April 21, 2010 (U.S. Coast Guard Photo)
Illustration 7: Deepwater Horizon Blowout Disaster in the Gulf of Mexico - Spring 2010
On 22 April 2010 the Deepwater Horizon offshore oil drilling platform burned and collapsed after a catastrophic blowout two days earlier. Located about 65 miles off the Louisiana coast, the collapse of the rig ruptured the well pipe at BP's Macondo well over 1500 meters below the surface. This caused an uncontrolled gusher of oil and gas to surge into the Gulf waters from the seafloor.
In anticipation of a potential environmental disaster, AAI immediately contacted the German satellite imaging company RapidEye AG to obtain spectral imagery of the environmentally sensitive shallow water and coastal areas that could potentially be affected by the oil. The RapidEye constellation can rapidly respond and acquire daily coverage at moderately high spatial resolution (5m pixels) in five spectral bands, allowing meaningful assessments of impacted water quality, shallow water bottom cover, and beach impairments using AAI's spectral image analysis technologies. The immediate goal was to collect imagery of the coastal region just prior to contact with the oil, which is critical to obtaining later meaningful environmental damage assessments of areas impacted by the oil using data acquired during a subsequent round of "after impact" image acquisitions.
The initial RapidEye imagery, acquired on 26 April 2010, focused on the visible surface oil feature and its immediate surroundings to enable the spectral reflectance properties of the oil-related material and water to be characterized. The second acquisition, on 27 April 2010, covered an expanded area extending from the accident site to the Louisiana coast. The third acquisition, on 29 April 2010, substantially expanded coverage to include some of the most environmentally sensitive barrier islands, coastal areas, wetlands, estuaries, and land areas in Louisiana and Mississippi. It included a 100km x 325km swath, comprised of an array of 65 images, each covering a 25km x 25km area at 5m spatial resolution. Its timing was critical, providing coverage immediately before impact, minimizing the contribution to the observed effects of other environmental factors unrelated to the spill. A mosaic of the 29 April imagery, calibrated to units of material reflectance using AAI's iCee™ application, is shown in Figure 1 below.
In addition to providing a meaningful baseline for environmental damage assessment, the imagery from 29 April also provided key information about the nature and characteristics of the disaster during its first week, including key early oil state and transport insights. This was needed to plan the timing and locations of the future image acquisitions, as well as provide needed insight for developing the strategies for identifying, characterizing, and quantifying the damage.
Figure 1. Mosaic of 65 RapidEye images acquired on 29 April 2010. The image boundaries and locations are shown on the right. The imagery was calibrated to units of material reflectance using AAI's iCee™ image restoration application, yielding the reflectance mosaic on the left.
Central Oil Feature
A portion of the relatively prominent large surface oil feature centered over the accident site can be seen on the eastern edge of the mosaic in Figure 1 above. A close-up view (see Figure 2 below) revealed that it generally has the characteristics of a stable water-in-oil (w/o) emulsion. The high-resolution imagery revealed the feature to be an apparently viscous and stable mass, with spectral reflectance and structural properties that differed from those of either a spreading slick or an unstable or meta-stable oil-in-water (o/w) emulsion.
Figure 2. Portion of the surface oil feature just west of the Deepwater Horizon accident site
In particular, the image-retrieved mean pixel reflectance of the oil feature varied across the mass, but it averaged around 6 percent. AAI's SHORZAN Land / Water Interface application indicated that the oil feature was largely submerged, indicating that the pixel reflectance was likely being influenced by reflections from the sea surface, as well as attenuation by the water in which it was floating. AAI's QSC water quality and Material Mapper applications were used to remove the surface reflections and attenuation effects, and to retrieve the material spectra of the submerged oil feature component. The retrieved mean visible material reflectance of the floating mass was found to be around 1.5 percent with a spectral shape consistent with an optically thick and stable w/o emulsification. It did not have a spectrum consistent with a red or black meso-stable emulsion.
The physical structure of the mass was also consistent with a stable w/o emulsion. Unlike a spreading slick or an unstable or meso-stable oil-in-water (o/w) emulsion, it displayed a prominent shear structure characteristic of a viscous self-adhesive mass (Figure 2). Also unlike a slick, the mass caused little or no detectable suppression of the sea surface wave amplitudes or velocities, suggesting minimal changes in water surface viscosity and shear stress due to surface spreading. This can be seen in the close up view of the surface mass in Figure 3 below, where details of wave motion are revealed by the red and blue "motion" indicators derived from the RapidEye data. The red bars are on the leading edges and the blue bars are on the trailing edges of the moving waves. The sizes and densities of red and blue wave velocity indicators on the floating mass are the same or greater than most of those in the surrounding waters. The patches of water with suppressed wave velocities to the left of the mass in Figure 3 are likely due to films of dispersant applied to the mass.
Figure 3. Close-up view of the surface oil feature, showing surface wave direction and velocity indicators derived from the RapidEye imagery. Red features represent the leading edge and blue the trailing edge of moving objects, including surface waves. Note, for illustration, the red bows and blue sterns of the three boat hulls: the sizes of the bars indicate the middle boat is traveling at a higher velocity than the other two, which are basically drifting.
These spectral and physical characteristics of the central oil feature provided an early indication that at least a portion of the crude oil discharging from the wellhead on the seafloor contained enough emulsifying agents, such as crude oil resins, asphaltines, and natural wax, to form a stable w/o emulsion before reaching the surface. Generally such emulsions evolve as weathering products from slicks, but the central oil feature appeared to be a cumulative deposit of oil already in the stable emulsified state.
The accelerated pre-emergent formation of the stable emulsion was likely the result of the turbulence associated with the high-pressure injection of the oil into the seawater from the ruptured well pipe. The formation of stable w/o emulsions involves a turbulent agitation-driven dispersion of water droplets in the 1-10 µm size range into the crude oil, where they are held in a rigid structure by the asphaltines, resins and/or waxes. The dispersed phase droplets can be as high as 60-80 percent of the volume, and they are prevented from coalescing and separating from the oil by the strong visco-elastic oil-water interface produced by the asphaltines in particular. This, rather than the loss of the low molecular weight and volatile components from the oil, gives rise to its high viscosity. It also gives rise to a grey, rather than black or red coloration. Its grey coloration is a consequence of enhanced reflection from the dispersed water droplets (multiple scattering) within the inherently black intact crude oil material. The enhanced reflectance of the oil in the central feature may also have been the result of additional entrained gas bubbles, since escaping natural gas and methane were driving much of the turbulence. The red and black coloration of the less stable emulsions arises from the loss of the low molecular weight and volatile components and separation of the water from the oil, respectively, each of which are largely retained in a stable w/o emulsion. The amount of entrained water also produces a significant increase in its volume, and the presence of entrained gas could sufficiently lower its density to account for its buoyancy and efficient emergence at the surface.
The high-resolution imagery further revealed a process of disintegration and dispersal of the material at the terminus of the mass and along its edges. This can be seen in Figure 2 above. A substantial amount of material can be seen shearing off and flowing away from the central mass under the influence of a local current. This detachment and escape of material from the peripheries of the central mass at least partially offset the continual addition of material at the center of the mass situated above the wellhead. Furthermore, because the material flowing away from the feature is more dispersed than the integrated central mass, its contrast with the water is reduced due to radiance averaging (known as the mixed pixel effect), making it difficult to detect. This is particularly true in coarser scale imagery like NASA's MODIS imagery. Because of this continual truncation of the intact portion of the central oil feature, its visible area significantly underestimates the cumulative volume of oil that has accumulated on the surface since the onset of the incident. Initial estimates of the flow rate of oil from the well into the Gulf were based largely on the size of the central oil feature. Those estimates were subsequently challenged, following public release of live video feeds of the oil and gas escaping from the ruptured wellhead pipe on the seafloor. The apparent near-stationary character of this continually truncated central oil feature has also given the deceptive impression that the oil is staying away from the shoreline.
Surface Oil Transport from the Central Feature
The high-resolution imagery revealed that surface currents in the Gulf have been transporting surface oil from the central feature toward the Louisiana coast since before the 29 April date of the RapidEye imagery discussed here. This can be seen in Figure 4 below. By April 29, surface oil dislodged from the flow lobe of the central feature had migrated about 25km west of the central feature, where it was then apparently captured by north-flowing currents. At least two prominent generally parallel streams of dispersed surface oil can be seen to have flowed about 50km northward, where they were apparently captured by another current that drew the oil to within a few km of the southeastern Louisiana coast. There it appears to have been captured by another current that carried the surface oil another 25km southwestward along the Louisiana inner shelf. A close-up view of this loop to the southwest is shown in Figure 5 below, and a close-up view of the flow of oil along the Louisiana inner shelf is shown in Figure 6.
Figure 4. Transport of surface oil from a flow lobe of the central oil feature toward the southern Louisiana shoreline.
Figure 5. Close-up view of the two north-flowing streams of surface oil being captured by a current, likely the Louisiana Coastal Current, causing it to loop to the southwest and flow along the Louisiana inner shelf.
Figure 6. A close-up view of the surface oil flowing southwestward along the Louisiana inner shelf. The path apparently followed by this oil, shown in Figures 4 and 5, is over 100km from its apparent source in the central oil feature.
Image of Submerged Oil
AAI's QSC water quality and Material Mapper applications were used to probe the water column to determine whether the surface oil in the currents was accompanied by additional oil beneath the surface. Of particular interest were the currents transporting oil from the central oil feature toward the southern Louisiana shoreline. It was discovered that there was indeed an unexpectedly large submerged oil component accompanying the visible surface oil component in the surface currents. This can be seen in Figure 7 below.
The left half of Figure 7 shows a zoomed-in portion of the Figure 4 reflectance image, focusing in on a 25km x 25km area that includes the westward flow of surface oil from the central feature to the area where the surface oil was being captured by the north-flowing currents. On the right side of Figure 7 is the image-derived visible reflectance of materials submerged in the waters below the surface. This is a down-looking view through the uppermost part of the water column to the deep ocean below. In shallower waters this produces a reflectance image of the materials on the sea floor. In the deeper waters here, however, it is a reflectance image of the materials suspended in the ocean below about 2.5 meters depth.
The Figure 7 results revealed that the same surface currents that were transporting the surface oil from the flow lobe of the central feature westward into the north-flowing currents were also transporting a black material beneath the surface. The spectral reflectance properties of the black material were consistent with those of submerged crude oil, likely in the form of a meso-stable water-in-oil emulsion that has undergone at least partial separation of the entrained water. This conclusion is based on the image-retrieved reflectance spectrum of the black suspended material, shown in Figure 8 below. Also included for reference are image-retrieved reflectance spectra of the surface oil from the central feature, and a spectrum of the deep-water background in an area presumed to be relatively unaffected by the oil. The spectral shape of the submerged oil is generally similar to that of both the surface oil and deep-water background spectra. The two oil spectra show enhanced near-infrared reflectance relative to the background, however. This can be seen more clearly in the reflectance ratio spectra in Figure 9, where the two oil spectra were divided by the background water spectrum. Both ratio spectra are relatively flat and unstructured in the three visible wavelength bands (consistent with their black and grey coloration), but show enhanced relative reflectance in the near-infrared band, a characteristic of a water-in-oil emulsion. Neither shows enhanced reflectance in the red band, however, indicating a minimal loss of the low molecular weight and volatile components in both forms of oil. The primary difference between the two forms of oil was a generally wavelength-independent factor of 1.4 drop in reflectance at the visible wavelengths (red spectrum in Figure 9), consistent with a significantly lower water content in the darker submerged oil. This suggests the submerged black oil was an early stage meta-stable oil-in-water emulsion from which the water had coalesced and separated, but there was minimal loss of its low molecular weight and volatile components.
Figure 7. Image retrieved reflectance properties of submerged materials accompanying the flow of surface oil from the central feature westward, where it is captured by the north-flowing currents. On the left is the reflectance image (see bottom of Figure 4 for reference), and on the right is the visible reflectance composite of the submerged materials.
Figure 8. Image retrieved four-band reflectance spectra of representative samples of the black submerged oil (black), surface oil from central feature (yellow), and the deep-water background (cyan).
Figure 9. Reflectance ratio spectra, showing the relative spectral characteristics and color of the two states of oil. The yellow and black spectra reveal the spectral properties of the surface oil and submerged oil, respectively, relative to the deep-water background. The red spectrum shows the spectral properties of the surface oil relative to those of the submerged oil.
An expanded look at the Material Mapper results for the area to the north of the area in Figure 7 is shown in Figure 10. The results revealed that the flow of surface oil to the north was accompanied by a submerged component of black oil. Apparently being driven by the same current, a portion of the subsurface oil also looped into the current flowing southwestward along the Louisiana inner shelf. Much of the submerged oil continued its northward flow, however.
Figure 10. AAI's Material Mapper results revealed that the surface oil flowing north was accompanied by a subsurface flow of submerged black oil. A portion of the submerged oil similarly looped into the current flowing southwestward along the Louisiana inner shelf. The black area on the bottom left is a portion of the image not included in the composite shown here.
This look beneath the surface revealed that the amount of suspended oil entering these currents was surprisingly large, considering the early date of the image and the early estimates of the flow rate of oil from the wellhead into the Gulf. The image in Figure 7 covers a 25km x 25km area, and the submerged black oil occupied over 80% (~500 km2) of that area alone. This suggests that the volume of submerged oil being injected into the north-flowing currents that feed the Louisiana Coastal Current posed a substantially larger threat to the Louisiana coastal marshlands, estuaries, and fishing areas west of the Southwest Pass of the Mississippi River than inferred from the visible surface oil component alone.
Seafloor Deposition of the Submerged Black Oil
Close-up views of the submerged oil flowing along the Louisiana inner shelf are shown in Figures 11 and 12. In Figure 11, the reflectance image (left) and Material Mapper results (right) for a zoomed-in area on the eastern edge of the shelf is shown. Here the waters on the shelf are shallow enough for Material Mapper to retrieve the reflectance spectra of the seafloor materials. Variations in the seafloor sediments are clearly apparent. The figures revealed that some of the submerged black oil emulsion was apparently drawn from the current along the shelf (bottom right) into the shallow waters of the shelf, where there are scattered deposits of black oil on the seafloor. The seafloor deposits of black oil appear darker than the suspended emulsion in the deeper waters along the shelf, and they formed as discrete patches rather than as a generalized stain. This suggests that the oil may not have been in a fully dispersed state, but included some discrete congealed masses. Many of the masses were elongated and aligned in a fashion that suggests they were being influenced by flow shear and the spatial frequency of the surface waves. This suggests that they were not yet fully fixed to the seafloor material, but still retained some alignment mobility. There also appeared to be some topographic control of the deposition by the seafloor morphology in the areas affected.
Figure 11. Reflectance image (left) and Material Mapper results (right) for a zoomed-in area on the eastern edge of the shelf.
In Figure 12, the proximity of the submerged black oil to the entrance of the river can be seen. It is uncertain whether any oil had entered the waterway at the time this image was acquired. Although none is visible in the waterway, earlier black oil deposits on the substrate may have been masked by subsequent waterway sediment deposition.
Figure 12. Close-up view of the Material Mapper results, showing the proximity of the submerged black oil to the mouth of the waterway.
Evidence for a Widely Dispersed Weathered Emulsion
The extent of submerged black oil already entrained in these currents after only seven days further suggests that the submerged component had likely been spreading throughout the Gulf more rapidly and efficiently than initially expected based on surface oil observations. This large-scale spread of oil is further supported by the anomalous red coloration and refractive banding of the deep water over much of the RapidEye image mosaic. In Figure 13 is presented the northern portion of the mosaic. Rather than having the normal deep blue coloration in the deep waters eastward of the Chandeleur islands and southward of the barrier islands off the coast of Mississippi, the water has a distinct red coloration with subtle banding parallel to the orbital path of the sensor. This is also true in the southern portion of the mosaic, shown in Figure 14, particularly north of the central oil feature. The water is notably bluer in the southwestern portion of the mosaic, and there is little or no visible banding there.
Figure 13. Northern portion of the 29 April 2010 reflectance mosaic of 65 RapidEye images, revealing the anomalous red coloration and subtle banding of the deep-water portion of the Gulf.
Figure 14. Southern portion of the 29 April 2010 reflectance mosaic of 65 RapidEye images, also revealing the anomalous red coloration and banding of the deep water portion of the Gulf to the north of the central oil feature. Note the traditional lighter blue coloration of the deep water and lack of visible banding in the southwestern portion of the image.
The origin of the anomalous red coloration of the water north of the central oil feature appears to be due to the widespread presence of a red material in the water column. This was deduced from the ratio reflectance spectrum in Figure 15. Pixel spectra of water near the central oil feature, but away from areas directly affected by either the grey surface oil or the submerged black oil, were retrieved from the image and a mean spectrum was created. Also retrieved were pixel spectra of water outside of the anomalous red area, and a mean spectrum of that water was created. The ratio of these two mean spectra, shown in Figure 15, reveals the presence of a red material in the anomalous red waters.
Figure 15. Evidence for the presence of a red suspended material in the water column near the grey surface oil and black submerged oil components.
The ratio spectrum of the red material is consistent with the known properties of meso-stable water-in-oil emulsions from which a portion of its low molecular weight and volatile components have been lost by weathering. These red emulsions are not as fully weathered as brown mousse, however, which would have had a higher 555.7nm relative band reflectance in Figure 15. The subtle banding in the water showing the anomalous red coloration was notably oriented parallel to the orbit of the sensor. The spatial pattern of the banding indicated that it was likely the expected optical consequence of the red emulsified component in the seawater. Although banding can be observed in surface films of oil having thicknesses in the 0.0003-0.002mm range, the spatial scale of the banding is much different. The large spatial scale of the bands (several km-scale widths) are more consistent with the combined refractive (variation in optical penetration depth with variation in sensor look angle) and dispersive (diffractive and refractive color separation) optical scattering effects of a red emulsified component in the seawater. The presence of a suspended red emulsified component in the seawater can also explain the later widespread appearance of red floating streamers of oil across large expanses of the Gulf in this region. Turbulent transport of a suspended emulsified material to the surface, particularly materials as adhesive as emulsified and partially weathered southern Louisiana crude, naturally tend to coalesce at the surface under the influence of the surface tension effects there. This leads to the formation of floating cohesive masses. Because the material is non-rigid, these masses can coalesce into the kinds of red-orange oil streamers that appeared over time at ever increasing frequency and extent.
Evidence for Increasing Toxicity
The presence of a growing submerged suspension of oil in the seawater raises the specter that the water could become highly toxic. Southern Louisiana crude oil contains relatively high levels of light aromatic hydrocarbons and low molecular weight n-alkanes. As a submerged suspension injected at depth, the oil gradually releases these toxic low molecular weight hydrocarbon compounds directly into the seawater, giving rise to the red weathered emulsion product described above. The released low molecular weight hydrocarbon compounds can reach higher cumulative concentrations and affect a larger volume of water than from a traditional surface spill, where much of it is either released to a thin surface layer of the water or it evaporates directly into the air. With the continual injection of fresh crude into the seawater from the runaway well, the concentrations and effective residence times of these substances in the water could potentially build to unprecedented levels of toxicity and lethality.
Figure 16. Image-retrieved concentration of colored dissolved organic carbon in the water column for the same area as shown in Figure 7.
The concentration patterns of CDOC in Figure 16 reveal a strong correlation with the pattern of submerged oil (Figure 7). Concentrations in the eastern half appear to be higher and more uniform, and likely reflect the added effects of the dispersants dropped onto the central oil feature located to the east. The changes in CDOC concentration in these areas can be used to quantify the increase in toxicity caused by the application of dispersants. A larger-scale view of the pattern of CDOC concentration is shown in Figure 17 for the same area covered in Figure 10. Again there is a strong correlation with the pattern of submerged oil in Figure 10. This suggests that the submerged black oil component may be the primary source of the released low molecular weight component, and that the dispersed red emulsion may largely be the weathered byproduct of that release.
Figure 17. Concentrations of colored dissolved organic carbon, reflecting the release of low molecular weight and volatile hydrocarbon compounds to seawater from the submerged oil. The black area on the bottom left is a portion of the image not included in the composite shown here.
Evidence for the lethality of the dissolved hydrocarbons was revealed by the observed increase in suspended chlorophyll concentrations in the area containing the submerged and suspended oil. This is shown in Figure 18 for the same area shown in Figure 17.
Figure 18. Image-derived concentration of suspended chlorophyll associated with phytoplankton. The black area on the bottom left is a portion of the image not included in the composite shown here.
Similar events have occurred during previous spills. During those spills, there was an initial temporary reduction in primary production immediately following the spill. The concentrations of green and blue-green algae were then observed to increase within a few days to a week of a spill in open waters, the point in time observed here. The increased growth in open waters has been attributed to the death of zooplankton in the toxic waters. Zooplankton competes for available phosphorus in nutrient-starved open waters, and it normally controls the species composition of phytoplankton. The death of the zooplankton by the toxic volatile fractions of oil in the seawater makes more phosphorus available, triggering the observed primary production of the algae.The lethality of the waters in the Gulf is likely to be significantly higher than for traditional surface oil spills. During surface spills the highly toxic volatile aromatic hydrocarbons typically undergo dissolution from the oil and evaporation from the water within about 48h. Where the water is covered by a layer of oil, the rate of evaporation from the underlying water is inhibited, but the residence time of the toxic components in the water is still quite short. Because of these short time scales, the relative toxicity of these components on aquatic organisms typically show species variability. The toxicity for a species is frequently expressed as the dissolved hydrocarbon concentration that produces mortality in 50% of a population over a specified time period, usually either 48h or 96h. Toxicity varies from species to species due largely to differences in their tolerance to the relatively brief exposures to elevated dissolved hydrocarbon concentrations. With the Gulf spill, fresh crude oil has been continually injected into the water since 22 April 2010. Unlike a surface spill event, fresh oil is constantly replenishing the source of the toxic water-soluble component. Dispersants are also being added to the oil at the source on the seafloor, accelerating the release of the toxic substances. This would be expected to induce higher accumulated dissolved hydrocarbon concentrations, and sustain them over substantially longer (several weeks) periods of times. These periods of exposure would surpass the tolerance limits of most organisms, virtually eliminating species-to-species differences in lethality. The geographical area and volume affected by the spill is also very large compared to typical surface spills, forcing long-term exposure and likely an unprecedented scale of lethality for the many and diverse populations living in the spill area.
Environmental Impact Assessment and Damage Claims Support
The information being retrieved from this and other early imagery is providing critical insight into the nature of the spill-related materials and governing processes that are and will be impacting the open waters and coastal environment of the Gulf. It is also providing key baseline conditions for identifying and assessing short-term and long-term changes to the water quality, shallow water seafloor, and coastal land cover conditions in the affected areas. This will be important not only for assessing the nature and extent of environmental damage, but also for establishing meaningful restoration targets and monitoring progress against those targets.
A key feature of AAI's technologies being brought to bear on the problem is the unique ability to image materials beneath the surface of the water. Detailed imagery of the seafloor can be acquired in shallow waters, and in deeper waters imagery of submerged materials at several meters depth can be acquired. It also provides the ability to retrieve key water quality characteristics beneath the surface, including several measures of water clarity and composition in high spatial detail.
A second key feature of the technology is the ability to accurately characterize and identify the materials in the water column and on the seafloor based on their retrieved spectral reflectance characteristics. The technologies are fully traceable to the underlying radiative transfer physics, and the accuracies of the retrieved material properties have been verified by ground truth. Subtle differences in material characteristics can be discriminated and characterized, allowing the water column to be assessed for toxicity and sensitive benthic habitats to be surveyed and assessed for stress and damage. The subpixel technology employed in Mixed Material Classifier can provide an accurate measure of the area covered by each material of interest in each pixel, even when mixed with other materials. This allows quantitatively accurate measurements of areal coverage of each material to be made. This can be done over broad areas fully remotely without the need for ground truth, allowing vast remote and inaccessible regions to be efficiently characterized and assessed with the same level of detail and accuracy as fully accessible ones. It also allows imagery acquired on different dates to be used in combination to characterize and monitor changes in the conditions of habitats and environments over time.
A third key feature is the unique ability to utilize image data from a wide range of available commercial and government sensors on both satellite and airborne platforms, and maintain comparable levels of discrimination performance for each. This flexibility to take advantage of different sensors of opportunity interchangeably significantly increases the chances of successfully obtaining imagery of an area of interest at a critical point in time with suitable image quality (e.g., low cloud cover). It also provides an opportunity to select imagery with spatial resolution and coverage characteristics that are appropriately tailored to the particular problem being addressed.
These features of the technology are allowing information with high actionable utility to be retrieved from this and other imagery of the Gulf. The image-retrieved information from the open waters, used in conjunction with strategically sampled field data, will, for example, allow meaningful assessments of the nature and extent of impact on fisheries stocks throughout the Gulf to be made and monitored over time. This will be important both for identifying areas where the fisheries can remain open or be re-opened, and for providing high utility data to support damage claims estimation and justification by the recreational and commercial fishing-related industries. Results such as those shown in Figure 10, for example, will also allow the sources of oil creating damage and affecting those industries to be substantively traced. The information will provide valuable data to support damage claims and restoration targets based on the Clean Water and Oil Pollution Acts.
Because of the broad area coverage and detailed spatial resolution of the imagery being collected, the technologies will also allow comprehensive surveys of environmental damage to be made both on land and the shallow water seafloor. Information is retrieved from every pixel, allowing surveys to be performed with complete coverage over the entire area included in the image. The spatial detail of the retrieved information will vary depending on the sensor used to collect the imagery. With the subpixel technologies employed, however, sub-meter-scale damage should be detectable and meaningfully characterized and quantified for much of the available imagery. The coverage and spatial detail will be critical for pinpointing the locations impacted by oil, and for accurately characterizing and quantifying the extent of damage in the regions of interest. For areas targeted for restoration, restoration targets (pre-spill conditions) can be derived and spectrally characterized using pre-spill imagery. Progress against restoration targets can then be quantitatively monitored over time using repeat coverage imagery and change analysis approaches.
An illustration of the use of imagery from two different sensors to pinpoint and characterize the seafloor locations impacted by oil is shown in Figure 19. On the left is a Material Mapper image of the seafloor for an area on the shelf retrieved from the 29 April RapidEye image. On the right is the comparable seafloor image retrieved from a QuickBird image acquired about a month later on 24 May 2010. The lighter color materials on the seafloor in the left image are mobile recently deposited river sediments. The darker tan materials are more stable previously deposited river sediments. On the right, the patterns of the seafloor materials have shifted, and deposits of the black oil are widespread. The differences in the patterns of the seafloor materials are due largely to differences in the tidal state, as revealed in patterns of suspended sediments in Figure 20. The mobile sediments are entrained by the moving water, and the concentrations of suspended sediments therefore provide an effective indicator of flow rates and directions. The tide is coming in on the left, and receding on the right. Note that the patterns of seafloor oil in the right image in Figure 19 appear to be influenced by the tide, suggesting that the oil was still mobile on that date.
Figure 19. Before and after Material Mapper images of the seafloor for a small area on the shelf. The figure on the left is a natural color image of the seafloor from the 29 April 2010 RapidEye image, acquired immediately before contamination by the oil. The figure on the right is the corresponding Material Mapper image of the seafloor from a 24 May 2010 QuickBird image, showing significant contamination (black and dark grey).
Figure 20. Image-retrieved concentrations of suspended sediments for the 29 April 2010 (left) and 24 May 2010 (right) images. Entrainment of the mobile seafloor sediment reveals the differences in tidal states on the two dates. The tide was coming in at the time of the left image, and receding at the time of the right image.
A close-up view of an area located in the bottom left corner of Figure 19 is shown in Figure 21, showing enhanced spatial detail of the complex pattern of contamination at that site. This level of detail can be used to meaningfully document the extent of contamination of specific sites, such as shrimp beds, and larger-scale fisheries, which can provide valuable defensible information and evidence for estimating and justifying damage claims by affected parties. It can also provide actionable reconnaissance information for planning and guiding clean-up operations, and measuring progress over time against restoration targets.
Figure 21. Close-up view of the bottom left corner of the Figure 19 image, showing enhanced spatial detail of the contaminated areas.
The results presented here serve as an illustration of the information that can and will be retrieved from this and other imagery as the environmental disaster in the Gulf proceeds. AAI continues to gather critical imagery in support of upcoming environmental impact assessments and claims estimation and justification efforts.
Return to the Incident Response Page >>
Contact AAI