Spectral unmixing library for satellite land cover mapping
Project description
The Earth Library
earthlib
is a spectral library and a set of software tools for satellite-base land cover mapping. import earthlib as eli
The spectral library contains several thousand unique spectral endmembers representing green vegetation, soil, non-photosynthetic vegetation, urban materials, and char. The reflectance cover the visible to the shortwave infrared (400-2450 nm) at 10 nm band widths.
The software tools resample these data to match the wavelenths of popular satellite and airborne earth observing sensors and to run spectral mixture analysis in Google Earth Engine via the earthengine
python package.
This work is in development and has not yet been formally described.
Table of Contents
Installation
via pip
This library can be installed via pip directly from Github.
pip install git+https://github.com/earth-chris/earthlib.git
You can also clone the source repository and install it locally.
git clone https://github.com/earth-chris/earthlib.git
cd earthlib
pip install . -r requirements.txt
via conda
I recommend working with earthlib
in conda
(download from here). The environment.yml
file in this repository contains a series of packages that are either required (earthengine-api
) or just convenient (jupyter
, folium
) to have.
git clone https://github.com/earth-chris/earthlib.git
cd earthlib
conda env update
using iPython defaults
If you're interested in using our custom ipython
profile, which contains a few plotting defaults, you can set an environment variable to do this for you. From the base earthlib
directory, run the following:
conda activate earthlib
conda env config vars set IPYTHONDIR=$PWD/ipython
Data Sources
This package uses spectral library data from a range of sources. These include:
- The Joint Fire Science Program
- World Agroforestry (ICRAF) Global Soil Spectral Library
- UCSB's Urban Reflectance Spectra
- UW/BNL/NASA HySPIRI airborne calibration spectra
- USGS Spectral Library Version 7
- Vegetation spectra modeled using PROSAIL (using PyPROSAIL)
Contact
All (c) 2018+ Christopher B. Anderson & Lingling Liu. This work is supported by the Stanford Center for Conservation Biology and the Natural Capital Project.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.