Old-growth forests show low canopy resilience to droughts at the southern edge of the taiga.

Forest mortality and resilience driven by drought disturbances have attracted tons of attention. However, the acquisition of continuous spatial-temporal data is generally enslaved to the conventional field investigations. In this study, the resilience of semiarid forest was characterized with canopy dynamics from remote sensing observations, combining the variations in canopy greenness and water content. We integrated dense normalized difference vegetation index (NDVI) and normalized difference infrared index (NDII) time series from Landsat datasets, intending to assess the canopy resilience in 24 conifer patches along a climatic aridity gradient at the southern edge of the taiga in northern Mongolia and southern Siberia of Russia. The results exhibited four patterns of coordinated NDVI-NDII variation trends, indicating that the canopy water content of coniferous forests may decrease at first during a drought period, and sustained water loss may, in turn, induce an accompanying reduction in canopy greenness. Meanwhile, the patches with canopy recovery growth after initial declines were considered to have resilience to climate change. We further observed the combined effects of aridity degree and tree age on canopy resilience, and all seven patches with no resilience corresponded to the old-tree group (the oldest trees reached or exceeded the age of 90). The observations indicated that the old-growth forests in semiarid regions were less likely to show canopy resilience, which corresponded to a higher risk of sustained decline.