Pittsburg, NH — You Are Here
As a kid, I remember looking at maps we had in my family’s rustic log cabin in northern New Hampshire and getting a lay of the land of the largest town (in area) east of the Mississippi. Pittsburg, NH is known as the snowmobile capital of the East and has vast expanses of wilderness that draws visitors to the area.
You can buy recreational maps in the local stores but none of them had the detail I first saw on that old map.
After some research, I discovered one of the old maps was made by the Canadian government — not the USGS. They claim it to be “As the most detailed country-wide map series, the 1:50,000 paper topographic series also known as topo or topographical is ideal for any purpose that requires good detail of an area. These topographical paper maps include contour lines, water bodies, roads, trails, population centres, treed areas, etc. Printed Topo maps are great for recreational activities such as hiking, camping, biking, or fishing.”
The most current version of that map is now over 30 years old so I figured it was time for an upgrade. As a GIS professional, I started researching available data. Roads were where I started but came to a quick realization that there was no good/accurate data source, particularly for the dozens of logging roads. At a proverbial dead end, I shifted from map creation to data collection. I decided to use OpenStreetMap as my database for everything except for polygons and hydrography and started updating existing OSM data and digitizing where it was missing in the Pittsburg area.
Useful tools I used were RapID which uses artificial intelligence to predict features on high-resolution satellite imagery and JOSM to aid in the creation of building footprints where the Microsoft Buildings database missed them. Yes — I tried to digitize every building in the 291.3 square mile town. That’s what all those little black dots (just like the old USGS maps) are on the final map.
After I had updated and digitized everything I could do remotely, I shifted to ground truthing and field verifying roads, trails, names, builds, emergency helicopter landing zones, gates, etc. I did not drive every road in the town — I mostly focused on the west side of Route 3.
I downloaded the OSM data first onto my phone and used the OsmAnd app since that was the best app I could find to try and edit OSM data offline. I don’t understand why there isn’t a better app for this — maybe there is but I’m not aware of it. After a day of driving around and adding dozens of notes in the app (with the intention of making the edits afterward), a fellow OSM editor saw them and started editing the map by trying to decipher my notes. Turns out my definition of a camp (i.e. hunting camp out in the boonies) was interpreted by the New Jersey-based editor as a campground. I appreciated the help but I ended up having to fix the edits.
After field collection, I shifted to mapping the 200+ miles of snowmobile trails in the area. When I ride there with my cousins, they know every turn of every trail at 50mph and the way across the lakes. Otherwise, you have to rely on the paper map each snowmobile club makes every year (trails close due to logging or land use issues so new maps are needed each year). Recently, the New Hampshire Snowmobile Association released an app and website with all the state-recognized trails.
Instead of reinventing the wheel, I reached out to see if this data was publicly available but was turned down and told to contact each club. With this news, I took the latest Pittsburg Ridge Runners paper map (2022) and georeferenced it in ArcGIS Pro and started updating OpenStreetMap with this information. It took a while but I was able to create OSM relations for each corridor and primary trail along with their associated attributes. After I had done this, the 2023 map came out and apparently they change corridors to other trails so I had to review my work and make the appropriate edits.
With the OSM data finally in a state I thought was good enough to make a map, I started the process of consuming the data in ArcGIS. Enter FME. Since OSM data isn’t structured like a normal GIS dataset (think of point, line, or polygon tables with defined fields, types, widths, etc.), it’s a simple data structure consisting of nodes, ways, and relations with tags containing the key and value. Mapbox has a good write-up about it here. What this all means is that ArcGIS (the application I used to create the map) can’t natively/easily consume this data. To get around this, I used FME from Safe Software to extract, transform and load into a file geodatabase that ArcGIS Pro can use. The FME workbench calls the Overpass Turbo API for a defined extent (i.e. Pittsburg) and downloads the OSM file containing every node, way, and relation in the extent. The next trick is to convert from OSM structure to Esri structure and spit out a new dataset to use for map rendering. There is some dark magic involved but end result was worth the development since the task is now automated and can be used for other areas other than Pittsburg (or when I get to ground truthing the east side of Route 3).
Logging road gates posed a challenge since I wanted to render them on the map as lines 90 degrees to the road. In OSM, they are captured as a node. I ended up calculating the angle to the neighboring road and adding 90 degrees.
Speaking of gates, not all of them were field verified/mapped based on the list from the state. Very little info is available other than vague road names and numbers. As you can see, the ones with check marks next to them were verified.
After the data was ready, the actual map creation process was able to start. I’m no cartographer but I tried to create a nice map. Being partially color blind doesn’t help so I was often asking my wife or kids if the colors looked ok. What was missing from the map right out the gate was all the hydrography and land cover. Since Pittsburg borders two states (Vermont & Maine) and the province of Quebec, trying to find data that was consistent was a challenge. The landcover source from LANDFIRE was a huge find since I was looking for a uniform tree cover raster. Contours was the next find but were disappointed when the feature service limited the coverage into Canada at the scale I was using in the map. I was never able to find a good hydrography data source for Maine since every hosted service didn’t allow me to tweak the colors or apply definition queries (let me know if you know of one!). The National Hydrography Dataset was an option for NH, VT, and ME but the data NH GRANIT provided was better for this area so I used theirs.
In the home stretch, I wanted to add some metrics to the map to summarize the miles of road by type along with the snowmobile trails (even though they aren’t on this map — more to come on that one!) I once again used FME to calculate the statistics and ArcGIS Pro to render them into a simple chart. With this done, I focused on labeling and had no desire to convert to symbology so I just tweaked a bunch of labeling settings to get close to where I was happy. I know it’s not perfect but for the first draft, it’s good enough for me.
The last chapter in the saga of making this map was when I brought it to the local UPS store to print. After 20 minutes of working with the UPS guy, we gave up and left empty-handed. Apparently, I exported without converting the map to a graphic first and Adobe has an issue opening the PDF if it contains a PNG (thanks to that chart I made!). After some troubleshooting, I was able to find an answer on Stack Exchange and re-export. I’m planning on going back to UPS tomorrow to see if it prints….
In hindsight, I spent way more time on this one map than I ever expected for a side project. It was a lot of late nights and with my desk right outside my son’s room, he complained of the mouse clicking when he was trying to fall asleep. Was it worth it? I’ll let you know once I can actually print the dang thing and hang it up in the cabin and maybe one of my kids 30 years from now will want to update it…
Well if you made it this far, here’s the final map. Hope you like it and be kind in your cartographic critique!
Here’s the same area — the left is from a map produced by NH Department of Safety and on the right is the one I created.
