Introduction

Hello Everyone! I have arrived at my next destination in Buffalo, Wyoming! The last few months after my Wenatchee, Washington internship have been extremely busy! I have traveled to all of the western states except for Montana and I have traveled to Australia to explore the Great Barrier Reef. My friend, Heather (CLM Alumni), notified me of a great opportunity back in Buffalo! This experience would be a remote sensing/GIS based internship for the BLM (Bureau of Land Management). I would be working with Janelle Gonzalez and Diane Adams on the PRBR (Powder River Basin Restoration) project. Primarily, I will be looking at aerial photographs for areas of cheatgrass (Bromus tectorum) using remote sensing techniques. There will many other GIS projects that I will be doing during the late Winter/ Early Spring season. All of this will be explained in detail below!

Opportunities
I have been at this internship for almost a month and I have been extremely busy with all kinds of GIS and office activities. At the very beginning, I had computer access trouble, so I did a variety of interesting jobs. I cleaned out the range files area and organized many of the maps and notes, so the overall area would not be cluttered. I helped out with billing and I learned about allotments in the Buffalo Field Office area and about the people who work with the BLM. I did two different kinds of data entry. I updated the Eagle Survey Database and I inputted AIM data into Microsoft Access and produced reports for the fire ecologist!

When I had access to my computer, I was able to do a variety of GIS tasks regarding the PRBR project. (I will explain more about this later on.) Another job that they wanted me to do was to document and organize all of the very old aerial photographs from 1954-1969. These huge aerial photographs were surprisingly heavy and it took me multiple trips to carry all of the photographs back to my office from cold storage. I think it was around seventy five pounds worth of aerial photographs. I had to look at the top of the photographs and type into a computer database the date and the serial number of each photograph. Then I had to look in the Wyoming NOC database to see if these aerial photographs have been scanned and digitized into a file. If they were not scanned in, I had to make a note about which of the photographs had to be scanned. Some of the photographs from the 1950s were eerily beautiful. There were clouds in the photographs and some of the landscapes had some light fog, so the pictures looked like they were taken during prehistoric times. Some of the forested areas looked very haunted in north central Wyoming. O___O

A collection of 1954-1969 maps! They were very heavy and were difficult to carry around. It was worth the effort! Looking at these aerial photographs was an interesting experience.

I had to write down the information on the top of each map in a Microsoft Excel file and document them. In the future, I will see if these maps have been digitized or scanned into the computer.

Another small opportunity was to attend a regional GIS meeting in Buffalo with Diane. I was able to talk with other GIS people in Wyoming and learn about GIS projects and activities that were going on in the Powder River Basin. I learned all about the GIS tools that were being used for the energy sector in Wyoming, which was very interesting to learn about. I also was fortunate to learn about the GIS programs that were being used for sage grouse monitoring!!

Remote Sensing

For the majority of my winter time in Buffalo, I will be working on remote sensing classification of landscapes in the Powder River Basin study area. In 2014, the BLM worked with the Department of Transportation to take aerial photographs of a large section of the Powder River Basin in north central Wyoming. The plane took very detailed aerial pictures and digitized them as four banded orthorectified photographs for computer use. The pixels size for the photographs were six inch by six inch. This meant we could get detailed and accurate images of sagebrush and individual plant colonies (even individual cows). The plane took pictures during a time when cheatgrass was drying up and becoming senescent. During this time, the cheatgrass would show up as dark red in a green landscape. Since the orthorectified photographs have four bands, I could easily use various remote sensing techniques to isolate the cheatgrass light signature and make a specific raster file for cheatgrass for the BLM to use for future restoration work!

There were about one thousand mosaic tiles of landscape imagery of the Powder River Basin. Each mosaic tile incorporated a few sections in a township. I would process each mosaic tile with a few tools. The first part would be to look at various signatures in the landscape. I would take samples of each signature in order to do a classification at a later point in time. Using an image classification tool in ArcGIS I drew polygons around various signatures I wanted and made a small template of sample signatures. I drew polygons around areas of cheatgrass (1), cured cheatgrass (2), roads (4), bare ground (7), riparian shrubs (10), scoria hills (12), grey badlands (13), sagebrush (16), deciduous trees (18), conifers (19), riparian grasses (20), shadows (23), upland grass (26), algae reservoirs (28), muddy reservoirs (29), and areas of no data (30). Once I developed a sample template with each group of signatures, I did an interactive classification method to see what the map would look like if each of the signatures were assigned a color. For example, all the areas of cheatgrass would be assigned a red color, all the sagebrush would be turquoise color, and all of the other signatures would be assigned a different color. The final product would show a raster map with ten to sixteen different colors representing all of the signatures.

I am using a classification tool in this picture. I drew polygons around certain signatures and assigned them a class. For example, I drew a few polygons around shadow, riparian grass and sagebrush signatures in the above photograph. The small box you see are the class signatures I created. Each color represents a different class signature. I will use this signature template for the image classification process in my next step.

After gathering my signatures, I used the interactive classification process. The process quickly looks at the signatures files and  projects the estimated results. You can see various patterns and images in the landscape. The Class 1 Red colored represents the cheatgrass signature. The image classification process is only used to see if different signatures results matched with the aerial orthophotographs. If there was a red area that shows up on image classification and it was not a cheatgrass area in the aerial orthophotographs, then I would have to readjust my signature sampling process.

If all the signatures were correct and the raster colors matched with the aerial images locations, then I would proceed with the maximum likelihood classification for the signatures. The maximum likelihood classification would look at the raster signatures and evaluate each pixel on the map. For example, the highest likelihood of a pixel with a similar signature to the cheatgrass sample would be assigned under a cheatgrass signature. The computer looks through all of the raster signatures samples and calculates which pixels belong to which raster signature class. The end result would show a raster signature map with a more accurate representation of each of the ~sixteen signatures that were sampled on the map. When you zoom in, all of the areas of cheatgrass would be assigned a specific color and the same goes with all of the other signatures. The overall product would show a very colorful mosaic of colors.

(Side note: Some signatures may not be present on the aerial photograph, so the number of signatures per mosaic tile may vary.)

The maximum likelihood classification process would fine tune the signature results creating a more accurate classification.

The final step would be to isolate the cheatgrass signatures into a another raster file. I would use a Spatial Analyst tool known as Con to isolate the cheatgrass signatures. After running the Con tool, the computer would extract the cheatgrass signatures only and create a raster file showing only areas of cheatgrass. When I turn off all of the layers and show the aerial image, I could add the extracted cheatgrass raster on the map and show specifically where all the cheatgrass areas are on the aerial image. The final product would show all of the areas in the Powder River Basin that have cheatgrass areas. Later on, I would evaluate density levels of the cheatgrass and show areas that might need to be treated for future restoration projects.

The yellow on this map shows where all of the cheatgrass areas are on the mosaic tile. The Con tool extracts the specific cheatgrass signatures and creates a raster file that could be laid on top of an aerial orthophotograph.

The final result would be to process all the cheatgrass signature files for each mosaic and combine them into a large raster file. Anyone in the future could use this data to look at cheatgrass densities and find out where there needs to be treatment. It is good to look at all the raster files together to make sure everything lines up and that you did everything correctly when creating the signature files.

There were a few small issues that I did encounter with detecting cheatgrass. Some of the signatures were a little askew due to the quality of the aerial photographs. Some of the photographs around the perimeter or ends of the study area were lighter in color, making cheatgrass detection a little harder. Scoria hills were orange to red and showed very similar light signatures to cheatgrass, so I had to be careful when choosing my sample signatures.

Some of the interesting views I see from these aerial photographs were astounding! When working with these photographs for a long period of time, I pictured myself looking at a different alien landscape. Some badland or riparian areas looked very exotic and from the air, it did not look like a landscape you would see in Wyoming. The only dominant lifeforms on these alien landscapes were cows. Some of the cows were so big I could actually see what they were doing. Some were sleeping, laying on their side, climbing up very steep topography, and some were standing in water. I would actually see their trails and daily activities. Some of the red angus cows show up as a cheatgrass signature, fortunately they are rare and could be easily detected apart from the cheatgrass when looking on the aerial photograph. There were some areas in the Powder River Basin where the red to purple cheatgrass was so dense it looked like Mars! I would not want to walk through that area when the grass is seeding. My socks would be ruined!

A small screenshot of a very dense population of cheatgrass!! D: This cheatgrass population is surrounded by sagebrush, riparian grass, cured cheatgrass, and upland grass areas.

Preparing for NISIMS
A very small project that I will be working on during my BLM internship would be doing NISIMS! I did NISIMS for the Wenatchee, Washington BLM and I loved looking for and documenting different kinds of invasive plants. I put together a small power point and I am researching some of the invasive plants that were in north central Wyoming. Later, when I am ground truthing my remote sensing results for the PRBR project, I will also be documenting different invasive plants that I have encountered on the landscape such as cheatgrass, leafy spurge (Euphorbia esula) and various thistles (Cirsium). I am very excited for the NISIMS opportunity and I can’t wait to go out in the field and begin classifying invasive plants again!
———————————————————————————————————————————-
Those are the events for this month! Stay tuned for more adventures, CLM Interns and Family Members!!

Have a Great Adventure Everyone!

Justin Chappelle
BLM: Buffalo Field Office
Buffalo, Wyoming

And Now….. Your Moment of Zen

Sunset at the Grand Tetons!

PS. Time for a Secret A-HA Moment! \(>_>)/
Be warned about eating berries in the sagebrush steppe or badlands. I have seen people eat berries from Astragalus and Sambucus and get really sick from them. You have to be aware of the properties of the plant before consuming a berry. Some berries have to be cooked and prepared. Different plants growing on badlands or salt flats may contain harmful alkaloids and selenium, which would be stored in the seed pods/ seeds. Please check with your supervisor or book about consuming berries.