Processes; rain; rain intensity; Rainfall; Rainfall distribution; Rainfall intensity; Rainfall measurement; Stations; Statistics for Engineering; Stochastic models; Stochasticity; Waste Water Technology; Water Management; Water Pollution Control; Kenya; GCM; Downscaling; Crop model; Food security; Mutual information
• ISSN: 1436-3240; 1436-3259
• DOI: 10.1007/s00477-012-0667-9

We applied a simple statistical downscaling procedure for transforming daily global climate model (GCM) rainfall to the scale of an agricultural experimental station in Katumani, Kenya. The transformation made was two-fold. First, we corrected the rainfall frequency bias of the climate model by truncating its daily rainfall cumulative distribution into the station’s distribution based on a prescribed observed wet-day threshold. Then, we corrected the climate model rainfall intensity bias by mapping its truncated rainfall distribution into the station’s truncated distribution. Further improvements were made to the bias corrected GCM rainfall by linking it with a stochastic disaggregation scheme to correct the time structure problem inherent with daily GCM rainfall. Results of the simple and hybridized GCM downscaled precipitation variables (total, probability of occurrence, intensity and dry spell length) were linked with a crop model for a more objective evaluation of their performance using a non-linear measure based on mutual information based on entropy. This study is useful for the identification of both suitable downscaling technique as well as the effective precipitation variables for forecasting crop yields using GCM’s outputs which can be useful for addressing food security problems beforehand in critical basins around the world.