Figure 9. - Artist's conception of convection column, then about 31,000 ft. high, made from slide taken by 

 jet pilot over fire near 2000 hours. 



prevailing wind and falling under the gravita- 

 tional influence (8)' ; 



2. By being carried aloft in the convection 

 column, then entering the influence of a vor- 

 tex street and being stabilized in it. The vor- 

 tex street moved on downstream of the strong 

 convection column (1). (For a more detailed 

 discussion of this point, see the section below 

 titled "Firebrand Activity.") 



As the man on Roman Nose Lookout re- 

 treated to a rockslide, where he stayed until 

 the fire had abated the following day, a 

 crowning fire moved up Zuni Creek out of 

 Pack River and carried over into Ruby Creek. 



2200. By this time the fire had moved 

 from Ruby Ridge into Ruby Creek and addi- 

 tional spotting occurred on Caribou Ridge 

 (see foldout map). The fire front had become 

 broken, probably for several reasons, includ- 

 ing the facts that more area had been logged 

 over a long period of time, so little slash was 

 present, and that the spot fires had not con- 

 solidated into a single front. Hence there was 

 poor fuel continuity and the main fire front 

 was deprived of fuel. Sporadic flareups were 

 occurring, indicating the meeting of heading 

 and backing fires or burnout of pockets of 



^Numbers in parentheses refer to Literature Cited. 



fuel. At this time the fire was primarily a 

 brush fire, with considerable crowning. 



2300 on into September 2, 1967. The fire 

 became more affected by prior burnout and 

 became a ground fire as it moved into Caribou 

 Creek. The rate of spread had decreased al- 

 though some crowning took place. The main 

 front had become poorly defined because of 

 reduced fuel continuity. Nearly 60 percent of 

 the area could now be considered affected by 

 previous burning. 



A spot fire northeast of the main fire zone 

 was reported at 2345. This probably resulted 

 from the spot fire at the mouth of Snow 

 Creek. Various portions of the fire became 

 individual fires that resulted in flank increases 

 and an occasional minor run. This type of be- 

 havior continued until the fire was controlled. 



Weather conditions continued to improve 

 during the night (see table 2, p. 16). Observa- 

 tions made in Ruby, Highland, and Falls Creeks 

 at 0200 indicated the fire was cooling down. 



One of the greatest concerns at this time was 

 the possibility of a wind shift to the northwest 

 or southeast. A fire front 20 miles wide would 

 have resulted; however, the wind held steady 

 and the acreage increased only slightly, 4 per- 

 cent, before control was achieved. 



9 



