WRF simulation and downscaling of local climate in Central Asia

Progresses in understanding climate variation and its ecosystem responses and feedback in Central Asia have been limited by availability of high resolution and quality climate data. To ease this limitation, the Weather Research and Forecasting (WRF) model is implemented in Central Asia in this study after repeated yearly sensitivity evaluations. The optimized WRF is used to downscale the ERA-Interim reanalysis data, producing a new regional dataset of 20 × 20 km spatial resolution covering 1980–2015. The downscaled precipitation and daily maximum and minimum temperatures of seven years from 1980–1986, when the observational data are the most abundant, are compared in statistical measures to station observations and also contrasted with the comparison results between the ERA-Interim data and the observations. Results of those comparisons show that the downscaled data have overall comparable accuracy in precipitation and daily extreme temperatures to the ERA-Interim data. An outstanding improvement of the downscaled data is in their describing the variance in precipitation and temperatures at daily to annual timescales, while the downscaled extreme precipitation and temperature events also show more accurate comparisons to the observed at some sub-regions in Central Asia. The validated WRF model, which can be used to further downscale past as well as future climate data for specific research needs, and the downscaled high resolution and quality data present both a tool and a useful dataset for ecological and detailed climate research in Central Asia.