Digital Mapping

Old information from startup phases. See new stuff.

Background

It is getting increasingly hard to find cartographers who do it the old way by scrapping and adding information to color separated mylar sheets. The Trail Conference is investigating going digital. Volunteers will be necessary. If you are an expert or know someone who is and would like to volunteer, please send mail to wdaniels@bestweb.net. At first we will need experts in setting things up. Later we will need people with the patience and attention to detail to actually do the work.

Below are my findings so far. I welcome additions and corrections. We have tentatively decide to redo the West Hudson Maps set first because it is small and has lots of changes that need to be reflected on our maps.

What we know

Funding opportunities

• ESRI Conservation Program conservation geography, activism and multicultural social change
• CTSP, ESRI Conservation Program Resources General Conservation

GPS

GPS units are readily available at low cost. However, because of Selective Availability (SA), their accuracy is not good enough for mapping trails unless DPGS techniques are used with it. This increases the cost and complexity. Even many of the pay-for services do not reach many of the hiking trails. Accurate GPS data depends on being able to see 6-8 satellites continuously. This may be severely impeded on our trails by being under heavy tree cover and/or in narrow valleys behind a mountain which block the view of satellites that are low on the horizon. Some of these problems can be cured by going at the right times when there are suitable configurations of satellites high in the sky or at leaf-off season to avoid the trees. Another problem that must be dealt with is the limited storage capability of most GPSs, typically 200-400 points, before the data must be downloaded to a computer, i.e. long trails may require carrying a laptop computer.

There are at least four possible ways of getting DGPS corrections, each requiring different radio gear. Because of the expense of the radios, one has to pick one technology.

• The Coast Guard broadcasts differential correction data on 300 kHz beacons in the US and Canada. The service is free, but only available near coastlines and major bodies of water. It does not reach to many of our hiking trails. The two nearby stations are Montauk and Sandy Hook. The low frequencies give some coverage in valleys.
• various commercial companies provide differential corrections via a data subcarrier on certain broadcast FM stations. Coverage is limited to areas with participating transmitters. DCI uses WNEW xxx and covers 30-50 miles from the transmitter and is line of sight. It needs an RDS 3000 receiver - $375; 10 meter service - $75 per year; 5 meter service - $250 per year. These signals will be severely impacted by being in the radio shadow of a hill.
• some commercial providers distribute differential corrections via satellite. Coverage is worldwide, but the receivers are expensive and the data costs about $800 per year. The satellites are geo-synchronous and hence high in the sky so they should be receivable most places.
• you can do your own DGPS with a suitable base station GPS receiver and a radio link. Commercial base stations are expensive. See OEM GPS receivers. There are two suboptions here. One is to partner with someone and install our own base station somewhere, e.g. top of Bear Mt. The other is to actually have two mobile setups connected by short range radio, 5-10 miles. One would leave one unit in a car near the trailhead as a temporary base. There are problems of accurately finding the position of the temporay base which can shift all the readings.

It is also possible to buy correction data and correct information recorded after the fact. However the consumer grade GPS units do not record the pseudorange data needed to do the postprocessing.

Cartography

Cartography is the art of actually drawing the maps. There are several strategies for doing this digitally.

• One can do it entirely in the raster domain. This is most similar to the mylar techniques. The main problem is cleaning up the USGS base maps which have a lot of   stuff that is extraneous or even wrong. We could start from a scan of our current maps but this does not help in regions were we need to extend the coverage. If we decide to do 100 ft contours instead of the 20s that are on the quads, there is even more stuff to remove.
• One can do it in a pure vector mode. One has to spend a lot of time vectorizing the raster data. There are programs which aid in this process by following the curves and producing vectors, e.g. the contours. The problem is that the programs get confused when the contours get close together or have breaks, for example where the height is imbedded in the contour line. It requires a lot of hand touchup. It can follow individual colors so it is not confused by crossing streams and roads. In vector mode one would plot the trails directly from GPS data or draw the trail on a map and digitize it. I have not seen the output of a pure digital map at high resolution so it remains to be seen how good it is. Herbie has a lot of the 1:100000 maps and has produced some incouraging output. It has 20 meter contours and that may be a show stopper for those used to feet.
• One can produce countours from DEM data. NYS now has 10 meter DEM almost available. This is accurate enough to show most but not all features. The data can be used by a program such as Surfer or the Spatial Analyist extension of ArcView to produce countour lines at any spacing, e.g. 20 feet instead of the meters that the data format has. This process produces some artifacts which have to be cleaned up by hand
• One can do a mixture of raster and vector data, e.g. a base map of fixed up USGS quads as above with the trails overlayed with a vector representation.
• Question to ConsGIS expert:

  > 1. We foresee two uses, tracking land ownership along and adjoining trails
  > we maintain or will build as we get permission and as part of the process
  > of producing high quality, printed hiking maps whose sale is a major part
  > of our budget. I have not seen many examples of ArcView dealing with
  > contour lines and such. Are there builtin facilities that will help or do
  > we need other packages to go with it? I have used both Corel Draw and
  > Freehand to aid in map production. One of my concerns it the shear size of
  > the files. Can ArcView deal with small pieces and put them together
  > seamlessly? For example bring in one piece and then automatically scroll
  > to the next piece?

  AT A BASIC LEVEL ARCVIEW CAN HANDLE ANY KIND OF LINE DATA, AND YOU CAN DIGITIZE IN YOUR OWN LINES AS WELL AS NEEDED. YOU CAN ALSO BREAK YOUR DATA UP INTO WHATEVER CHUNKS YOU WANT AND IN ARCVIEW THEY WILL APPEAR AS A SINGLE SEAMLESS MAP AS LONG AS ALL OF YOUR CHUNKS ARE REGISTERED TO THE SAME PROJECTION AND COORDINATE SYSTEM.

  ArcInfo (not ArcView) has a companion program ArcScan which reportedly can cope with contour lines if they are color separated out so that they are the only thing on a page. The white paper pointed to by the link above is comprehesive (but large 62 pages). It points out that we will need 400-800 dpi scans of separated contours. This leaves an open question of getting the separated contours. In the past we got them from USGS. They are still available at $36 for the first color on a quad, plus $9 for each additional color. Our area is covered by Rolla office, 573-308-3500. In the past we have bought only the brown (contour) and blue (water) layers.

  We may need ArcPress for actual map production.

• Raster to vector conversion

Original question:

We are looking for solutions of raster-vector conversion to perform line tracing on scanned and geocoded drawings. If you have information, please write about adventages/disadventages, prices etc.

Bruce Bannermann and Robert Christie offered me the ProVec software at: http://www.abakos.com.au/ It looks a good software, you can download a DEMO version with 30 day licence and a limitation of A4 size.

Antonio Pestana provided me a very interesting FREE software (beta version), manufactured by U.N. statistical department. MapScan is really a great solution for free!!! http://www.un.org/Depts/unsd/softproj/software.htm 

Daivesh Sanqhvi offered me a professional GIS service: Based in Virginia, USA, we are one of the world's largest GIS/AM/FM conversion companies, employing over 1,600 people. We offer full-function GIS implementation services. Our intelligent conversion technology ensures highest quality levels and fastest turn-around times in the industry. I would encourage you to visit our website http://www.apexinc.com to get a corporate insight into our organization.

Paul Hastings and Aqus Salim offered me the R2V software, available at www.ablesw.com.

GIS

Geographic Information Systems are useful from several standpoints. At the simplest level, they have builtin facilities for dealing with drawing maps (the cartography part). Their strong point is, however, recording in a database all the information about a map, e.g. names of objects and their locations, land ownership, vegetation, pointers to pictures, etc. It is this second aspect where we might gain the support of a number of partners and get funding for both parts of the work. One suggestion is to use the Digital Orthophoto Quads (actually 4 per quad) as a background for GIS work. These are partially available in our region with more promised. Currently northwest NJ and some along the Hudson in NY are what is available.

Map Data

Digital Map data that is usable for high quality hiking maps is hard to find. There is lots of raster data available free or at low cost but it is all too low resolution. There is very little vector data, DLGs,  available at 1:24000 but complete data at 1:100000. The latter does not have quite enough detail and the contours are at 20 meters instead of 20 feet. USGS claims they are moving in the direction of DLGs but only through cost sharing arrangements. We have had good results of printing from the DLGs, but even that requires significant work placing names and labeling contours. NY and NJ are not high on their list. Much of the commercially available data has copyright restrictions which make it unusable for anything other than private viewing. It typically is also at too low resolution.

USGS Map Colors (RGB)

From GPS-L mailing list:

...

And I assume you know that USGS topo maps are accurate only within 40 feet, don't you? The only *really* accurate features are the benchmarks. Otherwise, roads, streams and other objects are moved within that 40 foot range to make a more visually pleasing product.

Dr. Richard B. Shepard, President
Applied Ecosystem Services, Inc.

------

Benchmarks are points of known elevation. The horizontal positions of benchmarks in the National Geodetic Data Base are known to within 6 arc seconds in Latitude and Longitude. (One arc second equates, I think, to about a hundred meters, about the same as a K-Mart GPS receiver.) Most are known to about one arc second or better, but the guarantee is good only to six !!! Benchmarks are established (in every country) by national geodetic agencies that do differential leveling (with levels and level rods), generally to first-order standards and specifications.

When plotted on a USGS 7 1/2 minute Quadrangle map, that's (1 arc second) about 0.15 inches on the printed paper map.

On the other hand, a Triangulation Station is a point of known position (3-25 mm on the ground). That's going to be plotted to an accuracy of 0.01 inches on the paper map. Triangulation stations were established (in every country) by national geodetic agencies that did classical triangulation (with theodolites), generally to first-order standards and specifications. Well-defined positions, such as road intersections, etc., are plotted to within 0.01 inches of their true position (20 feet). Benchmarks ain't Triangulation Stations, and vice-versa.

Clifford J. Mugnier (cjmce@uno.edu)
The Topographic Engineering Laboratory
Department of Civil and Environmental Engineering
UNIVERSITY OF NEW ORLEANS

Links and newsgroups of useful information

GPS resource list - updated semi-weekly - very complete, the entries below are mostly a selection from it.

Progress reports

Apply for CTSP grant  (Jan 5)

Accepted for second round of application process (due Feb 20)

2/10 Presented to TC Board for approval to go ahead with grant process. Approved.

2/17 Found ArcScan white paper - major improvement in understanding..

2/20 Submitted second round application

3/12 Pubs meeting confirmed West Hudson is first attempt at going digital

4/15 CTSP grant failed

5/1 Attended GISMO meeting (NYC GIS users group)

5/5 Attended CAPARC meeting (Capital Area ArcView/Info users group)
      10 meter DEM data for whole state available
      hydrography may be done by year end
      State run cooperative on GIS data - nominal charges for non-profits (I have forms to join)

5/14 Pubs meeting - set 6/30 as a drop dead date for making significant progress on WH or farm it out to cartographer

5/15 John Fox of OPRHP GIS - they keep data in ArcInfo but use CorelDraw to make printable maps since it is easier

Project Plan

Last updated: Saturday, December 08, 2001