Spire Measure revisited

[TWorth]

Here's an initial kick at the can at something of an "improved" Spire Measure. The programming wasn't as difficult as I was expecting, but the data crunching and downloading of NED data has taken a long time. Though more work is obviously needed, the results so far are interesting, and seem yield better scores for "spire-like" peaks compared to the original Spire Measure.

http://www.geocities.com/tworth87/calc.html

-Tim W

[JoeGrim]

That's really neat, Tim! I look forward to seeing the final list!

[John Kirk]

Tim,

Looks like we'll have some interesting discussion material for West Elk Peak's hike. I started building a SQL database to house the coordinate/elevation set for the whole state, but haven't gotten too far with so many other active projects. Best I've been able to do for mass downloads is nine or ten quads at a time, then using a Freeware tool called MicroDEM to extract the coords/elevation, one quad at a time. Makes for a painful process if doing the whole state. There is slight loss in data integrity with that too, because the tool requires a re-projection to spit out a table. I'm guessing you might have a "less touches" dataset extraction method?

[TWorth]

Joe: Thanks for the comment, I agree, it will be interesting to see which peaks score well. Highest score I've seen so far is for a peak in the Gore Range.

John: Yes, we'll have to go over some of the finer points. As you probably found, developing a set of "radius coordinates" isn't too difficult - simple trigonometry with an adjustment for longitude distance based on latitude. Getting the elevations for each coordinate pair - a bit tougher. I used a GIS program(http://www.globalmapper.com/) to export elevations from 1/3 second NED data in .BIL file format. It's a slow, multi-step process, and the ASCII output files are somewhat large(100+ MB), but its much easier than going quad by quad and the elevations seem to be accurate. The NED data is available at http://seamless.usgs.gov/website/seamless/viewer.htm

The third step is to plug the elevation differences into a formula with correct weightings. So far so good on that, though assigning the weightings wasn't simple, as adjustments had to be made for varying distances of the intervals(25m - 200m).

[John Kirk]

By the way, I'd really like to build a SQL database with elevations for coordinate pairs for the whole state (mountainous portion at least), so let me know if you still have the ascii files.

Looking forward to that Gunnison list - Currecanti Needle has to have a super high value.

[TWorth]

Ah, I forgot to link to Gunnison:

http://www.geocities.com/tworth87/gunnison.html


I have the ASCII files with all the coordinates - still making some changes at the moment but I'll send them along when they stabilize.

Some of the higher scoring peaks(Peak C, Sharkstooth, Elephant Head Rock) are going to have to be re-figured since the summit coordinates are sometimes a bit off which affects the numerical rating.

[JoeGrim]

Tim,

I'm curious why you weighted 59% to the 0-100m radius range and 41% to the 100-800m range. Is there a scientific/mathematical reason for this? Or are they numbers you arbitrarily chose to use? Likewise, how did you choose the radius intervals? Just curious how you decided to use them. Looking through the stats on your site is definitely interesting; it adds yet another reason to climb peaks.

[TWorth]

Joe:

Yes, those figures are arbitrarily chosen. The difference between this and the original spire measure is A) hard bounds are set on distance and B) drop values are looked at discretely.

This measure would rather use arbitrarily set bounds at the expense of mathmatical purity if it provides a more representative measure of the intended targets than a non-arbitrary scheme. The measure is still objective within the bounds chosen.

For distance, the 800m value is the key, the exact intervals within the 0-800m range are less important. I ran the analysis on different sets of intervals within 0-800m, with little change to the final results.

For weightings, I have changed the 59-41 to 50-50, again, with not much change in results. Other weightings being processed are 100% on 0-25m, 100% on 0-100m, and even weightings from 0-800m. I'll post those at some point. Will probably get to do more on this with winter approaching.

It's true, there are some obscure peaks high on the lists which would be fun to check out in person.

[JoeGrim]

Thanks Tim for the explanations and for taking all the time to compile the data to make the calculations!