Use of Radar Vegetation Index (RVI) in Passive Microwave Algorithms for Soil Moisture Estimates
Abstract
The Soil Moisture Active Passive (SMAP) satellite will provide a unique opportunity for the estimation of soil moisture by having simultaneous radar and radiometer measurements available. As with the Soil Moisture and Ocean Salinity (SMOS) satellite, the soil moisture algorithms will need to account for the contribution of vegetation to the brightness temperature. Global maps of vegetation volumetric water content (VWC) are difficult to obtain, and the SMOS mission has opted to estimate the optical depth of standing vegetation by using a relationship between the VWC and the leaf area index (LAI). LAI is estimated from optical remote sensing or through soil-vegetation-atmosphere transfer modeling. During the growing season, the VWC of agricultural crops can increase rapidly, and if cloud cover exists during an optical acquisition, the estimation of LAI may be delayed, resulting in an underestimation of the VWC and overestimation of the soil moisture. Alternatively, the radar vegetation index (RVI) has shown strong correlation and linear relationship with VWC for rice and soybeans. Using the SMAP radar to produce RVI values that are coincident to brightness temperature measurements may eliminate the need for LAI estimates. The SMAP Validation Experiment 2012 (SMAPVEX12) was a cal/val campaign for the SMAP mission held in Manitoba, Canada, during a 6-week period in June and July, 2012. During this campaign, soil moisture measurements were obtained for 55 fields with varying soil texture and vegetation cover. Vegetation was sampled from each field weekly to determine the VWC. Soil moisture measurements were taken coincident to overpasses by an aircraft carrying the Passive and Active L-band System (PALS) instrumentation. The aircraft flew flight lines at both high and low altitudes. The low altitude flight lines provided a footprint size approximately equivalent to the size of the SMAPVEX12 field sites. Of the 55 field sites, the low altitude flight lines provided measurements for 15 fields. One field was planted in corn; three were pasture; six were soybeans; three were wheat; and two were winter wheat. The average RVI for each field was determined for each PALS overpass, with sampled radar data confined to the field dimensions. A linear interpolation was conducted between measured values of VWC to estimate a daily VWC value. A linear regression was conducted between the average VWC and the RVI, for each vegetation type. A positive linear relationship was found for all crops, with the exception of pasture. The correlation between the RVI and VWC was strong for corn and pasture, but moderate for soybeans and winter wheat; however, the correlation for corn was not significant. The developed models were utilized to provide a calculated VWC which was inputted into a modified version of the Land Parameter Retrieval Model (LPRM) to determine the error associated with using a calculated VWC from the RVI versus measured VWC data. The LPRM outputs for both scenarios were compared to the PALS radiometer measurements of brightness temperature.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2013
- Bibcode:
- 2013AGUFM.B34B..02R
- Keywords:
-
- 0480 BIOGEOSCIENCES Remote sensing;
- 1855 HYDROLOGY Remote sensing;
- 1866 HYDROLOGY Soil moisture