Effects of climate change on lake area and vegetation cover over the past 55 years in Northeast Inner Mongolia grassland, China

Previous studies indicate that vegetation and lakes in arid and semi-arid regions are closely related to climate change. However, the responses of vegetation and lakes to climate-related variables have rarely been contrastively studied. The Hulun Lake (HLL) and its surrounding grasslands in Northeast Inner Mongolia of China comprise a semi-arid region that has experienced intense climate change over the last several decades. This study aims to understand the key factors that control the variations in the lake area and vegetation cover over the last five decades in the HLL region, as well as contrastively investigate the different responses of the lake area and vegetation cover to climate-related variables. Analysis results indicate that the variations with increasing change rates in the HLL area were mainly controlled by precipitation fluctuation. Moreover, an increase or decrease in the air temperature and relative humidity could affect the response time of the HLL area to the change in precipitation. The Normalized Difference Vegetation Index (NDVI) variations from 1990 to 1999 around the HLL region were mainly controlled by the intense local grazing rather than climatic restrictions. The fluctuation in NDVI after 1999 reflects the response of the vegetation cover to climatic change. The correlation analysis shows that the variations in the HLL area were closely related to the multi-year tendency of precipitation amount and to the Standardized Precipitation Evapotranspiration Index (SPEI) with a long timescale. However, NDVI is sensitive to changes in short-term precipitation amount, such as seasonal precipitation, and in SPEI with a monthly timescale. The soil moisture at shallow depths (< 0.1 m) was the key root–zone soil moisture that could influence NDVI, whereas soil moisture at depths of 0.4 to 1.0 m and the HLL area were closely related and had similar responses to climatic change.