|New Mexico Supercomputing Challenge|
Challenge Team Interim Report
Problem Definition: Our project this year is a simulation of a small closed body of water. We are investigating the effects that various organisms and chemical imbalances have on such a pond. The chemicals will include oxygen, phosphates, nitrates, and hydrogen ions. Other factors we are simulating include sunlight, temperature, and possibly evaporation.
Problem Solution: We have examined some different fluid flow models, and concluded that the best ones involve partial differential equations. We will start with a two-dimensional representation of a pond with some simple organisms and take into account fewer chemicals. Eventually we plan to expand our model into three dimensions with more complex organisms and to add more chemicals. We have decided not to include fish because of their wholly unpredictable influence on fluid dynamics. Initially, we will ignore some of the empirical aspects of ponds such as seepage and evaporation. Later, if we are able to complete the research, we will add these elements.
We also plan to divide the two dimensional pond into a grid. Once we implement the three-dimensional pond, we will divide it into cubes. Each cube will be responsible for tracking the chemical levels and organisms inside it. We will also include code that will be responsible for mediating interactions between each cube and its neighbor.
These aspects are interdependent with the organisms that are thriving and perishing in the different parts of the pond. This implies a great deal of complexity in our simulation. As a result, we most certainly need a supercomputer. We will simulate the pond in discrete time steps. One way to increase the accuracy of our model, given the resources, would be to monitor the different areas of the pond for high levels of activity; when we find such an area, we will dedicate additional processing power to that area and divide the area into even smaller cubes. This would provide a higher resolution for the information gathered about the pond.
Progress to Date: Our progress to date has mostly been in the planning and research stages. We have examined some fluid flow models and decided that we have to use partial differential equations to determine. We will begin coding soon, and hope to have our code completed and the data collected by early March.
Expected Results: Although our simulation cannot possibly accurately model every tiny detail of a real pond, it will be able to predict the large scale and important effects of chemical dumping and other natural and unnatural factors on a pond. Because of the complex interactions of the different facets of even a small body of water, our accuracy will increase drastically as we simulate more and more organisms and chemicals.
Applications: Our simulated pond can be used to investigate the effects of harmful chemical toxins or even heated water being poured into a pond. These kinds of dumping can harm the delicate ecosystem of a pond. We could determine whether or not to build a factory next to a pond by testing the effects of its output on a simulated pond. A heat source, such as a factory could be just the kind of harmful factor that could offset the balance of a pond. Gathering empirical data on a factory's effects on a pond defeats the purpose since the pond is ruined in the testing. Our simulated pond would be a better choice for testing.
For questions about the Supercomputing Challenge, a 501(c)3 organization, contact us at: consult1516 @ supercomputingchallenge.org
New Mexico Supercomputing Challenge, Inc.
80 Cascabel Street
Los Alamos, New Mexico 87544