NEW ENGLAND HURRICANE — BROOKS AND CHAPMAN 241 



Weymouth), Massachusetts Bay, the southeast tip of Maine (East- 

 port), lower New Brunswick (Pennfield Ridge, Blissville), and Sum- 

 merside and Charlottetown, Prince Edward Island, 6 and then across 

 Newfoundland. It finally merged into an extratropical cyclone 

 southeast of Greenland. The "eye" of the storm, with its lull and 

 starlit skies, was seen as the center passed Providence at 12 : 20 a. m. 

 on the 15th ; at South Weymouth at 1 : 15 a. m. (also at Milton, 10 miles 

 from the track) ; at Rockport at 2 a. m. ; and at 4 : 30 a. m. as it skirted 

 Portland, where the wind dropped from 30 to 10 miles an hour. The 

 central area of light winds was diffuse and extensive, reaching more 

 than 30 miles to the left of the path of lowest pressure. 



Details of the storm, as carefully observed and recorded in detail 

 by open-scale automatic instruments at Blue Hill Meteorological 

 Observatory, are presented in figure 2, to indicate its features after 

 some weakening and stretching as it got into middle latitudes. 



FORECASTING THE COURSE 



Predicting the course a tropical hurricane will take if it moves 

 into higher latitudes requires constant attention. The factors are 

 not to be fitted into a simple formula. They are, chiefly, surface and 

 upper air pressures, pressure gradients, and related wind fields ; tem- 

 perature and moisture gradients; and, if over land, the roughness 

 of the terrain. 7 These control the movement of the atmosphere in 

 which the hurricane is carried along and also the storm's supply of 

 energy and f rictional drag ; they thus determine the speed, direction, 

 and force of the storm. 



As early as the 12th the probability that the storm would enter 

 New England was thought to be high. Upper-air observations 

 showed a closed cyclone at 10,000 feet over the Southeastern States; 

 from this, one might have expected the hurricane, after going ashore, 

 to curve back toward the north-northwest or northwest, somewhat 

 as the 1938 storm did. But on the 13th the arrival of colder air in 

 Canada, with its more rapid decrease of pressure upward than in 

 the warmer air, steepened the northward pressure gradient aloft, 

 thereby opening the upper-air cyclone and making the wind at 10,000 

 feet turn southerly in the Middle Atlantic region and southwest to 

 west-southwest over the North Atlantic States. This indicated that 

 the storm would recurve rather sharply and follow a northeast to 

 east-northeast direction, passing, perhaps, just south of New Eng- 

 land. On the morning of the 14th it was evident from the shift of 

 the winds aloft more to the south that the arc of recurvature would 



•G. H. Noyes of the Boston office of the U. S. Weather Bureau, to whom the authors 

 are indebted, has compiled a schedule of the position of lowest pressure. 



7 See Wexler, Raymond, The filling of the New England hurricane of September 1938, Bull. 

 Amer. Meteorol. Soc, vol. 20, pp. 277-281, 1939. 



