Remote Sensing Rapid Identification And Driving Factors Analysis Of Grassland Resource Change Information In Xilingol

Grassland is an essential natural resource,and it plays an irreplaceable role in developing the animal husbandry economy and maintaining the stability of the biosphere.Therefore,it is of great significance to investigate the quantitative,qualitative,and spatial distribution characteristics of grassland resources and understand the coupling relationship between grassland and its external driving forces for the scientific management of grassland resources and animal husbandry's stable development.Selecting Xilingol grassland as a research area,this study used nearly 40 years of multi-source remote sensing data in combination with the ground survey data to rapidly extract the grassland types,grassland spatial pattern,and grassland productivity on a large scale by using the geospatial statistical methods.Quantitative analysis of the external driving factors leading to the change of the above characteristics provides a scientific basis for the rational utilization of grassland resources in Xilingol.The main conclusions are as follows:(1)According to the Chinese grassland classification system and the actual situation of Xilingol grassland and the requirements of remote sensing classification,a large-scale classification system for grassland remote sensing suitable for medium resolution images was built by combining the grassland types with similar habitat conditions,same constructive species,and same biotypes.The results showed that the average accuracy of this classification system for the object-oriented and random forest classification algorithm reached over 84%.It met the requirements of large-scale grassland remote sensing rapid classification.Regarding the importance of classification features,spectral features played a significant role in distinguishing different grassland types,followed by location features.(2)During the 1980s to 2010s,the area,fragmentation degree,and location of the main grassland types in Xilingol grassland all significantly changed.First,the Artemisia frigida steppe increased by14,900 km²,followed by the Cleistogenes squarrosa steppe,which grew by 2,800 km².The Leymus chinensis steppe decreased by 10,800 km².In terms of landscape change,each grassland type's patch number and patch density increased significantly,which indicates that both the number of trivial patches and the fragmentation degree of each grassland type increased.In terms of spatial shift,Stipa baicalensis steppe,Filifolium sibiricum steppe,Leymus chinensis steppe,Artemisia frigida steppe,and Stipa klemenzii steppe has been shifting east.In contrast,the Stipa steppe and Cleistogenes squarrosa steppe have been moving west.The driving forces of grassland type change in Xilingol in the past 40 years were:Annual total precipitation>Livestock numbers change>Changes in the GDP of the primary industry>Population change.It indicated that water condition and overgrazing were the main driving forces of grassland type change.(3)In 1988,1998,2008 and 2018,the grassland area of Xilingol was 176,600 km²,176,660 km²,176,300 km² and 173,200 km² respectively,accounting for nearly 86.00%of the total area.From 1988 to2018,approximately 11,000 km² of land was transferred from grassland,while about 8,200 km² was transferred into grassland,2800 km² was lost.Sandy land,saline-alkali land,and cultivated land were the primary transferred forms from grassland,accounting for nearly 44.57%,20.69%,and 19.39%of the total transferred area from grassland.In comparison,the transferred area into grassland from sandy land and saline-alkali land accounted for 44.22%and 11.46%of the total transferred area into grassland.It indicated that although it had a remarkable effect on controlling the sandy land and saline-alkali land,the sandy land expansion and grassland salinization is still severe.The driving factors of change in grassland spatial pattern in Xilingol from 1988 to 2018 were:Population changes>Changes in the GDP of the primary industry>Changes in the GDP of the tertiary industry>Annual total precipitation>Annual average temperature>Changes in the GDP of the secondary industry.It suggests that the driving effect of human activities on the change in grassland spatial pattern was greater than that of the natural environment.(4)The average annual Net Primary Productivity(NPP)of Xilingol grassland from 1982 to 2018was 251.13 g·C·m^-2·a^-1,which showed an increasing spatial distribution from west to east.The average NPP of meadow steppe,typical steppe,and desert steppe were 374.15,255.38,and 153.37 g·C·m^-2·a^-1.Among them,the average annual NPP was the highest in Chrysomellia sinensis steppe(423.35 g·C·m^-2·a^-1),and the lowest in Stipa klemenzii steppe(151.43 g·C·m^-2·a^-1).From 1982 to 2018,the NPP of Xilingol grassland showed a slight downward trend with an interannual variation rate of-0.42 g·C·m^-2·a^-1.The NPP of meadow grassland showed an upward trend with an interannual variation rate of 0.26g·C·m^-2·a^-1.The NPP of typical steppe and desert steppe showed a downward trend,with interannual variation rates of-0.59 g·C·m^-2·a^-1 and-0.48 g·C·m^-2·a^-1,respectively.From 1982 to 2018,the interannual change of precipitation was the dominant factor affecting the grassland productivity change in Xilingol.