Long-term Correlation For Daily Precipitation Records Over Southern China

In this study,the long-term correlation of precipitation records in southern China are investigated by using detrended fluctuation analysis method(DFA)based on daily precipitation data from the National Meteorological Information Center of China.Moreover,based on the long-term correlation characteristic of the observed precipitation records,the performance of three climate models(BCC_CSM1.1(m),CESM,and MIROC4h)and two reanalysis data(NCEP and ECMWF)in simulating precipitation over southern China are evaluated.Finally,we study the temporal and spatial changes in long-term correlation with increasing greenhouse gas emissions and reduced aerosol emissions in the late-21st century(2071-2100).The main conclusions are as follows:(1)The annual precipitation series in southern China shows strong long-term persistence with a scaling exponent value of 0.61.In addition,the DFA result for randomly shuffled precipitation records indicates that the scaling exponent is 0.5 and further verify the fact that the long-term correlation characteristic for precipitation series over southern China is caused by fractal characteristic.(2)It is seen that the scaling exponents of annual precipitation records show a decreasing trend from the coast to the inland over southern China.The results further illustrate that the precipitation in the coastal areas of southern China is more predictable on the time scale of 10-1000 days,while the precipitation series in the inland regions of southern China are less predictable on the time scale of 10-1000 days.(3)Compared with the observed data,the scaling exponents of ECMWF reanalysis data,BCC_CSM1.1(m),CESM,and MIROC4h models simulation data are close to observation in most parts of southern China,with 92%,90%,88%and 91%of the grids passing the 0.01 significant test.These results indicate that the ECMWF reanalysis data and three climate model can basically reproduce the long-term correlation of precipitation series over southern China.While the results of the DFA analysis of the NCEP reanalysis data are poor,with only 53%of the grid points passing 0.01 significant test.(4)Under the RCP4.5 simulation scenario,the scaling exponents of the precipitation records in the northeastern part of Hunan and the northwestern part of Jiangxi will be enhanced in the late-21st century,while the rest of southern China is consistent with the current condition(1987-2016).The long-term correlation of precipitation records in Huaihe river basin will strengthen during late-21st century(2071-2100)under RCP8.5 simulation scenario.However,The long-term correlation of precipitation records over all southern China regions are consistent with present condition(1987-2016)during late-21st century(2071-2100)under RCP8.5_FixA simulation scenario.In addition,the differences between the RCP8.5 and RCP4.5 and RCP8.5_FixA simulation scenarios indicate that whether the increase in greenhouse gas emissions or the decrease in aerosol emissions will increase the long-term sustainability of the precipitation series in the Huaihe river basin,especially the reduction in aerosol emissions in the future will increase the scale exponents of the precipitation series in the Huaihe river basin in summer by more than 0.1.