Multi‐month time‐lag effects of regional vegetation responses to precipitation in arid and semi‐arid grassland: A case study of Hulunbuir, Inner Mongolia

Abstract The 16 years of normalized difference vegetation index (NDVI) and precipitation data are used to analyze the time‐lag effects of the growing‐season NDVI response to precipitation at regional scales. This study focuses on the arid and semi‐arid Hulunbuir grassland dominated by perennials in northeast China. The multi‐month time‐lag effects are examined using simple statistical approaches, which can detect the two distinct time‐lags for four subregions with four major land‐cover types. A “positive” time‐lag effect of the growing‐season NDVI response to precipitation is observed at 1‐month (May in the current year) time‐lag and 13‐month (May in the previous year) time‐lag while a “negative” time‐lag effect is observed at 9‐month (September in the previous year) time‐lag. In addition, the prediction results of NDVI based on precipitation indicate that the NDVI prediction model considered the lagged monthly precipitation has good performance. Therefore, revealing the time‐lag effects is very important for accurately predicting the growing‐season NDVI and evaluating the vegetation dynamics.