Revisiting assessments of ecosystem drought recovery

Abstract The time taken for ecosystems to recover from drought (drought recovery time) is critically important for the ecosystem state. However, recent literature presents contradictory conclusions on this feature: one study concludes that drought recovery time in the tropics and high northern latitudes is shortest (<4 months) but another concludes that it is longest (>12 months) in these regions. Here we explore the reasons for these contradictory results and revisit assessments of drought recovery time. We find that the study period, drought identification method and recovery level definition are main factors contributing to the contradictory conclusions. Further, we emphasize that including droughts that did not decrease ecosystem production or using a period of abnormal water availability to define ecosystem recovery level can strongly bias drought recovery time estimates. Based on our refined methods, we find the drought recovery time is also longest in some tropical regions but not in high northern latitudes during 1901–2010. Our study helps to resolve the recent controversy and provides insight for future drought recovery assessments.