burnSeverity

Burn Severity Mapping with spaceborne multispectral imagery from Sentinel-2 MSI and Landsat-8/9 OLI sensors. The process has been broken down into 4 parts:

Part A - Prepare file perimeter vector file
Part B - Burn Severity Product Generation (barc, quicklooks and qa powerpoint)
Part C - Filter and Export
Part D - Map Generation (not described here)

Methodology: https://catalogue.data.gov.bc.ca/dataset/fire-burn-severity-same-year

The burn severity products can be generated either using the Jupyter Notebook (BurnSeverity_Mapping.ipynb) or the Python script (main_PartB.py). I recommend that first time users begin with the Jupyter Notebook implementation, especially if the goal is to map fewer than 10 fires at once.

Google Earth Engine Registration

Please follow the instructions in Getting_Started_with_GEE.md.

Part A

Manually prepare fire perimeter shapefile in a GIS software. Ensure that the shapefile contains 5 TEXT fields:
Fire_NUMBE: firenumber or some unique identifier
pre_T1: start of pre-fire image interval ("yyyy-mm-dd")
pre_T2: end of pre-fire image interval ("yyyy-mm-dd")
post_T1: start of post-fire image interval ("yyyy-mm-dd")
post_T2: end of pre-fire image interval ("yyyy-mm-dd")

You can change the field names if you like, just make sure that you update lines 42 - 47 in main_PartB.py.

Part B

Jupyter Notebook Version

Click on the Notebook (BurnSeverityMapping.ipynb)
Click Open in Colab
Follow the steps.

Automated Version

Installation

Please follow the instructions outlined here: https://github.com/SashaNasonova/geeMosaics. These scripts require gee, geemap and osgeo gdal packages.

Execution

Download repository and store it locally. To prepare:

Open config.yaml in a code editor (ex. Notepad++)
Adapt configuration file including the GEE cloud project name
Define inputs: root directory, fire perimeter shapefile and platform (S2, L5, L7, L8, L9)
Select processing parameters: cloud masking, evaluation mode only, export pre/post alternates (quicklook images of all available images), select post-fire image by AOT/cloud metadata export data (NBR, dNBR, dNBR_scaled), or override

Override can be used if you already know which image dates you want to use. The script will accept a dictionary of pre- and post-fire image dates along with the platform (S2, L5, L7, L8 or L9). It can also be used for reruns with the help of quicklooks in the alt folder if the first attempt isn't satisfactory.

To run:

In Anaconda Prompt activate the gee environment
Run the following commands to authenticate gee

python
import ee
ee.Authenticate()

Follow prompts to authenticate
Edit batch file to run script or just run the main_partB.py script (python main_PartB.py)

Please use the test data to get started. The shapefile is already formated and filled out with the correct dates.

Part C

Script to filter, smooth, vectorize, and save as a geodatabase.
To run:

Open main_PartC.py in a code editor (ex. Notepad++)
Edit lines 48 - 51
Copy and paste entire script into the Python window in ArcPro/ArcMap and run

SashaNasonova/burnSeverity