You are here: » Information Zones » Rspsoc Students » Student Members » 1167
RSPSoc Student Profile: Obinna C.D. Anejionu
Occupation: PhD Student
Interests: GIS and Remote Sensing application in Land Surveying and Environmental Modelling
Course: Remote Sensing and GIS
University: Lancaster University
Research: Remote surveys of gas flaring and modelling of the environmental and health impacts in the Niger Del
Details of Research: The Niger Delta region of Nigeria contains some of the most endangered ecosystems in the world due to pollution arising from oil and gas activities in the region. Gas flaring has been identified as a major contributor to the overall environmental pollution in the region. There are strong indications that flaring has caused widespread air pollution, heat stress, acid rain, and soil bacteria reduction. However, the specific environmental and health impacts of the continuous flaring of gas in the region have been the subject of much debate and speculation. Previous efforts made to evaluate the magnitude of the impact on the surrounding environment have been severely hampered due to lack of sufficient information on the location of the flares and flaring volume. Accurate identification of active flares is a prerequisite to modelling the impact of flares on the ecosystem. Earlier attempts made at identifying gas flares through remote sensing have been based on visual identification of flares from satellite images such as NOAA AVHRR and DMSP OLS; these approaches were limited by issues such as non-automatic detection of flares and misidentification or non discrimination of flares especially amidst other sources of lights (urban lights and biomass fires). This research is seeking to solve this problem by primarily developing reliable remote sensing techniques that will automatically detect flares, and estimate the quantity of gas flared from each source; consequently evaluating the environmental impacts, through various modelling techniques. The specific objectives are to produce a comprehensive spatial inventory of flares for an extended time series; to estimate the flare volumes at the various flow stations; to model air pollution from the flares and produce a series of pollution maps of the Niger Delta; and to model the overall impact of the pollutants on the various aspects of the ecosystem. The methods used so far have involved the determination of appropriate techniques for combinations of three Landsat bands (4, 7, and 6), which have demonstrated capabilities of detecting flares. The work has focussed on determining empirically the optimum threshold values for the discrimination of flares from detectable false alarms such as forest fires, sun glint, and hot natural and anthropogenic surfaces. A problematic issue encountered so far has been the spatial offsets between the various bands, which have been minimised through the spatial buffering of identified potential flares in one of the bands before passing it through another stage of the process. Preliminary results have demonstrated the capability of the techniques to accurately identify locations of active onshore and offshore flares. However, improvements in the techniques are being investigated in order to make the approach more robust and minimise errors of commission and omission, through temporal persistence analysis. Furthermore, ongoing work is developing techniques for the estimation of flare volume, through the fusion of the outputs from the analysis of Landsat data with information derived from MODIS, AVHRR, MSG/SEVIRI and DMSP data.
