Rainfall and temperature estimation for a data sparse region

Abstract. Humanitarian and development agencies face difficult decisions about where and how to prioritise climate risk reduction measures. These tasks are especially challenging in regions with few meteorological stations, complex topography and extreme weather events. In this study, we blend surface meteorological observations, remotely sensed (TRMM and NDVI) data, physiographic indices, and regression techniques to produce gridded maps of annual mean precipitation and temperature, as well as parameters for site-specific, daily weather generation in Yemen. Maps of annual means were cross-validated and tested against independent observations. These replicated known features such as peak rainfall totals in the highlands and western escarpment, as well as maximum temperatures along the coastal plains and interior. The weather generator reproduced daily and annual diagnostics when run with parameters from observed meteorological series for a test site at Taiz. However, when run with interpolated parameters, the frequency of wet days, mean wet-day amount, annual totals and variability were underestimated. Stratification of sites for model calibration improved representation of the growing season's rainfall totals. Future work should focus on a wider range of model inputs to better discriminate controls exerted by different landscape units.