INVESTIGADORES
SALVIA Maria Mercedes
congresos y reuniones científicas
Título:
Modeling multiangular and multipolarization backscattering of wetland marshes under various environmental conditions.
Autor/es:
GRINGS, FRANCISCO MATIAS; KARSZENBAUM HAYDEE; FERRAZZOLI, PAOLO; GUERRIERO, LEILA; SALVIA, MARIA MERCEDES; KANDUS, PATRICIA; JACOBO BERLLES, JULIO
Lugar:
Boulder, Colorado
Reunión:
Simposio; International Geoscience and Remote Sensing Symposium ( IGARSS); 2006
Resumen:
Wetlands are areas where the frequent and prolonged presence of water at or near the soil surface drives the natural system. Wetlands are important ecosystems that provide: (1) flood and storm control due to their hydrologic storage capacity, (2) protection of subsurface water resources and provision of valuable watersheds and recharging ground water supplies (3) pollution treatment by serving as a biological and chemical oxidation basin (4) erosion control by serving as a sedimentation area and filtering basin, absorbing silt and organic matter.Through monitoring of wetland regions, an improved understanding of the water balance and stability of regional watershed systems can be obtained.This paper is focused on the description and the simulation of junco marshes radar signatures. Junco marshes are one of the most abundant marsh types of the Lower Delta of Paraná River in Argentina. They cover 415 km2 (18% of the lower watershed) and they are one of the main responsible of the water buffer effect on this wetland. It is also a patchy ecosystem driven by a complex fire-forced dynamics. In thepast years, the Paraná site was extensively monitored using optical and microwave spaceborne instruments.More recently, we have particularly investigated the capability of SAR to map the coverage of juncos in normal conditions and monitor their evolution under particular events, such as burning, re-growth, flooding, etc. This work, based on data collected by ERS SAR and ENVISAT ASAR at low incident angles, made it necessary to model and explain the backscattering coefficient of these marshes. To this aim, the scattering processes on junco marshes have been simulated by the electromagnetic model developed at TorVergataUniversity. In its generalversion, the model describes the soil as a homogeneous half space with a rough interface and the vegetation as a discrete ensemble of lossy dielectric elements. Canonical shapes, such as discs and cylinders, are selected for the elements. Originally, the model was developed and tested for agricultural fields and forests. It has been adapted to wetland juncos, where the surface-vegetation double bounce is the main scattering component, since the underlying surface is generally covered by the strongly reflecting calm water.In its early version, the modelsuccessfully explained an ERS-2 VV series collected during the re-growth of junco marshes patches after an intense burn, based on restricted scattering and biophysical assumptions. More recently, we used an improved version of the model to interpret the radar response (VV and HH) of junco marshes in occasion of an extraordinary flooding event observed by ENVISAT ASAR in APP S1 mode. A water retrieval scheme was also developed. Additional radar data, provided through an ESA ENVISAT ASAR AOproject, includes observations of the junco marsh under different environmental conditions, at HH, VV and HV polarizations, and at steep and slant looking angles. To be able to explain these observations, the model required further refinements, both in its electromagnetic accuracy and in the characterization of the input data sets. The objective of this paper is to present the model and the modifications introduced to explain recent observations. We focus on some aspects,i.e. a more accurate characterization of soil-shoot interaction, a statistical description of shoot orientation based on ground measurements, and a more precise evaluation of shoot permittivity, based on dry matter densitymeasurements. We also show how these changes improve the model s ability to simulate the junco radar response for various environmental conditions such as fire-force dynamics, floodings, and phenology changes, and different ASAR modes. Results are shown as comparisons between simulated and measured C band signatures for several samples. The accuracy of the present model is critically discussed, as well as implications for its use for monitoring applications.