Title

Machine Learning Approach to Wetland Delineation

Submitter Information

Brianna Danielle LunardiFollow

Prize Winner

Healthy Great Lakes

Type of Proposal

Visual Presentation (Poster, Installation, Demonstration)

Start Date

22-3-2018 2:30 PM

End Date

22-3-2018 4:30 PM

Location

Atrium

Faculty

Faculty of Science

Faculty Sponsor

Chris Houser

Abstract/Description of Original Work

Machine Learning Approach to Wetland Delineation Wetlands are a natural resource protected under federal law, and understanding their boundaries are necessary for management and protection. This study examines the variability of wetland boundary delineation. Wetland boundaries have multiple definitions and delineations can vary by evaluator. Specifically the background and training an individual or group has will determine where they define the wetland boundaries and delineate inter-wetland ecological regions. Diagnostic features such as soil type, vegetation, and hydrology are often used to define where a wetland begins or ends. This interactive presentation will highlight the uncertainty of wetland boundaries. Data will be collected during the presentation and included as part of the study. The complete data base will be used in a machine learning analysis to determine the diagnostic features and scales that evaluators are using to delineate wetlands.

Grand Challenges

Healthy Great Lakes

Share

COinS
 
Mar 22nd, 2:30 PM Mar 22nd, 4:30 PM

Machine Learning Approach to Wetland Delineation

Atrium

Machine Learning Approach to Wetland Delineation Wetlands are a natural resource protected under federal law, and understanding their boundaries are necessary for management and protection. This study examines the variability of wetland boundary delineation. Wetland boundaries have multiple definitions and delineations can vary by evaluator. Specifically the background and training an individual or group has will determine where they define the wetland boundaries and delineate inter-wetland ecological regions. Diagnostic features such as soil type, vegetation, and hydrology are often used to define where a wetland begins or ends. This interactive presentation will highlight the uncertainty of wetland boundaries. Data will be collected during the presentation and included as part of the study. The complete data base will be used in a machine learning analysis to determine the diagnostic features and scales that evaluators are using to delineate wetlands.