The Use of Ultra/Hyperspectral Remote Sensing for Premature Detection of Botanical Change in Soil-Vegetation System as Indicator of Small Leakages of Liquid Hydrocarbons in Pipelines

Abstract

The project aims to use ultra/hyperspectral remote sensing data and techniques for the premature detection of botanical changes in the soil-vegetation system caused by small leaks of liquid hydrocarbons in pipelines. To achieve such objective, we propose to organize, process and analyze a set of spectral data collected from leaves and canopies of five plant species (B. brizantha, Neonotonia wightii, Phaseolus vulgaris, Zea mays and Saccharum spp) from an experiment in real scale, in which controlled, small contaminations of gasoline and diesel were applied to simulate sites of soil-vegetation with leakages unnoticed by current monitoring systems. It is postulated that specific changes caused by soil contamination with hydrocarbons can be detected early in the spectra of the plants under investigation. To the extent that this hypothesis is demonstrated, the techniques and methodologies used and tested here have a potential application in the operational monitoring of oil spills along the national pipeline network.