No abstract

During the oil extraction procedure, natural gases escape from wells, and the process of recuperating such gases requires important investments from oil and gas companies. That is why, most often, they favor burning them with flares. This practice, which is frequently employed by oil-producing companies, is a major cause of greenhouse gas emissions. Under growing demands from the World Bank and environmental defenders, many producer countries are devoted to decreasing gas flaring. For this reason, several researchers in the oil and gas industry, academia, and governments are working to propose new methods for estimating flared gas volumes, and among the most used techniques are those that exploit remote sensing data, particularly Visible Infrared Imaging Radiometer Suite (VIIRS) Nighttime Light (NTL) ones. Indeed, it is possible to extract, from such data, some physical parameters of flames produced by gas flares. In this investigation, a linear spectral unmixing-based approach, which addresses the spectral variability phenomenon, was designed to estimate accurate physical parameters from VIIRS NTL data. Then, these parameters are used to derive flared gas volumes through intercepting zero polynomial regression models that exploit in situ measurements. Experiments based on synthetic data were first conducted to validate the proposed linear spectral unmixing-based approach. Second, experiments based on real VIIRS NTL data covering the flare, named FIT-M8-101A-1U and located in the Berkine basin (Hassi Messaoud) in Algeria, were carried out. Then, the obtained flared gas volumes were compared with in situ measurements.

Selecting the decisive spectral bands is a key issue in unsupervised hyperspectral band selection techniques. These methods are the most popular ways for dimensionality reduction of original data. A compact data representation without compromising the physical information and optimizing the separation between different materials are the main objectives of such selection processes. In this work, a hyperspectral band selection approach is proposed based on linear spectral unmixing and sequential clustering techniques. The use of these two specific techniques constitutes the main novelty of this investigation. The proposed approach operates in different successive steps. It starts with extracting material spectra contained in the considered data using an unmixing method. Then, the variance of extracted spectra samples is calculated at each wavelength, which results in a variances vector. This one is segmented into a fixed number of clusters using a sequential clustering strategy. Finally, only one spectral band is selected for each segment. This band corresponds to the wavelength at which a maximum variance value is obtained. Experiments on three real hyperspectral data demonstrate the superiority of the proposed approach in comparison with four methods from the literature.

No abstract

No abstract