No abstract

Widespread aridification of the land surface causes substantial environmental challenges and is generally well documented. However, the mechanisms underlying increased aridity remain relatively underexplored. Here, we investigated the anthropogenic and natural factors affecting long-term global aridity changes using multisource observation-based aridity index, factorial simulations from the Coupled Model Intercomparison Project phase 6 (CMIP6), and rigorous detection and attribution (D&A) methods. Our study found that anthropogenic forcings, mainly rising greenhouse gas emissions (GHGE) and aerosols, caused the increased aridification of the globe and each hemisphere with high statistical confidence for 1965–2014; the GHGE contributed to drying trends, whereas the aerosol emissions led to wetting tendencies; moreover, the bias-corrected CMIP6 future aridity index based on the scaling factors from optimal D&A demonstrated greater aridification than the original simulations. These findings highlight the dominant role of human effects on increasing aridification at broad spatial scales, implying future reductions in aridity will rely primarily on the GHGE mitigation.

This study investigates reference evapotranspiration (ET0) trends in China from 1960 to 2012 based on the Penman–Monteith equation and gridded meteorological measurements. Under the combined impacts of factors influencing ET0 (i.e., net radiation [RN], mean temperature [TAVE], vapour pressure deficit [VPD], and wind speed [WND]), both seasonal and annual ET0 for the whole China and more than half of the grids decreased over the past 53 years. The attribution analyses suggest that for the whole China, the WND is responsible for annual and seasonal ET0 decreases (excluding summer, where RN is responsible). Across China, the annual cause of WND with the largest spatial extent (43.1% of grids) mainly derives from north of the Changjiang River Basin (CJRB), whereas VPD (RN) as a cause is dispersedly distributed (within and to the south of the CJRB). In summer, RN is dominant in more than half of the grids, but the dominance of VPD and WND accounts for approximately 90% of grids during the remaining seasons. Finally, the correlation coefficients between ET0 and the Atlantic Oscillation (AO), North AO, Indian Ocean Dipole (IOD), Pacific Decadal Oscillation (PDO), and El Niño Southern Oscillation (ENSO) indices with different lead times are calculated. For the whole China, annual and seasonal ET0 always significantly correlate with these indices (excluding the IOD) but with varied lead times. Additionally, near half of the grids show significant and maximum (i.e., the largest one between ET0 and a certain index with a lead time of 0–3 seasons) correlation coefficients of ET0 with PDO in spring and summer, ENSO in autumn, and AO in winter. This study is not only significant for understanding ET0 changes, but it also provides preliminary and fundamental reference information for ET0 prediction.

Given the key roles of the Indo-China Peninsula (ICP) in weather and climate systems, hydrometeorology, and ecology, the annual and monthly changes in the Food and Agriculture Organization-56 Penman-Monteith reference evapotranspiration (ET 0 ), which was calculated based on the Climatic Research Unit datasets, were investigated in ICP during 1961-2017. The annual ICP ET 0 significantly (p < .05) increased, with different increasing tendencies in most months. In particular, larger and more significant (p < .05) ET 0 values were found during October-December. The annual and monthly ET 0 changes showed evident spatial differences, characterized by increases in more than 50% of the ICP area except for decreases in around 70% of that area during March-May. A sensitivity experiment-based separation method was utilized to evaluate the contribution of each influential factor, and the corresponding determinants were identified by comparing the contributions. Results showed that the annual ICP ET 0 increase was attributed to the increased vapour pressure deficit (Vpd). However, the annual determinants varied spatially, with net solar radiation (Rn) in the southern region of ICP, wind speed (Wnd) in the northeast, and Vpd in the remaining regions. The monthly ICP determinant was Wnd in January, March-May and December, and Vpd for the remaining months. Despite different spatial patterns of monthly dominants, Vpd and Wnd were the determinants with the most extensive distributions over ICP (>75% of ICP in total). The results of this study can significantly fulfil the research gap regarding the ICP ET 0 changes and the underlying mechanisms.