Spatial Disaggregation of the 0.25-degree GLDAS Air Temperature Dataset to 30-arcsec Resolution

Abstract:

Air temperature is a key input variable in ecological and hydrological models for simulating the hydrological cycle and water budget. Several global reanalysis products have been developed at different organizations, which provide gridded air temperature datasets at resolutions ranging from 0.25º to 2.5º (or 27.8 – 278.3 km at the equator). However, gridded air temperature products at a high-resolution (≤1 km) are available only for limited areas of the world. To meet the needs for global eco-hydrological modeling, we aim to produce a continuous daily air temperature datasets at 1-km resolution for the global coverage. In this study, we developed a technique that spatially disaggregates the 0.25º Global Land Data Assimilation System (GLDAS) daily air temperature data to 30-arcsec (0.928 km at the equator) resolution by integrating the GLDAS data with the 30-arcsec WorldClim 1950 - 2000 monthly normal air temperature data. The method was tested using the GLDAS and Worldclim maximum and minimum air temperature datasets from 2002 and 2010 for the conterminous Unites States and Africa. The 30-arcsec disaggregated GLDAS (GLDASd) air temperature dataset retains the mean values of the original GLDAS data, while adding spatial variabilities inherited from the Worldclim data. A great improvement in GLDAS disaggregation is shown in mountain areas where complex terrain features have strong impact on temperature. We validated the disaggregation method by comparing the GLDASd product with daily meteorological observations archived by the Global Historical Climatology Network (GHCN) and the Global Surface Summary of the Day (GSOD) datasets. Additionally, the 30-arcsec TopoWX daily air temperature product was used to compare with the GLDASd data for the conterminous United States. The proposed data disaggregation method provides a convenient and efficient tool for generating a global high-resolution air temperature dataset, which will be beneficial to global eco-hydrological modeling and particularly useful for developing countries where high-resolution meteorological data are unavailable.