Thermal Methods to Monitor Soil Moisture in Different Study Areas

Thermal methods to monitor soil moisture in different study areas

CXu^a, JQu^a, XHao^a, LGutenberg^a,b, ZZhu^b, MCosh^c

^a Global Environment and Natural Resources Institute (GENRI), College of Science, George Mason University, Fairfax, VA 22032, USA

^b U. S. Geological Survey, Reston, VA 20192, USA

^c Hydrology and Remote Sensing Laboratory of USDA ARS, Beltsville, MD 20705, USA

Soil moisture (SM), especially the surface soil moisture (SSM), from the land surface to 5 cm depth, is a key factor to the hydrology circle, biochemical process, the climate change and many other environmental issues. It will also have a fundamental effect on drought, which will influence the agriculture, vegetation and plants growing condition, and will finally have an impact on food security and human’s health. From what have been listed above, we can see how important it is to study the SSM. Currently, in situ measurement is the most reliable way for SSM monitoring. Based on human, and financial reasons, in situ measurement is usually only conducted in small study area for short study period. If we want to get long-time, large study area SSM study, in situ measurement is not the most suitable way, remote sensing methods are currently popular for large scale SSM monitoring. The two most commonly used remote sensing methods are the microwave method and thermal method. Microwave remote sensing provides the capability for soil moisture estimation, and numbers of soil moisture data products from passive microwave remote sensing missions are available for global and regional applications, such as AMSR-E, AMSR-2, and SMOS, SMAP in recent years. However, these soil moisture products from microwave remote sensing usually have low spatial resolution at tens of kilometers, which are not adequate for regional and local agricultural applications, and is difficult to be combined with in situ measurement for further calibration and validation work. The thermal method is a potential soil moisture retrieval method using remote sensing measurements with moderate to high spatial resolutions, such as Terra/Aqua MODIS and Landsat TM/ETM/TIRS. Two of the most commonly used thermal methods are the thermal inertia method and the universal triangle method. In this study, due to the small scale of the study area, thermal methods were chosen.

MODIS and Landsat dataset will be fused to generate results at both fine spatial and temporal resolution. Although the thermal infrared channels of Landsat 7/8 are close to MODIS TIR channels, there are differences among spectral response functions. For consistency of data from different sensors, it’s still necessary to check and remove the inter-sensor biases before data fusion. The AIRS high spectral TIR measurements over the study area and period were convolved with the spectral response functions of selected Landsat 7/8 and MODIS channels to determine inter-sensor bias and generate consistent Landsat 7/8 and MODIS data for data fusion to get daily LST at 30m resolution. The accuracy of the retrieved soil moisture will be evaluated with in situ observations of surface soil moisture at 0-5cm depth. In this study, two study areas will be chosen to make a comparison, on study area is in the agriculture region in Iowa, the U.S., another is in the polar region in Alaska, the U.S. Both of the thermal inertia and universal triangle method will be used to study the difference between these two thermal methods, and some improvement work will be done according to different study regions. Analysis of soil moisture data at 30m resolution over the study area show details of spatial variation and temporal change, and demonstrated good potential for agricultural applications at local scale.