Gridded daily precipitation data for Iran: A comparison of different methods

Abstract

Study region: Iran Study focus: Gridded precipitation products are of great interest for hydrological applications. The inhomogeneous geography and uneven spatial distribution of rain gauges in Iran make it difficult to estimate valuable interpolated precipitation with daily or monthly resolutions. Therefore, we evaluated the performance of two empirical and four geostatistical interpolation methods. New hydrological insights for the region: Atmospheric circulation pattern (CP) classification was used to understand precipitation behavior and to improve interpolation. Based on 500 hPa geopotential fields, six CPs were identified, in order to explain large scale precipitation behavior. Variograms were normed and clustered to reduce the computational effort of the geostatistical methods.Leave-one-out cross-validation shows that the geostatistical methods outperform the empirical ones, and the differences among the geostatistical methods are small.The difference among all the methodologies decreased substantially for spatial aggregation to coarser resolutions. In contrast, temporal aggregation reduced the difference to a much lower extent.A large dataset consisting of 1561 locations with daily observations was used for this study. Comparison with the GPCC daily dataset shows that the data used for interpolation has a larger influence than the choice of the interpolation method.

Keywords

• Daily precipitation
• Iran
• Interpolation
• Kriging
• Clustering
• Atmospheric circulation patterns