USGS "Historical' topo maps

[steamboateng]

For a while I've been watching the USGS sites for news of their 'Historical' map site. This last week I found it to be up and running.
It is here: http://nationalmap.gov/historical/
This site offers USGS topo maps for all the US states; they range in dates from the mid 1800's to 2000, and are offered in a variety of scales. These maps are free to download. Included in the collection is a great selection of USGS 7.5 min. Quadrangle maps, which is good choice for Trainzers wishing to recreate US routes from a bygone day. Maps at that scale (1:24000) are great for identifying track layout, yards, industries, power plants, schools, churches, and public buildings, as well as industry locations. Owners of TransDem will appreciate that they are a perfect scale for DEM overlays or UTM tiles.
The maps are downloaded as a .pdf file. You will need to convert the Adobe format to a .png, jpg or .bmp to get the maps into TransDem.

 

[geophil]

The maps are downloaded as a .pdf file. You will need to convert the Adobe format to a .png, jpg or .bmp to get the maps into TransDem.

Yes. They call it GeoPDF. Since GeoPDF is regarded as a final document format, you are not supposed to convert it into anything else. Besides the image, GeoPDF contains georeferencing information. What I would like to see is a convenient and legal way to transform it to something more useful. gdal_translate - endorsed by USGS for other purposes - could be an option, but does not seem to work properly: It does not recognize the projection. It can convert to plain jpg or tif, but apparently not proper GeoTIFF. So it's down to manual georeferencing again. And I think that's by intention because of the copyright aspect. Fair enough for current editions of the maps but why for those where copyright has expired?

The GeoPDF sample I tried had UTM/NAD83 projection despite the legend claiming it to be geographic (or perhaps I misunderstood).

 

[steamboateng]

The maps convert easily enough to jpeg, but as you say, the internal geaoreferencing data is lost. Just finished generating a 60 mile route with these excellent maps; and yes, I had to manually georeference each, using the tried and true two corner method!
I used older maps; 1945-1955, I used NAD27 projection as a reference. They seem to have come out all right.

 

[geophil]

What kind of geopdf to jpeg converter do you use?

 

[steamboateng]

I have Adobe Photoshop Elements v.7.0. It does a great job converting the file; first converting it into its native format; then converting it into jpg. It takes a while!
I tried converting with PSP XI, but come up with error messages. Havn't tried Gimp yet; don't know if its got the converter or a plug-in for it!

 

[HiBaller]

What a great site! I've just pulled down about 20 7.5 charts of areas I am interested in. Most of them are from the mid-60's before railroad right-of-ways were converted into bike and hike trails.

Bill

 

[JCitron]

Thank you, Mike for the update on this.

I finally added the link to my favorites. The last time I was there, there were still numerous missing maps and information.

I too wish they were readable directly into TransDem with out having to use the corner method. That would make things a lot easier.

John

 

[steamboateng]

Well, the USGS site claims it has most maps included now. I checked a few states randomly and they seem to have a fairly complete selection. The manual georeferencing isn't so bad. Even a large 60 mile route rarely has more than 20 or so maps. An evenings work! Besides, laying out each map individually gives me a chance to have a closer look at them, gathering 'route ideas' as I go along.

 

[steamboateng]

If your a Gimp fan and like to collect USGS topo maps; then good news! Gimp will open the geo.pdf file and convert it to .....say .jpg. In Gimp, select 'open' file, and then hit the 'import' button from the pop-up. After the image appears, choose 'save as' and pick your favorite graphic format to save it as.
I used Gimp v.2.0.

 

[geophil]

IrfanView with Ghostscript installed can also do it. But we always lose the georeferencing information. Not that it would be of much use at the moment, but you never know...

There is an interesting document available at USGS: "Standards for Scanned U.S. Geological Survey Historical Topographic Quadrangle Collection".

In the summary it says:

Two georeferenced digital formats are used for distribution of the maps: (1) Portable Document Format (PDF) with a geospatial extension that is called Georeferenced PDF (GeoPDF®); and (2) Tagged Image File Format (TIFF) with an accompanying transformation file in Extensible Markup Language (XML) readable by open-source and proprietary Geographic Information Systems (GIS).

Has anyone seen the tiff variant yet?

 

[steamboateng]

Does that mean if it's imported into TransDem in TIFF Format it will open as a georeferenced map?

 

[geophil]

Does that mean if it's imported into TransDem in TIFF Format it will open as a georeferenced map?

I'm afraid it won't. Their TIFF is plain TIFF, not GeoTIFF. It would just save extraction of the image from the PDF envelope.

That USGS document says, there will be an additional xml file with the georeferencing information for the TIFF image which would follow GDAL conventions. The problem now is that georeferencing of the historical maps should adhere to the original projection. And that has been "polyconic" for decades. TransDEM never bothered with polyconic, because you will not find any polyconic projection coordinates on the map sheets. But they exist and the georeferencing information would reference them. I have the polyconic projection formulas but no clue about the parameters. There is this famous "Snyder" book, the USGS cartography bible, 1st edition from the 1930s, and it will probably explain it all, but I haven't looked into it.

 

[steamboateng]

The USGS 'Historical' site is a true 'gifthorse' (even though I pay for it each year.....and it's gettin' that time again, here...) to those of us who chase historical data for our routes. If manual georeferencing of the maps is neccesary, then so be it. I'm sure there are many of us (not just Trainzers) who simply appreciate the maps being made available.
Thanks for your always helpful input, Roland.

 

[geophil]

Michael,

the GeoPDF historical maps from this collection you have been working with so far: Which era are they? And what kind of projection are they using (see lower left corner in the map legend)?

I'm asking because if they are "polyconic" I might be able to find a way for at least partial automation.

 

[Thai1On]

Seems the maps of Cincinnati aren't that old :n:. The oldest one is from the early 50's.

Dave

 

[geophil]

I have spent some time programming over Easter (the weather is awful anyway, never ever before had to scrape snow and ice off the car windows on an Easter Sunday).

The next version of TransDEM will offer partial automation for georeferencing historical USGS topo maps in polyconic projection. There is still a manual step involved: to extract both the image and georeferencing information from the GeoPDF file. GDAL offers the necessary tool for this initial step. Then TransDEM can take over.

See here or here for details.

 

[JCitron]

I have spent some time programming over Easter (the weather is awful anyway, never ever before had to scrape snow and ice off the car windows on an Easter Sunday).

The next version of TransDEM will offer partial automation for georeferencing historical USGS topo maps in polyconic projection. There is still a manual step involved: to extract both the image and georeferencing information from the GeoPDF file. GDAL offers the necessary tool for this initial step. Then TransDEM can take over.

See here or here for details.

Thank you Roland for your hard work. This is greatly appreciated.

We had nice weather this year - highs in the mid-teens C/50s F. On this day back in 1997, we were hit by 20 inches/51 cm of heavy wet snow. Right now we're sending you some of our warm weather.

John

 

[steamboateng]

Having gobbled the last of my 'Cadbury' bunnies, I can continue with the more intense side of Trainzing.
Roland, I've been hanging around these maps too long, as I'm almost begining to understand what your talking about!
I assume the method you describe will also work for the more informative 7.5' maps.
Most of the maps I use were generated from the late 1930's to late 1940's, referencing both Polyconic Projection and NAD27.
Your ongoing work is much appreciated.

 

[geophil]

I assume the method you describe will also work for the more informative 7.5' maps.

Yes, it will.

According to one paper, the difference between Polyconic/NAD27 and UTM/NAD27 is negligible when it comes to printing a map. A fraction of a millimeter, the paper says. The projection coordinate values, however, are completely different.

 

[steamboateng]

I understand your point. (There I go again.....) I wouldn't want to build a real railroad with these maps, and they were never intended for that. I would think they are close enough for Trainz, however.