FireHR

High-resolution burned areas using Sentinel-2 data


Keywords
Burned, Areas, Wildfires, Sentinel-2, VIIRS, Deep, Learning
License
Apache-2.0
Install
pip install FireHR==0.1.6

Documentation

A Practical Method for High-Resolution Burned Areas Monitoring using Sentinel-2 and VIIRS

Install using pip

pip install FireHR

Install from repo

git clone https://github.com/mnpinto/FireHR
cd FireHR; pip install -e .

How to use

Configuration of Google Earth Engine API

Run the following python code and follow the link to generate the authentication file:

import ee
ee.Authenticate()

Once you are done you should see the message Successfully saved authorization token. and the file ~/.config/earthengine/credentials should exist.

Please refer to https://developers.google.com/earth-engine/guides/python_install for more information about the GEE Python API.

Command line utility to run FireHR for an event as outputed by BA-Net post-processing

firehr_from_banet_events ba100m_PT2020_218.tif

Optional arguments:

Argument default type description
path '.' str Path to save the outputs.
composite_days_before 120 int Time window size in days for the pre-fire data
composite_days_after 120 int Time window size in days for the post-fire data
max_cloud_fraction None float (0.0-1.0) Remove images with a cloud fraction higher than specified
use_least_cloudy None int Select the n least cloudy images in the time_window

Example using use_least_cloudy parameter:

firehr_from_banet_events ba100m_PT2020_218.tif --use_least_cloudy 5