Academy for Tech & Classics\Desert Academy (2)/Interim Report

Interim Report
http://mode.lanl.k12.nm.us/get_interim1112.php?team_id=2

Problem Definition
After a fire burns through a wooded area, tons of ash and debris is left in the environment. When rain falls, it washes this debris into rivers, lakes, and reservoirs, contaminating water resources. Along with Contamination concerns, a fire is capable of changing the ground cover of an area. This may result in flash floods or landslides. The job of a firefighter in these situations is to extinguish the fire to minimize debris and damage to the forest and forest floor. With current firefighting techniques, firefighters may not be able to complete this task or are insufficient in doing so. As a result, a fire may be left to further damage the forest and, in turn, affect human water resources.

Problem Solution
If we are able to accurately generate different forest types, fire behavior, and firefighting techniques, we will gain a better understanding of how to effectively treat forest fires to protect vital water resources. By looking at recent fire damage in the Pacheco Canyon, we will incorporate several factors of fire spread, like how slope and elevation, tree density, and wind affect the spread of a fire.

Progress to Date
currently, we are finishing up research and our forest models, and now are looking into the physics of fire and fire spread, and how firefighters are directed in a fire situation. Recently, we have been emailing Stephen Guerin, and discussing the possibility of our group using the sand table, which is a scale model of an area with a projector mounted above that is capable of showing fire spread in the modeled area.

Expected Results
At the end of the project, we expect to be able to confidently predict how fire will spread through different forest types, how to better protect forests to further protect water resources, and how rainfall or other forms of precipitation effect debris left after the burn.

Review
Hello friends,

I am Reffat Sharmeen, a graduate student at University of New Mexico. I worked three years in Bangladesh as a software developer. Here at UNM, I have been working two years as a programmer and web designer, also with database support. I got into touch with Supercomputing Challenge activities and Project GUTS last summer 2011 and feel honored to be a UNM cadre mentor now. I am supposed to write a review of your project’s interim report which drove me to explore your ideas and share my ideas too.

Based on your interim report problem definition, it seems that you are interested in modeling forest fires, why it occurs, what is the after effect and how to fight against it to protect natural water resources. I am glad to see that you already did some research to model forest fire and looked into some recent history of it. I am sharing some of my ideas that you may find useful for your project acceleration.

The forest fire is extensive in size, propagates at a high speed and changes its direction unexpectedly which makes it uncontrollable instantly. But if you can build up strategies to detect, prevent and suppress these fires, the negative ecological effects can be reduced. To do this, I feel you are right to find out the causes of forest fires first. As you are exploring forest fire history, I am sure you will find out that the natural causes of forest fire also vary from area to area. The human use of fire for agricultural and hunting purposes also contributes to forest fires.

Regarding fire behavior, it is wise to investigate how fire spreads. Usually flammable materials are the main factor here; the vertical arrangement also affects fire behavior.

As forest fires are caused by a combination of natural factors and human interactions, preemptive methods of reducing the risk of fires as well as lessening its severity are needed. Protection techniques may vary for reducing a fire’s rate of spread, fire intensity, flame length, and heat of the area. To lessen ecological imbalance, you can build a forest fire model that can be used to compare different prevention techniques.

To fight against fire, early detection is a key factor. Techniques people tried in last century were not very successful due to limitations in communication technology. Now you can explore the potential of using electronic systems like GIS or GPS in the high risk areas. Wireless sensor networks may also be helpful to detect temperature, smoke or other natural attributes.

To model forest fires, you can do statistical analysis with help of computational science. Analyzing past fire events will be useful to predict future fire behavior. Based on your data, models can be of various shapes also.

Forest fires can cause debris flow which is extremely destructive to life and property. Sometimes it is triggered by extreme rainfall but its high speed and large amount of woody debris inside it make it a problem in steep mountain areas. You can present a model to characterize debris flow and analyzing its topographic effect which will be really important to study forest fires.

My colleague, Celia Einhorn, from the Challenge, suggests that you look at a winning team's final report on the Control and Spread of Fires and perhaps use their work as one of your references. Here is that link http://www.supercomputingchallenge.org/archive/09-10/finalreports/99.pdf

It looks like Stephen Guerrin is your mentor. That is super.

You have a face to face evaluation on Saturday, February 11th at 9 o'clock in room 315 at Santa Fe Community College. Best of luck with that.

Do ask the judges any questions that you might have about your progress.

What language are you doing your models in? Do you need any assistance with that part of your project?

In the end, I would say, you chose a very interesting and important ecological topic and your progress rate is also satisfactory. I wish you to be successful in April challenge.

Team Comments
Hello,

This is Sara Hartse. Thank you for taking the time to write your comments and give us suggestions. Especially interesting is the idea that communications and spatial awareness of the fire is an important variable in effective fire fighting. We had our interim evaluation and received many suggestions. Right now we have a fairly satisfactory simple fire model (based on fuel load, moisture load, elevation and wind direction), simple fire fighter behavior a dynamic landscape generation program and a system for approximating the erosion effects of a particular fire. We are doing our programming in Netlogo. At the moment we are investigating ways we might incorporate GIS data into our model. We are also interested in looking at ways of optimizing the fire-fighter behavior through some kind of computational evolution program.

Thanks so much again for your time. My teammate Hugo would like to add that you are awesome.