Estimation Of Surface Solar Radiation And Analysis Of Its Spatio-temporal Variation Over Mainland China

Solar radiation is the primary energy source on earth.Solar radiation,which affects the transportation of materials and energys and drives the cycle of carbon,water,heat and nutrient in earth-atmospheric system,determines climate and environment parameter on earth.Due to the sparse distribution of solar radiation measurement stations,surface solar radiation predicting models are of vital importance for obtaining accurate and continuous surface solar radiation(SSR).In recent years,the global solar radiation is proved to be changed significantly in the worldwide.But variation of global solar radiation in different areas show different characteristics and the reasons for variations are complicated.Therefore,it is important to analyze the variation characterisctics of global solar radiation and detect its drivers in different areas.In this study,measured climatological data between 1994 and 2013 and remote-sensing image between 2004 and 2013 are used to develop several different solar radiation predicting models.On the base of several solar radiation predicting models,this work will choose the best model,ANFIS model,to develop a surface solar radiation dataset covering 53 years over 552 Chinese meteorological stations,and then investigate the spatial-temporal characteristics of SSR in China.In addition,the long-term variation of aerosol optical thickness(AOD),low cloud cover(LCC),total cloud cover(TCC)and water vapor content(WVC)are analyzed based on measured weather data over 420 stations during 1980-2013.Moreover,relationships between AOD/LCC/TCC/WVC and SSR are estimated to investigate the possible contribution of AOD,LCC,TCC and WVC to the long-term variation of SSR in the mainland China.The main conclusions of this paper are shown as follows:(1)The climate characteristics between 1961 and 2013 in mainland China are analyzed based on meteorological parameters in 592 stations.On the base of the regional climate charaters,the mainland China are divided into 7 regions.The long-term variation of annual mean temperature,sunshine duration,relative humidity and annual total precipitation on 8 meteorological stations over mainland China are analyzed.It is concluded that the temperature and precipitation show increased trends in mainland China during the years of 1961?2013.While the sunshine duration,relative humidity and surface solar radiation show decreased trends in mainland China.(2)Model performances of BC model,IBC model,YHM model,IYHM model,ANFIS model and MODIS model are compared and validated in this work with meteorological and solar radiation data during 1994-2013.Results show that RMSEs,MAEs in improved model(IBC and IYHM)are much lower than that in original models.The MODIS model has the lagerst statistical errrors among several models.The lowerst statistical error values are observed in ANFIS model,the RMSE is in the range of 1.42?2.82 MJ m-2 and MAEs range from 1.06 MJ m-2 to 2.04 MJ m-2.It can be inferred that the improved models(IBC model and IYHM model)are better than the original ones(BC model and YHM model),the higherst error values are observed in MODIS model,while ANFIS model shows highest predicting accuracy among all models.(3)A surface solar radiation dataset covering 53 years(from 1961 to 2013)over 552 China meteorological stations is built by ANFIS model.A daily solar radiation dataset covering 50 years(from 1961 to 2010),which is proposed by "Data Assimilation and Modeling Center for Tibetan Multi-spheres",is used to validate this dataset.Results show that the RMSEs are in the range of 0.40 MJ m-2-2.56 MJ m-2,the MAEs are in the range of 0.31 MJ M-2?2.26 MJ m-2 for 552 stations.It can be seen that the SSR dataset derived in this study has high precision.Thereafter,the spatial-temporal variations of SSR are analyzed.The spatial distribution of SSR in mainland China is uneven.If we draw a line along the Great Xing'an Mountains,the Taihang Mountains,Qinling Mountains,the west of Sichuan Basin and the bordering regions of Yunnan and Guizhou province.The SSR in east and south areas of the line are averagely lower than that in west and north areas.The highest SSR is observed in Tibetan area,and the lowest values are found in Sichuan Basin and southeast hilly areas.As for the long-term trend of global solar radiation,we can see that the long-term SSR in mainland China is decreased by-0.019 MJ m-2 year-1 during 1961-2013.(4)The long-term variations of AOD,LCC,TCC and WVC during 1980-2013 in mainland China over 420 stations are analyzed.The long-term trends of AOD,LCC,TCC and WVC are increased.Their correlation coefficient with SSR are-0.79,-0.48,-0.75,-0.18 respectively.Results imply that the increase of AOD is highly correlated with the decrease of SSR in Northwest and Central areas of China.The significantly increased LCC and TCC in the Northeast,Northwest,Inner Mongolia and Central areas of China are main reasons for the decrease of SSR in those areas.The variation of WVC seems correlated with the decrease of SSR in Northwest,Tibetan and Inner Mongolia areas.