Modules
This course is divided into separate modules, each designed to give you a hands-on experience working in teams to reproduce or examine a fundamental result in global change ecology. Each module will also introduce a different type of data and different tooling to handle it, as described in the summaries below. You will spend a few weeks working through each module in your team, largely at your own pace. Weekly readings and introductory live-code sessions will provide some necessary background, but real learning will happen only by doing. Most of the questions in your assignment for each module do not have 'right' answers, but involve open ended research to fully explore. Students without prior statistical or computational experience should be able to complete initial exercises, while teams with greater prior experience are expected to push those boundaries by going deeper into the analysis and presentation. Workflow and communication are central elements to each module. All work should appear in professional and well documented format using RMarkdown notebooks in the project GitHub repository and pass all automated checks on Travis-CI.
Climate Change
Tabular data
What is the evidence for a changing climate? In this unit, we will examine some of the most important indicators of global climate change:
including CO2 concentration, global mean temperature, sea level rise, land ice sheet melt,
and arctic sea ice cover. We will wrangle a wide range of tabular data file formats used by
NOAA and other agencies and introduce the fundamentals of data visualization with ggplot2
and data processing with basic functions from the readr, tidyr, and
dplyr packages as we
seek to replicate the principle results of climate change over the last decades and also the
past thousands of years.
Overfishing
Relational databases
The last fish in the sea?
In a paper in Science in 2006, Boris Worm and colleagues seized international headlines
and fueled an ongoing scientific controversy when they documented the persistent declines
of fisheries on a global scale. In this module, we will explore the evidence behind these
controvertial results using some of the best publicly available data from global fisheries
stock assessments. This will introduce us to techniques for working with relational databases,
where the data we need is spread across multiple tables. We will get deeper into `dplyr` and
related tools.
Environmental Justice
satellite Geospatial data
Residential segregation and systemic racism have substantial impacts on landscape heterogeneity and ecological processes in US cities. In August 2020, Christopher Schell and collegues published a review in
Science on "The ecological and evolutionary consequences of systemic racism in urban environments". Here we explore one aspect of Schell's analysis on historical redlining and residential segregation in relation to current vegetation patterns in US cities. This module provides an introduction to the fundamentals of working with spatial vector and
raster data in R. Our approach places an emphasis on the
Simple Features Access standard
(ISO 19125) and tidyverse-style workflow using the
sf package and emerging
ecosystem of
r-spatial tools.
Mass Extinctions
widgets Non-rectangular data
Are we experiencing the sixth mass extinction? In this unit, we will
attempt to estimate the current rate of species extinctions across major taxa and compare
this against the background rate of extinctions in the fossil record; replicating seminal
work from a series of papers by Tony Barnosky and colleagues. Though rectangular data frames
have formed the core unit of our analysis so far, to answer these questions we will learn to
work with non-rectangular data. This module will introduce the fundamentals of working with
non-rectangular data in the JSON format returned by a REST API. We will also encounter the use
of Regular Expressions for extracting information from unstructured textual data.
From Worm 
From Schell 
From Cerbellos et al., https://doi.org/10.1126/sciadv.1400253