Sunday, November 17, 2013

The Harriman Geographic Index System

A few months back I picked up a packet of US Army training regulations and manuals that were published in the 1920s and deal with mapping and aerial photography.  Army publications from this era don't often appear on eBay, and those dealing specifically with mapping, surveying and related topographic sciences are even rarer.  In fact, after years of hunting on eBay for historical publications dealing with these topics this was the first time I'd ever seen any from the inter-war period.  My guess is that virtually all outdated documents got heaved into the garbage can in the late 1930s as the Army was ramping up for war and new publications covering map production and map reading were introduced.

I was surprised to find myself in a small bidding war for these documents.  I'm sure it wasn't against anyone with a specific interest in Army topographic history.  The other bidder(s) were more likely motivated by the relative rarity of the documents.  In the end I paid about $30 for the packet and at the time I thought I'd bid too high.  As it turns out I think I made a good investment.

The packet included five Army Corps of Engineers publications:

  • Training Regulation 190-7, Topography and Surveying, Map Reading - The Harriman Geographic Index System (July 15, 1927)
  • Training Regulation 190-25, Topography and Surveying, Topographic Drafting (June 21, 1923)
  • Training Regulation 190-27, Topography and Surveying, Aerial Photographic Mapping (January 23, 1925)
  • Training Manual 2180-35, Topography and Surveying, Special Methods of Relief Representation (January 3, 1928)
  • Training Manual 2180-45, Topography and Surveying, Meridian Determination (April 16, 1928)

As a group these manuals represent an interesting view into the evolution of Army mapping activities that incorporate the lessons learned and the new technologies that emerged from our experience in WWI, particularly the use of aerial photographs as map substitutes and as base data for topographic map compilation.  In these documents you get the sense that the Corps of Engineers is starting to realize that it now has significant responsibility for providing standard map products to a modern Army with a potential world-wide mission.

Most of these publications cover topics I'm well familiar with, but the Harriman Geographic Index System is something I'd never heard of before.

Click on the photo to enlarge

Once I read through the document I realized that the Harriman system is designed to allow a Soldier to accurately locate himself or any feature on a map to within a few hundred feet anywhere in the world.  In essence it is an early worldwide grid reference system (although it bears no resemblance to the current Military Grid Reference System).

It is also very complex.  While the mechanics of the system were fairly easy for me to figure out, I can't imagine myself standing in front of a classroom of Soldiers trying to teach this system.  It might have been a useful tool for well educated officers working in the relative comfort and calm of a rear-area command post, but for a tired, cold and scared draftee with a 9th grade education who is sitting in a muddy foxhole trying to call for artillery fire support this system is all but unusable.

The Harriman system uses the South Pole as the origin point and the International Date Line as the meridian.  It successively divides up the Earth into smaller and smaller rectangles based on latitude and longitude.  Each of these rectangles get an index number in the Harriman unit system that, when combined, permit locating features to within about half an acre.

Harriman 'units'

Since the Harriman system is based on latitude and longitude it is projected onto a spheroid.  This means the land areas defined in this system vary with latitude.  The further away from the equator the smaller the land area encompassed by a Harriman system rectangle.  This also means the Harriman system is not a point identification system like the Military Grid Reference System, but is an area reference system that defines smaller and smaller rectangles on the face of the earth.  The smallest area that can be defined in the Harriman System, the Position Unit, is 2 seconds in latitude and 1 second in longitude.  This equates to about a 4,875 sq. ft. 'box' at 49 degrees latitude, or about a 70 ft x 70 ft area on the ground.  The Harriman system has the potential to identify a point feature such as a road intersection with a positional accuracy that is well within the map accuracy standard for a 1:50,000 topographic line map.  From a practical perspective Harriman's system is accurate enough.

Harriman Index Geograph

But it's the identification of these units that can get confusing.  The Harriman system requires the user to concatenate an ever longer string of numbers, separated by colons, semicolons and slashes, to identify locations.  The smaller the area the longer and more confusing the string.  For example, according to the manual the Harriman system ID for Battery Byrne at West Point would be designated as 2665:4515; 7792. Users of the Military Grid Reference System (MGRS) could argue that this system can be just as confusing. However, I'd counter that the MGRS use of the grid zone designation (ex: 18T) and the 100,000 meter grid zone ID (ex: WL) alpanumeric system takes a lot of confusion out of sending and receiving coordinate locations.  For example, the MGRS coordinate identifier for the same Battery Byrne location is 18T WL 8726 8316.  Perhaps it's my 30+ years of using MGRS that has me jaded, but I just think MGRS is less confusing.

Now the Harriman system isn't a bad system.  In fact, it's quite logical and it works well within its known limitations.  And I have to be honest, before the Harriman system there was... nothing.  The Chief of Engineers was quite clear about what the Harriman system is and is not:

"It should be realized that the Harriman index system is in no sense a method of map making or of chart building; still less is it a new system of projections.  It is merely a simplified system of using an established arrangement.  It may be used on any map or chart, regardless of projection or scale, provided the longitude and latitude of the southwest and northeast corners are available or can be determined by scaling on the map.  Since only arabic numerals are employed in location designation, this system is capable of use in any language."

In the late 1920s the federal government seemed quite enamored with the system and there are indications that a number of federal agencies had adopted it.  In 1928 Congress actually held hearings to decide whether to purchase an unlimited use license from its developer, George C. Harriman. But other than a few tangential references on the web I can't find any more discussion about it.  This training regulation is the only full reference I've found.  Even more interesting, when reviewing Army topographic references - training and field manuals - published beginning in the late 1930s as the US Army ramped up for war, I find no references to the Harriman system.

It appears Mr. Harriman's system was a flash in the pan, dropped by the Army in the 1930's as the Corps of Engineers realized it needed a better map coordinate system to address the exploding world-wide mapping requirements.  The Army needed a coordinate system that was logical, consistent, accurate and easy to teach to the millions of draftees about to be deployed to battlefields around the world.  It was out of this requirement that we got the Military Grid Reference System, a system stood the test of time and war.


Wednesday, November 13, 2013

The State of Mobile GIS Software

Over the past six months or so I've been doing a lot of casual testing of the various mobile GIS platforms available on the market today.  Right now is an ideal time to discuss the offerings because just in the past week we've had an update to a key application in this arena (Collector for ArcGIS), we are on the verge of having an interesting new hardware player enter the market (Garmin's soon to be released Monterra handheld GPS) and several vendors are dropping serious hints about where they see their products headed in 2014.

I was ramping up to do a lengthy blog post on this when I dropped by Alex Mahrou's always interesting RockyMountainGeo GIS blog and was surprised to see he had already done all my work for me.  Back in October Alex did a great overview of the current offerings in a posting titled Enterprise Mobile GIS Software Functionality.  All I can do is add minor updates to some of his information and add a few of my own observations.

I like the switchboard analogy!

The single biggest update is the newest version of Collector for ArcGIS (version 10.2) that was released last week for the iOS and Android platforms.  This version addresses one of the two biggest complaints about earlier versions of Collector - polyline and polygon data collection.  It also offers an improved user interface and well thought out workflows.  While the Android version still has some rough edges, the iOS version is a polished, smoothly functioning app that reflects ESRI's mature experience in developing for Apple's mobile operating system.  It is a very good app.

Where ESRI seems to be unnecessarily holding back is off-line data collection and editing, and data synchronization.  As Alex notes, ESRI informally promised that this feature would 'absolutely, positively' be incorporated into Collector before the end of 2013.  It now looks like we'll have to wait until sometime in early 2014, when ESRI plans for a significant overall upgrade to Collector, perhaps better positioning it within their enterprise software offerings.  In my opinion ESRI missed the ball on this one.  Incorporating off-line data storage and editing in the iOS and Android operating systems isn't hard to do; Trimble had it available almost six months ago in their initial release of TerraFlex.  I understand there are other issues at play here - background map data caching and the incorporation of operational layers (both something Trimble's offering lacks), but ESRI still could have incorporated basic off-line functionality in this new release and just built on it for the upcoming major release.

Trimble's TerraFlex is an app I tested back in June and was initially very impressed.  Where most of ESRI's mobile offerings (Collector, ArcGIS App, ArcGIS for Windows Mobile) require some expensive back-end infrastructure - ArcGIS Online, Portal or ArcGIS for Server - TerraFlex offers a far simpler mobile solution paradigm.  Everything is cloud based and single fee.  You pay your money and you get everything TerraFlex has to offer, and all for a relatively paltry price as compared to ESRI's mobile solutions in the same marketplace.  Of course, this easier to use solution comes at a cost (pun intended) - what the initial release of TerraFlex didn't offer was pretty extensive; no background map caching, no data editing either on the device or in the desktop interface, no operational layers, and some very limited data export options.  On the other hand, what TerraFlex does offer is pretty impressive given the price: off line data storage and sync, mature and stable apps not just on the iOS and Android platforms, but Trimble also had an app available for the Windows Embedded Handheld OS right out of the gate.  Trimble wasn't about to leave out the thousands of Trimble customers running their Juno handhelds who are still stuck with a dying Windows OS.  Kudos to Trimble on this.

Trimble indicates many of these shortcomings will be addressed in 2014, and Trimble seems poised to leverage what they do best - allow TerraFlex to incorporate high precision GNSS positions (including RTK-based solutions) into the data collection stream.  This could turn TerraFlex from a mere mapping grade data collector into a serious high precision data collection tool.

In his blog post Alex discusses Fulcrum.  To be honest, this is an application I've known about but have not had a chance to test.  Looks like I'll have to take it for a spin sometime soon.

So as 2013 draws to a close where are we at with mobile GIS solutions?  The best analogy I can think of is that of a ballplayer with a lot of potential who's just been called up to the majors.  His batting stats are getting better with each game, but he still has problems connecting with the ball.  The potential is there, he just needs more time. So it is with mobile GIS apps.  Most are still somewhat of a 'swing and a miss', but they are getting close to smacking the ball out of the park.  Whether it's off-line data collection with ESRI's offerings or TerraFlex's incorporation of cached maps, in-app editing or incorporation of high precision position feeds, 2014 is starting to look like the field will really mature and we'll get closer to the full promise of mobile GIS.

It'll be an interesting year!



Saturday, November 2, 2013

William J. Hudson's Pocket Transit History Site

Several times in this blog I've referenced an excellent pocket transit history site established by a gentleman by the name of William J. Hudson.  His site was the resource for historical information on the development of the pocket transit, its many variations and changes down through the years.  I considered it the best resource on pocket transits available anywhere on the web.

Unfortunately it looks like Mr. Hudson's site went off-line well over a year ago.  I first noticed it about six months ago while doing some research on a pocket transit I had just added to my collection.  Over the last few months I've been contacted several times either directly or through this blog about Mr. Hudson's site, asking if I knew when it might be available again.  Since I don't know Mr. Hudson and I've never communicated with him I had no way of contacting him to see when or even if he intended to reestablish his website.  It seemed a great resource for those interested in pocket transits was gone from the web forever.

Well, things don't really disappear from the web.  They just get archived.  I'm happy to report that I've stumbled upon an archive of the main page of Mr. Hudson's site hidden away in a dark corner of the internet.  Unfortunately none of the linked pages such as his serial number breakdown page were archived, so that resource appears lost.  But the main page is still chock full of useful information about the pocket transit's history, development and features.

I've linked to the internet archive for this page from the image below.  When you open the page you'll have the option to download the page and graphics as a zip file.

Click here to access the site archive

I'm glad to have access again to at least a portion of this great resource.  Mr. Hudson, if you read this, can we have your site back please?  Hundreds a couple of diehard pocket transit collectors really miss the information your site provided!