Subtask: Present Day Northern Gulf Coast Landscape Structure

In Louisiana, estimates of land loss due to recent hurricanes based on 30 m resolution Landsat images are known to be biased downward, because patchy, fragmented wetland break-up is not fully resolved by moderate resolution satellite sensors that may have a large number of mixed classification pixels (mixels). Tests over pilot areas in southern Louisiana have revealed that sub 5 m resolution commercial satellite sensors such as QuickBird and IKONOS have the potential for more accurate fractional water cover mapping that will likely lead to improved estimates of episodic land loss due to hurricane landfalls. A time series of fractional land cover maps can identify temporal changes in surface properties, such as flooding and scouring due to extreme storm events. This activity will test methodologies to effectively increase accuracy of fractional land cover mapping over large areas using multi-temporal, multi-spatial and multi-source remote sensing and geospatial data.

1. Provide improved estimates of Louisiana wetland land loss due to hurricane impacts since 2004 based upon a change detection mapping analysis that incorporates pre- and post landfall (Katrina, Rita, Gustav, Ike) fractional water classification of a combination of QuickBird, IKONOS, and Landsat images. This analysis will also support the historical aerial photography-based retrospective investigation of the processes of episodic land loss that is the focus of NGOM Project Subtask 3.6.
2. Develop a strategy to identify the best data sources and methods to map and monitor wetlands over large areas using multi-temporal, multi-spatial scale and multi-source remote sensing and geospatial data. The strategy created for wetland mapping and monitoring is to be adopted within NLCD 2011 as a new method for the consistent and accurate characterization of the nation's wetlands.