gusty winds which persisted for several hours. 

 These winds were ultimately dependent upon 

 the exceptional summertime pressure gradient 

 that existed in the lower levels of the tropo- 

 sphere. The extent to which winds of gradient 

 strength or greater were experienced near the 

 surface would, of course, vary, as previously 

 discussed. Turbulence commonly attending a 

 cold front may or may not have been an im- 

 portant triggering factor in this case. 



The strong pressure gradient and the cold 

 front, both associated with the approaching 

 trough, may be regarded as individual but 

 related parts of a weather system. Basic to 

 such a system is the juxtaposition of two dif- 

 ferent (warm and cool) atrmasses. The upper 

 tropospheric jet stream, whose core in this 

 case was at only 25,000 to 30,000 ft. m.s.l., 

 would also be part of this system. Moving 

 eastward, its axis at 2300 P.d.t. September 1 

 was about 100 miles northwest of Spokane 

 and the Sundance area. The windspeed over 

 Spokane was approximately 80 knots (inter- 

 polated), having increased from 60 knots at 

 1700 P.d.t. Such a jet could, through dynamic 

 processes, possibly effect a vertical transfer of 

 momentum (and other properties) to the 

 lower troposphere. This would occur within a 



frontal zone sloping below the jet's core, in the 

 form of a dry-adiabatic sinking of air from the 

 upper tropospheric portion of the zone (lo- 

 cated many miles upwind of its lower portion) 

 (6). Some evidence was found supporting the 

 presence of such an effect in the Sundance 

 case, but was regarded as inconclusive. 



The estimated hourly free-atmosphere (gra- 

 dient) windspeeds at 5,000 ft. m.s.l. in the 

 Sundance area, used in this fire report, are 

 given in figure 26. These speeds are based 

 upon the 6-hourly upperwind observations at 

 Spokane. Observed surface winds, referred to 

 earlier, were used as a guide in allowing for a 

 slight timelag in adapting a vertical cross- 

 section analysis of Spokane upper winds (not 

 shown) to the Sundance area. The lag, ad- 

 judged also according to spatial differences in 

 pressure gradient on the 700-mb. maps, did 

 not exceed 1 hour and was adjudged to be 

 zero after the time of maximum wind. 



Humidity 



Also plotted in figure 26 is a curve of esti- 

 mated "modified" free-atmosphere relative 

 humidity at 5,000 ft.. These humidity values 

 are a compromise between those in the free 

 atmosphere and those observed near ground 



I I I I I '''''' 



20 16 12 08 04 24 20 16 12 08 04 P d.t. 



SEPT. 2 SEPT. 1 



Figure 26. - Time graph of estimated free-atmosphere windspeed and "modified" free-atmosphere relative 

 humidity at 5,000 ft. m.s.l., Sundance Fire area, September 1-2, 1967. 



31 



