资源环境科技发展态势分析平台

Many spatial vegetative physiological and environmental impact studies demand consistent as well as fine resolution daily meteorological data, which is often unavailable. This problem is evident when studies and applications focus at the regional and local scale. Meteorological stations cannot provide the needed data due to their irregular locations and several associated shortcomings such as missing data. Climate data from global sources like the National Centers for Environmental Prediction (NCEP) and WorldClim are provided at a coarse resolution which makes them inappropriate to use directly at regional and local scales. In order to circumvent this problem, we downscaled daily WorldClim and NCEP global climate grids to produce weather data sets using a delta method to 0.0083 (approximately 1 x 1 km) spatial resolution for the Amhara region of northwestern Ethiopia. The downscaled data set includes daily precipitation, and minimum and maximum temperature from 1979 to 2010. The drizzle effect in the downscaled data set was first eliminated by using a rain threshold that removed values of < 1mm per day. We compared the downscaled values with our calibration data from 56 meteorological stations in the Amhara region. The comparison between the downscaled and the calibrated data from weather stations showed biases in the downscaled values. A delta-change bias correction technique on a 10-day average basis was used to correct the biases associated with the downscaled data. Simple additive and multiplicative bias correction methods were used for temperature and precipitation, respectively. We then validated the corrected downscaled daily weather data using ten independent weather stations from the region. We found that the downscaling and the bias correction methods have improved the NCEP values. The validation exhibited no bias. The full data set can be accessed freely under the following link: ftp://palantir.boku.ac.at/Public/ClimateDataEthiopia/.