geoLibro

2006.September.7

When GIS Works and That Sucks

Filed under: Uncategorized — geoLibro @ 1:33 am

So this document was left to its own devices for about two weeks and I see that nothing got done. I was busy building a web site devoted to our GIS efforts here (don’t laugh, it actually took a decent amout of work to get the php/mysql/rss combination to all flow together on a brand new dedicated Apache install), and another prototype for a site that endeavors to collect into one place some of the spatial analysis projects that are happening on campus. Plus, I hired a grad. student to help in the lab and with some other work, bought and listened to Dylan’s Modern Times, custom-built some hillshades for a professor (well, you know, adjusted some settings during a conversion from DEMs), and seemed to make some advances toward being a recognized support person on campus. And while I did all of that, this blog sat and produced nothing, which of course disgusts me.

But what I spent the most time on by far was a very interesting project that predated my arrival here by, I don’t know, years maybe. I can’t really say much about it, unfortunately. It’s not secretive or black helicoptor or anything, it’s just not my work and I haven’t asked permission to talk about it. But what’s interesting about it is that these researchers had been collecting and working on data using custom conversion algorithms and their own formulas for this and that but hadn’t really gotten around to putting it on a map yet. So after a considerable amount of time spent in Postgres and MySQL, we finally put some of these resultsets to their respective geographies (grid for one, U.S. counties for the other) and something very interesting came out.

One of the images below has some data attached to it and it all looks okay; there’s some pattern, but it’s not obvious to those unfamiliar with the data what’s going on. Good, most maps are like that. The other one has at least one very obvious problem (no, not the null values in NM). Do you see it? It’s shaped like Florida. Maybe you can’t see it on these tiny versions. The point is that Florida’s gridded values are all much higher than most of the rest of the country, indicating to the team that something was wrong somewhere.

The lesson? Well, if there has to be one I guess it’s either A) get your data to a GIS as early in the process as possible, or B) sometimes GIS can be more helpful than you want it to be.
mostly regular patterns
why you so dirty, Florida?

Advertisements

Leave a Comment »

No comments yet.

RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at WordPress.com.

%d bloggers like this: