Quantifying Changes in Wetland Area and Habitat Types in the Deepwater Horizon Louisiana Restoration Area 1985-Present with Remote Sensing
USGS researchers will quantify wetland change and wetland vegetation community type change through the analyses of aerial vegetation survey data and investigate potential relationships between Normalized Difference Vegetation Index (NDVI) and marsh elevation change.
The Science Issue and Relevance: Given the unprecedented temporal, spatial, and funding scales associated with the Deepwater Horizon (DWH) oil spill restoration effort, data to inform robust Monitoring and Adaptive Management (MAM) is needed to support restoration planning and implementation. The selection, monitoring, and management of coastal restoration projects requires a method of accurately and rapidly quantifying and assessing historical, current, and future predicted emergent vegetated wetland habitat area. The objectives of this effort are to develop geospatial datasets that can be used to assess historical trends in wetland area, assess restoration project impacts to wetland area and zonation of wetland habitat types that account for historical trends and interannual variability in those metrics, and to advance our understanding of elevation thresholds where vegetation productivity is sufficient to promote elevation gain rates that offset relative sea-level rise (RSLR).
Methodology for Addressing the Issue: This assessment proposes to analyze all cloud-free dates of Landsat (1985-present) and Sentinel-2 imagery (2015-present) (Data Sources: USGS Earth Resources Observation and Science Center, European Space Agency). Images will be classified to quantify the wetland area in each image. Aquatic vegetation will be identified in each image so that it may be accounted for and its effect removed from wetland area change analyses. Water level is known to be one of the most significant contributors to variability in wetland area estimates. However, the lack of spatially distributed gauges with a sufficient period of record has led to more simplistic means of accounting for variable water levels in past analyses. In this analysis, we will investigate the use of estimated water level based on relationships of specific locations to long term gauges. For each restoration project in the study area, without action (WOA) projections, including uncertainty ranges, will be created using the historical, pre-construction trends and compare those to post-construction observations to calculate wetland area benefit attributable to all constructed coastal restoration projects.
To quantify wetland change and wetland vegetation community type change, this effort will utilize aerial vegetation survey data, as well as patterns observed in multi-temporal, remotely sensed imagery to classify vegetation community types in coastal Louisiana. These classifications will be created annually for Landsat and Sentinel-2 imagery, and the resulting classifications will be used to analyze trends and change amongst coastal wetland vegetation communities through time.
Finally, this effort will investigate potential relationships between Normalized Difference Vegetation Index (NDVI) and marsh elevation change. For community types where significant relationships between NDVI, elevation change rate, and inundation are estimated, we will synthesize the two analyses to serve as the informative basis for delineating target elevation ranges for created marshes of various vegetation communities.
Future Steps: The datasets and information produced by this effort will be used as baseline monitoring for ecosystem restoration. These datasets will be used to assess project impact on land loss and habitat change in the context of background variability and trends and allow for adaptive management decisions for future restoration efforts. Additionally, the outcomes of this effort may be used to better inform the development of restoration objectives, and may enable resource managers to quantify appropriate elevations for different marsh vegetation types.
USGS researchers will quantify wetland change and wetland vegetation community type change through the analyses of aerial vegetation survey data and investigate potential relationships between Normalized Difference Vegetation Index (NDVI) and marsh elevation change.
The Science Issue and Relevance: Given the unprecedented temporal, spatial, and funding scales associated with the Deepwater Horizon (DWH) oil spill restoration effort, data to inform robust Monitoring and Adaptive Management (MAM) is needed to support restoration planning and implementation. The selection, monitoring, and management of coastal restoration projects requires a method of accurately and rapidly quantifying and assessing historical, current, and future predicted emergent vegetated wetland habitat area. The objectives of this effort are to develop geospatial datasets that can be used to assess historical trends in wetland area, assess restoration project impacts to wetland area and zonation of wetland habitat types that account for historical trends and interannual variability in those metrics, and to advance our understanding of elevation thresholds where vegetation productivity is sufficient to promote elevation gain rates that offset relative sea-level rise (RSLR).
Methodology for Addressing the Issue: This assessment proposes to analyze all cloud-free dates of Landsat (1985-present) and Sentinel-2 imagery (2015-present) (Data Sources: USGS Earth Resources Observation and Science Center, European Space Agency). Images will be classified to quantify the wetland area in each image. Aquatic vegetation will be identified in each image so that it may be accounted for and its effect removed from wetland area change analyses. Water level is known to be one of the most significant contributors to variability in wetland area estimates. However, the lack of spatially distributed gauges with a sufficient period of record has led to more simplistic means of accounting for variable water levels in past analyses. In this analysis, we will investigate the use of estimated water level based on relationships of specific locations to long term gauges. For each restoration project in the study area, without action (WOA) projections, including uncertainty ranges, will be created using the historical, pre-construction trends and compare those to post-construction observations to calculate wetland area benefit attributable to all constructed coastal restoration projects.
To quantify wetland change and wetland vegetation community type change, this effort will utilize aerial vegetation survey data, as well as patterns observed in multi-temporal, remotely sensed imagery to classify vegetation community types in coastal Louisiana. These classifications will be created annually for Landsat and Sentinel-2 imagery, and the resulting classifications will be used to analyze trends and change amongst coastal wetland vegetation communities through time.
Finally, this effort will investigate potential relationships between Normalized Difference Vegetation Index (NDVI) and marsh elevation change. For community types where significant relationships between NDVI, elevation change rate, and inundation are estimated, we will synthesize the two analyses to serve as the informative basis for delineating target elevation ranges for created marshes of various vegetation communities.
Future Steps: The datasets and information produced by this effort will be used as baseline monitoring for ecosystem restoration. These datasets will be used to assess project impact on land loss and habitat change in the context of background variability and trends and allow for adaptive management decisions for future restoration efforts. Additionally, the outcomes of this effort may be used to better inform the development of restoration objectives, and may enable resource managers to quantify appropriate elevations for different marsh vegetation types.