AGS has been using Landsat-8 multispectral imagery to detect the extent of significant hydrocarbon and produced-water spills associated with oil and gas production activities. The purpose of these investigations is to assist incident response and investigation with the detection of possible early signs of a spill, to increase situational awareness by obtaining land use/cover information for remote spill locations, and to monitor the spill remediation process.
Change detection analysis can be used to detect early signs of a spill, where stronger changes in the site vegetation are associated with spills and weaker changes are more likely the result of natural causes. Figure 1 shows the progression of vegetation loss over time due to a spill. Additional spill analysis methods include land use/cover classification and identification of materials associated with the spill.
Figure 1. Expansion of area of impact over time t2 to t6 due to a spill
NASA’s ozone monitoring instrument (OMI) and atmospheric infrared sounder (AIRS) data were utilized to identify the precursor of large hazardous gas release events over a number of days. These datasets have kilometre-scale spatial resolution (e.g., 13 km × 24 km) and include area statistics, column amount, and volume mixing ratio for the atmosphere’s planetary boundary layer and mid-troposphere. OMI and AIRS provide data on key air quality components like SO2, NO2, HCHO, and CO.
EO technologies have some limitations. Satellite revisit time and the availability of cloud-free imagery can limit the use of optical data. Combining multiple free satellite data sources (e.g., Landsat-8 and Sentinel-2) can minimize the revisit time limitation, and using the ESA’s Sentinel-1 SAR or the Japan Aerospace Exploration Agency ALOS-2 data can rectify the cloud cover issue, as these data do not come from optical sensors. High-resolution commercial imagery can be utilized for further analysis.