Abstract: Soil organic carbon (SOC) density is an important soil attribute that affects food security and agricultural decision-making. Previous SOC density prediction research has mostly been based on natural environmental variables. However, in areas with frequent human activities, anthropogenic variables can also affect soil properties to some extent. This study was conducted based on random forest (RF) selection of environmental variables and four human activity variables, including population density, built-up volume, road density, and hourly anthropogenic heat flux, to explore the importance of human activities in predicting SOC density in cultivated land in the Huang-Huai-Hai Plain(HHH plain). The results indicate that environmental covariates can only explain 35% of the changes in SOC density in cultivated land. In contrast, the model with human activity and environmental variables enhanced the prediction accuracy. The R² and LCCC rose by 37.14% and 19.67%, respectively, while the MAE and RMSE decreased by 8.47% and 9.88%, respectively, suggesting better performance and prediction. This indicates that human activity variables have a significant impact on the changes in SOC density in the Huang-Huai-Hai Plain region. In the relative importance ranking, the highest daytime land surface temperature is the most important predictor, followed by the standard deviation of daytime land surface temperature, ranking second in the prediction model. Among the four human activity variables for SOC density prediction, hourly anthropogenic heat flux are the most important predictor variable, accounting for 8.25% of the importance, followed by population density, built-up volume, and road density.