Funds are awarded to carry out a baseline assessment of the impacts on the ecology of the Gulf Coast marshes resulting from BP Deepwater Horizon Oil Spill. The associated coastal wetlands, estuaries and lagoon systems of the Gulf Coast are an irreplaceable national resource and comprise a unique suite of ecosystems of rich biodiversity and widely recognized ecological and economic value. Because of the toxicity of all weights of crude oil there is the potential for smothering, significant injury and mortality of the tidal wetlands. This RAPID project will address a number of critical scientific questions about the impact of this environmental disaster on the ecosystems of the Gulf including: quantifying the extent of the spill, ecological consequences, and environmental factors that contributed. The project will focus on the application of innovative and unique algorithms and computer programs to convert hyperspectral data from the Advanced Visible Infrared Imaging Spectrometer into measurements of coast marsh vegetation functional types and species; map species/functional type; estimate density of vegetation; evaluate evidence of vegetation stress by comparing contaminated vs. uncontaminated vegetation; and map the location of oil contamination in the wetlands. The research will be valuable to state and federal resource managers, policy makers, and citizens of the region. The data will establish a baseline to measure the longer term consequences to the oil spill.
Our project addressed early and longer-term impacts in the ecologically sensitive coastal wetlands of Barataria Bay, LA caused by the BP Deepwater Horizon oil spill (largest oil spill in U.S. history), using an advanced airborne remote sensing instrument, the Advanced Visible Infrared Imaging Spectrometer (AVIRIS), flown by NASA over the wetlands several times at different flight altitudes, between May 2010 (before impact) to October 2010 (after impact) and again the following summer, between July and October 2011. Barataria Bay is ecologically sensitive and known to have had extensive areas covered by oil. We addressed three questions related to the impact of the oil spill. First, whether we could map the actual location of oil in the wetlands using spectral bands in AVIRIS data that are sensitive to oil and second, whether there was any detectable effect of oil on wetland conditions, based on changes in reflectance that are indicative of physiological injury, and third whether we could use AVIRIS to map the species composition of the wetland area. We processed the AVIRIS data from the Baratara Bay region of Louisiana for two periods, September 14, 2010 (pixel resolution: 3.5 m), and August 15, 2011 (pixel resolution: 7.7 m) to study. We show (Kokaly et al., 2012) that oil contamination was successfully mapped along the shorelines of Barataria Bay using AVIRIS data acquired on July 31, September 14 and October 2, 2010. This was done by identifying the oil using hydrocarbon absorption features in the shortwave infrared region, centered at wavelengths of 1720 nm and 2300 nm. The high spectral resolution of AVIRIS was required to measure these features. We found more than 75% of the shorelines in Barataria Bay had oil present in the intertidal region by the fall 2010 (68.45 acres). Except for the heaviest crude weights, oil is toxic to plants, and for the heaviest weights, oil coats the plant surface and suffocates them. In the next phase of our study we used AVIRIS data from September 14, 2010 and August 15, 2011 to determine how far inland plants were injured by oil. This study found damage from oil penetrated an average of 16 meters (52.5 feet) into the marsh, as detected by remote sensing indexes that are sensitive to declining condition of the species. This distance is approximately at the extent of mean high tide, given average shoreline elevation. A comparison of index values in September 2010 and August 2011 for shorelines that were oiled in 2010, show they were not significantly different than unoiled shorelines, indicating that after one year, they were at or near recovery in terms of vegetation density.
