THE SHIFT OF THE STORM TRACK. 



By Charles J. Kullmer. 



Our only maps of the storm frequency of the whole of the northern hemisphere (we 

 have none of the southern hemisphere) are those of H. H. Dunwoody, published in 1893, 

 and covering the 10-year period from 1878 to 1887. The storm frequency is given in 

 separate maps for each month, and these are combined into a year map. Dunwoody 

 divides the map of the northern hemisphere into squares measuring 5° on a side and records 

 in each square the number of barometric depressions whose centers crossed it during the 

 period under discussion. For instance, the centers of 306 depressions passed through the 

 Lake Michigan square, or about 30 barometric depressions a year. Practically the storm 

 frequency was much greater than this, since 419 storms passed through the Lake Superior 

 square on the north and 177 storms in the square just south; but this need not concern us, 

 for we have here to do only with the tracks of the centers of depressions. In view of the 

 importance of storms in many phases of life, it has seemed worth while to reconstruct 

 the storm frequency maps for the United States for the latest available decade, 1899-1908, 

 in order to determine whether in the interval of 21 years any general shift of the storm track 

 has taken place. The material for this is available in the plotted tracks of barometric 

 depressions given in the monthly summaries of the "Monthly Weather Review." 



Dunwoody does not give the individual tracks, nor does he describe in detail the methods 

 used in constructing his charts, so that there is some doubt as to the extent to which his 

 material may be considered identical with that used in the present investigation. A com- 

 parison of the maps and a discussion of the results, however, seems to show that the material 

 is in general the same and that reliable comparisons are possible. 



The monthly maps of the United States for the new series of observations as compared 

 with the old are given in figures 52 to 63. In these the figures represent the number of storm 

 centers for each 5° square, those of Dunwoody for 1878-1887 being placed above, and those 

 of the later period, 1899-1908, below. The curved lines represent the number of storm 

 centers passing tkrough a given square each year, the earlier conditions being indicated by 

 dotted lines and the later by solid hues. In the first map, that for January, figure 52, 

 we at once find five important phenomena, which occur also on later maps and on the 

 year map; first, an increase in storm frequency over the southwestern States; second, a 

 decrease in the west in latitude 40° to 45°; third, a great increase in western Canada; fourth, 

 a general southerly movement of the lines of equal storm frequency ; and fifth (and perhaps 

 most interesting), the phenomenon of a double maximum in latitude 45° to 50° as com- 

 pared with the single maximum of the earlier period. The great increase in western 

 Canada is of such an amount as to suggest a difference in the observational material, but 

 this is not necessarily the case, as will appear later. 



The February map, figure 53, shows all the phenomena of that for January. The 

 southwestern increase in storm frequency is especially marked, the number of centers 

 passing through a single square having in one case increased from 9 to 32. Just to the 

 northeast of this square the western decrease in latitude 40° to 45° is also quite pronounced. 

 The increase in western Canada and the southerly shift of the lines as a whole are nearly 

 as marked as in the January map, while the double maximum is less distinct, although it 

 clearly exists. March (figure 54) shows most prominently the southwestern increase, 

 while the general southerly shift is visible, and the double maximum still appears, although 



14 193 



