Drylands, comprising land regions characterized by water-limited, sparse vegetation, have commonly been projected to expand globally under climate warming. Such projections, however, rely on an atmospheric proxy for drylands, the aridity index, which has recently been shown to yield qualitatively incorrect projections of various components of the terrestrial water cycle. Here, we use an alternative index of drylands, based directly on relevant ecohydrological variables, and compare projections of both indices in Coupled Model Intercomparison Project Phase 5 climate models as well as Dynamic Global Vegetation Models. The aridity index overestimates simulated ecohydrological index changes. This divergence reflects different index sensitivities to hydroclimate change and opposite responses to the physiological effect on vegetation of increasing atmospheric CO₂. Atmospheric aridity is thus not an accurate proxy of the future extent of drylands. Despite greater uncertainties than in atmospheric projections, climate model ecohydrological projections indicate no global drylands expansion under greenhouse warming, contrary to previous claims based on atmospheric aridity.

旱地，包括以水资源有限、植被稀疏为特征的陆地区域，通常预计在气候变暖的情况下会在全球范围内扩张。然而，这种预测依赖于干旱地区的大气替代指标，即干旱指数，该指数最近被证明对陆地水循环的各个组成部分产生了质量上不正确的预测。在这里，我们使用直接基于相关生态水文变量的旱地替代指数，并比较耦合模型比对项目第 5 阶段气候模型和动态全球植被模型中两个指数的预测。干旱指数高估了模拟的生态水文指数变化。₂ . 因此，大气干旱并不是未来旱地范围的准确代表。尽管比大气预测具有更大的不确定性，但气候模型生态水文预测表明，温室变暖下全球旱地没有扩张，这与先前基于大气干旱的说法相反。