Aridity regulates the terrestrial ecosystem productivity in water-limited regions. The aridity index (AI) is often defined as the ratio of annual precipitation to annual potential evapotranspiration. However, how the drying and wetting influence the vegetation response and its characteristic along aridity gradient remains unclear. Here, we examined trends of the AI and normalized difference vegetation index (NDVI) in the drylands of East Asia from 1982 to 2015, which denoted the drying and wetting trends and vegetation response, respectively. The results show that the variability in land area and AI from 1982–2015 was lower in the whole dryland than its subtypes including hyper-arid, arid, semi-arid, and dry sub-humid regions. Drying and wetting trends were observed in each AI interval along a spatial aridity gradient. Wetting trends are prevalent owing to their area and magnitude were twice those of drying trends. Spatial variation of aridity shaped the pattern of trends in vegetation response to drying and wetting trends; drier regions had smaller and narrower ranges of variation in NDVI trends relative to wetter regions. A shift in AI trends and NDVI trends along the spatial aridity gradient occurred at 0.4 of AI. Distinct patterns of vegetation response to aridity change were observed across the aridity threshold, and the transition region was identified in the studied drylands. The results suggest that changes in the subtypes might be masked by the entire drylands and then leading to failure in recognizing the transformation of the subtypes. This implies that terrestrial carbon storage variability prediction should consider the spatial aridity changes to avoid the uncertainties due to the divergent vegetation response to AI trends at different aridity levels.

干旱调节了缺水地区的陆地生态系统生产力。干旱指数（AI）通常定义为年降水量与年潜在蒸散量之比。然而，干燥和润湿如何影响植物响应及其沿干旱梯度的特征尚不清楚。在这里，我们研究了1982年至2015年东亚干旱地区的AI和归一化植被指数（NDVI）趋势，分别表示了干旱和湿润趋势以及植被响应。结果表明，从1982年至2015年，整个旱地的土地面积和AI的变异性均低于其亚型，包括高干旱，干旱，半干旱和亚湿润干旱地区。在每个AI间隔沿空间干旱梯度观察到干燥和润湿趋势。由于其面积和幅度是干燥趋势的两倍，所以润湿趋势是普遍的。干旱的空间变化塑造了植被对干燥和湿润趋势的响应模式。相对较湿的地区，较干燥的地区的NDVI趋势变化范围越来越小。AI趋势和NDVI趋势随空间干旱梯度的变化发生在AI的0.4处。在整个干旱阈值上观察到了不同的植被对干旱变化的响应模式，并在研究的干旱地区确定了过渡区。结果表明，亚型的变化可能被整个旱地掩盖，然后导致无法识别亚型的转化。