Mapping Coastal Marsh Erosion and Migration to Estimate Blue Carbon Fluxes
Courtney Di Vittorio, Wake Forest University, divittoc@wfu.edu (Presenter)
Yasin Wahid Rabby, Wake Forest University, yasinwr@wfu.edu
Dorothy M Peteet, NASA / GSFC, peteet@ldeo.columbia.edu
Anastasia Romanou, Columbia University / NASA GISS, anastasia.romanou@columbia.edu
The aims of this NASA OBB project are as follows: (i) Quantify historical areas of coastal marsh changes along the Atlantic and Gulf Coasts over the past 40 years using the Landsat database, (ii) Develop a US inventory of coastal marsh depth estimates, (iii) Use these geospatial datasets to study the drivers of marsh erosion and migration rates and estimate the associated blue carbon flux, and (iv) Assess the significance of this flux relative to the carbon fluxes represented in NASA GISS Model-E (a global coupled ocean climate model), and explore the implications for predictive modeling. This poster summarizes our research activities and findings during the first half of this three-year project. Our research thus far has focused on the first two tasks, as they involve the creation of new datasets that are needed to implement tasks two and three. The historical coastal marsh erosion and migration maps (task #1) are highly dependent on the training and validation data and the approach for detecting change within time series of remote sensing indices. We have therefore iterated on several training datasets and classification methods to balance accuracy, scalability, and utility for end-users. During this process, we compared several US wetland inventories and explored the differences between them in the context of multi-decadal change detection, and we developed a new multi-temporal classification methodology. While developing this new method, we explored existing change detection and classification tools within Google Earth Engine and adapted them to coastal marsh mapping, enabling large-scale implementation of our approach. Toward task #2, we have compiled and screened existing in-situ databases that contain information on marsh depth across the Atlantic and Gulf Coasts and have added new transect data that was collected by one of our Co-PI’s. Several undergraduate students at Wake Forest University and Columbia University have supported this project by collecting and organizing geospatial datasets and by compiling the training data needed for classification.
Poster: Poster_Di_Vittorio_3-29_58_35.pdf
Associated Project(s):
Poster Location ID: 3-29
Presentation Type: Poster
Session: Poster Session 3
Session Date: Thu (May 11) 3:00-5:00 PM
CCE Program: OBB