Development of an effective and potentially scalable weather generator for temperature and growing degree days

摘要: The main purpose of this study is to develop an easy-to-use weather generator (WG) for the downscaling gridded data point measurements at regional scale. WG applied daily averaged temperatures and annual growing degree days (GDD) wheat. This particular choice variables motivated by future investigations on temperature impacts as most important climate variable wheat cultivation under irrigation in Iran. proposed statistical relates large-scale ERA-40 reanalysis local GDD. Long-term observations Iran are used 16 synoptic stations from 1961 2001, which common period with data. We perform using two approaches: first a linear regression model that uses fingerprints (FP) defined squared correlation variability, second employs multiple (MR) analysis relate information neighboring grid points station Extending usual downscaling, we implement providing uncertainty realizations GDD adding Gaussian random noise. well represents variability. For 2-m temperature, FPs more localized during warm compared cold season. While MR slightly superior time series, FP seems best further assess quality WGs applying probabilistic verification scores like continuous ranked probability score (CRPS) respective skill score. They clearly demonstrate superiority deterministic downscaling.