|Finn, Michael - Us Geological Survey (USGS)|
|Lewis, David - Stennis Space Center|
|Bosch, David - Dave|
|Giraldo, Mario - Kennesaw State University|
|Yamamoto, Kristina - Us Geological Survey (USGS)|
|Sullivan, Dana - Turf Scout,llc|
|Luna, Ronaldo - University Of Science And Technology Of China|
|Kincaid, Russell - Stennis Space Center|
|Allam, Gopala - University Of Science And Technology Of China|
|Kvien, Craig - University Of Georgia|
|Williams, Michael - Us Geological Survey (USGS)|
Submitted to: GIScience and Remote Sensing
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 8/5/2011
Publication Date: 11/11/2011
Citation: Finn, M.P., Lewis, D., Bosch, D.D., Giraldo, M., Yamamoto, K., Sullivan, D.G., Luna, R., Kincaid, R., Allam, G.K., Kvien, C., Williams, M.S. 2011. Remote sensing of soil moisture using airborne hyperspectral data. GIScience and Remote Sensing. 48:(4)522-540.
Interpretive Summary: Soil moisture can be related to many fundamental physical processes, from plant growth to global warming. Soil moisture is typically measured with point measurements. However, the application of point data to larger spatial areas is problematic at best. Remote sensing is seen as a promising tool for measuring soil moisture over large spatial areas over large time frames. However, application of remotely sensed measurements of soil moisture requires considerable field testing. Here we compare in situ measurements of soil moisture to estimates obtained using reflectance values available from remotely collected hyperspectral data. Results indicate considerable promise for estimating soil moisture conditions at the land surface (within 5 cm), but more uncertainty related to estimating soil moisture with depth (greater than 5 cm). Further research is required to determine the precise measurements that will yield the most accurate results.
Technical Abstract: The Institute for Technology Development (ITD) has developed an airborne hyperspectral sensor system that collects electromagnetic reflectance data of the terrain. The system consists of sensors for three different sections of the electromagnetic spectrum; the Ultra-Violet (UV), Visible/Near Infrared (VNIR) and Short Wave Infrared (SWIR). Based upon previous research showing the value of measuring the long wavelengths of the electro-magnetic spectrum when attempting to detect soil moisture, a study was conducted to further refine applications of hyperspectral data for estimating soil moisture. The Short Wave Infrared (SWIR) sensor was used to gather data over an area in which soil moisture probes were located. We evaluated in situ soil moisture values and compared them to transformation results in order to correlate the transformations with field collected values of soil moisture at specific sampling stations. An airborne hyperspectral instrument with a SWIR sensor was flown twice over the Little River Experimental Watershed in Georgia, 2005 and 2007, to relate remotely sensed soil moisture to in situ measurements of soil moisture. A highly significant (R2 value of above 0.7 for both sampling dates) correlation to a soil moisture probe at 5 cm was computed. Models for the 20 cm and 30 cm depths were not able to estimate soil moisture to the same degree.