Large scale measurements of soil moisture for validation of remotely sensed data: Georgia soil moisture experiment of 2003

A series of soil moisture experiments were conducted in 2003 (SMEX03) to develop enhanced datasets necessary to improve spatiotemporal characterization of soil moisture and to enhance satellite-based retrievals. One component of this research was conducted in South Central Georgia of the US, from June 17th to July 21st (SMEX03 GA). This study analyzes measurements of soil moisture and temperature collected during SNEX03 GA. A network of in situ soil moisture measurement devices, established to provide validation data for the satellite collections and for long-term estimation of soil moisture conditions throughout the region, provided continuous measurements at 19 sites. Additional soil moisture and temperature validation data were collected daily from 49 field sites. These sites represented a diversity of land covers including forest, cotton, peanut, and pasture. Precipitation that occurred prior to June 22nd and from June 29th through July 2nd produced drying conditions from June 23rd to June 28th and gradual wetting from June 29th through July 2nd. Soil moisture in the top 0-1 cm of the soil was found to be more responsive to precipitation and to have greater variability than soil moisture at the 0-3 or 3-6 cm layers. Within different land covers, soil moisture followed the same trends, but varied with land use. Pasture sites were consistently the wettest while row-crop sites were normally the driest. Good agreement was observed between soil moisture measurements collected with the in situ network and the 49 SMEX sites. For the study period, soil moisture across the entire 50 kin by 75 km region and five of the six 25 km by 25 km EASE-Grids demonstrated time stable characteristics. Time stability analysis and statistical tests demonstrated the in situ stations had a small dry bias as compared to the SNMX03 GA measurements. These results indicate that the in situ network will be a good resource for long-term calibration of remotely sensed soil moisture and provide a new and unique source for future satellite product validation. (c) 2005 Elsevier B.V. All rights reserved.