Any Graphics folks interested in helping NWAC?

Craig Miller has proposed a three dimensional , colored version of the avalanche rose. Craig suggests a three-dimensional conical mountain with trees appearing in green. I might suggest creating a cone with N, NE, E, SE, S, SW, W, and NW sectors. The cone would show more information if each sector had one gully and one ridge extending above and below treeline. This way surface hoar and wet snow conditions could be displayed in the gullies and wind loading on the ridges.<br><br>Some of the complications are that end product will need to be several different cones (although all may not be displayed on a particular day): Olympics, N Cascades, S Cascades, Central Cascades, and East Slopes Cascades. It will probably also be necessary to display Pass level winds and slopes some way, since winds may load these slopes opposite slopes aspects than prevailing winds.<br><br>Elevations will be relevant in the Cascades and Olympics because of the effects of variable freezing levels.<br><br>It will also be necessary to have the 3-D avalanche rose display in some way as to reflect change in hazard with changes in weather conditions over time.<br><br>Finally, the system must be set up in such a way as it is easy for forecasters fill in the appropriate fields (or select the appropriate cones) and for users to understand the information presented. It must interface with the current NWAC system. <br><br>Mark Moore has said he would be interested in this product. He suggested that funds might be available through avalanche.org for creating this product. If the product is quite good, it would not be surprising the other avalanche centers would be interested.<br><br>If you are interested, I would suggest posting a reply here, then perhaps creating a mock-up of the initial product. I can help to get the ideas across. Once the initial mock-up is ready (that can be duplicated for other areas in the region, then I'll have you, whoever you are, work with Mark Moore directly in working the kinks out of the system. At that point I'll back out of the project. This would be after the conclusion of NWAC's forecasting season.

Hi Gary et al,<br><br>I want to make sure I understand. The purpose of this new graphic is to enhance the 'traditional' avalanche rose, example at the following URL:<br><br> geosurvey.state.co.us/avalanche/Default....=107&area=2&nav=last <br><br>The idea is to enhance the graphic with more detail such as:<br>- gullies and ridgelines for each aspect<br>- winds (pass and prevailing)<br>- wind loading<br>- changes over time<br><br>Requirements:<br>- end users (backcountry skiers, highway crews, 'bilers etc) must be able to quickly understand and use the information<br>- Must integrate well into the NWAC workflow and information system<br><br>I will think on this and try to provide a few sketches. My first impression is that this is more data than can be easily and quickly understood (without training of end users) on a graphic. Perhaps we should have two that sit side by side - one shows avy detail and one shows weather detail.<br><br>I would be curious what the Colorado Avalanche Information Center think are the weaknesses of their system.<br><br>Anyway - I will try to whip something up and post.<br><br>Michael

Michael,<br><br>Thanks for the reply. As to topofusion, I noticed that NW aspects were not visible because of the perspective. Also, on interfacing, a complex diagram would be difficult to program because of all the various sectors.<br><br> The idea is to enhance the graphic with more detail such as:<br>- gullies and ridgelines for each aspect<br>- winds (pass and prevailing)<br>- wind loading<br>- changes over time<br> <br>Requirements:<br>- end users (backcountry skiers, highway crews, 'bilers etc) must be able to quickly understand and use the information<br>- Must integrate well into the NWAC workflow and information system <br><br>That is pretty much correct. I'm going to stay with this project until we have a capable graphics person and a reasonable initial sample. Afterthat, I'll back out. Mark did volunteer that he thought funding would be available, but that is something the graphics person would have to work on with Mark (apply for a grant). <br><br>One possible idea is that there could be several maps and when you click on them, say 3 or 4 appear for that area (example E slopes Cascades), each depicting a different time period.<br><br>But brainstorming is welcome here. I'm only one person.

I work with web-based GIS. I've often wondered if there would be interest in creating an interactive online topo map of avalanche terrain overlaid with a color coding system based on parameters run through a model that uses DEM data. On the forecaster side, conditions data would be entered in database for the present time, +4 hours, +8 hours, etc. for each aspect in a number of elevation zones and geographic regions (and maybe even slope angles). This information would be run through a filter that color codes areas of a web-browsable topo map (a la Topozone).<br><br>Using an open-source solution like UMN MapServer and TerraServer topo data it could be done relatively easily, at a low cost, and would not require the pregeneration of images or storage of large datasets. The only data that would have to be stored locally would be the DEM data (there may even be a way around that). All parameters would be stored in a database and maps would be generated on the fly by the mapserver as web users request them. <br><br>From the user perspective, one would search for a location on the NWAC site, pull up a topo map, and then view the current and forecasted danger rose applied to the area they plan to visit. <br><br>This could conceivably be made "3D", however 3D mapping over the web would require a plugin and would make the user interface more complex. I would advise against it.<br><br>The hardest part will be coming up with an algorithm that sorts the DEM data into blocks useful for avalanche forecasting. A DEM is just a grid of points with elevations associated with them, what we want is a grid of points with elevation, aspect, and slope associated. That task could easily be done using commercial GIS software, but it may also be possible using free software as well. <br><br>So anyway, it is possible... the real question is whether NWAC wants to be publishing computer generated maps that say a particular slope is or isnt safe. A danger rose is much more generic, and probably safer from a liability standpoint. But conceptually, it's no different than a weather model...<br><br>Thoughts?

Jerm-<br><br>I had the same thoughts running through my head as well (I'm also a plotboy- er- GIS "professional"). I think we both agree that the DEM data is freely available- either from the UW Library data website or even through one of NOAA's websites (though it's 90m). I'm a little weary of processing time. Although the end product is the result of several "basic" overlay operations, the mere fact that you're going to run it on an area as large as the Cascades makes me wonder how long it would take to run & if that could be done on a daily basis (especially since that DEM should *probably* be 10m or at most, 30m resolution).<br><br>Maybe a sample image is in order?<br><br>I think your biggest point is whether or not the NWAC would want to post "maps" of danger zones like you described..

Using a WMS mapserver you would only have to process what is actually being viewed, when it's viewed. As long as you don't turn on the overlay function at really small scales, the dataset actually being worked on at any given time would be pretty small. Of course, if NWAC's traffic went through the roof it would require some processing power (each map starts it's own process), but my guess is that wont be a problem. We handle pretty big traffic on a relatively weak machine, with many layers being overlaid on a 24 bit image (using UMN MapServer).<br><br>Do you have access to spatial analyst? I do not. If you can make me a point shp file (or three seperate GeoTiff files) with attributes for elevation, aspect (strike), and slope (dip), I can set up a sample map page...<br><br>edited to add: Actually, it looks like you can do this using GRASS (an open source GIS client) ... which I was planning to install on my machine anyway. I'll give it a shot tonight.