Since the "two-carbon goal" was proposed, the research on the spatial-temporal variation characteristics of land use carbon emissions is of guiding significance for the formulation of carbon emission reduction policies. Based on FLUS and grey prediction model, land use carbon emissions in 2030 were predicted in Urumqi, and the spatial-temporal pattern of carbon emissions from 2000 to 2030 was simulated by combining with previous years. The results were as follows: (1) During the land use change in the study area from 2000 to 2020, the forest land, unused land, water area and construction land showed positive growth, while the arable land and grassland showed negative growth. The area of grassland and cultivated land was mainly transferred to unused land and construction land, and the area change of water area and woodland was relatively stable. (2) The total carbon emissions in the study area showed an increasing trend from 2000 to 2020, and construction land was the main determinant. The spatial distribution of carbon emissions in the study area showed an increasing trend from south to north, with the highest carbon emissions in Midong District and the lowest carbon emissions in Daban District. (3) In the calculation results of carbon emissions under different scenarios in 2030, the total carbon emissions under cultivated land protection scenario were the highest (27.736 7 million t), while the carbon emissions under ecological environment scenario were the lowest (27.671 5 million t). (4) From 2000 to 2030, most areas of the study area tended to be stable, with large aggregation and sporadic distribution of the changing areas. The increasing area of carbon emission was mainly distributed in the central area of the study area and Midong District in the north, while the decreasing area was mainly distributed in Urumqi County and Dabancheng District in the south. The area with the largest increase in carbon emission was in the cultivated land protection scenario, accounting for 25.4% of the study area. The area with the largest decrease in carbon emission was in the ecological environment scenario, accounting for 13.3% of the study area.