Surface Climate and SnowWeather Relationships of the Kuparuk Basin
increasing at all locations, March record low temperatures do not differ much
from those in February, the result of late cold outbreaks as the global atmosphere
mixes its large reservoir of "aged" polar air equatorward. Weather events at this
time are dominated by the equatorward plunges of these cold and dry polar air
masses, so the months of March and April are typically relatively cloud free
across much of northern Alaska, with precipitating storms at a minimum.
Following the DC trend, higher-elevation IMN is consistently the warm spot
and also tends to have the largest diurnal variation. The northernmost sites, BET
and FRA, remain the coolest sites, as they continue to be influenced by the shal-
low but frigid polar air masses that persistently dominate the coastal region. As
LC progresses and the length of day increases, the diurnal range in temperatures
decreases across the network. IMN consistently achieves maximum temperatures
above freezing by late April, almost a month ahead of the other sites, and not
surprisingly the initiation of snowmelt generally occurs there first as well. Well
before breakup the increasing insolation creates small regions of bare ground in
wind-exposed areas, such as ridge crests, that had thin seasonal snow cover.
Generally, however, it is not until minimum temperatures consistently remain
above freezing that the breakup of the snowpack begins in earnest (Kane et al.
1997), usually occurring at any given location in about a two-week period. The
running average procedure used in Figure 5 masks this rather episodic behavior
that occurs first at the most inland sites during mid-May to early June and pro-
gressively later as one moves toward the coast. After breakup has occurred, inter-
site temperature differences along the network are reduced.
In general, winds are a much more difficult variable than temperature to
measure and quantify. This is especially true in the Arctic cold season, where
physical conditions (e.g. riming and icing) adversely impact the mechanical wind
instrumentation. Hence, wind data from these sites tend to be less robust than the
temperature record. Additional factors, such as curious bears and other wildlife,
create problems in the warm season. Since the sites used in this study are
unmanned for all but a few brief visits per year, instrument performance can be
degraded for long periods of time. Fortunately winds were measured at several
heights on each tower, providing some redundancy in measurement. Whenever
possible, wind measurements at 3 m were used, with a correction factor based on
a logarithmic profile assumption for each site being applied to the wind speed as
necessary if other heights (usually 10 m) were used to replace missing 3-m val-
ues. In practice the difference in wind speeds was usually less than 10% with
height for wind speeds greater than the 1-m s1 threshold.