We use a numerical dynamic soil–landscape model to study one aspect of the spatio-temporal soil–landscape evolution process, the effect of climatic fluctuations on soil grading distribution in space and time in response to the interplay between physical weathering and surface erosion (soil mineralogical fluxes). We simulate a synthetic soil–landscape system over the middle and late Quaternary (last 400 ka). The results show that (1) soil–landscape response to climate change is non-linear and highly spatially variable, even at hillslope scale; and (2) soil–landscape adjustment to climate change can lag tens of thousands of years and is both spatially and temporally variable. We propose that the legacy of past climatic condition (i.e. last glacial maximum) in modern soil–landscape systems vary considerably in space. This implies that the spatiotemporal uniformity in which soil is typically described in Earth system modeling and analysis (e.g. carbon cycle) grossly underestimates their actual complexity.
