A Singularity in North Carolina Temperatures

Southeastern Geographer Vol. 26, No. 2, November 1986, pp. 126- 134 A SINGULARITY IN NORTH CAROLINA TEMPERATURES Peter J. Robinson and Lori Peterkin INTRODUCTION. Meteorological singularities, the regular occurrence of specific types of weather on or around specific dates, have been noted for various parts of the world. (J) In the eastern United States several temperature singularities have been identified. The best known are the mid-October "Indian Summer" and the "January Thaw", although additional anomalies regularly occur in both early and late November . (2) Recent studies of the temporal variability of temperature have emphasized long-term trends associated with "climatic change" and have not considered whether such persistent short-term anomalies continue to exist. (3) In North Carolina, folklore supports the existence of a warm spell in January, while the term "Indian Summer" is loosely used for any warm period in late fall, whether preceded by a cold spell or not. This paper investigates the current climatological basis for these perceptions and assesses some potential causal mechanisms for the one singularity which appears to be persistent and widespread in the state. IDENTIFICATION OF SINGULARITIES. A singularity is a feature which must have wide areal extent, persist for several days, and recur around the same date in numerous years. (4) Hence long-period records for a group of stations are required for their identification. For North Carolina, stations with continuous records and no site changes for the 1948—1978 period were used. Since the State is conventionally divided into three physiographic provinces, which are largely reflected in the climate, two stations from each province were chosen. The stations provide a rough transect across the state (Fig. 1). To ensure that anomalies which persisted over several days were identified, weekly average temperatures were used, being calculated from the records of daily maximum and minimum values. (5) The initial identification of potential singularities was undertaken using the 30-year average weekly temperature for each station (Fig. 2). Dr. Robinson is Associate Professor of Geography at the University of North Carolina in Chapel Hill, NC 27514. Mrs. Peterkin is an environmental consultant in Albuquerque, NM. Vol. XXVI, No. 2 127 Mt Airy • Asheviile MOUNTANS /•Hickory PIEDMONT • Nashville New Holland COASTAL PLAIN Fig. 1. Distribution of North Carolina temperature stations used in the study. Although the general shape of the curve is smooth, as expected from climatological considerations when using long term average values, several periods and stations display deviations which represent potential singularities. Their significance was tested using the method of Wahl. (6) This method incorporates the temporal autocorrelation of temperatures by assuming that temperatures change between weeks in a linear manner. Thus the expected value for a given week i is the mean of the values for weeks i — 1 and i + 1. This expected value was then compared to the actual value using a Student's t test. After a further test for skewness in the data, those weeks with values significant at the 0.10 level were retained. Thereafter, the time of the anomalies at all stations were compared. Those years when at least 5 stations showed anomalies at the same time were considered to be singularity years. The final step was to determine the number of such singularity years. The results of this analysis revealed several anomalies for individual stations, such as that in late December at New Holland, but only three periods when the effect was sufficiently widespread to represent a potential singularity (Table 1). Of these three, only one was frequent in occurrence and widespread in impact (Table 1 and Fig. 2). Although the January anomaly suggests a "January Thaw", it occurs throughout the state in only approximately a quarter of the years. It is characterized by a few years with marked anomalously high temperatures statewide, several years with warm conditions in some areas, and some years with no places in the state showing an anomaly. Similarly, although the first winter cold snap coincides approximately with the late November sin- 128 Southeastern Geographer 0) L D -P 0 L 0) a e o 70 60 50 40 30 20 NEU HOLLAND HIGH POINT ASHEVILLE NASHVILLE MT. AIRY HICKORY H—KH—I—I—h-H- H—I—li...