New Mexico Supercomputing Challenge
 I About I Resources I Help I Archive I News Flash I

Registration

Proposals
Interims
Final Reports

Dates

Kickoff
Evaluations
Expo
STI
Wiki

School Map

Mail

Challenge
Technical Guide

Past Participant Survey

GUTS

Challenge Team Interim Report

 Team Number: 049 School Name: Los Alamos High School Area of Science: Earth Science Project Title: 3-D Erosion modeling
 Abstract Interim Final Report

Our goal is to construct a geological model of erosion caused by water (fluvial erosion) that reflects the change of the landscape over time. We want this model to incorporate the effects of the groundUs mineral composition and its resistance to erosion.

To solve this problem, we would start with storing the information needed in a computer data structure, of a to be determined type. This requires a two-dimensional array of altitudes to describe the shape of the landscape and and the erodability (resistance to erosion) factor for a given depth of the ground composition under the surface. For this information, the best storage would probably be a three-dimensional array (length by width by height) of landscape filled with integers standing for erodability of the matter in question. For example, a 0 would stand for nothing, 1 would stand for easily washed out material like sand , and above that the integers up to 10 would stand for a range of rock types. A necessary part of fluvial erosion is to find the river tree, which will show how the water will flow over the land and transport the eroded sediment to lower altitudes. This could be done using another two-dimensional array of slope gradients derived from the altitude array and and find the courses of the streams. To find out how much water flows in the rivers over a time interval (water rate of flow), a two-dimensional array of the rainfall pattern in that landscape will be randomly generated to be dependent on altitude. Finally, the transportation of the soil can be simulated by modifying the three-dimensional landscape array to move a member RcubeS from its original slot down the river path to a new slot where it would be deposited as sediment. The entire model will function in iterative cycles where each cycle would represent a period of geological time. Each cycleUs landscape will be saved into a big file to be represented as frames in an animated image.

So far, our progress has been mostly to finish all of the paperwork and non-scientific organization required for the contest such as talking over sponsorship issues with the high school, filling out and sending necessary paperwork, and assembling a team of five members, a teacher, and a mentor. Also we negotiated with the school administration for use of a laptop computer. Starting research about how fluvial erosion functions was done by looking at the geology and chemistry textbooks available at the school and talking with several teachers. We have looked at existing commercial, agricultural, municipal, and scientific models on the internet. Now we have gone to the Los Alamos National Lab's Otowi library and found and read books on geomorphology. We have completed and initial computer program which generates a random canyon-like landscape based on plasma fractals and then finds the peaks and river paths. The source code and a sample landscape are available at . For graphical output, we are using Mathworks MatLab 5.3.

Our anticipated result is an animated landscape being gradually eroded. Each frame will be a plot of the previously mentioned three-dimensional landscape array.

Team Members

Team Mail