IAFE   05512
INSTITUTO DE ASTRONOMIA Y FISICA DEL ESPACIO
Unidad Ejecutora - UE
congresos y reuniones científicas
Título:
Monitoring Spartina Marshes in the Argentine Coast: integrating Biophysical Parameters, Hyperspectral Field Data and Satellite Observations.
Autor/es:
GONZÁLEZ TRILLA, G; KANDUS, P; MARCOVECCHIO, J.
Lugar:
Orlando
Reunión:
Conferencia; 9 th INTECOL International Wetlands Conference WETLANDS IN A COMPLEX WORLD; 2012
Institución organizadora:
Wetlands
Resumen:
In South America, coastal marshes have their great expression along the humid temperate northern coasts of Argentina, providing recognized ecosystem services. Because functions and values depend on the ecological integrity, there is a growing interest in using environmental indicators to quantitatively determine changes in health of coastal ecosystems. Commonly used and accepted parameters for ecosystem assessment are biomass (Bio) and leaf area index (LAI). However, a deeper understanding is needed about the interactions between biophysical field parameters and the electromagnetic signal from remote sensing. In this paper we attempt to integrate LAI and Bio acquired at local field scale with satellite remote sensing data from Landsat TM sensor. We calibrate FieldSpec® field radiometer observations with LAI and Bio field data (Biofield and LAIfield) of S. alterniflora stands at Bahia Blanca Estuary,(Argentina). Biofield and LAIfield were estimated by a nondestructive method using permanent sample plots. performed during a two years period (TYP). Field spectral profiles at different biomass levels was recorded under natural conditions and manipulating standing biomass by means of successive harvestings. We performed regression equations relating Bio and LAI (EqLAI and EqBIO) to vegetation indices (VI) derived from field hyperspectral data. The Normalized Difference VI (NDVI) showed logaritmic adjustemts (EqLAI R2 = 0.77 and EqBIO R2 = 0.72) whereas the modified soil-adjusted vegetation index (MSAVI) had significant linear regressions (EqLAI R2 = 0.85 and EqBIO R2 = 0.82). Samples of pixels were extracted from 16 Landsat TM scenes callibrated to surface reflectance of the same TYP and the same sites where Biofield and LAIfield measurements were performed. TM surface reflectance were transformed into MSAVI (MSAVIsat). Bio and LAI were estimated applying EqLAI and EqBIO on the MSAVIsat .As a result, Bio and LAI derived from satellite data (Biosat and LAIsat) were estimated and were compared to Biofield and LAIfield. MSAVIsat seems to followed Biofield and LAIfield temporal behaviour of S. alterniflora, showing its seasonal pattern. In addition, Biosat and LAIsat derived from MSAVIsat data correlated positively with Biofield and LAIfield, but LAI has better adjustment (R2 = 0.79 and R2 = 0.43 respectively). LAIfield and LAI estimated from satellite obsevations (LAIsat) did not significantly differ (α = 0.05) for none of the 16 dates analized, showing best fits during summer. Results obtained in this work indicate that LAI patterns of Spartina alterniflora could be monitored by satellite remote sensing data.