Pygaarst: geospatial analysis and remote sensing tools for Python

Pygaarst is a Python package that is designed to make it easy to access geospatial raster data such as remote sensing imagery, and perform frequently needed processing steps in a human-friendly way.

Loading data, accessing the intended band or dataset, converting it to meaningful units and calculating standardised indices should be easy. As should be plotting the imagery on a map or combining raster with vector datasets, such as ESRI shapefiles.

Some examples:

>>> from pygaarst import raster
>>> basedir = "path/to/data"
>>> landsatscene = "LE70690142004201EDC01"
>>> sc = raster.Landsatscene(os.path.join(basedir, landsatscene))
>>> swirband = sc.band7
>>> type(swirband.data)
<type 'numpy.ndarray'>
    >>> type(swirband.radiance)
<type 'numpy.ndarray'>
>>> swirband.radiance.shape
(8021, 8501)
    >>> ndvi = sc.ndvi

As of the moment this documentation is written, pygaarst is under development and not all of the API design should be considered stable. The following capabilities are supported:

  • Python 2 (2.6+)
  • GeoTiff base class (in pygaarst.raster), with Landsatband, ALIband and Hyperionband inheriting from it
  • Landsatscene, Hyperionscene and ALIscene to represent a scene directory as retrieved from the USGS data portal
  • VIIRSHDF5 for NPP VIIRS SDS dataset files in HDF5 format, as retrieved from NOAA’s portals
  • Reading georeference files or metadata and calculation of pixel-center and, for gridded data, pixel-corner coordinate references
  • A metadata parser for USGS-style MTL files in pygaarst.mtlutils
  • Multiple helper methods, functions and properties for frequently repeated tasks: transformation to at-sensor radiance and reflectance, NDVI and NBR (for Landsat) as well as generic normalized difference indices, radiant temperature from thermal infrared bands, calculation and export of radiance spectra for Hyperion, LTK cloud masking algorithm for Landsat...
  • A client for the NASA’s modaps data download API in pygaarst.modapsclient

The following capabilities are planned, roughly in order of priority:

  • HDF4 (EOS) swath datasets (MODIS and ASTER Level 1B)
  • Basic geometric and statistical operations involving raster and vector structures such as overlays, find-nearest, is-in...
  • Python 3 support
  • MODIS gridded data products

A step-by-step example of using pygaarst.raster to access and plot VIIRS and Landsat data is worked through in this IPython Notebook.

Release:0.1
Date:March 04, 2015

Contents:

Indices and tables