Drought has become a global concern, and timely and accurate drought monitoring has an important guiding significance for the regional ecological security, economic development, and agricultural production of countries along the "New Silk Road Economic Belt" in the context of global warming. Since natural disasters have occurred frequently around the world in recent years, the drought has also posed a great threat to the global ecological environment, particularly to the long-term social and economic development of Central Asia and the surrounding countries. In this study, a Temperature Vegetation Dryness Index (TVDI) model was established using the MODIS Land Surface Temperature (LST) and Normalized Difference Vegetation Index (NDVI) in Central Asia from 2001 to 2020. Based on trend analysis, MK test, and geographic detectors, the spatial and temporal variations, of drought and drought changing patterns in Central Asia in the last 20 years, the drought evolution pattern and the influence of different factors (temperature, mean annual precipitation, land use type, population density, DEM, slope, aspect) on TVDI were also investigated. The results show that: 1) There was an obvious regional pattern in the spatial distribution of TVDI in the study area, which were mainly characterized by regional and local droughts. The frequency and degree of droughts in the southwestern part of Central Asia are than those in the northeastern part, with a decreasing trend from the southwest to the northeast. For all the study area, 66% of the area showed a significantly increasing trend of drought (P<0.05), whereas, 34% of the area showed a significant decreasing trend of drought (P<0.05), indicating a pattern of "dry in the north and wet in the south". Overall, Central Asia shows an increasing trend of aridity, with the most severest drought in the northwestern part of Kazakhstan. 2) The degree annual of drought slightly increased over the past 20 years, with the severest drought occurred in 2012. There was a significantly different spatial distribution of TVDI, which with a clear seasonal change. The drought degree was ranked as follows: summer>spring>autumn. 3) The single-factor detection results showed that the contribution of the influencing factors to drought was ranked as follows: temperature > DEM > slope > precipitation > land use type, with the controlling factors of temperature and DEM for the drought (all q values were greater than 0.45). The multi-factor analysis showed the two-factor enhancement and non-linear enhancement, with t major contribution from the four combinations, e.g., — temperature and DEM, temperature and slope, temperature and land use, and temperature and precipitation. Under the background of global climate warming and increasing human activities (vegetation deterioration, urbanization and construction, cultivation, overgrazing, industrial pollution, etc.), the air temperature gradually increased and the precipitation decreased in Central Asia over the past 20 years. These changes led to vegetation degradation, which in turn resulted in ecological degradation and increased drought degrees. This finding can provide a scientific reference for the decision-making on the drought prevention and control, particularly for the ecological risks and geo-security impacts. [ABSTRACT FROM AUTHOR]