1. Introduction

Since the arrival of colonial settlers, anthropogenic activities have transformed the landscape of the Florida peninsula. Much of the transformation within south and central Florida occurred during the twentieth century, when the pace of urbanization over coastal areas and agricultural production over the interior substantially increased. During this period, natural surface hydrologic features were altered and diverted for irrigation and domestic use, as well as for flood prevention, particularly in the Everglades and the Kissimmee River basin. Prior to the construction of canals and levees, much of the surface runoff of the peninsula drained into the wide, perpetually inundated floodplain of the Kissimmee River, which in turn drained southward into Lake Okeechobee. Natural flow over the south rim of Lake Okeechobee spilled into the wide expanse of the Everglades. The surface hydrologic budget of the Everglades was dominated by evaporation and transpiration, with surface runoff into Florida Bay composing a relatively small percentage of the annual rainfall (Kushlan 1990). Following the major water diversion projects of the early and middle twentieth century, much of the Kissimmee River basin had been drained, and the south shore of Lake Okeechobee had been dammed.

Pre-1900s	1993
Fig. 1. USGS land-cover data for (left) pre-1900 natural land cover and (right) 1993 land use. [click on images above to view larger version]

A comparison of maps that represent pre-1900 natural land cover and 1993 land-use patterns (Fig. 1) clearly demonstrates the significant changes that have occurred throughout Florida over the twentieth century. Within the Everglades, agriculture and extensive saw grass have displaced the natural saw grass plains and the freshwater marsh/slough areas. The area of freshwater marshes around and to the north of Lake Okeechobee (in the Kissimmee River basin) has been drained and converted to mixed agriculture and pasture. The current land-use patterns in the interior, central portion of the Florida peninsula, including extensive mixed agriculture, cities, roads, residential areas, and urban complexes, have collectively supplanted much of the predominantly pine forest areas of the natural landscape. The question of interest for this paper is whether these land-cover changes have affected the warm season climate of the region.

Several previous studies have addressed the impacts of land-cover change on warm season climate over spatial scales similar to the Florida peninsula, with a particular focus on surface-forced circulations and their relationship to the spatial distribution of cumulus convective rainfall (Anthes 1984; Wetzel 1990; Segal and Arritt 1992; Dalu and Pielke 1993; Pielke et al. 1999; Pielke 2001; Weaver and Avissar 2001). Over the Florida peninsula, the dominant mode of convection during the warm season is associated with the sea-breeze fronts (Byers and Rodebush 1948; Pielke 1974; Michaels et al. 1987; Simpson 1994). Because the sea breezes are driven primarily by contrasting thermal properties between the land and adjacent ocean, it is possible that alterations in the nature of the land cover of the peninsula have had impacts on the physical characteristics of these circulations (Pielke and Cotton 1977; Gannon and Warner 1990; Baker et al. 2001). This mechanism could have important implications for the distribution of sea-breeze convective rainfall. Furthermore, it is possible that changes in land cover may have impacted other (inland) mesoscale circulation features (and related convective rainfall), as well as the diurnal cycle of surface thermodynamics (e.g., surface energy fluxes and shelter-level temperature).

In this paper, the previous study by Pielke et al. (1999) is revisited and significantly expanded upon by presenting additional simulations of the warm season weather of the Florida peninsula. These simulations were performed using the Regional Atmospheric Modeling System (RAMS; Pielke et al. 1992). In the previous study, the authors presented three simulation spanning the period July-August 1973 wherein the model configuration was identical, with the exception of the definition of land-cover. One each of the three simulations employed a land-cover database valid for 1900, 1973, or 1993. However, those simulations were limited in scope, primarily because they were performed only for a single warm season period. Furthermore, the impacts of land-cover change were not examined in the context of other model factors that are known to influence strongly the numerical simulation of warm season convective rainfall and near-surface sensible weather. In this current work, additional simulations were performed for three separate warm season periods, and the results were subjected to a number of model sensitivity factors. Furthermore, new land-cover datasets were constructed for use in the new simulations presented in this paper. These datasets are highly detailed and considerably more realistic than those used by Pielke et al. (1999).

Go back to Home | Go ahead to Land-Use Data