Comparison of interpolation calculation of soil carbon storage and its influencing factors in Horqin Sandy Land

Accurate estimation of soil carbon stocks is crucial to the realization of carbon neutrality and response to climate change. This study used different interpretation methods including inverse distance weighting (IDW), ordinary kriging (OK), and regression kriging (RK) to interpolate soil organic carbon stock and calculate regional soil organic carbon storage based on the organic carbon data of 310 sites in the 0－100 cm soil layer in Horqin Sandy Land, which is located in the eastern end of the semi-arid farming-pastoral ecotone in the north of our country. Mean error (ME), mean normalized error (MSE), and root-mean-square error (RMSE) were used to evaluate the interpolation results, and the optimal interpolation method for soil carbon storage calculation was verified with measured SOCS. The results showed that: 1) in the estimation of SOCS in semi-arid sandy land, the environmental variables that enter the regression prediction model are annual accumulated temperature, precipitation, sand content, and altitude in order of influence degree from large to small. SOCS showed significant negative correlation with altitude, annual accumulated temperature, precipitation, and sand content. 2) the soil organic carbon stocks calculated by different interpolation methods were significantly different. The soil organic carbon stocks calculated by RK, IDW, and OK were 150.34, 158.03, and 140.11 Tg, respectively. Compared with the measured values, OK and IDW underestimated and overestimated soil organic carbon stock, highlighting the importance of the interpolation method in the estimation of soil organic carbon stock at the regional scale. 3) the results based on RK interpolation were optimal (ME = 0.208, MSE = 1.211, RMSE = 1.973, R² = 0.370 6) and improved the interpolation accuracy by 6.0% and 22.4% compared with OK and IDW, respectively. This study shows that RK interpolation in semi-arid sandy land can still improve the accuracy of interpolation, leading to more accurate estimation of soil organic carbon storage.