Journal Article

Spatiotemporal Dynamics of Cropland Topsoil Organic Carbon in Changchun, China, Based on Machine Learning and Multi-Source Geospatial Data

Jingyao Xia, Yadi Yang, Haoming Li, Huiqing Wen, Mingchang Wang, Xiaoyan Li

2026

Overview

What are the main findings? * QRNN outperformed RF and XGBoost and was selected as the optimal model for SOCD prediction. * SOCD showed a non-monotonic pattern with alternating decline and recovery phases, with consistently low values in the southwest. QRNN outperformed RF and XGBoost and was selected as the optimal model for SOCD prediction. SOCD showed a non-monotonic pattern with alternating decline and recovery phases, with consistently low values in the southwest. What are the implications of the main findings? * Latitude, elevation, and mean annual temperature were stable dominant drivers of SOCD in Changchun croplands. * Multi-source data and stage-based analysis improve SOC monitoring and support sustainable soil management. Latitude, elevation, and mean annual temperature were stable dominant drivers of SOCD in Changchun croplands. Multi-source data and stage-based analysis improve SOC monitoring and support sustainable soil management. Soil organic carbon (SOC) of cropland is a key indicator of soil fertility and contributes to climate regulation and carbon storage. The understanding of SOCchanges in cropland in Northeast China still lacks high-precision long-term empirical evidence. This study is of great significance for ensuring national food security and regional sustainable development. Taking Changchun, a representative black soil region, as the study area, this study integrated 953 field samples with 19 predictors to estimate cropland soil organic carbon density (SOCD) from 2000 to 2022. The performance of quantile regression neural network (QRNN), random forest (RF), and extreme gradient boosting (XGBoost) models was compared. QRNN showed the best overall performance (R[sup.2] = 0.74, RMSE = 0.57 kg/m[sup.2], MAE = 0.40 kg/m[sup.2], and RPIQ = 2.46) and also exhibited greater stability in temporal-stage validation. Results indicated that SOCD exhibited an overall declining trend with intermittent recoveries, decreasing from 3.72 kg/m[sup.2] in 2000 to 3.36 kg/m[sup.2] in 2005, then increasing to 3.55 kg/m[sup.2] in 2010, slightly declining to 3.46 kg/m[sup.2] in 2015, and recovering to 3.63 kg/m[sup.2] in 2022. Spatially, SOCD remained low in the southwest, fluctuated markedly in the north, and was relatively stable in the central region. The analysis of the optimal parameter geographic detector (OPGD) showed that Y-latitude, elevation, and mean annual temperature (MAT) were stable dominant factors, while precipitation (PRE) and remote sensing variables showed stage-dependent effects. Interactions among multiple factors further enhanced the explanation of SOCD variations. These findings provide theoretical support for enhancing soil carbon retention and promoting long-term cropland sustainability in black soil areas.

• Publisher: MDPI AG
• Subject: Accuracy / Agricultural land / Carbon / Carbon content / Carbon sequestration / Changchun / China / Climate change / Comparative analysis / Crops / Deep learning / Electronic data processing / Elevation / Food security / Latitude / Machine learning / Methods / Neural networks / optimal parameter geographical detector / Organic carbon / quantile regression neural network / Regional development / Regional planning / Remote sensing / Soil fertility / Soil management / soil organic carbon density / Soils / Spatial data / spatiotemporal differentiation / Sustainability management / Sustainable development / Technology application / Topsoil / Trends