Notes concerning ATLSS Model Runs for NSM Comparisons
Louis J. Gross and Jane Comiskey
The Institute for Environmental Modeling
University of Tennessee, Knoxville
Knoxville, Tennessee 37996-1610
(Copyright University of Tennessee - 2003)

As part of the Comprehensive Everglades Restoration Plans (CERP), a series of future water management regimes have been proposed and simulated in the Everglades and Big Cypress landscapes using the South Florida Water Management Model (SFWMM). The Natural Systems Model (NSM) has been used to simulate pre-management hydrologic conditions over the same model area. We have used outputs from these hydrologic models to perform analyses using ATLSS Spatially-Explicit Species Index (SESI) models, which provide yearly analyses of the effects of hydrology on the availability of breeding and/or foraging habitat for the species modeled across the model area. We here discuss limitations to the conclusions drawn from these simulations, relative to conclusions drawn from ATLSS comparisons of other CERP scenarios.

The District hydrology models operate at the 2-mile scale of resolution and produce water depth outputs at that scale. In previous ATLSS comparative evaluations of hydrologic scenarios performed as part of the Restudy, water has been redistributed to 500-m x 500-m pixels using the ATLSS computer-generated high resolution topography map layer. Generation of this map layer requires an accurate high-resolution representation of vegetation for the relevant time period. For evaluation of CERP SFWMM scenarios, ATLSS uses the representation provided by the Florida Gap Analysis Vegetation Map (FGAP) version 2.1, which is based on classification of 1992/1993 Landsat TM imagery. This is combined with USGS HAED and Lidar data to produce a high resolution elevation map. No vegetation map comparable to FGAP is available for the historical period simulated by the NSM Hydrologic Model. For the NSM comparisons, therefore, we do not redistribute water depths over a simulated topography to achieve a higher scale of resolution. For purposes of identification of habitat types, we utilize the FGAP vegetation map for all scenarios.

Drainage and impoundment, and the subsequent development of urban areas and intensive agriculture, have had profound impacts on Everglades biota. Spatial extent of the Everglades has been reduced by half, and the remaining areas have been affected by major alterations in hydropattern. Analysis of hydroperiod maps reveals the substantially larger extent of inundation under NSM, extending northward over former sawgrass prairies now drained and subsumed in the Everglades Agriculture Area, and eastward over natural areas now converted to urban centers. Cycles of drought and flood have been exacerbated, and timing and predictability of water flows have been affected as well. Changes in soil, vegetation, and water quality, as well as hydropattern alterations, have impacted breeding and foraging potentials and population distributions. Invasions of exotic species, changes in water temperature, dissolved oxygen, and nutrient loading, increased human activity on the landscape, and a multitude of other changes have accompanied water management.

The Natural Systems Model simulates hydropatterns before canals, levees, dikes and pumps reshaped the landscape. The NSM reflects frequency, duration, depth, and spatial extent of water inundation under the natural system. To the extent that ATLSS index models are hydrologically driven, we can compare NSM biotic impacts with those of SFWMM scenarios within areas that remain in natural vegetation under the SFWMM. However, each index model also has a habitat suitability component that excludes pixels identified as habitat types that are unsuitable for breeding of the modeled species (e.g., urban and agricultural types).

Pixels that have been excluded based on their classification in the FGAP vegetation map from SFWMM hydrology simulations must also be excluded from NSM simulations. Although some of these areas would in fact have provided suitable breeding habitat during the historical period of the NSM, we have no way to assign historically correct vegetation types without a vegetation map contemporaneous with NSM hydrology. For this reason, we cannot capture the biological effects of loss of spatial extent of breeding habitat under the managed system.

Therefore, in addition to other sources of model uncertainty for ATLSS Model outputs for the Initial CERP Update, the NSM evaluations include limitations imposed by changes in habitat type between the NSM modeling time period and that of the SFWMM.