This fork aims to remove the arcpy dependency for this script and replacing it with rasterio. This will remove the requirement of an ESRI license. This repository contains a python script for automatically differentiating areas of rock outcrop using Landsat-8 data. The python script applies the method of Burton-Johnson, et al. (2016) to automatically identify rock outcrop areas from top of atmosphere corrected Landsat-8 tiles from Antarctica. Relevant modifications should be made for application to other Landsat datasets where band numbers may change.

1. Clone or fork this repository
2. Clone or fork landsat-utils

3. Download supplementary material from publication | (Direct download link for tc-10-1665-2016-supplement.zip)
4. Unzip supplementary material and locate "Landsat Tile IDs - Differentiating snow and rock in Antarctic.txt" (File location in zip = "'Supplementary\ Materials'/'Landsat Tile IDs - Differentiating snow and rock in Antarctic.txt'")
5. Write a quick script to load the text file into memory (it's a tab-delimited file) and return a list of scene ids (the values in the first column). This serves as the input for the landsat-util/landsat/downloader.py

6. Create a virtual environment in your base directory

virtual env "env_name"

7. Activate your virtual environment and install pip requirements NOTE: install requirements from this repository, not landsat-utils"

source "env_name"/bin/activate
pip install -r requirements.txt

8. Install homura (landsat-util/landsat/downloader.py requires it). landsat-utils/requirements.txt includes versions of some packages that are no longer available.

pip install homura

You should now have everything you need to download images, correct them, and generate output from the model.

This proof of concept example is designed to be run on a local machine with reasonable resources (>=4 GB memory, >= 50 GB storage)

1. Create a script that creates an instance of the landsat-util/landsat/downloader.py Downloader class.

• The only necessary parameter is the download_dir which should be the string of a path to a directory with at least 2GB of storage. The image correction and model scripts require that your image directory have the structure path/to/images/"SCENE_ID of a single image"/ e.g. path/to/images/LC82201072015017LGN00 that contains .tif files named "LC82201072015017LGN00_B#.TIF

2. Call the Downloader.download(["scene_id"]) method on your class instance passing a list containing a single scene id from the Supplementary material.

3. Untar the downloaded file (This might take a while)

untar xf "filename"

You now have a directory with a .tif file for each band and a .txt file containing the correction constants.

4. Create an instance of the LandsatTOACorrecter class from image_correction.py

• This class should be instantiated with the path to the directory containing the landsat tile bands.

5. Call the LandsatTOACorrecter.correct_toa_reflectance() method and the LandsatTOACorrecter.correct_toa_brightness_temp()

This will create a new directory "corrected" within your scene's directory with a copy of the band .tif files that have corrected values These corrected bands are the input to the model

tile

1. Create a textfile with a single line containing the scene id (this step will be removed in future versions of the model)

2. Open the jupyter notebook burton_johnson_model.ipynb

3. In Cell 3,

• Change the values of landsatTileList to the name of the file created in step 1

• Change the value of landsatDirectory to the directory containing the corrected bands "path/to/images/corrected/'scene id'"

4. Run all cells in order except for:

• Cell 7, first line:

schema = {'geometry': 'Polygon', 'properties': {'area': 'float:13.3', 'id_img': 'int', 'id_ADD':'int'}}

• Cells 10-14, first line of Cell 10:

with shp.Reader(coastMaskShpfile) as coast:

The final two cells write the output file to the corrected image directory and a full color image of the extent for which the model generated. These files have suffixes "_burjo_output.tif" and "full_color_seg.tif" respectively

1. Fork it
2. Create your feature branch: git checkout -b my-new-feature
3. Commit your changes: git commit -am 'Add some feature'
4. Push to the branch: git push origin my-new-feature
5. Submit a pull request

• 11 May 2016: Uploaded to GitHub
• 03 Nov 2015: Created ArcPy script

Burton-Johnson, A., Black, M., Fretwell, P. T., and Kaluza-Gilbert, J.: A fully automated methodology for differentiating rock from snow, clouds and sea in Antarctica from Landsat imagery: A new rock outcrop map and area estimation for the entire Antarctic continent, The Cryosphere Discuss., doi:10.5194/tc-2016-56, in review, 2016.

Code released under the GPLv3 license.