The 1986 Growing Season Drought in the Southeastern United States: Spatial and Temporal Patterns

Southeastern Geographer Vol. 28, No. 2, November 1988, pp. 80-96 THE 1986 GROWING SEASON DROUGHT IN THE SOUTHEASTERN UNITED STATES: SPATIAL AND TEMPORAL PATTERNS* Peter T. Soulé INTRODUCTION. Drought is a recurring element of climate that can have devastating impacts on natural ecosystems and human activities over large regions of the earth. In 1986 drought conditions for most of the Southeastern United States were considered to be the worst in the last century. (I) The drought began during the winter when long-wave troughs in the upper-level westerlies failed to follow their normal path through the Gulf Coast states. As a result, fewer cyclonic storms developed in the Southeast, and cumulative yearly rainfall amounts began to fall below normal. The synoptic situation worsened during the late spring and summer when an upper-level trough developed off the South Atlantic coast. It drew cyclonic storms out over the Atlantic and displaced the Bermuda High well to the west of its normal position. (2) The persistent dry, descending winds of the Bermuda High worsened the severe drought conditions and brought record low rainfall coupled with a record heat wave. For much of the Southeast, severe drought conditions continued until early autumn. The negative effects of long droughts are experienced by water users at different times. For example, soil-moisture deficiencies in 1986 were devastating farmers long before declining reservoir levels forced many municipalities to impose restrictions on water use. This illustrates one of the main dilemmas in drought research: how to define and monitor drought conditions. Within the climatological literature, it is recognized that drought has many different forms (e.g., agricultural, meteorological, hydrological). (3) Although each of these forms (or drought subtypes) can exhibit highly variable temporal and spatial patterns, analyses of more than one drought subtype are not generally performed. This lack of integration is a major gap in drought research. *Funds for the preparation of graphics were provided by the Department of Geography, University of Georgia. This support is gratefully acknowledged. Mr. Soulé is a Ph.D. student in Geography at the University of Georgia in Athens, GA 30602. Vol. XXVIII, No. 2 81 This study uses both agricultural and meteorological drought severity indices to examine the spatial and temporal patterns of drought during the 1986 growing season (late April to early November) in the Southeastern United States. Its primary objective is to identify spatially homogeneous drought regions as defined by the two severity indices and to determine whether these regional patterns are consistent across the different drought types. Secondary objectives are 1) to compare temporal patterns of drought development during the study period among the identified drought regions and 2) to demonstrate how maps of cumulative or average drought severity may be misleading when describing regional drought patterns. DATA. Data for this study consist of biweekly values of the Palmer Drought Severity Index (PDSI) and weekly values of the Crop Moisture Index (CMI) for 69 climatic divisions within the Southeastern United States between 26 April and 1 November, 1986 (Fig. 1). (4) The data were extracted from a series ofmaps featured in the Weekly Weather and Crop Bulletin. Although the National Weather Service (NWS) calculates both indices on a weekly basis during the growing season, maps of PDSI values usually are published only at biweekly intervals. Palmer defined drought as a period of moisture deficiency which deviates from climatic normals for the area under investigation. (5) Since the negative impacts of a drought may be experienced at different times by water users, it is appropriate to use multiple measures of drought. The two indices used in this study are derived from modulated soilmoisture budgets, based upon principles of supply and demand for moisture. They take into account both current and historical (climatic) values for water budget components (e.g., actual and potential évapotranspiration , soil-water recharge, runoff, etc.). The indices generally range between +7 and —7, with negative values indicating drought conditions (Tables 1 and 2). The PDSI is usually considered a measure of meteorological or general drought intensity, although when calculated in real time (as the NWS does during the growing season) it may more accurately depict conditions of hydrological drought severity. (6) Palmer designed...