Rating of vegetation growth state using radar and optical images

Abstract

The state of vegetation growth plays an important role in the ability of all biological aspects of a region to remain diverse and productive in their environment. There were used 3 types of satellite imagery to assess the state of vegetation growth in the Soberanía National Park in Panama. A TerraSAR-X image in polarization HV and an incidence angle of 29° was filtered by 15 different filters in order to reduce speckle and then each of the results was evaluated using assessors Noise Mean Value (NMV), Noise Variance (NV), Mean Square Difference (MSD) and Equivalent Number of Looks (ENL); and according to the criteria established for these evaluators was elected the best result of the filters used to subsequently obtain the backscattering coefficient ( ) of the image. A Landsat 8 image was used equally for the Normalized Difference Vegetation Index (NDVI) and the Enhanced Vegetation Index (EVI), which measure the state of the vegetation, to be correlated with the backscattering coefficient and get an exponential regression depending on such variables. There were obtained therefore two exponential functions where the results were  and  from the backscattering coefficient and were the representative variables of the state of vegetation growth throughout the study area. The correlation between these variables was low (R² = 0.447 and R² = 0.424 for NDVI and EVI respectively), making validation of results with the NDVI and EVI products from MODIS similarly low.