Utilizing the NOAA/AVHRR data of July and Auguest of 2006, according to the temperature vegetation dryness index(TVDI), the Sichuan hot-drought in 2006 has been monitored and estimated. The article adopted a new filtering-cloud composing technology used to monitoring drought,analyzed the property of Ts—NDVI characteristic space and the TVDI's indicating drought ability, put forward the TVDI compute model, grade standard, and the impact estimation method. The result indicated: 1) the decade-composed data obtained by the maximum surface temperature principle has better ability than the composed data obtained by the maximum NDVI on monitoring drought; 2) in the Ts-NDVI characteristic space,the dry side and the wet side almost are horizontal lines when the NDVI is on the small side, the temperature difference is about 45℃; 3) the TVDI has inequable level on monitoring drought as its value, the less TVDI indicates without drought occurrence, the bigger TVDI shows the drought certainly happened, the unsure drought conclusion mainly comes from the middle TVDI; 4) the monitored results of Sichuan hot-drought in 2006, obtained by the remote sensing and by the climate index, are basically consistent, the estimating error on the bearing drought crop area and on the drinking trouble population is within 10%.