Darren Gee
Topographic Land Surveyors

First, you can define a new coordinate system, using the preserve settings for editing option of the new key name dialog. This prevents having to re-enter all of the numbers again.

Second, you can add/change a value in the Registry on your system to disable the protection scheme as indicated in the manual/help file.

If either of these solutions is insufficient for your, let me know.

I'll figure out another work around for you.

He is also asking about units in ROT_X, ROT_Y, ROT_Z? These do NOT seem to be degrees, Radians or Grads... I have no idea about this one. All I know is this is used for BURSA only. Is BURSA a datum? What is BURSA?

Any assistance is appreciated. Thanks,

-Daniel

This selection of units is a de-facto standard. What is not standard, is the sign of the numbers.  CS-MAP uses the convention of "what is required to convert the datum being defined to WGS84", which is the more commonly used convention. Some folks, especially in Europe & Australia, will use the convention of "what does it take to convert WGS84 to the datum being defined". Essentially, this all means that the sign gets reversed.

As a result, if you define a datum and find your results are off significantly (i.e. 50 to 200 meters), first thing to do is change the sign of all the numbers and try again.

Bursa is not a datum. It is a transformation algorithm, i.e. a technique. Bursa is better than Molodensky, but requires more information which is not always available. When the required additional information is available, better results are achieved by using Bursa.

The Seven Parameter is a more rigorous implementation of the mathematics underlying Bursa. It is often preferable to Bursa. We support both Bursa and Seven Parameter so that folks can match precisely the numbers generated by other systems which use Bursa.

The algorithm needs to be pretty short and a program won't do as everything has to all be coded by me. Please help!

>From Snyder, page 58:

B = cos (lat) * sin (lng - lng0)

x = k * R * arctanh (B) + x0

y = k * R * {arctan [tan (lat) / cos (lng - lng0) - lat0]} + y0

where:

k = 0.9996012717

R = radius of the sphere (Airy 1948 - 6377563)

lng0 = central meridian (2 deg West)

lat0 = origin latitude (49 deg North)

x0 = false easting (400000.0 meters)

y0 = false northing (-100000.0 meters)

This would provide an approximate mapping from lat/long to X and Y for the Ordinance Survey of Great Britain. Since you would be approximating the ellipsoid with a sphere, you might want to use a different radius.

Check our last two newsletter articles for information on this. Otherwise, to calculate the distance and azimuth between two lat/longs on a sphere:

>From Snyder, page 30:

cos(c) = sin(lat2) * sin(lat1) + cos(lat2) * cos(lat1) * sin(lng2 -

lng1);

where c is the angular distance between the two points. Convert c (in

radians) to linear units by multiplying by the radius you have chosen to

use for your sphere.

For the Azimuth (Az):

tan(Az) = cos(lat1) * sin(lng2 - lng1) /

[cos(lat2) * sin(lat1) - sin(lat2) * cos(lat1) * sin (lng2 - lng1)];

Notice that this formula is well suited for the ATAN2 function. There are ellipsoidal forms of all these formulas, but they get quite involved. From your question, it appears that they are out of the question for now.

Hope this helps.

-John

If the DXF file has latitude where the X's normally appear, and longitude where the Y's normally appear, there is a bit of a problem. I believe Tralaine would be able to handle this, however. Using Tralaine's coordinate system editor, make a new coordinate system modeled after the appropriate LL?? system. That is, if the original Lat/Longs are NAD27, then make a new coordinate system based on LL27.

Once the new coordinate system has been made, set the Quadrant feature of the General tab to -1. This instructs Tralaine to swap the axes. Then use this new coordinate system definition as the coordinate system for the input. The results should be a .DXF file in whatever the selected output coordinate system is.

Please call (888-ASK-NORM) if you have a problem with this. Sounds like an interesting problem to work.

Also, are you aware of anything that will translate jgw file into something usable?

Intergraph software products use an alternative technique for georeferencing a tiff image. Mathematically, their technique is better, but is not the technique used by the USGS. Thus, my first shot at a raster image program does not support the Intergraph technique. I suspect that a future revision of this program will.

I am not aware of what a .jgw file is. Probably very similar to a twf file. In fact, probably the same, just related to a .jpg file. If this is the case, the .jgw file would be an ordinary text file with useful information in it. The problem is "what do you need to do with the associated .jpg". Perhaps my guess about .jpw's being a world file for a .jpg image is wrong. Could you send me one of these files?

Where can I get school district polygons for these four States for free?

Free maps are hard to come by, as they are not easy to produce. You can get DLG/SDTS data from the Internet for free (ftp.edcftp.usgs.gov), and there are free programs which will convert this data to DXF form  (http://mcmcweb.er.usgs.gov/sdts/). But if you've never done this before, its a royal pain. If you succeed, you would then need to transfer the boundaries to this "free" map.

Best advice is to spend some money. Contact ADCI (www.adci.com) or WhiteStar (www.whitestar.com) for information and a quote. Expect to pay several hundred dollars if you need all four states.

You might check with the US Census Bureau. They produce TIGER files, and these files may include school district boundaries. TIGER files were not free, maybe they are now. I'm not sure that there is a "free" program to convert Tiger data to something useful. Check GISTools

(www.gistools.com) for some reasonably priced programs which will convert TIGER files to various formats. They might be able to tell if the Tiger files include school boundaries, and the best way to get the appropriate Tiger files.

Sorry I could not be of more assistance. 

Norm, 

Certainly, no need to apologize. This is an excellent response to my question. I understand that we will probably have to pay something for the coverage/layer/theme/table whatever. But, I am just trying to avoid paying $1,000 per State (which is the price quote I have now from MapInfo) when we may be able to get them for "several hundred dollars" as you say. 

Incidentally, I have done some fairly complicated translations between MapInfo, AutoCAD, Smallworld and Microstation using CorpsCon, MS Excel, MS Access, FME et cetera via DXF. And, you are right, it can be a pain. But, if I can save the $2,800 by investing a day of my time, I will.

Thanks for all the great links! Unfortunately for me I have not made adequate use of your website yet, but I intend to check it regularly now.

The question from the guy who wanted the distance between two coordinate pairs including a design angle is a problem I just finished tackling (and documenting exhaustively) using Trimble GPS, MS Access and MS Excel. I thought your answers on it were good. 

The company I work for is installing a Fiber Optic network throughout

the Northeast. We usually build our basemaps from D.O.T. maps that

include grid ticks based on "some" coordinate system. Then our GPS guys

go out into the field and bring back actual points that they have

surveyed. What we can't figure out is how, using AutoCadMap, do we get

our map to align with the GPS points which are on a different coordinate

system? The GPS info is arriving in DXF file format. Should we be using

AutoCad Land Development instead? Or is there something we're not doing

right in AutoCadMap?

First, one would need to figure out what the "some" coordinate systems

are. THis shouldn't be too difficult as I'm sure the DOT uses a

standard grid system and the GPS guys can probably give you either WGS84

lat/longs, or anything you ask for.

The grid ticks on the DOT map are probably either state planes or

UTM's. If the Y coordinates are in the order of 4 or 5 million, and the

X's are a couple of hundred thousand, then you have UTM's. (UTM's are

usually in meters.) Otherwise, the best guess would be State Plane.

Download the evaluation edition of TRalaine and use its Test Coordinate

SYstem feature to convert some points to lat/long to verify the guesses

above. Note that the difference between datums will be small (about 30

meters in that part of the country), of a pretty definitive test point

would be required to determine the datum involved.

GPS systems are designed to function on WGS84. In fact, WGS84 was

invented for GPS. Thus, the lat/longs produced by the GPS unit is self

(before any post processing) are WGS84 lat/longs. Many GPS receivers

provide the results in Degrees, Minutes, and fractional portions of

minutes, rather than degrees, minutes, and seconds. Watch out for

that. If the GPS boys are giving you anything besides lat/longs, the

coordinates are being processed by software similar to our Tralaine.

Therefore, they should be able to give you whatever you want; usually

this would be what the the original DOT maps are based on.

You can use AutoCAD map queries to convert drawings from one coordinate

system to another. Exactly how queries work, I don't know; I just know

that that is the process you must go through. If you have a lot of

these things to do, you may find our Tralaine product to be more

convenient.

IN any case, you need to covert one data set to the same coordinate

system as the other data set. Then the two will overlay properly. If

you have trouble determining the coordinate system of the DOT maps, send

me a coordinate, and the approximate lat/long and I'll give you some

educated guesses. Obviously, the more precise the lat/long, the more

definitive the guess will be. But if a degree or so is the best you can

do, it is likely that that will be sufficient.