Microwave remote sensing studies of De La Plata basin in Argentina

H. KARSZENBAUM, F. GRINGS, M. SALVIA,P. FERRAZZOLI, F. MOCCIA. A.SOLDANO, D. GONIADZKI. G. PARMUCHI, C. MONTENEGRO,P. KANDUS, M. BORRO

Oxnard, California, USA

Workshop; Microwave land surface hydrology workshop; 2008

Nasa

The De La Plata Basin is located in South America and covers about 3.6 million km2. It is the fifth largest basin in the world with extensive and important native forest, very productive agricultural areas, wetlands, important human settlements and infrastructure developments. The principal sub-basins are those of the Paraná, Paraguay and Uruguay Rivers. International projects addressing the hydroclimate of the basin, such as La Plata Basin (LPB) project, are right now in progress. Also, several projects addressing smaller areas within the basin are currently in progress related to radar remote sensing in wetlands, and more recently an ESA AO project related to data exploitation of forthcoming SMOS satellite has been accepted.Within this frame, this paper presents the work done and in progress related to radar applications in wetlands of the Paraná sub-basin, and the contribution of passive microwave measurements from the AMSR-E radiometer acquired over the whole basin.1. We discuss ENVISAT ASAR observations of two different types of marshes (with different structures) under different environmental conditions (normal, flooded, and dry). This analysis is done considering a site specific approach where vegetation structure is well known and also at land cover scale where the backscattering coefficient reflects the variations in vegetation structure and soil condition.2. We show how radiative transfer models can be used to interpret radar observations, to aid in the retrieval of water level below vegetation and in the determination of the buffer effect in wetlands. These models show a good fit with observations, implying that the vegetation modeling scheme, although being simplified, can explain the main characteristics of marsh radar response. Using the physical model, the backscattering coefficient trends observed at VV and HH polarizations for the type of marshes studied have been interpreted by considering the different processes leading to the overall backscatter (plant soil double bounce, volume scattering, and soil backscattering).3. Passive microwave signatures of AMSR-E radiometer collected over De La Plata basin, including the Chaco forest, were analyzed. In particular, the multi-temporal behavior of two normalized indexes was considered. Selected indexes were the normalized polarization difference (PI) at 6.9 GHz and the normalized difference between brightness temperatures at 36.0 Ghz and 6.9 GHz (FI). Within the forest, the PI decreases with increasing forest biomass, as expected. However, also the soil properties play an important role, especially in winter and spring, and a strong rainstorm produced an appreciable increase of PI and FI, also under a homogeneous and dense forest cover. Along the Parana river, the periodical or unexpected variations of water flux produce evident and well correlated variations of both indexes. This result is promising in view of an operational use of passive signatures for monitoring of river and floodplain status.