Analysis The Cause Of Reference Evapotranspiration And Simplified Method In Jiangxi

Jiangxi is in a subtropical monsoon climate zone with abundant hydrothermal resources,but distribution of rainfall was uneven in time,and the red soil water storage capacity is limited,which can easily induce seasonal drought.In recent years,the seasonal drought in has been frequent in Jiangxi province,and the annual precipitation has not changed significantly,and the change law of evaporation capacity and its genesis have become the key factors to explore the causes of drought.The study on the change rule and causes of evaporation capacity has become the key factor to explore the causes of drought,while meteorological stations are less,and lack of long sequence meteorological observation data has become an important obstacle.The analysis of the causes and simplified models of reference crop evapotranspiration in Jiangxi is helpful to the optimal allocation of agricultural water resources and the impacts of climate change on water resources.The results shows that:?1?FAO Penman Monteith-formula recommended by FAO was used to calculate the daily ET₀ from 1961 to 2015.Mann-Kendall trend test and mutation test analysis was applied to study the ET₀ variation characteristics and mutation of time of Nanchang.On the basis of using Path Analysis to research the causes of ET₀ variation.The result shows that Nanchang ET₀ monthly presented single-peak curve,7⁸ month is highest,which accounts for about 30%throughout the year.?2?From the annual variation,ET₀ and sunshine time has no significant trend,only present a significant decline in August.There were a significant reduction in the relative humidity and wind speed,while the highest and lowest temperature rise significantly.Sunshine duration and ET₀ yearly accumulative anomaly curve had the closest curve shape,mutations in time was also the most close,which were in 1990 and 1992,respectively.Path coefficient of sunshine duration,wind speed,relative humidity and lowest temperature was0.513,0.466,0.425 and 0.513,respectively,shows that sunshine duration was the most sensitive factor,followed by wind speed,relative humidity and the lowest temperature.Sunshine duration through relative humidity affects the ET₀ was also strong?the Indirect path coefficients was 0.236?,which was higher than the lowest temperature of direct action.?3?Meanwhile,12 methods were calibrated and compared with FAO-56Penman-Monteith formula in Jiangxi province,including Pristley-Taylor?P-T?method,Makkink method,Hargreaves-Samani?H-S?method,Irmak-Allen?I-A?method,Penman1948?48PM?method,Penman-Van Bavel?PVB?method,Baier-Robertson?B-R?method,modified Baier-Robertson?M-B-R?method,Schendel?Sch?method,Turc method,Jensen-Haise?J-H?method and Brutsaert-Stricker?B-S?.The daily meteorological data were extracted from 26meteorological stations in Jiangxi province and 17 stations in surrounding provinces from1961 to 2010.The results showed that Turc method performed best before parameter calibration.The average root mean squared error?RMSE?of all stations in Jiangxi province were 0.39 mm/d,the normalized root mean squared error?NRMSE?were 0.157,while the accuracy of Turc method was improved slightly after parameters calibration.Hence the Turc method could apply directly with default parameters?besides Nanchang?.?4?Spatial distribution of NRMSE showed that all methods have the largest errors in the north of Jiangxi province.The applicability of B-S method and 48PM method were best after parameter calibration.the RMSE were 0.08⁰.27 mm/d and 0.06⁰.34mm/d,NRMSE were0.017⁰.120 and 0.026⁰.154,d-index were nearly 1,absolute error of average ET₀ in all stations were less than 30 mm.P-T method was suitable for Jiangxi Province when lack of relative humidity and wind speed data.The parameter of?was from 1.05 to 1.18,RMSE was from 0.21 to 0.66 mm/d and NRMSE was from 0.08 to 0.30.RMSE?d-index and NRMSE of P-T method,I-A method and PVB method were nearly equivalent in all the stations,and higher than Mak method in terms of accuracy because it was obviously overestimated in ET₀in July and August by Mak method.RMSE of H-S method,B-R method,M-B-R method and J-H,and Sch method were over 0.75mm/d.It was concluded that these methods were unsuitable in ET₀ estimating in Jiangxi province.Addition to the 48PM and B-S method,other methods gave obvious changes in ET₀ around 2003.This may be due to obvious changes of some meteorological factors had been taken place in those years.