Objective The purpose of this study was to explore land use recessive morphology and its impact on land use carbon emissions, to provide practical guidance for spatial optimization of land use based on the "dual carbon" goals.

Method The Middle reaches of Yangtze River was taken as the research area, land use recessive form indicators were selected, spatial-temporal variation characteristics of land use recessive morphology and land use carbon emission were analyzed using carbon emission coefficient method and geographic detector model, based on land use data from 2010 to 2020, and revealed the correlations of land use recessive morphology and carbon emission.

Result ① The level of land use recessive morphology in the Middle Reaches of Yangtze River continued to increase, and the spatial pattern evolved from the "Middle triangle" with single provincial capital to the "Middle three-axis" with connected provincial capital. ② Construction land was the most important carbon source in the study area, accounting for more than 96% of carbon emissions. Forest land was the main source of carbon sink, accounting for 89% of carbon absorption capacity, and the net carbon emission from land use showed a rising trend, with an annual growth rate of 11.56%. High value areas of carbon emissions were mainly concentrated in Wuhan, Nanchang, Jingmen, Yichang, Jiujiang cities along the Yangtze River and Pingxiang, Xinyu, Xiangtan cities in the old industrial base. ③Geographic detector showed that the output value of secondary industry, the output value of tertiary industry, the average power consumption of land and the urbanization rate of land were the key driving factors affecting the net carbon emissions of land use. The q value of driving factors was above 0.6.

Conclusion The hidden recessive morphology of land use increased land use carbon emissions. The spatial distribution of net carbon emissions was formed by the interaction of multiple factors, and the urbanization rate of land had the highest explanatory power for carbon emissions in each index layer.