Using GIS for Avalanche Forecasting on Mt Rainier

[TheNumberNine]

I'm in an intermediate GIS course at UC Santa Cruz right now and I thought it would be fun to create a data model for avalanche forecasting on Mt Rainier for my senior exit project... I'm writing up my proposal right now.

 

My plan is to create a index of the mountain based off of an elevation and land cover datasets and input "scenarios" of weather patterns to see if I can predict where avalanche danger might be highest given that scenario and pre existing conditions. It's not exactly modeling real time avy danger but I think it will be a fun project.

 

Deets will be shared as I continue to work on it...

 

Midterms suck...

Edited February 8, 2012 by TheNumberNine

[DPS]

This is a great idea. In my previous job I did a lot of mapping of environmental data for Web based apps. I toyed with the idea of developing a client side scripted map (i.e. Google Maps) to map avalanche risk in near real time.

 

In my opionion, one would need to do the following:

 

-Create polygons on the map using he same rubric the avalanche forecasters use. Aspect, elevation, slope angle, position, etc.

 

-Include telemetered data including temperature, recent snowfall, wind, etc to fine tune the existing forecasts.

 

-Avalanche forecasters would need to be consulted to come up with a workable algorithm.

 

-The polygons would be shaded with different colors to indicate varying risk levels (green, yellow, red).

 

You could need a web service or at least do a web scrape to get data for a near real time map that would be usable to climbers, skiiers and other winter back country travelers.

 

I would personally not use ESRI (user hostile software as one collegue called it). Google maps have enough data to do this, although drawing the polygons would slow the map rendering a lot.

 

 

 

Edited February 8, 2012 by DPS

[ryanb]

This is a cool idea and I agree with DPS about avoiding ESRI.

 

My site hillmap lets you find slopes steeper then 30 degrees in google maps by doing a pixel by pixel analysis of raster dem data loaded from a usgs webservice in javascript and overlaying the results on google maps. I know other people are working on server side google map layers that incorporate aspect as well...feel free to hit me up if you want to talk about technical details.

 

Also check out http://avalanchesafety.blogspot.com/ ... he has done some really cool terrain analysis including slope, surface regularity, line of site, and local wind pattern that suggest a bunch of things that contribute to avi danger more then general aspect.

 

Edit: one thing i would try to be really careful about is preserving uncertainty. We don't have a great model for avalanches so you don't wan't to get too specific. Turbulence is hard to analyze/predict and you don't want to indicate that a SE slope is safe when wind eddying around a terrain feature and blowing against the prevalent direction could have wind loaded it. Or hoar frost could have formed on a south aspect in a clear cut shaded by trees. There is lots of great info about this on the blog I linked and in the Avalanche Handbook.

 

 

Edited February 8, 2012 by ryanb