Detection of petroleum hydrocarbons in continental areas using airborne hyperspectral thermal infrared data (SEBASS)

Abstract Remote sensing tools have proven to be effective for direct detection of petroleum hydrocarbons (PHC) in continental areas. PHC yield diagnostic absorption features in the near, shortwave, midwave, and thermal infrared (NIR-SWIR-MWIR-TIR) ranges. Thus far, the NIR-SWIR range has been primarily exploited for remote sensing PHC detection, whereas the application of MWIR and TIR data has remained underexplored. In this work, we aim to close this gap by using the Spatially-Enhanced Broadband Array Spectrograph System (SEBASS) (7.6–13.5 μm) to map PHC-impregnated soil substrates. A laboratory experiment was performed to identify the features in pure crude oil samples, in dry/wet mineral substrates, and in their physical mixtures using attenuated total reflectance (ATR) and directional-hemispherical reflectance (DHR) measurements. A similar setting was designed in an open environment and imaged by the SEBASS sensor. Close-range spectroscopy indicated that PHC features between 7 and 15 μm are subtle in comparison to the features of the mineral substrates, thus hindering direct identification of the PHC spectrally. However, the PHC coating of the grains increased the emissivity of the background features - a finding that was supported by SEBASS data processing using a matched filtering technique. The SEBASS data indicated that besides an increase in the emissivity of the underlying substrates, the presence of PHC also induces a change in the temperature of the contaminated targets. The combination of these two factors highlighted the contaminated sites indirectly. Since the mapping of clean, moist and PHC contaminated sites with TIR data depends on variations in the emissivity and temperature of the targets, the application can be extended to orbital sensors with lower spectral resolution (such as ASTER and ECOSTRESS), enabling the monitoring of larger areas.