Hyperspectral image processing to detect the soil salinity in coastal watershed

Abstract

Hyperspectral images are spectrally over determined, which means that they provide ample spectral information to identify and distinguish spectrally unique materials. With the development and perfection of the hyperspectral remote sensing technologies, it has been the major technique applied for various advance mapping including soil salinity mapping. Brackishwater aquaculture is generally regarded as an environmentally sound practice so far as resource utilization of the marginal land and water of coastal ecosystem goes. However However, impact of shrimp farming on the coastal ecosystem is very often debated. Amongst the biophysical factors, environmental concerns for the salinization of soil, and water have been raised and this issue needs to be scientifically addressed. With this background for this study hyperspectral analysis was conducted to ascertain the salinity status both near the shrimp farming region(Site A-4 location) and perennial brackishwater creek(Site B-2 location). The electrical resistively survey was also conducted in the same locations to ascertain the results of hyperspectral analysis. The study which integrated both these method, shows that spectral reflectance and wavelength at site A location which is near to shrimp farming area are emitting equivalent type of reflectance and soil properties at all the four locations in the site A are similar. By electrical resistivity survey high resistivity values were observed in those areas indicating there is no salinization. But in site B location, (near to creek)The conspicuous high spectral wavelength and features at 500 nm, 760nm-780nm &1220 nm and the triple feature 1520-1600 nm has been recorded. The resistivity data shows low resistivity which means high conductivity, indicating salinization. This study clearly elucidate different spectral reflectance at both the location.