Influence of Surface Heterogeneity on L-band (1.4 GHz) Measurements at Various Spatial Resolutions: Some Preliminary Results of the NAFE/CoSMOS 2005 Field Campaign
J Grant, A Van de Griend, J Wigneron, J Walker, J Kalma, K Saleh and R Panciera
Power Point Presentation
ESA’s Soil Moisture and Ocean Salinity (SMOS) mission, planned for launch in 2007, will carry a multi-angle interferometric L-band (1.4 GHz) radiometer for monitoring soil moisture and ocean salinity at a global scale. Mission requirements include the retrieval of soil moisture with a precision of 4% by volume. Spatial resolution of the instrument is ~35 km at nadir view, which means that most pixels of the earth’s surface will be inhomogeneous, consisting of a mixture of forest, crops, grass and bare soil. In the tropical, temperate and boreal zones, forests make up a large area of the land surface and will therefore certainly influence the microwave emission at a global scale. In order to understand the influence of forested areas on the soil moisture retrieval of inhomogeneous pixels, further research is necessary, especially based on experimental data. Almost all existing knowledge of this subject at the moment is based on modelling studies and represents rather optimistic cases in terms of retrieval and error. In these studies (e.g. Van de Griend et al., 2004) it was concluded that ignoring the a priori knowledge of the forest cover fraction (a) gives large errors in soil moisture retrieval if a = 10%, but if a is known and = 50%, soil moisture in the non-forested area can be determined with a precision better than 4%. Results of the CoSMOS/NAFE’05 field campaign will be used to validate these findings, and to try to improve data analysis for mixed vegetation pixels.
The CoSMOS/NAFE’05 field campaign was held during the month of November 2005. The experiment was conducted in the Goulburn River catchment in south-eastern Australia (see www.sasmas.unimelb.edu.au and www.nafe.unimelb.edu.au), approximately 300 km north-west of Sydney. Several focus farms were selected within the catchment for intensive ground and aircraft monitoring of soil moisture. The ‘Roscommon’ farm was chosen as a focus area for this particular study of heterogeneity effects, as it combined areas of native grass (grazing) with tree-covered areas dominated by eucalypt species. Soils in the area were predominantly sandy, with medium to high rock fraction in the tree-covered areas. Over Roscommon, dual-polarisation airborne measurements at L-band were done twice-weekly by the NAFE aircraft (PLMR radiometer) and once, on 15th November, by the CoSMOS aircraft (EMIRAD radiometer). PLMR measurements were done at four different altitudes (10000, 5000, 2500 and 625 ft), thereby offering the possibility to investigate the effect of heterogeneity at various spatial resolutions. EMIRAD measurements were done at incidence angles of 0º and 40º, PLMR measurements at angles of either 7º, 21.5º, or 38.5º. Ground measurements at Roscommon were done on the majority of the flight days and included soil and tree temperatures, top 5 cm soil and litter moisture content, and average LAI of the tree canopy.
Research and data analysis planned for 2007 will investigate the influence of the percentage tree cover on the soil moisture retrieval at several spatial resolutions, both with and without a priori knowledge of the surface cover type. Soil moisture retrieval will be done by inversion modelling of the t-? model and retrieved values will be compared to ground measurements. The effect of different moisture conditions of soil and litter will also be investigated, as will the effect of using various incidence angles.
This poster will give an overview of the experimental set-up at the Roscommon farm, and show some preliminary results of the available data. Soil moisture values found for homo- and heterogeneous pixels at different spatial resolutions will be compared to give an idea of the orders of magnitude involved. Also, the planned method of research and data analysis will be presented.