| As the most serious soil erosion area,red soil region of southern China shows the main erosion mode in water erosion.Rainfall is the most important dynamic factor which leads to water erosion,while rainfall erosivity is the physical quantity which characterizes the potential erosion ability of rainfall.Study on characteristics and changing trends in spatial and temporal distribution of rainfall erosivity,shows positive meaning for soil and water loss prediction and soil conservation planning.According to the 30 years’ daily rainfall data of 1987-2016 collected by 175 weather stations in southern red soil region,this paper analyzed spatial-temporal evolution of rainfall and its erosivity in the past 30 years,and built the spatial and temporal estimation models of rainfall erosivity based on multi-year average and annual rainfall data respectively.The main conclusions are as followed:In space,the average annual rainfall of the 175 sites in 7 provinces of the southern red soil region from 1987 to 2016 was 1643.8mm,of which the maximum annual rainfall reached 2706.4mm in Dongxing weather station of Guangxi Province;the minimum value was 1002.1mm in Dongfang weather station of Hainan Province.The CV value was 16.6%,which showed a moderate degree of spatial variability.The average rainfall space in the southern red soil region showed a decreasing trend from southeast to northwest during the 30 years.Due to the influence of different topographic features,the spatial distribution of rainfall presented a multi-center feature,with two peak centers in the south of Guangxi province and the southeast of Guangdong.In time,the highest rainfall in the southern red soil region during the period of 1987 to 2016 was 2019.3mm in 2016,22.8%more than the average,and the minimum value was 1271.3mm in 2003,22.7%less than the average.The CV value of coefficient of variation was 11.7%,which revealed moderate variability of annual rainfall in the southern red soil region for 30 years.Mann-Kendall Trend test Value Z=1.21,indicating that the annual rainfall changes into an upward trend.In addition,the average rainfall of the 175 sites was 667.8mm in summer,accounting for 40.63%of the year,followed by spring 31.24%,autumn 16.91%and winter 11.22%,with rainfall in spring and summer accounting for 71.87%of annual rainfall.The results of years of average seasonal rainfall Mann-Kendall showed that,with the exception of autumn,the 90%significant trend tests were not passed in the remaining three quarters.It shows that the rainfall in the red soil region of southern China has a significant upward trend in autumn alone,the trend of rainfall growth in summer is not obvious,and the rainfall in spring and winter has not changed significantly during the 30 years.In space,the average annual rainfall erosivity of the 175 sites in 7 provinces of the southern red soil region from 1987 to 2016 was 5327MJ·mm·/(hm2·h·a),of which the maximum annual rainfall erosivity reached 12323 MJ·mm·/(hm2·h·a)in Dongxing weather station of Guangxi Province;the minimum value was 3051 MJ·mm·/(hm2·h·a)in Wugang weather station of Hunan Province.The CV value was 29.9%,which showed a moderate degree of spatial variability.The average rainfall erosivity in the southern red soil region during the 30 years showed a decreasing trend from southeast to northwest.It formed three distinct central high value zones in the spatial distribution,respectively,in the southern part of Guangxi province,the Pearl River Delta and the central part of Guangdong,and the northeast of Jiangxi.The spatial distribution of seasonal rainfall erosivity was generally consistent with the spatial distribution of annual rainfall erosivity,which was gradually decreasing from southeast to northwest.The spatial variation coefficient of rainfall erosivity in four seasons was 34.9%,38.7%,65.5%and 29.4%,all of which were medium variations.And the CV of rainfall erosivity in autumn were significantly higher than other seasons,which indicated that the spatial variability of rainfall erosivity force in southern red soil area in autumn was the greatest.In time,the the highest rainfall erosivity in the red soil region of southern China was 6586 MJ·mm·/(hm2·h·a)in 2016,23.6%more than the average,and the minimum value was 4020 MJ·mm·/(hm2·h·a)in 1991,24.5%less than the average.The variation coefficient of rainfall erosivity in the southern red soil region was 13.7%,which indicated that the annual rainfall erosivity in the southern red soil region showed moderate variability in time during the 30 years.Mann-Kendall Trend Test Z=2.68,the annual rainfall erosivity showed an upward trend,and the change trend was tested by the significant trend of reliability 99%,which indicated that the annual rainfall erosivity in the southern red soil region has shown a significant upward trend during the 30 years.In addition,the highest average rainfall erosivity of 175 sites was 2434 MJ·mm·/(hm2·h·a)in summer,accounting for 45.65%of the annual proportion,followed by spring 27.83%,autumn 20.04%,winter 6.48%.And the rainfall erosivity in spring and summer accounted for 73.48%of the total year.The results of multi-year average seasonal rainfall erosivity Mann-Kendall showed that the change of rainfall erosivity in the four seasons was on the rise.However,except autumn,the other three seasons were not passed 99%significant trend tests,which indicated that the rainfall erosivity in the southern red soil region has only shown a significant upward trend in autumn,while the other three seasons of the growth trend was not obvious.The spatial estimation model of rainfall erosivity based on multi-year average rainfall R=4.6158P-2259.4(R2=0.6249)and the time estimation model based on the average rainfall of space R=1.6869P1.0879(R2=0.8324)were established by correlation regression analysis of rainfall and rainfall erosivity in the red soil region of southern China during the 30 years.It’s proved that the RMSE of the spatial estimation model of rainfall erosivity is 1218.2MJ·mm·/(hm2·h·a),the NSE is 0.462,and the RMSE of the time estimation model of rainfall erosivity is 336.8MJ·mm·/(hm2·h·a),the NSE is 0.809.Thus,the spatial estimation model and time estimation model of rainfall erosivity have high accuracy.And it can meet the needs of actual and scientific research precision,which can be used as a simple algorithm for annual rainfall erosivity in the red soil region of southern China. |
PDF Full Text Request