Cartopy

Cartopy
Name	Cartopy
Developer	SciTools
Released	2010s
Programming language	Python
Operating system	Linux, macOS, Microsoft Windows
License	BSD

Cartopy is an open-source cartographic library for geospatial plotting in Python used in scientific and operational visualization. It provides projection transformations, feature handling, and map rendering that integrate with Matplotlib and tools used by researchers at institutions such as Met Office, NASA, and ECMWF. Cartopy is employed in workflows across projects like CMIP, Landsat, and Copernicus for producing publication-quality maps and operational displays.

Overview

Cartopy arose to address needs in atmospheric science and geoscience communities, aligning with software used at University of Oxford, University of Reading, and University of Bristol. It complements libraries such as Basemap (now deprecated), GeoPandas, Shapely, and PROJ for coordinate operations and integrates with datasets from NOAA, USGS, and ESA. Developers and users include members of Python Software Foundation, contributors from GitHub, and research groups participating in IPCC assessments.

Features and Architecture

Cartopy's architecture centers on projection objects and geometrical transforms tied to PROJ and GDAL. It exposes interfaces for creating maps using projections like Mercator, Lambert conformal conic, Transverse Mercator, and Robinson. Cartopy integrates with Matplotlib's Artist API and supports handling shapefiles from Natural Earth, OpenStreetMap, and custom datasets created by users at National Geographic. It uses Shapely for geometric operations, pyproj for CRS definitions, and interacts with raster sources used by MODIS, Sentinel-2, and Landsat.

Installation and Compatibility

Cartopy is distributed for platforms including Ubuntu, Debian, Red Hat Enterprise Linux, macOS, and Microsoft Windows. Binary packages are available through package managers and repositories used by Anaconda and pip, though installation often requires system libraries such as PROJ, GDAL, and GEOS. Continuous integration and build processes reference services from Travis CI, GitHub Actions, and CircleCI. Compatibility matrices often mention support with NumPy, SciPy, Pandas, and Matplotlib versions used in research at MIT and Caltech.

Usage and Examples

Typical usage patterns follow examples found in tutorials from SciPy, notebooks authored by researchers at University of California, Berkeley, and visualizations employed by teams at NOAA and NASA. Users create axes with projections, add coastlines and political boundaries from Natural Earth, overlay gridded fields from ERA5, NCEP/NCAR products, and annotate maps with labels used in publications by AMS and RMS. Integration examples include combining Cartopy with Matplotlib for contour plots, with Xarray for multi-dimensional climate datasets, and with ImageMagick workflows for generating animations used in outreach by NWS and EUMETSAT.

Development and Community

Development is coordinated via repositories on GitHub and contributions from institutions such as University of Exeter, Monash University, and Met Éireann. The project follows issue tracking and pull request workflows familiar to contributors to NumPy and Pandas. Documentation and examples have been presented at conferences including AGU Fall Meeting, EGU General Assembly, and PyCon, with community support via mailing lists and chat platforms used by Python Software Foundation. Funding and collaboration have involved agencies like UKRI and programs linked to Horizon 2020.

Performance and Limitations

Cartopy performs well for publication graphics and moderate-scale mapping tasks but depends on the performance of dependencies like GDAL, PROJ, and Shapely for heavy geoprocessing. Rendering large vector datasets from OpenStreetMap or high-resolution raster imagery from Copernicus can be memory- and CPU-intensive; practitioners often resort to tiling strategies employed by Mapbox or pre-processing pipelines used by ESRI products. Limitations also include complexities in CRS handling that users familiar with WGS 84 and EPSG:4326 must navigate, and occasional build issues on continuous integration services like Travis CI and AppVeyor.

Category:Geographic information systems