Derivation of Indicators from Satellite Observations of Vegetation Essential Climate Variables Reclaimed Well and Mine Sites in Alberta, Canada

:warning: This is a work in progress. Expect frequent changes to the code and functionality.

:globe_with_meridians: https://ronnyhdez.github.io/reclaimed_sites_ab/

Datasets

Repository structure

Repository is organize in:

reclaimed_sites_ab/
├── leaftoolbox/
│   ├── __init__.py
│   ├── leaf.py
│   ├── module1.py
│   ├── module2.py
├── notebooks/
│   └── abandoned_wells.qmd
│   └── gee_filtering.qmd
│   └── land_cover.qmd
│   └── leaf_process.qmd
│   └── negative_buffer_check.qmd
│   └── reference_buffers.qmd
├── data/
│   └── dataset.csv
├── scripts/
│   └── create_sampler_assets.py
│   └── download_data.py
│   └── flagging_assets.py
│   └── run_polygons_date_filter.py
│   └── run_sampler.py
│   └── shp_exports_for_assets.py
├── utils/
│   ├── __init__.py
│   ├── utils.py
├── .gitignore
├── README.md
└── Pipfile

• leaftoolbox: Modules
• notebooks: Documented analysis and data exploration
• scripts: Code to generate assets in GEE

Data Pipeline

This project involves several stages to process and analyze data effectively.

0. Download Data

The pipeline begins with downloading the necessary datasets. Run the script scripts/download_data.py to retrieve all the datasets required for the analysis.

1. Export Assets to Google Earth Engine (GEE)

The first stage of the data pipeline involves exporting the processed assets to Google Earth Engine (GEE). All the compute processing will be done in GEE.

2. Flagging assets

The second_stage consist of several steps to flag the data. These flags will help later for filtering the abandoned wells, which will later be used with the LEAF-toolbox sampler.to be used with LEAF-toolbox sampler

3. Create sampler assets

This third stage will take the flagged assets and based on the research objectives, will filter polygons accordingly. The resulting assets will be used finally to run the LEAF-toolbox

Running the code

WIP dev notes

In summary, steps to recreate the results:

• Create first assets in GEE with scripts/export_assets.py
• Flag assets and create buffers with scripts/flagging_assets.py
• Prepare the asset to be sample with LEAF-toolbox: scripts/create_sampler_asset.py
• Run the LEAF-toolbox sampler on selected abandoned wells: scripts/run_sampler.py
• Read the pkl files and check time series in abandoned_wells_leaf_toolsbox.qmd

References

Alberta Biodiversity Monitoring Institute and Alberta Human Footprint Monitoring Program. ABMI Human Footprint Inventory (HFI) for Alberta 2021 (version 1.0). Geodatabase. Last modified August 1, 2023.

Canadian Forest Service. 2021. Canadian National Fire Database – Agency Fire Data. Natural Resources Canada, Canadian Forest Service, Northern Forestry Centre, Edmonton, Alberta. https://cwfis.cfs.nrcan.gc.ca/ha/nfdb

Chowdhury, S. (2021): Land-use/Land-cover classification of Alberta, derived from 2020 sentinel-2 multispectral data (image data, TIFF format); Alberta Energy Regulator / Alberta Geological Survey, AER/AGS Digital Data 2021-0019.

Fernandes, R. et al., 2021, "LEAF Toolbox", Canada Centre for Remote Sensing, https://github.com/rfernand387/LEAF-Toolbox/wiki, DOI: 10.5281/zenodo.4321298.