Forest and non-forest demonstrate comparable biophysical temperature effects but distinct asymmetry patterns during cropland conversion

Cropland changes impact surface properties and influence land surface temperature (LST), yet uncertainties persist due to their fragmented pattern. This study examines the actual biophysical LST effect resulting from cropland conversions with well-studied forests and less-explored non-forests. Analyzing data from 2003 to 2019 for global croplands undergoing fractional changes (+1.3%/4y), we find that the biophysical LST effects (-0.01℃/4y in summer; +0.002℃/4y in winter) account for an average of -55.6% and +1.4%, respectively, of the satellite-observed total LST trends in summer (+0.018℃/4y) and winter (+0.138℃/4y). Cropland conversions with forest vegetation (FV) and barren land (BAR) exhibit comparable summer cooling (-0.005℃/4y and -0.004℃/4y), while those with FV and non-forest vegetation (NFV) result in winter warming (+0.002℃/4y and +0.001℃/4y). Notably, cropland conversion with BAR shows the most pronounced LST sensitivity, followed by FV, NFV and wetland (WET). The sensitivity magnitude of cropland degradation to FV surpasses that of cropland expansion, revealing reversed asymmetry patterns for conversions with NFV, BAR, and WET. The findings caution against incorporating all types of natural landcovers in cropland conversions for land-based mitigation.