Comparison of infiltration in hydrological soil groups along the Soutpansberg Mountain

Abstract

Infiltration is a key hydrological process that controls the movement of water from the soil surface to the vadose zone and the groundwater system, thus influencing surface runoff, soil moisture availability, erosion processes, and groundwater recharge. Accurate estimation of infiltration rates is crucial for effective management of water resources, particularly in mountain regions characterised by complex terrain, heterogeneous soils, and various uses of land. The Soutpansberg Mountain Range in Limpopo Province, South Africa, is a designated Strategic Water Source Area and plays a critical role in regional water security. Despite this importance, limited site-specific studies have evaluated the performance of infiltration models in different hydrological soil groups within the region. This study aimed to investigate and compare the infiltration rates measured in the field with those simulated using commonly applied infiltration models across representative soil textures along the Soutpansberg Mountain Range. Field infiltration tests were conducted at seven selected sites representing clay, sandy, sandy clay, sandy loam, sandy clay loam, and clay loam soils. Infiltration rates were measured using the Cornell Sprinkle Infiltrometer, while soil texture classification and physical properties were determined by mechanical sieve analysis and interpretation using the USDA soil texture triangle. The Horton, Kostiakov, and Green–Ampt infiltration models were applied to simulate infiltration behaviour for each soil type, and model parameters were calibrated using observed field data. The performance of the model performance was evaluated using statistical indicators, including the Root Mean Square Error, the Mean Absolute Error, and the Mean Bias Error, to assess the level of agreement between the observed and simulated infiltration rates. The results revealed marked spatial variability in the infiltration rates between soil textures and hydrological soil groups, reflecting differences in soil structure, pore connectivity, and hydraulic properties. The performance of the model varied significantly with the type of soil, indicating that the suitability of the infiltration model is strongly dependent on the texture of the soil and the site-specific conditions, and no single model performed optimally on all soils. The study provides an improved understanding of infiltration processes in fractured mountainous environments and offers practical guidance for selecting appropriate infiltration models for hydrological assessments. The findings are relevant for catchment hydrological modelling, land use planning, groundwater protection, and sustainable water resource management in the Soutpansberg Mountain Range and similar settings.