editor_options	markdown wrap 72

ucsb-ratlas

A geospatial R example atlas of the UCSB campus

This atlas started as an effort to mimic all of the steps of the Carpentries' Intro to Raster and Vector Data workshop using data local to UCSB.

In the years hence, that lesson got stagnant, and Dutch Bicycle Carpentry was under a lot more active development.

And then AI happened.

Now this is an atlas heavily edited by Gemini.

One set of scripts runs parallel to the old Geospatial Data Carpentry episodes (ep_01.r ep_02.r ... ep_13.r) . Those episode scripts were backwards designed from maps that we were calling 'suitable for publication' (map01.r, map02.r, ... map12.r) as laid out in Maps 1 thru 12 below.

The goals were to produce nice atlas pages of campus that use all of the techniques covered in the Intro to GeoSpatial R Carpentry lesson. Now it's a lot more free-form

We created a Data Dictionary to help us keep track of object names and files. Gemini still needs to update that.

Getting Started

1: Clone this repo.

2: Run scripts/data_prep_new.r This script downloads a folder from our Carpentry Google drive into a downloaded_data directory), and unzips it into source_data

3: Now you can run run_every_map.r and run_every_episode.r (in that order) to produce output from these data sources. Any data that an episode writes is placed in output_data and any formatted maps are placed in images. Both are .gitignored. final_output has .png files used in the readme.

4: Script away! Feel free to tackle issues, express issues, or just do work if you see work that needs to be done.

5: Episode scripts produce a number of maps, but not particularly well formatted. They are formated as in the Lesson, with the addition of ggtitles to keep track of where they are generated.

All outputs from map scripts should have a 3 tall x 4 wide aspect ratio, except where noted.

Fall 2026 changes

This is a gratuitous edit to force a rebuild of the new web page

Check out our progess

Map 1. A wide view of campus with:

• Extent should be the same as #3 inset of map 7.
• NCOS – for now the new lagoon habitat shapefile
• Water
• Bathymetry and elevation in one layer
  • hillshade
• bike paths
• buildings – for context
• vernal pools:
  • vector data to be create via analysis from DEMs
  • this will come later

Something like this:

Map 2. A stylized thematic map with trees, water, and bikeways

• ArcGIS Online: Water:
• NCOS upper lagoon shapefile of bathymetric topo lines or polygons is it this bird habitat file? – yes bird habitats
• Shows as dark mode (and I like it)

Something like this:

Map 7 An atlas page layout with 4 insets:

Map 6 is the top tryptic = Maps 3-4-5 zoom-in.

Map 3. California Overview

Portrait 3x4 Western US Does it want water?

• remove tic marks

Map 4. The Bite of California

Portrait 3x4

• Needs to be further zoomed in?
• populated places needs to be better visualized
• needs water!
• shows as dark mode and I don't like it
• remove degrees

Map 5. Extended Campus

(issue #14) Landscape 4x3 * extended campus will have maptiles background? *

• shows as dark mode and I don't like it

Map 6. The tryptic

Wide Landscape 9x16ish? * A stripped down version of map #1, symbolized to match the look of 3-4-5

Map 7: The Layout

Map 7 sketch:

We used to have a jpg of the whiteboard here.

Map 8 RS Imagery

For starters, this will be one 8-band image visualized several different ways. Which is an expansion of episode 5. Something like this:

9 Analysis: Find landscape depressions on Campus DEM

ie: identify vernal pools. Find elevations > sea level that are surrounded by nearby neighbors that are higher.

10 Analysis: Find bike paths that cross water?

Map 11: Look at a Dibblee

Map 12 / Episode 12: 12 months of NDVI Raster Stack

UCSB Carpentry

Original lesson -- Introduction to Geospatial Raster and Vector Data with R