Abstract. The denudation history of active orogens is often interpreted in the context of modern climate gradients. Here we address the validity of this approach and ask what are the spatial and temporal variations in palaeoclimate for a latitudinally diverse range of active orogens? We do this using high-resolution (T159, ca. 80 × 80 km at the Equator) palaeoclimate simulations from the ECHAM5 global atmospheric general circulation model and a statistical cluster analysis of climate over different orogens (Andes, Himalayas, SE Alaska, Pacific NW USA). Time periods and boundary conditions considered include the Pliocene (PLIO, ∼ 3 Ma), the Last Glacial Maximum (LGM, ∼ 21 ka), mid-Holocene (MH, ∼ 6 ka), and pre-industrial (PI, reference year 1850). The regional simulated climates of each orogen are described by means of cluster analyses based on the variability in precipitation, 2 m air temperature, the intra-annual amplitude of these values, and monsoonal wind speeds where appropriate. Results indicate the largest differences in the PI climate existed for the LGM and PLIO climates in the form of widespread cooling and reduced precipitation in the LGM and warming and enhanced precipitation during the PLIO. The LGM climate shows the largest deviation in annual precipitation from the PI climate and shows enhanced precipitation in the temperate Andes and coastal regions for both SE Alaska and the US Pacific Northwest. Furthermore, LGM precipitation is reduced in the western Himalayas and enhanced in the eastern Himalayas, resulting in a shift of the wettest regional climates eastward along the orogen. The cluster-analysis results also suggest more climatic variability across latitudes east of the Andes in the PLIO climate than in other time slice experiments conducted here. Taken together, these results highlight significant changes in late Cenozoic regional climatology over the last ∼ 3 Myr. Comparison of simulated climate with proxy-based reconstructions for the MH and LGM reveal satisfactory to good performance of the model in reproducing precipitation changes, although in some cases discrepancies between neighbouring proxy observations highlight contradictions between proxy observations themselves. Finally, we document regions where the largest magnitudes of late Cenozoic changes in precipitation and temperature occur and offer the highest potential for future observational studies that quantify the impact of climate change on denudation and weathering rates.

LGM mean annual precipitation is 200-600 mm/yr lower over on the Tibetan Plateau. Model and proxy data comparison shows a good agreement for the LGM, but large differences for the MH. Large differences are also present between MH proxy studies near each other. The precipitation weighted annual mean δ 18 O p lapse rate at the Himalaya is about 0.4h/km larger during the MH and 0.2h/km smaller during the LGM than during the PI. Finally, rainfall associated with the continental Indian monsoon (between 70 • E-110 • E and 10 • N-30 • N) is about 44% less in the LGM than during PI times. The LGM monsoon period is about one month shorter than in PI times. Taken together, these results document significant spatial and temporal changes in temperature, precipitation, and δ 18 O p over the last ∼21 ka.These changes are large enough to impact interpretations of proxy data and the intensity of the Indian monsoon.