leohs 0.1.7

Landsat ETM+ OLI Harmonization Script

🌍 LEOHS: Landsat ETM+ OLI Harmonization Script

LEOHS is a Python package for harmonizing Landsat 7 and 8 imagery using Google Earth Engine (GEE). The harmonization functions generated from this tool can be used for long term landsat time series analysis. LEOHS is designed specifically to create harmonization functions optimized for user-defined study areas, time periods, and sampling parameters.

---

🔧 Requirements

• Python 3.10 — LEOHS must be run in its own environment
• an active Google Earth Engine account
• an area of interest shapefile

---

📦 Installation

1. Create a clean Python 3.10 environment (recommended name: leohs_env)

conda create -n leohs_env python=3.10 -y
conda activate leohs_env

2. Install LEOHS

pip install leohs

🚀 Example Usage

Before using LEOHS, you must import both ee and leohs, then authenticate and initialize your Google Earth Engine (GEE) session. Only after that should you call leohs.run_leohs(...).

import ee
import leohs
ee.Authenticate()   #need to authenticate GEE
ee.Initialize()     #need to initialize GEE

leohs.run_leohs(
    Aoi_shp_path=r"E:\Canada.shp",
    Save_folder_path=r"E:\Canada_output",
    SR_or_TOA="SR",
    months=[6,7,8],
    years=[2017],
    sample_points_n=100000)

🔧 run_leohs Parameters

• Aoi_shp_path (str):
  Path to your input AOI shapefile.

• Save_folder_path (str):
  Path to the output folder where results will be saved.

• SR_or_TOA (str):
  Type of Landsat imagery to process. Choose "SR" or "TOA".

• months (list of int):
  List of months to include in image filtering (e.g., [1,2,3,4,5,6,7,8,9,10,11,12]).

• years (list of int):
  List of years to include in filtering (e.g., [2013,2014,2015,2016,2017,2018,2019,2020,2021,2022]).

• sample_points_n (int):
  Number of sample points to generate (e.g., 100000). Max: 1,000,000.

• maxCloudCover (int, optional, default=50):
  Maximum cloud cover (%) for image filtering.

• Regression_types (list of str, optional, default=["OLS"]):
  List of regression models to run. Valid values: "OLS", "RMA", "TS".

• CFMask_filtering (bool, optional, default=True):
  Whether to apply CFMask filtering (cloud, water, snow masking).

• Water (bool, optional, default=True):
  Allow water pixels (only effective if CFMask_filtering=True).

• Snow (bool, optional, default=True):
  Allow snow pixels (only effective if CFMask_filtering=True).

🛰️ Outputs

The following files are exported to the specified Save_folder_path:

• Text log (TOA_LEOHS_harmonization.txt):
  Contains regression equations for each band, processing time, and diagnostic logs.

• Heatmaps (.png files):
  Visualizations of pixel distributions between Landsat 7 and 8 for each band.

• Pixel and pair data (.csv):
  Sampled pixel values and image names for all matched images.

🐛 Known Issues

• Google Earth Engine computation timeout
  Occasionally, you may encounter an ee.EEException: Computation timed out error. This can happen when GEE servers are under heavy load.
  🛠️ Recommended fix: Simply wait a bit of time, and re-run leohs.run_leohs(...). The issue usually resolves itself on retry.

📂 Additional Resources

Additional scripts for applying LEOHS, as well as global harmonization equations, can be found in the companion repository:
🔗 https://github.com/galenrichardson/LEOHS

📑 License

This project is licensed under the
GNU General Public License v3.0 or later (GPL-3.0-or-later)
© 2025 Galen Richardson

See the full license text in the LICENSE file or at gnu.org/licenses/gpl-3.0.

---

📬 Contact

Author: Galen Richardson
Email: [e-mail]

Feel free to reach out for questions, bug reports, suggestions, or collaboration ideas.