Downscaling of Seasonal Precipitation for Crop Simulation

Abstract

A nonhomogeneous hidden Markov model (NHMM) is used to make stochastic simulations of March–August daily rainfall at 10 stations over the southeastern United States, 1923–98. Station-averaged observed daily rainfall amount is prescribed as an input to the NHMM, which is then used to disaggregate the rainfall in space. These rainfall simulations are then used as inputs to a Crop Estimation through Resource and Environment Synthesis (CERES) crop model for maize. Regionally averaged yields derived from the NHMM rainfall simulations are found to correlate very highly (r = 0.93) with those generated by the crop model using observed rainfall; stationwise correlations range between 0.44 and 0.74. Rainfall and crop simulations are then constructed under increasing degrees of temporal smoothing applied to the regional rainfall input to the NHMM, designed to exclude the submonthly weather details that would be unpredictable in seasonal climate forecasts. Regional yields are found to be remarkably insensitive to this temporal smoothing; even with 90-day low-pass-filtered inputs to the NHMM, resulting yields are still correlated at 0.85 with the baseline simulation, whereas stationwise correlations range between 0.18 and 0.68. From these findings, it is expected that regional maize yields over the southeastern United States will be largely insensitive to year-to-year details of subseasonal rainfall variability; they should be downscalable, in principle, using an NHMM from climate forecasts archived at daily resolution, with the important caveat that the latter need to be skillful enough at the 90-day time scale. As a by-product of the analysis, subseasonal-to-interdecadal summer rainfall variability over the southeastern United States is interpretable in terms of six discrete weather states indicative of a monsoonlike climate regime. Low-simulated-yield years are found to be associated with delayed summer rainfall onset.