• How does R plot maps?
• What is a straight line, anyway?
• Large spatial datasets with R
• Lifecycle of R Spatial packages
• R Spatial community

Slides: https://edzer.github.io/UseR2021/

R Markdown at: https://github.com/edzer/UseR2021

• Unless you are in the business of actually creating globes,

… any two-dimensional plot of Earth-bound surface data involves projection

plot(map)

maps::map("world")
maps::map("usa")

?map
projection: character string that names a map projection to use.  See
          ‘mapproject’ (in the ‘mapproj’ library).  The default is to
          use a rectangular projection with the aspect ratio chosen so
          that longitude and latitude scales are equivalent at the
          center of the picture.

https://xkcd.com/977/

https://xkcd.com/977/

Projection 101

• For small areas, projection is usually no big issue (Earth is nearly flat for small areas)
• Projections cannot preserve distances; only distances to a single location
• Projections preserve area or shape or direction or some compromise, or …
• There are many of them!

• there is no need for
  • North to be up
  • Europe to be in the middle
• but a projection of a randomly rotated Earth is often harder to read!

For simple features, some definitions:

• feature: abstraction of real world phenomena, with geometrical properties
• simple feature: feature with all geometric attributes described piecewise by straight line or planar interpolation between sets of points

Straight lines, after (re)projection, are usually no longer straight. So, when are they straight?

GeoJSON IETF standard:

• A line between two positions is a straight Cartesian line, the shortest line between those two points in the coordinate reference system. […]
• The coordinate reference system for all GeoJSON coordinates is a geographic coordinate reference system, using the World Geodetic System 1984 (WGS 84) [WGS84] datum, with longitude and latitude units of decimal degrees.

• but do GeoJSON users realise that?

This example is contrived, but relevant for every line except sections of meridians or the equator.

We can add nodes (st_segmentise), or remove them (st_simplify).

For data with ellipsoidal (long/lat) coordinates, sf 1.0-0 switched to using spherical geometry (\(S^2\)) rather than Cartesian geometry (\(R^2\)). Straight lines are now great circle segments on a sphere.

As the Earth’s shape is closer to a sphere than to a flat plane, this is a good thing, but surprises are going to show up for a while due to Cartesian habits for some 50+ years.

To get the “old” (pre-sf 1.0) behaviour, use

• sf_use_s2(FALSE),
• st_set_crs(NA), or
• project to to +proj=eqc.

More discussion on this in the (upcoming) Spatial Data Science book by Roger Bivand and me; also do follow Dewey Dunnington.

• too large to
  • hold in memory (raster, terra, stars, sf, …)
  • keep in local storage
  • download (weather: ERA5, climate: CMIP6, earth observation: Landsat/Copernicus)
• Google Earth Engine (and similar) don’t allow you to reproduce analysis, independently
• openEO and openEO Platform are part of a larger initiative for allowing reproducible (open source, vendor independent) computing on large, cloud-based data archives, using R and other platforms
• In that same context, we contributed to the STAC (spatio-temporal asset catalogue) specification, a modern, light-weight approach for for discovery of images and image collections

• Many data scientists will run some day into challenges with spatial data
• R Spatial is an open and friendly community of people using the R package ecosystem for handling and analysing spatial data
• It uses and interfaces a lot of software used by a much larger community (e.g. OSGEO Foundation)
• sf now uses spherical geometry:
  • straight lines may need noding in the projection where they are (supposed to be) straight,
  • simplifying, if needed, should be done after projecting
  • automated noding may be needed at some stage
• we should reconsider projections used in default plots (after all, stringsAsFactors is also no longer TRUE)
• analysing large spatial data sets is, and will remain a challenge
• rgdal and rgeos will retire Jan 1, 2024:
  • sf, terra, stars, … provide alternatives:
  • UseR!s and developers will have to migrate