The application of satellite-derived datasets and geospatial analysis techniques in ecology and conservation has grown substantially over the last decade. With the emergence of cloud computing platforms that facilitate big data analysis, researchers, resource managers, and remote sensing enthusiasts are now able to interrogate petabyte-scale datasets with ease. Owing to new server-based infrastructure built by Google, anyone with an internet connection and a basic computer can conduct sophisticated spatial analyses. Via online lectures, hands-on practicals, and discussion sessions, this short course will teach you the foundations of applying Google Earth Engine to answer a range of ecological and conservation questions.
What is Google Earth Engine?
This course aims to train students, researchers, and practitioners in the application of Google Earth Engine (GEE) to conservation science. Specifically, it seeks to familiarize participants with the basic operation of the GEE environment,focusingon visualization, analysis, and automated detection of biological patterns and processes. The course will begin with a brief review of the fundamental theory behind remote sensing and geospatial analyses, followed by a series of tutorials on the following topics:
Introduction to remote sensing
What is remote sensing?
Approaches of capture and associated data characteristics
Atmospheric effects, corrections, and its value
Raster vs. vectordata models
Resolution and their trade-offs: spatial, spectral, temporal, and radiometric
Introduction to Google Earth Engine
Data catalog – satellite products useful for conservation science and ecology
Earth Engine editor
Google Earth EngineFundamentals
Understanding and developing your code
Indices and atmospheric correction
Binary change detection
Extract time-series data (e.g. NDVI)
What is supervised classification and how is it accomplished using GEE?
Training data vs. test data vs. validation data
Understand the capture of optical satellite imagery
Understand the trade-off between spatial, spectral, and temporal resolution
Compute and interpret spectral indices
Search, filter, visualize, upload, and download geospatial data using GEE
Obtain time-series environmental data for further analysis
General understanding of supervised classification
Perform change analysis and demonstrate the use of GEE in a conservation science framework
Basic familiarity with remote sensing and/or coding background (for example QGIS/ArcGIS, or R) is highly recommended. A background in biological sciences will be beneficial for practical examples and case studies. However, this is not essential. Before the course begins, participants will need to create a Google Earth Engine account (at least two weeks prior).
Further details of how to prepare for the course will be sent directly to registered applicants.
Course Dates: April 19th – April 30th, 2021. Sessions: 2-hour daily sessions (Monday – Friday) Session times: 6pm – 8pm (Standard South African Time)
Each session will consist of a theoretical introduction, demonstration of code, and self-learning practicals. Recordings of the live sessions will be made available to course participants.
Tuition is $600. Limited partial scholarships are available for students with demonstrated financial need. If you are interested in being considered for a partial scholarship, please make sure to include a scholarship motivation in your application. We will assess your situation individually and determine your eligibility for a scholarship if you are selected for the course.
Please note, seats are limited.
Dr Sandra MacFadyen is a landscape ecologist interested in macroscale ecosystem dynamics with an emphasis on applied spatial statistics for biodiversity conservation. Based in the Kruger National Park as a postdoctoral researcher with BioMath, Stellenbosch University, her research interests focus on exploring the links between patterns and processes to develop a more holistic understanding of ecosystem dynamics in large protected areas.
Dr Joseph White is a postdoctoral researcher at the University of the Witwatersrand, South Africa, working on species distribution shifts and disrupted ecosystem services in response to global change using occupancy models and remote sensed products. He is interested in spatial ecology and using earth observation to provide ecological and conservation insights.
Geethen Singh is a Ph.D. candidate at the University of the Witwatersrand, South Africa. He is interested in applying earth observation and machine learning within the domain of ecology and is currently working on better understanding invasive species spread and the invasive risk posed using earth observation-based monitoring.