A Scientific Perspective of the 2010 Gulf of Mexico Oil Spill
The Rosenstiel School
Following a tragic explosion, an oil gusher at the bottom of the Gulf of Mexico, latitude 28.74°N and longitude 88.39°W, started to spew out large amounts of oil at a depth of 1500 m (5000 ft) on April 20, 2010. Various estimates of the flow vary by an order of magnitude, 200,000 gallons per day to 2,000,000 gallons a day.
The gushing, light crude oil is a complex mixture of hydrocarbons and other trace compounds with a mean density of about 0.95 g/cm3, lighter than the surrounding seawater that has a density of about 1.03 g/cm3. Denser seawater over less dense oil is an unstable situation. The buoyant plume is also driven by natural gas that is gushing out of the broken pipes. The density difference between the oil/gas mixture and the water drives a buoyant plume.
As this plume rises through the water column, a number of different physical and chemical processes affect it. First, ambient water is entrained into the oil/gas mixture through turbulent mixing, slowing it down and increasing its relative density. This combined oil/gas/water plume eventually reaches neutral buoyancy at a depth that is determined by the background stratification of the Gulf. This exact depth where this occurs is unknown, and is not easy to predict. After this initial mixed-phase plume terminates, gas bubbles and oil droplets begin to rise upward toward the surface, leaving the entrained water behind. Most, if not all, of the gas becomes dissolved in the water column before reaching the surface. The rate of rise of the remaining oil droplets depends on their size (large droplets rise faster, small droplets rise slower), and the average size is also very difficult to predict.
Once the oil hits the surface it starts to disperse outwards, as seen in the video and in the images below. This low density fluid anomaly on the ocean surface is under the influence of both gravity and the earth’s rotation. The balancing of these two accelerations leads to the oil spill forming a blob on the ocean surface.
If there was no continuous source of oil or weathering processes, theory and observations predict that the diameter of the blob would be on the order of 80 km (50 miles). The blob can be seen in the satellite image of the spill from April 30th and the blob’s size is in good agreement with theory.
Once on the surface, the oil is both moved and mixed by ocean currents and by wind, as well as, undergo further physical and chemical changes called “weathering” due to the wind- and wave-enhanced mixing, evaporation, and sinking. Waves will mix significant amounts of oil down to depths of about twice the wave height. The weathering process will eventually cause some of the oil to sink and some of it to form tar balls. Evaporation removes somewhere between 40 to 80% of the oil during the first week after the oil surfaces and then evaporation becomes a slower process. All these processes change the shape of the blob, as well as, the continuous source of oil gushing upwards from the ocean bottom. During the first few weeks of the spill, predominant winds out of the southeast were opposite to ocean currents many times in the oil spill region, and this slowed down the spreading (“dispersion”) of the oil. We can see the effects of the ocean eddies on the oil spill in a time series of satellite images of the oil spill from May 5, 10 and 17. The fractal-looking Mississippi Delta is just visible on the left and the northern Gulf Beaches on the top of this May 5 image that shows the main oil blob has been dispersed by the eddy field as evident in the long filaments of both oil and of cleaner water being entrained into both the blob and other patches of oil, there are many wind lines of oil near the center of the image, and that the long filament of oil in the right center is being entrained into an eddy. On May 10, besides the main blob of oil being fed by the gusher, isolated patches and thin filaments shaped by coastal shelf dynamics are hitting the barrier beaches, and there is a patch of diluted oil between the main blob and another concentrated patch in the lower right of the image. On May 17, the main blob is still being fed by the gusher but with a reported reduced flow, a very nice example of a filament is flowing out of the main blob due to a cyclonic eddy, and there is a “line” of missing data through the center of the image.
The oil spill looks like it has grown tentacles; oceanographers call these features filaments. Eddies are primarily responsible for the rapid growth of the filaments seen in these satellite images of the ocean surface. Eddies are flow features with clockwise or anti-clockwise flow with speeds on the order of 10–100 cm/s (1/2 to 2 mph). Though their velocity field resembles a very small, slow hurricane, eddies have much different dynamics. These can be detected in satellite images and are shown schematically in the next image from ROFFStm.
Image courtesy of Mitch Roffer, ROFFStm.
The eddies can be seen in this analysis of ocean measurements by the Upper Ocean Dynamics Laboratory in the left image below. This sequence shows the evolution of the surface pressure field and the resulting geostrophic currents in the Gulf of Mexico during the time of the oil spill. The strong Loop Current and its eddy field are clearly visible in this sequence. On the right, temperature profiles in the Gulf of Mexico from the air launched probes deployed at different latitudes at a longitude of 88.5°W as part of cooperative research between AOML and RSMAS. A warm diurnal layer above the 30–40 m deep mixed layer, a deeper well-mixed region called a thermostad in some profiles around 100 m depth, and the top of the thermocline are the main features evident in this data from the surface to a depth of 300 m (> 900 ft). This data will be used for numerical model studies and to analyze the dynamics of the Loop Current and the Gulf of Mexico eddy field.
Figures provided by Dr. Nick Shay
Data assimilation enhances the information in the data by combining data with dynamical models using discrete mathematics and fast computers. One of the goals of science is to predict the outcome of complex events such as the weather. In the 1960s, Ed Lorenz of MIT told us that a butterfly flapping its wings in Brazil can effect the weather in Texas due to the nonlinearity of the atmosphere. Ocean dynamics are also nonlinear and the error in predicting where anything drifts in the ocean, under the influence of winds and currents, increases rapidly in time. The combined effects of nonlinearity and uncertainties in wind forcing (oil will move in a direction of the wind with speeds that are 3% of the wind speed) and currents results in a forecast time of only 3–4 days for predicting the movement of water particles with any amount of skill. The prediction of the movement of oil is even harder due to oil weathering.
A large number of high-resolution, numerical model simulations of the effects of current advection and wind forcing were performed by the Coastal and Shelf Modeling group using the source, initial conditions based on satellite data, the HYbrid Coordinate Ocean Model that has assimilated a suite of ocean data, and the Connectivity Ocean Modeling System.
Satellite observations show the oil blob dispersing in all directions, southward towards the Loop Current, westward along the LA coast, into the Mississippi Delta, and towards the northern Gulf beaches. The projected movement of the subsurface oil is also primarily due to eddies, but the flow is weaker, and it tends to align with the topography.
Our numerical model prediction shows that the oil is getting entrained into the Loop Current.
The left panel shows the initial position of particles in red for a numerical model simulation of the effects of ocean currents on the path of surface particles. The panel on the right is the results of a regional model 7 day simulation of ocean currents that predicts the movement of particles. Red denotes regions where we expect the highest concentrations of particles and the cooler colors, like magenta, are regions where we expect to find low concentrations of the particles. This model simulation shows particles being entrained into the Loop Current and spreading in all directions in the northern Gulf of Mexico.
The prediction work and mathematical analysis of the of the flow in the Gulf of Mexico by the nonlinear group suggests that the oil entering the Loop Current will be elongated and quickly transported into the Florida Current, and then further north into the Gulf Stream.
These 3 images are from a nonlinear analysis of the velocity field at 3 depths; the surface (0 m), the top of the main thermocline (200 m), and in the main thermocline at 700 m by Josefina Olascoaga using the output of a numerical circulation model from the date May 20, 2010. The trajectory of the particles are in black and you can see much less dispersion at depth than at the surface because of weaker currents. The colors denote different dispersion regimes. The red contour acts as a boundary to flow and particles spread rapidly along this contour.
Click on image for larger version
The surface oil further “weathers” as it travels in the strong currents with speeds of 100 to 200+ cm/s (2–4 mph) and our best guess is that tar balls will blown by the easterly winds into the FL Keys, Biscayne Bay, and the beaches of southern Florida. The tar ball distribution will be uneven or “patchy” with higher concentrations in some areas and many areas with none or very little. Further up the coast and into the Mid-Atlantic Bight, tar balls may be ejected from the Gulf Stream into “spin-off” or frontal eddies (first discovered by our professor emeritus Tom Lee in 1970) and end up on more northern beaches along the eastern coast. Previous studies of the circulation of the Gulf of Mexico show very complex circulation patterns because of the energetic eddies and for example, waters in the northern Gulf of Mexico ended up offshore of Mexico. We believe that areas of beach that have historically seen a lot of plastics on the beach will probably get more tar balls.
The tar balls, with gooey oil in the center, will probably end up in very sensitive marine environments such as estuaries, coral reefs, mangroves, and sargassum weed lines known as the nursery grounds of the Northern Atlantic, Gulf of Mexico, and Caribbean Sea pelagic fisheries. The larval stage of the important gamefish that breed at this time of the year in the Gulf of Mexico and the straits of Florida is shown in these pictures from samples collected by UM/RSMAS scientists in the Florida Current.
Courtesy of Bob Cowen and the Larval Fish Group
Wind and current dynamics leads to the formation of ocean convergences and weed lines form. Phytoplankton, zooplankton, fish larvae, shrimp, and crabs are also concentrated in the convergence zones.
Unfortunately, the dispersant-laden, weathered oil will also convergence in the convergence zones and the offshore weed lines. The effects of this oil on marine life is not well understood. Inshore, tar balls will end up on beaches and some of these tar balls will end up in sensitive mangrove habitats, the nursery grounds of our inshore fisheries. Waves and weathering will push the nearshore oil subsurface and in contact with coral reefs.
One question that everyone asking is how bad will this oil spill be for the state of Florida. The ecological impact of the oil spill will be a strong function of the combined toxicity of weathered oil and dispersants, a function that is not well known at this time. It will also depend on how much oil will eventually end up along Florida’s coastlines. This is very hard to predict and depends on winds, if and when the Loop Current forms a warm core ring or not, and the strength and position of the Florida Current. There are some models saying a ring will form, and other saying it will stay attached. If the Loop Current forms a Loop Current ring, a lot of oil will be trapped in the ring and move westward into the western Gulf of Mexico, instead of the state of Florida. A Loop Current warm core ring is a large eddy, about 300 km (200 miles) wide with clockwise rotating flow.
If a major hurricane hits, the whole transport process will be put into fast-forward and time scales would be an order-of-magnitude faster and the penetration of oil into the bays, canal systems, and estuaries would extend much further inland since booms would be useless. Hurricanes would also re-suspend significant amounts of oil from tar mats on the bottom at depths less than 100m. The worst-case consequence could be a major shock to Florida’s already stressed fragile ecosystem whose recovery would be on time scales of decades.