Abstract: Drought is one of the most serious natural disasters with the widest distribution, the highest frequency, and the greatest economic losses in the world. A grave threat has been posed to agricultural production, food security, and socio-economic development. It is of great significance to timely and accurately monitor the drought at the regional scale. In this study, the NDVI-LST feature space was constructed using the MOD11A2 surface temperature 8-d synthetic product and MOD13A1 vegetation index 16-d synthetic product. The data was collected from the land surface temperature (LST) and normalized difference vegetation index (NDVI). Temperature vegetation dryness index (TVDI) was captured from the Guanzhong areas in a 20-year time series from 2001 to 2020. The spatial and temporal distribution of drought was analyzed to determine the causes and future trends of natural disasters. The results show that: 1) It was feasible to carry out drought monitoring on a regional scale using the TVDI remote sensing index. The average TVDI was 0.57 over the past 20 years. The drought type was in a normal state in general. The highest annual average TVDI was found in 2013 and the lowest in 2006. The average TVDI value fluctuated greatly during the 20-year period. Drought also shared strong regional characteristics, indicating a spatial distribution pattern of wet in the southwest, dry in the center and northeast, and normal in the north. 2) The spatial and temporal trends of drought showed that 82.00% of the areas had relatively stable droughts, and only 8.05% of the areas were with improved droughts. As such, there was a relatively small variation in the drought condition; Combined with the Hurst index, the future trend of drought was characterized by a relatively small amount of change. There was an anti-sustainable trend of drought in the future; Combined Hurst index and TVDI trend, 7.71% of the study area was continuously wet in the future, and 36.24% of the area was changed from drought to wet. 3) The TVDI value was positively correlated with the air temperature, while negatively correlated with the precipitation. There was a similar spatial distribution pattern of the 20-year average TVDI and the annual average air temperature. The air temperature and precipitation shared a similar distribution in the spatial pattern. The degree of drought depended mainly on the air temperature and precipitation. Meanwhile, the degree of aridity was negatively correlated with both elevation and slope. The arid and more arid types were mainly distributed in the central plains with low elevations and small slopes. While the moist and more moist types occupied a larger area in the mountainous areas south of the Qinling Mountains with high elevation. The TVDI values varied with months in the different land use types in the study area. The cropland and construction land presented the highest TVDI values, with the annual mean values of 0.76 and 0.75, respectively, whereas, the woodland was the smallest TVDI value, with an annual mean value of 0.45. Therefore, the woodland and grassland cover can be expected to positively mitigate the drought. The TVDI can be used to quantify the drought conditions from the regional scale. Remote sensing drought monitoring can also provide a theoretical foundation and scientific basis for the formulation of drought countermeasures in Guanzhong areas.