| Soil erosion is one of the severe global ecological and environmental problems.It seriously threatens the survival and development of human beings.The management of soil erosion must be based on the quantitative assessment of regional soil erosion.The pan-third pole is the sensitive and fragile area of global climate and environmental change,but the observation and mapping data of soil erosion in this area are scarce,and cannot meet the needs of regional soil and water conservation and ecological environment construction,thus hindering the implementation of the green Belt and Road Initiative.In order to meet the demand for the data of soil erosion in the pan-third pole region,improve the method of sampling survey of soil erosion further,estimate the actual rate of soil erosion quickly and accurately,and provide a data basis for regional soil erosion survey mapping,Therefore,the sampling survey method for regional soil erosion has been studied which are based on high spatial resolution remote sensing images.The sampling design has been developed using a stratified variable probability systematic sampling method,which considered the spatio-temporal characteristics of soil erosion and conservation,and the accumulation of practices in sampling surveys of soil erosion in China and the United States.Through the visual interpretation of freely high-resolution remote sensing images,the sampling survey of land use and soil and water conservation measures in the Pan-Third Pole region was obtained.This method can provide the support for the acquisition of the accurate data rapidly in regional soil erosion survey and evaluation,moreover,the interpretation results will provide a data base for the quantitative assessment and mapping of regional soil erosion.The interpretation of 31,900 sampling survey units have been completed,and detailed information on land use and soil and water conservation measures was obtained.The interpretation results were analyzed from the following aspects: the field investigation and the accuracy of the unit information extraction,the regional representation of the unit information,and the applicability of the interpretation results.(1)Based on the visual interpretation of finer resolution remote sensing images accessible in Google Earth,sampling survey of regional soil erosion can be completed rapidly.The main contents of this sampling survey include the positioning of sampling units,delineation of the spatial scope of units,interpretation of land use and soil and water conservation measures,field investigation of interpretation results,and calculation of soil loss rate of sampling units.This sampling survey method is mainly based on the visual interpretation of remote sensing images,which can make the information’s extraction rapidly,and can meet the needs of soil erosion survey and mapping in large areas.(2)The interpretation accuracy based on the sampling survey method is more than 80%at the unit scale,therefore,detail information of land-use and soil conservation measures can be collected rapidly and accurately for a larger region.The accuracy of this method was 80.1%,81.9%,85.8%,and 88.3% in Tibet,Xinjiang,Thailand,and Pakistan respectively.This accuracy is slightly higher than the remote sensing interpretation accuracy of the China Soil and Water Loss Dynamic Monitoring Project(64%-91 %,and the area weighted average is 76.86%).For the typical first-class and second-class deviation problems in interpretation,through field inspections,prior knowledge and experience summary,and taking into account the characteristics of historical images and seasonal changes,the accuracy of interpretation can be improved further.(3)The interpretation results of sampling units based on remote sensing(Remotely Sensed Sampling Unit,RS-SU)have a good representativeness for the land use structure The 11 large sample sites were selected in the Pan-Third Pole which represents main landscapes of the region,including block soil undulated hilly in northeast China(China),loess hilly in northwest China(China),red soil hilly in southern China(China),mountain in eastern Qinghai(China),valley mountain in southeast Tibet(China),dry mountain in eastern Xinjiang(China),wet mountains in south Asia(North-Central Thailand),the Great Lakes region(Central Kazakhstan),the Middle East mountain(Central Iran),rolling plain in eastern Europe(Central Ukraine)and cold plain in Russia(Central Russia).The comparison of RS-SU with global land cover datasets(GLC1km,GLC300 m,GLC30m and GLC10m)shows that,RS-SU is more similar to higher-resolution land cover data,the HS average value between sampling survey and GLC10 for land use can reach 0.83 especially.Therefore,RS-SU can represent the structural characteristics of regional land use nicely,which can be used for quantitative evaluation of soil erosion.The difference between RS-SU and global land cover datasets are mainly in grassland,shrub,unused land and cropland.In the process of remote sensing interpretation,we have made comprehensive corrections to different types of deviation problems.(4)Water erosion loss rate for sampling unit can be calculated accurately based on the results of sampling interpretation and the regional soil erosion factor data products.The calculation results of the soil loss rate in typical unit show that,soil erosion mainly occurs in cropland and bare land,it also exists in some sparse forests and grasslands.The soil erosion intensity at the typical units based on the interpretation results is basically consistent with the literature reports,and spatial variation for a sampling unit can be represented.The average value of the potential soil erosion rate(Ap = RKLSB)is higher than the average value of the actual soil erosion rate(A = RKLSBET),and this is particularly evident in cropland.Therefore,soil loss rate calculated by the sampling survey method is closer to the actual situation on the surface.Shortcomings of the traditional remote sensing mapping,especially that just represent the potential of soil erosion,can be overcome.So the approach is hopefully can be a large area or even a “global soil erosion sampling survey method” and provide valid support to global soil erosion mapping after some necessary modification. |
PDF Full Text Request