A High-Resolution Analysis of Daily Precipitation Climatology Reflecting the Impact of Topography across China

In this study, the daily precipitation observations collected by 2,014 Chinese surface stations are used to construct a gridded climate background field (CDC_CLM_PRE025) for 30 years (1971-2000) at a 0.25° latitude and longitude resolution, with the involvement of geographical factors (latitude, longitude, elevation, slope, and aspect), using thin plate smoothing splines. The analysis field reaches a fine consistency with the observed fields on annual cycles. Cross-validation shows that the majority of stations reported a relative error rate between the analyzed and observed precipitation under 20%, enjoying a correlation coefficient higher than 90%. Comparing with other gauge-based gridded precipitation datasets, that is GPCC, PRISM and CRU, CDC_CLM_PRE025 is in a better position to depict the impact of topography on surface precipitation, with a reduced relative error by 12%-23%. GPCC, PRISM, and CRU tend to underestimate the precipitation in mountainous area, compared with the observed one. A further analysis of the gap between the analyzed and observed field under different terrain conditions demonstrate that CDC_CLM_PRE025 offers an analysis field that depicts the relationship between precipitation and terrain elevation in a more accurate manner, compared with the observed field, closer to the reasonable distribution of precipitation on the ground. Comparison tests under different interpolation schemes and observation network densities show that interpolation schemes, reflecting the effects of topography or not, would substantively affect the results of precipitation analysis, compared with observation network density. Meanwhile, the value derived from the precipitation analysis reflecting the effects of topography would be universally higher than the one stemmed from the scheme excluding such effects, particularly in an area with a grid box elevation higher than 2,000m.