Long-range transport of black carbon from South Asia to the Tibetan plateau and its deposition on glaciers directly enhances glacier melt. Here we find South Asian black carbon also has an indirect effect on the plateau’s glaciers shrinkage by acting to reduce the water supply over the southern Tibetan plateau. Black carbon enhances vertical convection and cloud condensation, which results in water vapor depletion over the Indian subcontinent that is the main moisture flux source for the southern Tibetan plateau. Increasing concentrations of black carbon causes a decrease in summer precipitation over the southern Tibetan plateau, resulting in 11.0% glacier deficit mass balance on average from 2007 to 2016; this loss rises to 22.1% in the Himalayas. The direct (accelerated melt) and indirect (mass supply decrease) effects of black carbon are driving the glacial mass decline of the so-called “Asian Water Tower”.

As a typical arid and semi-arid area, central Asia (CA) has scarce water resources and fragile ecosystems that are particularly sensitive and vulnerable to climate change. In this study, dynamic downscaling was conducted to produce a regional dataset that incorporated the time period 1986–2100 for the CA. The results show that dynamic downscaling significantly improves the simulation for the mean and extreme climate over the CA, compared to the driving CCSM4 model. We show that significant warming will occur over CA with 2.0 °C and 5.0 °C increasing under the RCP4.5 and RCP8.5 scenarios, respectively by the end of twenty-first century. The daily maximum temperature, the daily minimum temperature and the annual total number of days with a minimum temperature greater than 25 °C will also increase significantly. The annual total number of days with a minimum temperature less than 0 °C will decrease significantly. Long-term trends in the projected winter precipitation under different emission scenarios exhibit robust and increasing changes during the twenty-first century, especially under the RCP8.5 scenario with an increasing about 0.1 mm/day. Significant differences are shown in the projection of precipitation-related indices over CA under different emission scenarios, and the impact of emissions is apparent for the number of days with ≥ 10 mm of precipitation, the density of precipitation on days with ≥ 1 mm of precipitation, and particularly for the maximum consecutive number of dry days that will increase significantly under the RCP8.5 scenario. Therefore, reduced greenhouse gases emissions have implications for mitigating extreme drought events over the CA in the future.

Due to the substantial domain gaps in Remote Sensing (RS) images that are characterized by variabilities such as location, wavelength, and sensor type, Remote Sensing Domain Generalization (RSDG) has emerged as a critical and valuable research frontier, focusing on developing models that generalize effectively across diverse scenarios. However, research in this area remains underexplored: (1) Current cross-domain methods primarily focus on Domain Adaptation (DA), which adapts models to predefined domains rather than to unseen ones; (2) Few studies target the RSDG issue, especially for semantic segmentation tasks. Existing related models are developed for specific unknown domains, struggling with issues of underfitting on other unseen scenarios; (3) Existing RS foundation models tend to prioritize in-domain performance over cross-domain generalization. To this end, we introduce the first vision foundation model for RSDG semantic segmentation, CrossEarth. CrossEarth demonstrates strong cross-domain generalization through a specially designed data-level Earth-Style Injection pipeline and a model-level Multi-Task Training pipeline. In addition, for the semantic segmentation task, we have curated an RSDG benchmark comprising 32 semantic segmentation scenarios across various regions, spectral bands, platforms, and climates, providing comprehensive evaluations of the generalizability of future RSDG models. Extensive experiments on this collection demonstrate the superiority of CrossEarth over existing state-of-the-art methods.