Enhancing Soil Carbon Storage: Developing high-resolution maps of topsoil organic carbon sequestration potential in Taiwan

Abstract

Developing methods to estimate Soil Organic Carbon sequestration potential (SOC_sp) at the regional scale is essential for quantifying the additional stable carbon that soils can sequester for climate change mitigation. This study investigates methods to estimate SOC_sp across Taiwan. It evaluates three methods for estimating SOC saturation in fine fraction soils (less than 53 µm) (Hassink, boundary line, and quantile regression) using high-density sampling (more than one site per square km in cultivated areas) and digital soil mapping. Results show that SOC_sp derived from the boundary line approach was about 2.1 times higher than those estimated by the quantile regression (QR) approach. Estimates using the QR approach could be regarded as the technically achievable goal as most of the samples collected from cultivated soils had low SOC content. A moderate spatial model for SOC_sp was achieved using random forest model (R²: 0.54–0.57) on validation samples. Soil type, temperature and elevation were important controlling factors of SOC_sp in topsoil. Regions with high SOC_sp were mainly located in intensively cultivated and clay-rich soils, with paddy soils exhibiting greater sequestration potential compared to other land uses. In contrast, most forest soils in slope and mountainous areas had low SOC_sp, as their SOC levels were near or above estimated saturation in the fine fraction soils. This study offers practical and strategic insights for scientific and policy-oriented efforts to improve soil health, boost carbon sequestration, and address climate change through optimisation of land use.