From Paper to Digital and Back Again: A Map’s Journey
Hopefully by now everyone is convinced of the importance of paper maps as an essential component of your GIS toolbox. Our generation of GIS professionals has been blessed with a rich digital database. Even in the age of instant access to massive amounts of digital geospatial data, our base maps are often derived from analog sources.
I’d like to explore various scenarios to discuss how we enrich our data during the cycle of paper to digital and back again. Let’s start with some scenarios that illustrate the challenges and opportunities of the paper-digital cycle. This certainly isn’t all-inclusive, but I have found these three to be the most common that I have encountered in my career. After the scenarios, we’ll look at ways to address these challenges and opportunities.
1. Handwritten notes on a fresh-off-the press GIS map.
One of the most common uses of a paper map is in a meeting of some sort. It could be a public meeting with dozens or hundreds of people, or it could be between you and a colleague. Everyone is writing on the same base map, which comes from the GIS database.
One of my first GIS jobs was building a series of zoning maps for cities in a rural county. I’ll spare the boring details, but we were able to link text files of zoning to a digital parcels map. Unfortunately, the text files were woefully out of date, so to update the zoning we took the printed maps to each and every city, and with city staff, corrected the maps to reflect current conditions.
2. Handwritten notes on a historic map.
Another scenario is discovering a map with a known base map but no documented digital source. Mining drawers in the archives, I discovered a large Mylar map with handwritten notes next to little colored paper dots, which represented falcon sightings from decades past. This was rich data. Fortunately, I was familiar with the base map; it was old, but the boundaries of the study area hadn’t changed. Using a large format scanner, I turned the Mylar map into an image, then georeferenced it to known points in the GIS database.
3. A paper map with no known spatial reference.
Sometimes paper maps don’t originate from digital sources, or if they do, the digital source is of unknown origin or lineage. How many of us have been handed as-builts or CAD drawings of a project area like a building or pipeline and been asked to get it into GIS? The text is blurry, there is no coordinate system, and identifiable features may have changed or may be outside of the drawing area.
Putting It All Together
The one thing that all of these scenarios have in common is the need to turn analog information into digital data. The key to this is to understand your client and their needs, as well as the limitations of the data itself. Being a GIS specialist isn’t just about reading maps and knowing the software, it’s about reading people. It is about speaking different languages and translating those into the common language of a map.
Consider the first two components of GIS: geography and information. Geography is the spatial aspect of the system; information is the data attached to the spatial features.
We’ll start with the geography part. After all, this is where our forebears began, scanning 9x9 inch air photos and digitizing USGS topo maps to make DLGs. If they could do it with their nascent technology, so can we!
Making It Spatial
In the first scenario, the zoning maps, the “G” is pretty straightforward, as is the “I.” The notes themselves don’t necessarily need to be recorded in the GIS database. As changes are made digitally in the GIS map, one parcel turns from “residential, high density” to “light commercial,” and so on. Even without digitally recording the change, that paper map itself is a record. I always note the date of the meeting, and with whom it was, on the map. Then as I bring the changes into GIS, I check them off on the paper map, in a different color of ink or pencil than the notes.
In the second scenario, I have a big image file. Using the georeferencing tools offered in various GIS applications, I can line up the scanned map to our base GIS data, in this case, a study area base map. Regardless of the original projection, once I georeference it to our base data the image will be in our standard coordinate system.
We could go into discussion about spatial accuracy and RMS values, but that is beyond the scope of this article. This is where knowing your client and their needs becomes critical. In my previous article, I mentioned a student project on Indian Removal. That story is told at a continental scale, so precise locations aren’t as critical as showing the vast scope of the tragedy. Using a simple base map, we can show many general locations — the start of the journey, the end — with coarse spatial accuracy.
What about the falcons, though? This is a much smaller area, about 183,000 acres, and richly mapped from the late 1800s through the present. The falcon sightings were recorded in pre-digital days, so there are no UTM coordinates, but the paper dots and handwritten notes indicate very specific places. By creating new feature classes I can heads-up digitize these points into GIS features.
In the third scenario, an old blurry CAD drawing, it becomes more challenging. I may need to break out the GPS or mine some other more recent CAD drawings. If I can find a minimum of three identifiable points on the ground, or in the GIS database, I can georeference the drawing to GIS — after I’ve scanned it, of course.
Consider also the dates of the data and not only what you are georeferencing it from, but also what you are georeferencing it to. Even a recreational grade GPS is more accurate than a 1:24k topo map. As long as you have the datum and coordinate system set correctly, everything should line up.
Again, we come back the necessary scale and locational accuracy. There is already a good digital zoning map, so comments specific to a parcel can be added easily, but when someone discusses a rare plant sighting, a water valve or an archaeological site, spatial accuracy is critical. Further research is required, whether by mining old records, looking at someone’s GPS, or personal interviews…or maybe…by looking at someone’s treasured map that they have carried in their field pack for years.
Turning Information into Data
So in all these scenarios, using the GIS tools at our disposal, we now have a georeferenced map and hopefully some GIS features. We’ve taken care of the “G,” but what about the “I”?
A paper map may have lines, points, polygons and text written upon it. Some of the text may be associated with a feature, but not always. How do you attach notes to a place?
Start with making point features. If there is a lot of text, make a new point in the middle of the text. I suggest a very long “COMMENTS” or “MAP TEXT” field in your attribute table, maybe both. These can always be parsed into other fields once you come up with a final data model. The initial goal is to get all of the information into a digital format.
Some of us are “lumpers,” putting all the data into one place and then separating it later. Some of us are “splitters,” perhaps putting each note into a separate feature class then merging them later. A paper map with written notes on it is at first a lumping exercise. Then it is up to the GIS specialist to lump or split the notes into a digital database.
Like RMS values, the various aspects of annotation and labeling are beyond the scope of this piece. The first step is attaching text to a place on the map. After that, we can make new features as lines, points, polygons or annotation. Annotation and labeling are also beyond the scope of this piece, but I suggest researching other resources for your particular GIS software. Professional cartographers have advised me that even with all of the tools at our disposal, putting text on a map is still one of the most challenging elements of map production.
Paper maps can be ephemeral, capturing a snapshot in time, or they can carry decades of experiences and memories. Even today, I discover new data in paper maps that have long been in drawers and filing cabinets. We have many tools to turn these paper maps into digital data, but utilizing these tools in the proper context is as vital as knowing how to use them in the software package. Just because you can operate a table saw doesn’t mean you can build a cabinet. True, nobody has ever cut their finger off with the georeferencing toolbar, but bad maps have sent people over cliffs.
Read the first article in the Paper Maps series, "Death of the Paper Map ‘Greatly Exaggerated’."