Depiction Runoff Element

Nov 19, 2009

The Runoff element allows simulation of water flow from a point source. The simulation can proceed with a limited supply of water or, alternately, the maximum inundation zone (assuming unlimited supply) can be determined. Flow velocity, general hydrodynamics are not modeled. Soil absorption is modeled indirectly as will become apparent. The basic premise here is that topological data alone, derived from Digital Elevation Models (DEM), contains sufficient information to model watersheds and drainage networks to a fairly high degree of accuracy, with the resolution of the DEM, of course, dictating the end result of the simulation.

For more details get the PDF file.

This simulation element will be included in our 1.01P version that is slated for a year-end release. This will be a free update for our current users. Likely, this will supplant the somewhat experimental Fluid Flow element.

As something of a political junkie, I've always wanted to be able to make detailed maps of election results. These days, it's pretty easy to get county-level election maps for most statewide contests--but not all contests are statewide, and a county is a pretty big chunk of land, and usually not as homogenous as the county maps make them look. What I've always wanted to do was to take the basic precinct-level voting data that counties release, and turn it into a map.

There are some very nice, and very expensive tools for doing this kind of mapping, and there are companies that do very well making precinct-by-precinct maps of election results for various organizations. If you are comfortable doing some programming, you can do some very nice stuff using Google Earth. These folks did several states that way using 2004 election data.

But neither of these processes are at all easy--and even using Depiction hasn't been easy. It's very easy to load precinct shapefiles that are available from many county GIS pages into Depiction. And once it's in there, you can add properties like candidate vote percentages, and then go through and update them. The problem is that, when you have hundreds of precincts, it doesn't make sense to go through and insert votes for each of them by hand. And there hasn't been an easy way to link massive amounts of outside data to shapes in Depiction.

Until now! Our whiz-kid intern Daniel has built a fantastic Excel macro that will match vote data to precinct numbers in a GML file (the kind of file exported by Depiction). It could probably also be used for things like census data and such, but like I said, I'm a political junkie.

So I built this map using the election results from Fairfax County, Virginia, from Tuesday's gubernatorial election there.

At some point I'll write a detailed tutorial on how I did it, but here's the short version: 

• I downloaded the precinct shapefile from the Fairfax county GIS page.
• I got the precinct-level election results from the State of Virginia and put them into a spreadsheet (this was the hardest part, because they aren't presented very conveniently, but some creative copy/paste/filter work made it work).
• I imported it into Depiction, then added two fields to the precinct shapes: McDonnellVote and DeedsVote.
• Then, I exported it all to a GML file, and used the GML Data Inserter macro to insert the McDonnel percentages, and then the Deeds percentages, into the GML file.
• Then, I imported it back into Depiction and fixed a couple errors.
• Last, I selected all the new shapes, and used the 'thematic mapping' colorization option to shade the precincts based on their votes.

And there you have it! All in all, this probably took less than two hours to do. And the beautiful thing is, once this data is in a depiction, there's so much more you can do with it! I could add more voting data and compare the past to the present. I could bring in other juristiction shapefiles and look for useful voting patterns within them. Or, best of all, I could use the same depiction to plan the logistics of an entire campaign, with all the important vote data right there.

I have had a few questions lately about the quality of our imagery and maps. There is some confusion on how Depiction handles maps and imagery, actually. When you select a map or image source in the Quickstart Data list, you are getting a static image at the zoom level of your depiction area. If you're using 1.01.n, they will be put into Revealers and you will still be able to see the active map or image source you've selected for your background in the lower left corner of the Depiction screen. If you don't have "N" yet, I highly recommend you run the Update program from the Help menu or the Depiction folder on your Windows Start Menu.

The background map or imagery is a live feed from a web server that changes as you zoom in and out, so that it will get better as you go in: not fuzzy as when you have a static image. Depiction also stores all of the tiles in a cache file, so the more you zoom and move around in a given area, the faster the performance will be.

I will add some images later, and I hope this will help clarify things for anyone reading.

Happy Depicting!
George'
Depiction Support Manager

 The question of positional accuracy of Depiction comes up often since our app allows exceptional capabilities to the user to integrate data from diverse sources and coordinate systems. In addition, by allowing these data to be visualized in revealers Depiction allows the user to expose positional inaccuracies between datasets more so than with any other app.

 Do you want to see how accurately (or not so) the NED elevation data (30m resolution) lines up with satellite imagery, street maps, or topography data? Depiction allows you to do that via revealers. See this posting for example: http://depiction.com/node/308#comment-136

 As a simple experiment, I took a geographic (latlong) coordinate and decided to compare it visually between Depiction and Google Earth. The location was: 47.64540N, 122.30141W

Here it is in Google Earth (pushpin at the NW corner of the building at the center):

And, here it is in Depiction (satellite imagery backdrop is from Terraserver, which is originally in UTM and Depiction reprojects it to Mercator). Can you tell any appreciable difference in location?: