Abstract: The subcloud secondary evaporation of raindrops which happens during their falling process from the cloud base to the ground affects the stable isotopic ratios in the raindrops.It is of great significance to understand the changes of the stable isotopes in precipitation in the study of water cycle in a watershed.In this paper，the isotope data in precipitation were acquired from the Global Network of Isotopes in Precipitation（GNIP）and the related literatures across the Yellow River Basin.The relationship between subcloud secondary evaporation and meteorological elements was analyzed using local meteoric water line（LMWL）.Based on an improved Stewart model,the change of the evaporation remaining ratios of raindrops (f) and the difference of Dexcess from subcloud to ground（Δd）were quantitatively calculated.In this model,the heights of falling raindrops were estimated as lifting condensation level（LCL）using the insitu meteorological measurements.The stations cover the main climate zones across the Yellow River Basin,which can represent an integrated pattern of subsecondary evaporation in the study region.The results showed that the LMWL of the Yellow River Basin isδ2H=7.01δ18O+1.25（n=293, R2=0.92）,and both the slope and the intercept of LMWL are smaller than that of the global meteoric water line（GMWL）,which indicates that precipitation isotopes are affected by the subcloud secondary evaporation.The precipitation of 0~10 mm per month indicates a significant subcloud evaporation effect.And the higher air temperature (or the lower vapor pressure and relative humidity) may result in the greater subcloud secondary evaporation.There was an obvious seasonality of evaporation remaining ratio and Dexcess changes in the basin.From spring to winter, f and Δd were gradually increased,and subcloud secondary evaporation was gradually decreased.According to the climate zoning in China,the larger interannual variability in subcloud evaporation was exhibited in the subregions of Inner MongoliaGansu,Central Inner Mongolia,ShanxiShaanxiGansu and part of Weihe River (especially Xi’an).In contrast,smaller interannual trend was seen in the subregions of northern Qinghai,Qilian MountainsQinghai Lake,part of Weihe River (Pingliang,Changwu and Huashan),and ShandongHuaihe River.The linear relationship between Δd and f was different under various meteorological conditions as well as geographic locations,and the regression coefficient in the empirical formula should be treated according to local meteorological backgrounds.

Key words: stable isotopes of hydrogen and oxygen, subcloud secondary evaporation, precipitation, the Yellow River Basin