Variation Characteristics Of Pan Evaporation And Potential Evaporation In China And Their Relationship With Meteorological Elements

Evaporation is the main link between land and atmospheric water heat exchange,and it has an important impact on the surface water balance and energy balance.It plays an important role in agricultural and surface hydrological processes.This article mainly focuses on the characteristics of changes in pan evaporation(PE)and potential evaporation in China and the main meteorological factors that cause them.The Mann-Kendall method is used to analyze the trend of Chinese PE from 1961 to 2013,and the spatial variability is further studied by EOF decomposition of the annual average PE anomaly.Then,by studying the partial correlation coefficients between PE and 5 meteorological elements,the meteorological factors that have the greatest impact on PE are identified.Based on station observation data,the performance of 10sets of potential evaporation formulas in nine major river basins of China is ranked according to three different statistics(the standard deviation ratio between PE and potential evaporation,the correlation coefficient,and the ratio of the two trends)to select the most suitable potential evaporation formula for each basin.Finally,the sensitivity analyses are conducted on the selected formulas in each basin to determine the main meteorological factors of potential evaporation changes in various regions.The primary coverage which this article studies as follows to show:(1)Both annual and seasonal mean PE of all stations show a significantly downward trend,and an abrupt change occurs in 1978.The stations with significantly downward trend of PE mainly locates in the North China Plain,Xinjiang,Guangdong,Guangxi,and Hainan,while PE shows significantly increasing trend in Fujian,Zhejiang and Guizhou provinces.Then we apply the EOF analysis to annual PE anomalies.For the first mode(eof-1),the time coefficient changes from positive to negative in 1981,and its spatial pattern is similar as that of PE trend.The eof-2presents an opposite pattern in south and north China,and since 2002,the PE decreases in the north China,but increases in the south China.(2)We calculate the partial correlation coefficients between PE and five climate elements,including precipitation,temperature,surface wind speed,relative humidity and sunshine duration,respectively.The results show that except for precipitation,other four variables have very good correlation with PE.The correlation between wind speed and PE is significantly positive,and the regions with the largest correlation are consistent with those with the largest eof-1 variability.The correlation between humidity and PE is significantly negative.The correlation between temperature and PE are overall positive with the largest values appearing over areas where PE increases.The correlation coefficients between the sunshine duration and PE are greater than 0.6 in three seasons except for spring.Moreover,we find that the linear trends of both wind speed and sunshine duration have greatly impacted on their relationships with PE.Thus,we may conclude that the decreasing trend in PE should be largely affected by the declining in wind speed and sunshine duration.Furthermore,when the drought occurs,PE increases significantly,and the precipitation,temperature,relative humidity and sunshine duration changes also have significant contribution to PE increasing.Thus,PE could be a good indicator of drought.(3)Potential evaporation(PET)values computed from ten models with observed PE data are compared for nine Chinese river basins for the 1961-2013 period.The modeled PETs show larger biases than the PE in both magnitude and annual trend over most stations.PE in most stations are greater than 4 mm/d,while modeled PETs are less than 3 mm/d.In 7 of the basins(except for Southeast and Southwest China),PE shows decreasing trends with magnitudes ranging between-0.01 mm·d^-1 a^-1 and-0.03 mm·d^-1 a^-1 while the decreasing trends in modeled PETs are less than-0.01mm·d^-1 a^-1.This finding implies that the differences between PE and PET are more pronounced in the arid region.And in the Northwest China,FAO model is the only model that can well capture the decreasing PE trend.(4)A ranking scheme is applied to rank the statistical quantities(ratio of standard deviations,correlation coefficient and ratio of trends)between PE and modeled PETs in different river basins.The results suggest that,PET_Ham1 is the best model in the Pearl River basin,PET_Ham2 outperforms other models in the Huai,Yangtze and Yellow River basins,and PET_FAO is the best model for the remaining basins.(5)Sensitivity analyses reveal that wind speed and sunshine duration are two important factors for decreasing PET in most basins except in Southeast and Southwest China.In the four river basins calculated by the Hamon1/Hamon2 model,the contribution of sunshine duration to the PET change is always greater than 10%.While in Hai,Songliao river basins,and Northwest China,the wind speed contribution exceeds 20%.In addition,in the Songliao river basin,the daily minimum temperature also has a relatively obvious contribution,which can reach 16.41%.The increasing PET trend in Southeast China is mainly attributed to the reduced relative humidity.