A Late November Singularity

Abstract

A Late November Singularity Jerome Namias" Asummary of CLiMATOLOGiCAL data for Washington, D. C, appears each month on the back of the U.S. Weather Bureau's printed Daily Weather Map.t Our interest is frequently aroused by graphs thereon showing the highest temperatures ever recorded on each day, such as the one for November shown in Figure 1. Despite the long period of record used in compiling these data (1871-1960), there appear to be certain periods when the highest temperature changes level abruptly. For example, in Figure 1, abrupt changes occur between November 1 and 6 and between November 21 and 25. These abrupt changes might suggest that the annual course of daily maximum temperatures is composed of step functions. Figure 1 thus raises the following questions with regard to the late November change: 1.Does the suggested sharp decline from November 21-25 also characterize the mean of a long series of mean maximum temperatures , or is this merely a chance fluctuation in an "unstable" statistic (extreme daily values from a sample of 89 years)? 2.Is this phenomenon local or is it associated with large-scale changes in the regional and general circulation, so that singularities at this time might be found elsewhere? Numerous factors to be cited point in the direction of a real singularity at this time of year, and that it is associated with largescale (regional and planetary) features. We shall not deal with the statistical reality of this singularity because of the great difficulty * Director of long-range forecasting in the U. S. Weather Bureau for many years, Mr. Namias is now Chief of the Extended Forecast Division of the National Meteorological Center, USWB-ESSA, Washington, D. C., 20233. t The U. S. Weather Bureau Daily Weather Map referred to here was discontinued in April 1968. 55 56ASSOCIATION OF PACIFIC COAST GEOGRAPHERS of dealing with time series problems, but rather stress logical synoptic aspects. The history of the study of singularities goes far back in meteorology —certainly to 1869 when the British meteorologist Alexander Buchan wrote his classical paper, "Interruptions in the Regular Rise and Fall of Temperature in the Course of the Year."1 Considerable work on this subject was later performed by A. Schmauss during the 1930's.2 In America, where the subject of singularities receives a quizzical eye particularly from statisticians, perhaps the best known singularity is the January Thaw which has been studied in some detail by Wahl3 and later by Duquet* Temperature analysis I have chosen to work with maximum rather than with minimum or mean temperatures because the former are probably more representative of the lower troposphere. This is so because nocturnal cooling of the layer of air lying near the surface increases the vertical stability and with the usually accompanying light winds a limit is not imposed on the cooling by turbulent mixing. On the other hand, the maximum temperature is limited because the instability generated by high surface temperatures frequendy leads to heat transfer aloft. The Washington, D.C., mean unsmoothed temperature maxima during the years 1872-1950 were computed for two roughly 40-year periods, as shown in Figure 2. It is quite apparent that a sharp drop in maximum temperature between November 21 and November 25 has characterized the period during 1911-1950 but not the years 1872-1910. Assuming the earlier records are as reliable as the later ones, one might seriously question the "reality" of the late November 1 Alexander Buchan, "Interruptions in the Regular Rise and Fall of Temperature in the Course of the Year," Journal of the Scottish Meteorological Society, Vol. 2 (Edinburgh, 1869), pp. 4-15. 2 A. Schmauss, "Synoptische Singularitäten," Meteorologische Zeitschiß, Vol. 55 (1938), pp. 385-403. 3E-W. Wahl, "The January Thaw in New England—An Example of A Weather Singularity," Bulletin of the American Meteorological Society, Vol. 33, (1952), pp. 380-386. * R. T. Duquet, "The January Warm Spell and Associated Large-Scale Circulation Changes," Monthly Weather Review, Vol. 91 ( 1963), pp. 47-60. VOLUME 30 YEARBOOK 1968 57 WASHINGTON. D.C RECORD HIGHEST 1871—1960 Q 65 NORMALMAXIMUM 1921 — 1950 NOVEMBER Figure 1. (Upper) Highest temperatures observed at Washington, D...