MEDS Courses

Quantitative skills and understanding are critical when working with, understanding, analyzing and gleaning insights from environmental data. In the intensive EDS 212 course, students will refresh fundamental skills in basic math (algebra, uni- and multivariate functions, units and unit conversions), derivative and integral calculus, differential equations, linear algebra, and reading, writing and evaluating logical operations.

This course will teach students how to store and manage environmental information. The course will focus on relational database structure, schemas and data relationships, and introduce SQL as a means to create and query databases. This course also covers the concept of metadata as well as archiving data products on data repositories to make them available to the broader community.

The generation and analysis of environmental data is often a complex, multi-step process that may involve the collaboration of many people. Increasingly tools that document and help to organize workflows are being used to ensure reproducibility, shareability, and transparency of the results. This course will introduce students to the conceptual organization of analytical workflows (including code, documents, and data) as a way to conduct reproducible analyses. These concepts will be combined with the practice of various tools and collaborative coding techniques to develop and manage multi-step analytical workflows as a team through students' first group projects. Students will also be introduced to the use of a remote server to conduct data-intensive analyses.

This course teaches the fundamentals of programming in Python. Students will learn foundational skills and concepts including data structures, programming basics, and how to clean, subset, aggregate, transform and visualize data. Course materials demonstrate the application of these techniques for environmental data analysis and problem solving.

Introduces students to the broad range of data sets used to monitor and understand human and natural systems. Course will cover field and station data, remote sensing products, and large-scale climate datasets including climate model projections. Skills will include evaluating data collection and quality control methods used in existing datasets, and working with existing databases of time-series and spatial information including cloud computing databases and new repositories of environmental datasets. Students will learn basic workflows for selecting, obtaining, and visualizing datasets, and best practices for reliable data intercomparisons. Students will gain hands-on experience with an environmental dataset of their choice by developing tutorial Jupyter notebook materials for a relevant use case.

This course teaches key scientific programming skills and demonstrates the application of these techniques to environmental data analysis and problem solving. Topics include structured programming and algorithm development, flow control, simple and advanced data input-output and representation, functions and objects, documentation, testing and debugging. The course will be taught using a combination of the R and Python programming languages.

This course teaches a variety of statistical techniques commonly used to analyze environmental data sets and quantitatively address environmental questions with empirical data. The course covers fundamental statistical concepts and tools, including sampling and study design, linear regression, inference, and time series analysis, as well as foundational concepts of spatial and space-time dependency and associated impacts on inference.

This course introduces the spatial modeling and analytic techniques of geographic information science to data science students. The emphasis is on deep understanding of spatial data models and the analytic operations they enable. Recognizing remotely sensed data as a key data type within environmental data science, this course will also introduce fundamental concepts and applications of remote sensing. In addition to this theoretical background, students will become familiar with libraries, packages, and APIs that support spatial analysis in R.

Computer-based modeling and simulation for practical environmental problem solving and environmental research. The course will cover both the selection and application of existing models and best practices for designing new models. Topics include conceptual models, static and dynamic models, and models of diffusion, growth and disturbance. Techniques include sensitivity analysis, calibration and model scenario design.

Machine learning can help process big/complex data and extract knowledge. It forms one of the foundations in data science. This course provides a broad introduction to machine learning and statistical pattern recognition. Topics include supervised learning (decision tree, random forest, support vector machines, neural networks) and unsupervised learning (clustering, dimensionality reduction, deep learning). Problems and exercises are framed within environmental science applications. The course will use programming languages like R and Python to support learning how to do advanced scientific programming to solve real environmental problems.

Effectively communicating your work in a responsible, accessible and visually-pleasing way is often (if not, always) a central part of data science. This course will focus on the basic principles for effective communication through data visualization and using technical tools and workflows for creating and sharing data visualizations with diverse audiences. By the end of this course, learners should be able to: (1) Identify which types of visualizations are most appropriate for your data and your audience (2) Prepare (e.g. clean, explore, wrangle) data so that it’s appropriately formatted for building data visualizations (3) Build effective, responsible, accessible, and aesthetically-pleasing, visualizations using the R programming language, and specifically {ggplot2} + ggplot2 extension packages (4) Write code from scratch and read and adapt code written by others (5) Apply a DEI (Diversity, Equity & Inclusion) lens to the process of designing data visualizations (6) Assess, critique, and provide constructive feedback on data visualizations

This course will present state of the art program evaluation techniques necessary to evaluate the impact of environmental policies. The program evaluation methods presented will aim at identifying and measuring the causal effect of policies, regulations, and interventions on environmental outcomes of interest. Students will learn the research designs and methods for estimating causal effects with experimental and non-experimental data. This will prepare the students for interpreting and conducting high-quality empirical research, with applications in cross-sectional data and panel data settings.

This course will focus on ethical considerations in collecting, using, and reporting environmental data, and how to recognize and account for biases in methodologies and algorithms. Students will also examine the human and societal implications of these issues within environmental data science.

Advanced, special topics in environmental data science. May be repeated for credit with changes in content.

Having a polished online presence is essential for showcasing your skills, projects, expertise, and (importantly) your personality. Learners, collaborators, and future employers alike will look to your public online profiles to glean information about you and your work. As data scientists, this often means that they’ll head to one of two places: GitHub and/or your personal website. This course is designed to help you both lay the foundation for creating and maintaining these profiles in an organized, informative, and visually-appealing way.

As climate change becomes a more and more urgent concern, there is a pressing need for trained professionals able to understand and work with data generated from global climate models, across a wide range of industries. However, the datasets involved can be enormous - on the order of many terabytes. This makes climate data an excellent way to learn 'big data' techniques while also gaining highly marketable skills in analyzing and visualizing information from future projections generated with climate models. Students in this course will learn how to access large datasets via cloud computing services (Amazon Web Services), and become familiar with using network Common Data Format (netCDF) data and the terminology associated with the Coupled Model Intercomparison Project (CMIP). Students will also gain skills in dimension reduction techniques for data visualization and multivariate geospatial statistics including regression and covariance maps and principal component analysis.

The {shiny} package provides a framework that allows R users to build interactive web applications and dashboards, and has become a popular tool for sharing data analyses and data-derived outputs with broad audiences. In this two-part short course, students will learn the fundamentals of reactivity, how to customize an application user interface (UI), best practices and workflows for developing shiny apps, and how to deploy their apps online via RStudio’s hosting service, shinyapps.io.

First quarter of a two-quarter group study/analysis of how to apply data science and tools to an environmental problem. In this quarter, students are expected to work with their project client to finalize project plans, assign individual roles and responsibilities, develop a project design plan and deliverables, and make significant headway on implementing those plans.

Second quarter of a two-quarter group study/analysis of how to apply data science and tools to an environmental problem. In this quarter, students are expected to complete all project plans and deliverables, develop and submit a project repository and technical documentation, give an oral defense of the project, present the research to a general audience.