Using GIS (maps) to find Roxx: A Guide

Community,

I was curious if anybody else out there is taking a GIS/Remote Sensing approach for locating new areas. I am located in an area that is not that well geologically endowed, Kansas City. Traditionally the most popular area for a weekend trip is Northwest Arkansas (Sam's Throne and HCR) but Jackson Falls/Wichita mountains are possible for a true weekend warrior. Missouri has been somewhat overlooked by the masses as they usually opt for Arkansas sandstone, however there has been plenty of development in the state and some areas are only becoming more popular, especially with the recent release of a new guidebook.

I'd like to share some of my methods and see if anyone has done anything similar. Maybe others can chime in and share tips and tricks and pitfalls to avoid. I use a fusion of many different tools, data sources, and even social media to help identify potential areas from my desktop. I realize that some of these methods may vary depending on your geographic area of interest, and often times you may go try and ground truth your finds, and it will end up being total choss.

Background: I want to find new crags/boulders to develop within a 2 hour radius of Kansas City, I know there is plenty of limestone out there but I'm mostly interested in sandstone.

Basic Approach: For those of you that don't have access to a software package such as ArcGIS you may end up using this approach. Your tools are Google Maps, Bing Maps, Google Streetview, Google Earth, and if you are GIS savvy you can usehttps://viewer.nationalmap.gov/viewer/ Use the terrain feature on google maps to look for significant elevation changes, in the midwest these are usually in river valleys, once you've identified significant slope (i.e. elevation change) try to SEE the rock on imagery. Shadows, Trees, distortion (which usually signifies significant slope) or just piss poor image resolution may hinder you from seeing the rockstone. Check all your different imagery sources. Use Bing Birds eye view, Google Earth historical imagery or whatever else you can find, I've noticed that even the oldest black and white imagery on google earth historical can help yield clues because often the rock has a very high reflectance value and this is highlighted in the black and white imagery. If for some reason your rock is located next to a major road fire up google street view to try and see it, there is also sometimes historical streetview.

Social Media and Ancillary Sources: This is where you get creative, lets say you found some rock that is near a conservation area, state park, or any kind of "named feature" fire up Facebook graph search and see if people have publicly posted any pictures at these areas, it is a very brute force approach and you need to choose your keywords carefully, but often times it can yield real photos of your area of interest. Do the same using google image search. Another interesting piece of information is soil surveys, often times these are very old but can provide interesting insight into rock type. You need to creative with these sources, for instance I found a webpage that documents natural arches in each state, this has nothing to do with climbing but it yielded a bunch of pictures of nice rock features, and where there is one rock, there is bound to be more.

REAL GIS APPROACHES: If you are a real deal GIS user then I'm not going to go into detail about how to use this data, I'd rather just talk about data sources
DEM's - The higher quality the better, however often times we are limited. Many of my targets are in rural areas so we are limited to 10 meter resolution elevation models which should have coverage for the entire contiguous US, this is generally enough to find significant cliff bands however it will likely be lacking if you are looking for nice boulders
SLOPE - duh, derive this from the DEM
IMAGERY - In college I messed around with extracting reflectance values, usually from NAIP imagery, often times rock features would have bright reflectance values, on any RGB band, however there is no standard spectral signature
SSURGO - Soils data was huge for me in differentiating areas of sandstone vs limestone, I had several ground truthed areas of sandstone that correlated highly with a certain soil type whereas the limestone correlated highly another, this is no big surprise as the parent material of limestone is more basic, where sandstone is more acidic.

Other tips: Context clues, I found that in my areas sandstone cliffs (versus limestone) were much more likely to produce sizable boulders at the base of the cliff which could be seen on imagery even though the cliff may be heavily distorted, I also found that in many areas pine trees were more likely to grow above sandstone cliffs because they are more tolerant to acidic soils (i.e. sandstone parent material)

Land Ownership: So lets say you think you've found the next big supercrag or bouldering destination. It's probably a good idea to CONTACT THE LAND OWNER. Sometimes it can be difficult to find a landowner, especially if there is no houses or people nearby. To obtain ownership information you need go to the COUNTY ASSESSORS office or if they have significant resources they may have a web interface map that can tell you who the land owner is at a given area. Usually a simple google search like "Boone County Assessor GIS" will reveal if the county has a mapping system. If they don't have a mapping system you may have to visit or contact the local assessor to find out the ownership. Familiarize yourself with the land grid system (sorry eastern states, this won't work for you) which is based on Section-Township-Range, if you can provide that information to the assessor it will make them happier to assist and cut down on the amount of time needed to find a landowner.

Thanks for reading, I hope this aides you in finding your new supercrag, you may find nothing or it may be choss, but hey at least you spent a nice day wandering around the woods. Please respond with any comments or questions!

On a side note I'm looking for a new GIS job, anywhere. I have extensive experience with all types of GIS data and my last position was for a utility scale wind and solar energy developer. Please feel free to contact if you have a lead :) Also check out my website to see some of my drone related stuff as well.
www.surfacespatial.com/

Don't forget Lidar as an elevation source. It's likely to have better resolution than nationwide DEMs. Some of the state-wide projects have pretty great resolution as well. I think someone in KY was using Lidar to identify clifflines, especially overhanging ones, for a Master's project.

Interesting topic for sure.

en.wikipedia.org/wiki/Natio…(United_States)

earthexplorer.usgs.gov/

Nice write up, I use ArcGIS and all sorts of aerial imagery as a civil engineer. It wasn't until I got into climbing that I realized that I was way better at reading topographic maps than any of my climbing buddies.

If I could add anything it would be that older land surveys will sometimes note cliff line. For example here in Kentucky there are a lot of older surveys to find the best areas for oil wells and coal seams. These surveys will sometimes date back as far as the 1800's but may not the location of clifflines as it was kind of a big deal to traverse a cliff in the 1800's. A lot of these older maps are publicly available through local university GIS programs. Here in KY NKU is a great resource for this information. Beyond that if you are looking for boulders in an area with a lot of tree cover, power line easements will sometimes reveal areas where there are boulders.

I agree, Lidar derived DEM's are very nice, although in my study areas they never seem to have coverage.

Sandbagger Vance wrote:Nice write up, I use ArcGIS and all sorts of aerial imagery as a civil engineer. It wasn't until I got into climbing that I realized that I was way better at reading topographic maps than any of my climbing buddies. If I could add anything it would be that older land surveys will sometimes note cliff line. For example here in Kentucky there are a lot of older surveys to find the best areas for oil wells and coal seams. These surveys will sometimes date back as far as the 1800's but may not the location of clifflines as it was kind of a big deal to traverse a cliff in the 1800's. A lot of these older maps are publicly available through local university GIS programs. Here in KY NKU is a great resource for this information. Beyond that if you are looking for boulders in an area with a lot of tree cover, power line easements will sometimes reveal areas where there are boulders.

Are you referencing General Land Office surveys? Used primarily during the homestead act and westward expansion, they generally only walked the section lines, a 1 mile grid, but I do remember there being many field notes that were assocated with township level maps that could be neat to dig into.

I developed an algorithm in undergrad to automatically georeference these GLO maps for the state of Kansas.

I have a story based on this. I'm a student at Arkansas Tech, and we have access to programs such as ArcGIS. I had the same idea you had and downloaded the Arkansas Topo map from the Arkansas GIS office webpage.

Loading up and searching both visually for something grey in a forest or by setting a parameter for highlighting areas with steep elevation change. I think my parameter was something along the lines of a greater than 35% grade (using the assumption that mountains would yield a more probable crag than a random fissure in a field). I also ignored all mines. A ping for a bluff came up to the north of Russellville, about 15 mins south of the Sam's Throne proper area. Decided to travel there and GPS lead me to a private land.

Owner was there, she didn't like the idea of someone walking around her 'yard'.

Buuuut, I want to say that ArcGIS could be used to find new rock, just check who owns it

I got bored one day and wrote some code to pull elevations from google maps. I than built a color coded map based on height differences between the locations. I tested it at crags I had climbed at and it highlighted the areas where you climb pretty nicely.

I thought about just running it over large areas to try to find new possible cliffs to climb.

Here are 2 samples. The color is just scale from the lowest to the highest elevation point and the black marks are based on difference in height compared to the points next to it. So the greater the distance between the 2 points the darker it gets. Points that are say under say 25ft will get no marks (forgot what the value I used when I made these it was a while ago.

The wall for this area is around 50-120ft. The area on the bottom right is about 200ft but is off limits for climbing.

On this lower photo the light area kinda in the middle top is the image is a boulder field. The main cliff line is around 200-300ft.

ViperScale wrote:I got bored one day and wrote some code to pull elevations from google maps. I than built a color coded map based on height differences between the locations. I tested it at crags I had climbed at and it highlighted the areas where you climb pretty nicely. I thought about just running it over large areas to try to find new possible cliffs to climb. Here are 2 samples. The color is just scale from the lowest to the highest elevation point and the black marks are based on difference in height compared to the points next to it. So the greater the distance between the 2 points the darker it gets. Points that are say under say 25ft will get no marks (forgot what the value I used when I made these it was a while ago. The wall for this area is around 50-120ft. The area on the bottom right is about 200ft but is off limits for climbing. On this lower photo the light area kinda in the middle top is the image is a boulder field. The main cliff line is around 200-300ft.

This is awesome! Care to share more about this technique?

Basically just set a max and min GPS location. The code than just loops through and pulls the elevation from google maps based on what the distance setting of how often you want to to check the next point. These values were output into a csv file.

The csv file with elevations were than loaded into an excel spreadsheet and conditional formatting color scale was applied. I than added some additional conditional formatting to apply the black highlights on top of that to figure out where there was big differences between values next to each other. I than shrunk down the cell size so you could not see the text and it made a nice even square box and made a screenshot of it.

ViperScale wrote:Basically just set a max and min GPS location. The code than just loops through and pulls the elevation from google maps based on what the distance setting of how often you want to to check the next point. These values were output into a csv file. The csv file with elevations were than loaded into an excel spreadsheet and conditional formatting color scale was applied. I than added some additional conditional formatting to apply the black highlights on top of that to figure out where there was big differences between values next to each other. I than shrunk down the cell size so you could not see the text and it made a nice even square box and made a screenshot of it.

So basically you are brute force converting point elevations to raster, it would be super handy to know the actual raster cell size that google is using, I'm betting in extremely rural areas it will still sometimes be NED 10M, but since google are the master data holders they probably have a lot of nice high resolution lidar data.

Ryan,
Morgan Paterson has a special program that points out vertical relief. He is easily found by going to the Conneticut pages & has shared the program with me in the past.

Sure everyone knows what this was taken from.

bump for more discussion

Ryan I am considering doing my GIS project on this on Indiana. Mountain project only has 36 routes but with how hill it can get there, theres gotta be more What tools did you use? could you potentially post up the mdb file?

Luke Nitcher wrote:

Ryan I am considering doing my GIS project on this on Indiana. Mountain project only has 36 routes but with how hill it can get there, theres gotta be more What tools did you use? could you potentially post up the mdb file?

Luke,

I would start by obtaining a Digital Elevation Model (DEM) of your target area, convert this to slope, start identifying areas with significant slope. Look at other areas where there IS known climbing to use as training data. If your GIS class has not covered how to use the raster data model this might be too advanced of an exercise. 

Michael P wrote:

I don't have much LIDAR experience but I played around with local data without too much success.  Luke you probably have already seen it but just in case:

http://opentopo.sdsc.edu/lidarDataset?opentopoID=OTLAS.062012.4326.1&minX=-86.608&minY=38.074&maxX=-86.503&maxY=38.135

Here is a link to Indiana 1m resolution LIDAR data

That is pretty great if you have statewide 1m lidar coverage, however you may want to start with a coarser resolution DEM for your initial identification of steep areas, especially if you are attempting to do this on a state wide scale.

http://gis.iu.edu/datasetInfo/statewide/in_2011.php

I grab my lidar data from here if I'm working in Indiana. OGRIP has some pretty good info as well.

Another good resource that I haven't had the chance to explore at work is http://libremap.org/ 

Seems like a promising repository, I have been meaning to use it to find some bouldering areas in KY, just haven't found the time. Plus my work computer craps out whenever I bring miles of topo into AutoCAD

ViperScale wrote: Basically just set a max and min GPS location. The code than just loops through and pulls the elevation from google maps based on what the distance setting of how often you want to to check the next point. These values were output into a csv file. The csv file with elevations were than loaded into an excel spreadsheet and conditional formatting color scale was applied. I than added some additional conditional formatting to apply the black highlights on top of that to figure out where there was big differences between values next to each other. I than shrunk down the cell size so you could not see the text and it made a nice even square box and made a screenshot of it.

That's pretty impressive, would you feel comfortable sharing the sourcecode? Or if I gave you some lat longs would you run the program in some areas of the Red? 

Justin Barrett wrote:  Buuuut, I want to say that ArcGIS could be used to find new rock, just check who owns it

I haven't done any GIS project for a few years, but when exploring new lands I would typically bring in an ownership layer that allowed me to choose land based on parameters (easily weed out privately, state, federal, Native American owned lands).  Overlay that to whatever area you are searching through and you can avoid getting shot at :)

Here is the file I used for Utah: https://gis.utah.gov/data/sgid-cadastre/land-ownership/

Played around with elevation models and ground surface types to identify possible climbing areas a number of times.  No new development came out of it, but it made my hours in the GIS lab much more enjoyable.  Really fun to see what others are using to discover new spots.  

It might be worth it to load in a layer of local geology with the units listed, that way when you get a result for a large elevation change or slope, you can see what unit it's in. A lot of time there's local information on the geologic units in your area so a quick search can give you some insight into rock quality. 

Source: (soon to be) geology graduate 

Sandbagger Vance wrote:

That's pretty impressive, would you feel comfortable sharing the sourcecode? Or if I gave you some lat longs would you run the program in some areas of the Red? 

Do you have the spatial analyst toolbox?  You can calculate the range of elevations within a 3x3 (or whatever grid you want) window of cells. 

http://pro.arcgis.com/en/pro-app/tool-reference/spatial-analyst/focal-statistics.htm