Assessing the effects of land cover configuration on climate with a regional climate model and stylized landscapes

作者：Marco Ferro, Thomas Heckelei, Silke Hüttel, Stefan Kollet, Stefan Poll, Jan Boerner · 发表于：Environmental Research Climate · 年份：2026 · DOI：10.1088/2752-5295/ae71fd · 研究领域：Remote Sensing in Agriculture、Plant Water Relations and Carbon Dynamics、Species Distribution and Climate Change

Abstract Land use and land cover change affects two climate-relevant landscape dimensions: land cover composition and its spatial configuration. While the climate impacts of changes in land cover composition are well documented, the role of changes in land cover configuration remains poorly understood. Here, we examine how configuration changes in cropland-forest landscapes affect the ecosystem´s climate-regulating functions by conducting a multi-scenario experiment. We use the Terrestrial System Modeling Platform (TSMP), a modeling framework to run coupled high-resolution regional climate models (RCMs) with a focus on terrestrial hydrology. We employ stylized landscapes as input for TSMP and combine machine learning with SHapley Additive exPlanations values, an explanatory AI technique, to disentangle direct and mediating contributions. Our results indicate that changes in land cover configuration influence topsoil and 2 m air temperatures when composition is kept constant. Greater forest fragmentation is associated with overall warming, particularly over cropland during the growing season. This can mainly be attributed to reduced wind speed, lower latent heat flux, and higher sensible heat flux. These findings support the hypothesis that increasing forest fragmentation reduces the landscape’s capacity to mitigate local climate. However, we also discuss potential biases towards warming effects in contemporary RCMs. Our study demonstrates the potential of RCMs to address key ...