Flutterby™! : Geography

Next unread comment / Catchup all unread comments User Account Info | Logout | XML/Pilot/etc versions | Long version (with comments) | Weblog archives | Site Map | | Browse Topics


2004-12-06 17:36:29.086601+00 by Dan Lyke 8 comments

So the lesson of late last week, and what I have to solve today: Lattitude and longitude have somewhat nebulous definitions, altitude is even worse, and a phrase that I formerly thought was born of ignorance, "GPS coordinates", is actually more precise than saying "lat, lon and altitude". More explanation on what all of that means when I figure it out. I'm sure Eric is snickering about now...

I just hope this planet and its close environs is the only place this software ever has to model...

[ related topics: Software Engineering Space & Astronomy Maps and Mapping ]

comments in ascending chronological order (reverse):

#Comment Re: GPS coordinates and mapping. Confusion. made: 2004-12-07 03:36:32.982585+00 by: Lisa Williams

Hey, Dan -- I have a GPS too, and I'm always really confused by how different the results are when I type in the coordinates from my GPS into something online like Terraserver.

For instance, I'll type in the coordinates I get from my GPS for the location of my house into Terraserver, and it will return a result quite far from my house.

I notice my GPS has many "standards," like NMEA etc for reporting coordinates. I have no idea if this has anything to do with it.

If you figure any more of it out, please blog it and remove my ignorance. If I figure it out I'll blog it too.

#Comment Re: made: 2004-12-13 04:35:34.798842+00 by: ebradway

Hmmm... I managed to miss this discussion - and yes, I was smirking ;)

Welcome to the world of cartography. The issues you are trying to cope with are as old as time. Here are some basics:

  1. All maps are based on a geometric transformation. That transformation takes the surface of a relative sphere (the Earth) and maps it on a 2 dimensional piece of paper.
  2. All transformations trade-off accuracy in feature shape, area, and distance. A map that you can accurately use a ruler to determine distance on will have feature shape and area all out of whack. Mercator's famous projection is a great example. You can use it with a ruler to accurately determine distances along the Trade Winds, but look at Greenland!
  3. Surveying and platts are all based off of established control monuments. Similarly, GPS has created it's own points of reference - the satellites that beam the microwaves to your receiver. They are generally very good at reporting the amount of error in their position (about 10m in Long and Lat and usually 30m in altitude) but when you save the location to a file, the accuracy information gets lost. Ideally, you would save the long, lat and alt with accuracy info for each. But the satellites themselves to create controls.
  4. Longitude and Latitude are just one coordinate system, there are dozens of others. And longitude and latitude are sometimes represented in difference formats: decimal degrees, degrees-minutes-seconds, decimal minutes, etc.

Lisa: Let me see what you typed into Terraserver. I'll let you know what's up.

#Comment Re: made: 2004-12-13 06:35:29.870918+00 by: Dan Lyke

All true, but #4 is an understatement and also ties in to #3 a bit. WGS84, which is the coordinate space that GPS eventually reports in (internal calculations are done in ECEF), longitude and lattitude are one coordinate system, but there are at least two other coordinate systems in extensive use the U.S. that use lattitude and longitude that can differ by 500 feet from each other and from WGS84, and in the world there are closer to 1800, all of which just say "lattitude and longitude". And a good number of the explosion of those permutations occur in the middle east. You see why I might be aware of them.

And #1 and #2 have implications on local positioning systems which are often very accurate, but which report their coordinates in X,Y meters. Mapping that back to a map transform is... fun.

#Comment Re: made: 2004-12-13 06:39:16.336066+00 by: Dan Lyke

And, oh yeah, let's not even start on altitude. Let's just say that saying "highest mountain" around modern cartographers can start gales of laughter and debates about lumps you never heard of, and Everest and K2 may or may not make that list.

#Comment Re: made: 2004-12-13 10:09:38.290117+00 by: Mars Saxman

ehh, Chimborazo?

#Comment Re: made: 2004-12-13 16:37:18.403422+00 by: ebradway

Differential GPS units record all the ECEF readings for all satellites they can pickup. They do this to allow you to remove some of the error introduced by atmospheric effects (the effects of "bounced" signals can't be removed). The readings are corrected based on a nearby base-station reading (that knows it's precise location and can calculate the shift). Probably the most accurate system would be to just retain as many ECEF readings as possible.

I just finished writing a paper for one of my classes about liability issues in map making and GIS. I focused on the issues of data accuracy. The way we bat around co-ordinates right now is very scary. Fortunately, the military uses their own coordinate system (military grid) that has been tried and tested. It's designed to work well for guys on the front line to call in coordinates for artillery. But the coordinate system is too cumbersome for small-scale (large extent) areas. That is, it works great if you're trying to take out your neighbor, but giving the location across the city or state starts becoming a rather involved process.

#Comment Re: made: 2004-12-13 18:01:33.701616+00 by: Dan Lyke

Aha! Thanks, the term "military grid" let me pull together a couple of other things I've run across and tie what people have been saying about "meters in X and Y" to UTM and UTP, and discover the next big thing I need to search for reading on: MGRS. I've been using UTM as my "this is good enough until I get told exactly what I should be doing", it turns out that was the right decision.

#Comment Re: made: 2004-12-13 19:28:10.489222+00 by: ebradway [edit history]

Military grid uses a series of numbers and letters. Each number and letter denotes a progressively smaller unit of measure within the grid. Here's a pretty good reference And here another.