|High Resolution Numerical Modeling of the Gulf of Mexico|
|Thursday, 06 September 2007 06:07|
Understanding the physical conditions in the waters of the Gulf of Mexico is a vital part of understanding almost everything about the Gulf - from the movement of storm systems through the sustainability of its fisheries. But the physical conditions of the water - its temperature, salinity, surface height, and the movement of its waters - are affected by many, many factors and these factors interact in complex ways. The driving forces of rainfall, wind, heat flux, and evaporation are complemented by the input of fresh water from 30 rivers and the large-scale movements of water into and out of the Gulf through the currents running through the Caribbean. And if these effects were not sufficiently complicated in their own right, waters of different salinities have different densities and there are complicated features of layering and mixing that can themselves affect the other measurements.
And yet it is vital to understand how these myriad features combine to create the physical conditions in the Gulf. Tropical storms propagate faster when surface temperatures are higher and the distribution of certain fish species can depend upon water currents in particular ranges of depth.
The high-resolution modeling being performed by Professor O'Brien and his colleagues is designed to allow scientists to perform virtual experiments on the Gulf. For example, scientists can look for the likely consequences of increased freshwater input from increased rainfall or the potential effects of decreased freshwater input that might result from increased withdrawals of river water upstream for urban development or agriculture. The "virtual Gulf" in this model is really an enormous system of equations that describe how all of the features of wind, heat flux, fresh water input, and others interact with one another to create the conditions in the Gulf. The "virtual Gulf" has a resolution of 5 km and includes 60 layers of water depth in its calculations. This model, which requires enormous levels of data memory and high computing power, is a classic type of application of supercomputing capability.
To learn more about this research, visit the web site at
Professor James O'Brien, Dr. Steven Morey and colleagues