This study uses the Carnegie-Ames-Stanford approach model to simulate and explore the spa-tiotemporal pattern of net primary productivity in the grassland ecosystem at the county scale in Qinghai Province from 1982 to 2018.It simulated and analyzed the productivity and livestock carrying capacity of the grassland ecosystem,selected theoretical livestock carrying capacity as an evaluation index,and combined it with actual livestock carrying capacity to determine the degree of livestock overload and its changing characteristics.The results indicated that the overall carrying capacity of grassland ecosystems at the county level in Qinghai Province showed a pattern of being high in the east and low in the west.Compared with 1982,the overall carrying capacity of grassland ecosystems in there showed an increas-ing trend in 1990,2000,2010,and 2018,with the annual average theoretical livestock carrying capacity per unit area increasing from 0.420 SHU/hm2 to 0.456,0.497,0.570,and 0.555 SHU/hm2.From 1982 to 2018,the overall balance of forage supply and livestock requirement had been in an overloaded state,and as of 2018,the overall overload was 34.5%.There are 24 counties with severe overloading in the red warning zone,and it is urgent to adjust their actual livestock carrying capacity in a timely manner.
grassland ecosystemtheoretical livestock carrying capacityactual livestock carrying ca-pacityCASA model