geoglue

geoglue is an open-source project developed by the Kraemer Lab at the University of Oxford, designed to streamline the process of fetching and aggregating geospatial data to various administrative levels. The tool is particularly useful for researchers and practitioners working in fields like epidemiology, climate science, and public health, where spatially resolved data is essential. Some of the methods are inspired by methods in the R {terra} package that offers easy to use tools to work with rasters in-memory.

The core functionality of geoglue evolves around automating the (i) retrieval of geospatial datasets (boundary data from GADM, climate data from ECMWF) and aggregating them according to administrative boundaries such as countries (admin0), states or provinces (admin1), and counties or districts (admin2). The aggregation of raster data, such as temperature or total precipitation to zonal boundaries is termed zonal statistics. We use the exactextract package to perform zonal statistics that allows consideration of raster pixels that have a partial overlap with a zonal polygon.

geoglue has several modules:

• geoglue.cds: Methods to work with data downloaded from ECMWF’s cdsapi

• geoglue.region: Fetch geospatial boundaries data from GADM or geoBoundaries

• geoglue.memoryraster: Read and work with GeoTIFF datasets in memory.

• geoglue.resample: Resample data using the Climate Data Operators library.

• geoglue.zonal_stats: Perform zonal statistics on climate data, such as those downloaded from cdsapi via the geoglue.cds module