52 S. L. Dingman et al. 



E - net evapotranspiration 



R - runoff 



/ = net infiltration 



AS = change in surface storage of water in the area considered. 



As with the energy balance, it is assumed that surface storage is 

 neither increasing nor decreasing, so that the average annual value of A5 

 is zero. In addition all the water that infiltrates eventually ends up as run- 

 off or evapotranspiration, so that 1 = for long-term average condi- 

 tions. Precipitation and runoff are estimated using standard techniques. 

 For average annual data, evapotranspiration is then found by subtrac- 

 tion. For shorter periods, information on evapotranspiration is based on 

 measurements of evaporation from pans or small ponds, on the energy 

 calculated to be available for evaporation via the energy balance equa- 

 tion, or in a few cases on observations of soil moisture (which may also 

 be used to estimate I). Data on short-term changes in storage are in some 

 cases available in the form of records of the changes in elevation of tun- 

 dra ponds (Hobbie 1980). 



The average annual precipitation recorded (1941-70) at the Barrow 

 National Weather Service Station is 124.1 mm (Table 2-1), but it is 

 known that this recorded amount is less than the true amount. Compari- 

 sons of the water equivalent of the tundra snowpack (Black 1954, Benson 

 1969) show that actual snowfall exceeds the amounts recorded at the Bar- 

 row gage because of the effects of wind on the gage catches. A summary 

 of data for six winters (Table 2-5) suggests that, on the average, the true 

 value is 1 .6 times the recorded value. Although data have indicated that 

 summer precipitation measurements do not need to be adjusted for wind 

 effect on gage catch. Brown et al. (1968) found that a correction should 

 be made for the effects of traces. Traces, recorded when precipitation is 

 less than 0.13 mm in a measurement period, have been summed as zero 

 values (Table 2-1). Brown et al. (1968) found that the measured summer 

 precipitation should be muhiplied by 1.1 to give the actual value. 



Thus, the average annual precipitation value of 170 mm is estimated 

 by multiplying the National Weather Service data by 1.1 for the months 

 June through August and by 1 .6 for the other nine months, when precipi- 

 tation is assumed to be in the form of snow (Table 2-1). The estimate of 

 total precipitation is consistent with most detailed studies of the region's 

 precipitation and water balance, including those of Black (1954), Mather 

 and Thornthwaite (1958) and Brown et al. (1968). Corrected values 

 (Table 2-1) show a precipitation maximum in August and a secondary 

 maximum in January. On the average, about 63% (106 mm) of the an- 

 nual precipitation falls as snow (September through May) and 37% (64 

 mm) as rain (June through August) at Barrow. At Barter Island, an aver- 

 age of 68% (167 mm) falls as snow and 32% (80 mm) falls as rain. The 



