Assessing spatial variability of soil water content through thermal inertia and NDVI

AVHRR (Advanced Very High Resolution Radiometer on board NOAA satellites) data are considered here to evaluate the possibility of using the surface temperature as an indicator of the soil/canopy water content at the short time-scale. This is obtained by means of an indirect approach based on a simplified soil-atmosphere energy balance. The techniques provide sufficiently detailed coverage of the processes in terms of the time and spatial scale with respect to hydrological applications. Two different approaches have been tried: the first based on thermal inertia measurements (Xue & Cracknell, 1995)(1) through ATI (Apparent Thermal Inertia), the second based on surface temperature (LST) and vegetation indices (NDVI), with the TVDI (Temperature Vegetation Dryness Index) suggested by Sandholt et al. (2002)(2). Both techniques were used in detecting moist areas in a single image (or day/night images), and in multitemporal multitemporal applications. In particular, a new cloud detection algorithm, based on the bimodal frequency distribution of the infrared brightness temperatures (Ch 5 AVHRR) when clouds affect the image, has been proposed for ATI. As regards TVDI, a modified technique has been proposed for fixing the warm edge of the triangle based on the detection of the extreme dryness conditions on a monthly basis. The modified TVDI has been tested in comparison with an antecedent precipitation index (API) for moisture detection in single images. The substitution of day/night land surface temperature differences instead of noon temperatures in the "triangle method" has been also tested with good results in the multitemporal approach. Application of the proposed techniques can allow one to track the evolution of soil moisture in space and time and to improve the knowledge on the relationship between vegetation (NDVI) and soil moisture dynamics.