Landscape modeling of coastal habitat change in the Mississippi delta

A landscape model was developed to investigate and predict the environmenta l factors affecting wetland habitat change within the Barataria and Terrebo nne basins of coastal Louisiana, USA. The model linked an overland-flooding hydrodynamic module, using cells of 100 km(2) in size and operating at a 1 -h time step, and a spatially articulated ecosystem module, resolving habit at type and change for 1-km(2) cells in daily time steps. Integration acros s different temporal and spatial scales was accomplished with interpolation routines and averaging algorithms. Forcing functions included dominant reg ional processes, such as subsidence, sedimentation, and sea-level rise. Hyd rologic functions were calibrated against existing climate and hydrologic t ime series, while habitat information was compared to maps prepared by the United States Fish and Wildlife Service (USFWS) for 1978 and 1988. Spatial calibration was done by initializing the landscape pattern of the m odel to a 1978 USFWS habitat map. After a 10-yr simulation, the results wer e compared against a 1988 USFWS habitat map. Simulated maps had an accuracy of 85-90 (out of a maximum of 100), based on a multiple resolution fit alg orithm. For validation the model was initialized with a 1956 USFWS habitat map, and the results from a 32-yr simulation were compared to the 1988 USFW S habitat map. The landscape model produced reasonable regional agreement, despite the fact that small-scale processes and features were not included. The validation runs produced land-loss rates that matched historical trend s with an accuracy fit above 75. The model simulated 30 years into the future, starting in 1988, testing for long-term climate variability under diverse scenarios. Results indicated t hat weather variability impacts land-loss rates more than replication of ex treme weather years. Even when extreme dry and wet years were repeated, the model predicted lower land loss when compared to historical records. This is indicative of the ability of the simulated plant communities to adapt to repetitive climatic forcing functions.