The official US govenrment site contains excellent documentation of sightings and a list of sofgtware tools including:
– GNOME:, the “General NOAA Operational Modeling Environment) is the oil spill trajectory model used by OR&R Emergency Response Division (ERD) responders during an oil spill.”.
In fact, you can actually download a copy of the model from this link, and use it yourself (with a Location File, and you can download those as well.)
The site also has a link to ADIOS2, “an oil weathering model that incorporates a database containing more than a thousand crude oils and refined products, and provides quick estimates of the expected characteristics and behavior of oil spilled into the marine environment.” Once again, you can download the software freely.
The BP crisis website has a long list of sampling procedures and sampling results, but does not appear to mention using a simulation to predict the movement of the spill.
Other organisations are getting in on the act:.UCAR/ NCAR has had a lot of publicity for a model whose objective “is to estimate a range of possible trajectories, based on the best understanding of how ocean currents transport material. The oil trajectory that actually occurs will depend critically both on the short-term evolution of the Loop Current, which feeds into the Gulf Stream, and on the state of the overlying atmosphere.”. They have produced a set of six simualtions, based on different states of the Loop Current, which “provide slightly different
scenarios of how the oil might be dispersed. The simulations all bring the oil to south Florida and then up the East Coast. However, the timing of the oil’s movement differs significantly depending on the configuration of the Loop Current.”
Sensibly, they stress that “are not a forecast because it is impossible to accurately predict the precise location of the oil weeks or months from now. Instead, the simulations provide an envelope of possible scenarios for the oil dispersal.” Full dtails and videos are here.
Im very impressed by the way the US government make the actual modelling software freely availble, with data and instructions.
Compare the UK system. According to SEPA, “The MCA has an
Oil/Chemical spill modelling system which uses basic oceanographic information to work out spill movement. In more complex cases the MCA can call on the modelling group at the Centre for Fisheries and Aquaculture Science (CEFAS) to conduct more detailed modelling. CEFAS has a UK wide 8km grid linked to a spill model. They have higher resolution models throughout much of England and Wales, but not Scotland.”
However, I can find little information available via Google about the CEFAS or MCA models:
– this poster, describing in very broad terms the GESAMP model, which is described at more length here and seems to be an IMO model anyway. It is perhaps nearer to ADIOS2 in scope than to GNOME.
– the CEFAS annual reports -eg the report for 2007 which says “Cefas has considerable experience modelling the movement of spilled oil”, but does not even name the model used..
– slightly more information about other models – eg of cumulative impacts of human pressures on the marine environment:.
– a page about Premiam, which sounds as though the process is at a very early stage indeed. This is dated 2010 and refers to ” Engag[ing> with the wider scientific, logistics and marine sampling communities to understand all potential ontributions during impact assessment and monitoring” and “Develop[ing> a web-based portal to promote the PREMIAM approach and to allow the wider community to interactively engage with the process.”
There does not appear to be any reference at all on the MCA website to oil spill simulation models.
There are obvious democratic advantages to openness – US citizens can run US models for themselves whenever they want. The amount of skill required is not trivial, but there are plenty of people nowadays who can do it, and any good PC should be able to handle the most complex model. In the UK, on the other hand, it is not easy even to find out which model is being used; and highly unlikely that the model is available for download, or that details have been published in an academic paper.
There are further advantages of openness, for the simulation process itself.
– if you can do it yourself, the process is demystified. Rather than an oracular pronouncement by experts, it becomes something which is understandable.
– It is even possible that some of the non-experts using the model may be able to suggest ways of improving it. (Sadly, all too often in the British system, Nanny knows best.)
– users will reach slightly different results. This brings home to all concerned the nature and proper use of a simulation – ie, it is not a prediction, but a means of narrowing down the possibilities so you can best prioritise your efforts.