REMOTE SENSING ANALYSIS OF LAND LOSS PROCESSES IN THE MISSISSIPPI RIVER DELTAIC PLAIN

The wetlands comprising the Mississippi River Delta Plain (MRDP) on Louisiana’s coast have been disappearing over the last century due to river management decisions, such as levee and dam building. This problem is immense given the cultural and economic value of the MRDP. While the occurrence of land loss is well known, it is not yet clear what processes drive land loss. Here we use remote sensing data to investigate whether land loss occurs by edge erosion from waves, or relative sinking from subsidence. To do this we track changes in shape of 795 lakes on the MRDP over the past 30 years. We use Landsat imagery in Google Earth Engine to create composite images. Composite images are created by calculating the cloud free, median value on a pixel-by-pixel basis over a three-year time period. On each composite image we selected all lakes within the Atchafalaya, Terrebonne and Barataria Watersheds which were detected in each composite image and were larger than sixteen pixels. We track the change in centroid location and area of each lake through the composite images, taking care to filter out erroneous data and those lakes with area changes less than 3600 m² between 1983 and 2016. Additionally, we eliminate lakes that merged during this time period, resulting in shared centroids. In the Atchafalaya Watershed half of the lakes are contracting and half are expanding. In Terrebonne and Barataria however, 90% of the lakes are growing. Overall, 81% of the lakes are growing in size, indicating as expected erosion is more dominant than accretion. Histograms of centroid movement direction for each basin are bi-modal with a dominant peak corresponding to NW and secondary peak to SSE. These two peaks are consistent with dominant wind-directions measured from buoys in offshore Gulf of Mexico. Additionally we find a positive correlation between lake area and magnitude of centroid movement, which recent theory suggests is a sign of wave-driven edge erosion since bigger lakes usually have longer fetch. These two results suggest to us that wind-driven wave edge erosion may be an important process of land loss on MRDP. Future work includes comparing changes in lake centroid position and area with data from the Coastwide Reference and Monitoring System in Louisiana.