GEMS Learning - Species Distribution Models (SDMs): Parametrizing, Modeling, Evaluation, and Interpretation

Location & Details

Nov. 9, 2023 to Dec. 14, 2023
9:00am to 12:00pm
About This Event: 

GEMS Learning provides modular non-credit digital and data science education for working professionals and students in food, agriculture, and natural resource application areas. Across the curriculum, instructors have built their course content from their own work executing large-scale data science projects to solve agricultural problems. 

Series: Spatio-Temporal Accounting of Biotic Threats

Have your research, studies or work required you to examine the geographic distribution of various species? This could be for crop protection, forestry, environmental protection, environmental impact assessment, urban and natural landscape design and development, or simply an interest in any given species, whether insects, pathogens or weeds? Then you are at the right place, species distribution models allow us to understand the potential and realized distribution of various species across our landscapes at different scales.

GEMS Learning - Species Distribution Models (SDMs): Parametrizing, Modeling, Evaluation, and Interpretation

Successful completion of this course will equip any student, researcher, practitioner, or extension worker with the ability to conduct sound and robust species distribution modelling. It starts from choosing the right SDM approach for the kind of biological information, occurrence dataset or environmental predictors to which we have access. Even though the focus will be on correlative SDMs, the discussion also will cover mechanistic SDMs. Throughout this practical course we will learn best practices that will help optimize developing, parametrizing and running species distribution models (SDMs). R data analysis software will be the major platform for most of the SDMs, however we will also learn how to run MaxEnt models for presence-only data.   We will also learn how to evaluate our SDM results.  We will use external evaluation data to check on accuracy of the modelled species distribution maps. In addition, many SDMs already provide a geographic projection of the species habitat/climate suitability, however more value can be added by crossing these data with other relevant thematic datasets for example with location of where the host crops for the modeled species grow. We will also learn how to interpret habitat suitability results. 

Fee: $1,049 (no charge for U of M affiliated)

The other course in the Spatio-Temporal Accounting of Biotic Threats series is:

Scholarships are available (see the link on the registration page). See the full line-up of courses and register.