Terrestrial Oil Spill Detection and Environmental Impact Monitoring using Multi-Source Earth Observation Data

点击翻译按钮获取中文摘要

Concerns about terrestrial or land-based oil and waste water spills have increased considerably in the past several years in the areas such as Alberta, Canada. For example, in spring 2012, a pipeline leaked 9.5 million liters of industrial waste water, the North America’s biggest spill in recent history, about 20 kilometres of Northeast of Zama City in Northern Alberta [1]. Many ecologically important ecosystems including rivers, lakes, wetlands and vegetation areas are threatened by such spills. With increasing establishment of pipelines and shipment of oil by other means such as rail (Railway Association of Canada) and higher possibility of oil spills, there is an immediate need to develop methods for early detection of spill, extraction and delineation of impacted areas, and monitoring the effectiveness of remediation process. Remote sensing is a rapid and cost effective technology that can be utilized for oil spill detection and environmental impact monitoring.While the value of earth observation (EO) data has been well reported and operationally applied in the case of maritime spills, it has been much less applied for land-based events partially because of the perceived limitations in the utility of EO data and techniques for terrestrial spill mapping. The objective of this research program is to develop effective methods for detecting terrestrial oil/waste water spill and monitoring its impact on the environment using time-series (before and after the event) multi-source EO data including satellite-based high resolution optical, polarimetric synthetic aperture radar (SAR) and ground-based electromagnetic survey measurements. The proposed research will build on earlier works investigating object-based image analysis to use multi-source EO data. The proposed research will use SAR data provided by Canadian Space Agency (CSA) through the Science and Operational Application Research (SOAR) program acquired over a known spill site near Zama City in Alberta. This work will be performed by a PhD student to be recruited. The anticipated outcome will be a framework including image processing techniques for terrestrial oil/waste water spill detection, delineation of impacted area and monitoring of affected ecosystems using EO data. It, also, establishes a base for future research on oil and gas related activities monitoring using satellite-based EO techniques.