Table 3. --Time trend analyses by step-wise regression, north slope 



Snow characteristic/ 

 plot 



Variable 



value 



Probability 



F 



Peak snow accumulation 

 Forest 



Recovery 



Spring snowmelt rate 

 Forest 



Recovery 



1 -1.3382 



Intercept 

 Winter 



temperature 

 Winter 



precipitation 2 0.3709 

 Time trend 3 -0.0405 



Intercept 47.4722 

 Winter 



temperature -1.3828 

 Winter 



precipitation 0.4521 

 Time trend -0.1512 



39.1890 0.534 



Intercept 

 Spring 



temperature 1 

 Time trend 



Intercept 

 Spring 



temperature 

 Time trend 



-1.7558 



0.0527 

 •0. 0078 



0.717 



-1.2085 0.654 



0. 0353 

 -0.0016 



0.874 



6.20 



3.11 

 0.16 



9.44 



6.60 

 3.19 



0.03 

 0.11 



0. 70-<-rei ected 



0.01 



0. 03 

 0.11 



11.77 0.01 

 0.50 0.50^rejected 



25.27 0.001 

 11.68 0.01 



^ean daily temperature for 5 months, November through March. 

 2 Total precipitation for 5 months, November through March. 



increasing numerical value with years since 1940 harvesting (i.e., 1 94 1 - - 1 , 

 1974--34). 



4 Mean daily temperature for period beginning on day of peak snow accumulation 

 and ending with day of snow disappearance on forest plot. 



For the forest plot, the winter temperature coefficient was negative, indicating a re- 

 duction in peak W.E. associated with warm, thawing winters. The time-trend variable 

 was unimportant, suggesting that change in the structure of the unmanaged forest from 

 young mature to mature over 34 years of record had no detectable effect on peak snow 

 accumulation . 



On the recovery plot, the two climatic variables responded in the same manner 

 (table 3). However, the negative time-trend coefficient was accepted as important 

 (P<=0.11). Apparently the managed, sapling stand was effecting a gradual reduction in 

 peak W.E. unexplained by winter temperature or precipitation input. The gradual decline 

 resulted probably from increases in interception loss associated with the expanding 

 canopies and slightly greater winter release of water from the snowpack. An earlier 

 study conducted near the headquarters station revealed that individually weighed sapling 

 trees (Douglas-fir and western white pine] will lose by evaporation to the atmosphere 4 

 to 5 percent of the total snowfall (Satterlund and Haupt 1970). 



10 



