Detection and Quantification of Total Petroleum Hydrocarbons in Soils Using Vegetation Optical Properties

A multiscale study was carried out under natural conditions for assessing the potential of vegetation reflectance to quantify soil Total Petroleum Hydrocarbons (TPH). The spectral signature of Rubus fruticosus L. was measured at leaf and canopy scales on an oil-contaminated site, and used for retrieving leaf pigment contents through inversion of PROSPECT and PROSAIL models. Leaf chlorophyll content was strongly related to soil TPH concentrations (R² ≥ 0.8), which allowed predicting TPH with good accuracy (RMSE <; 4 g·kg^-1). The quantification of TPH was also achieved at canopy scale (RMSE = 2.85 g·kg^-1). The proposed method shows promising perspectives for oil prospection and contamination monitoring over vegetated areas. This method is under assessment on airborne hyperspectral images for applications over large scale.