subzero air temperature were at A, coincident
ground. The Connecticut River Basin consistently
with the greatest mean morning temperature in
produced inversion structure in the lowest 100 m
the data set, and the greatest GSD of temperature
of air when more than 30 cm of snow cover was
difference with respect to the reference tempera-
present, in agreement with Magono's findings
ture at river level.
over deep snow in the Moshiri Basin.
The reference temperature at P systematically
diminished as snow thickness increased and snow
cover became more uniform. The diminution of
CONCLUSIONS OF PART III
temperature was evidenced by a diminished mean
The number of subzero (<17.8C) mornings
morning temperature, a diminished median tem-
perature, and more frequent subzero morning
varied by one-third, among observation points
temperature coincident with increasing snow
separated by a distance of less than 6 km, in our
cover.
158-day experiment. The number of mornings less
than 25C had even greater variation. This dif-
The morning air temperature observed in a
hamlet could not be systematically differentiated
ference is attributed to earlier formation and later
from those measured on nearby flats of similar
persistence of near surface inversions over small
elevation. Temperatures observed along a nar-
basins, and more persistent exchange of warmer
row creek valley, and those observed where side
air from aloft to slopes and ridges, in agreement
streams entered the river basin, were insignifi-
with the theory and observations of Magono et al.
cantly different from adjacent temperatures at
(1982) and Maki et al. (1986). We were not able to
similar elevation on mornings with inversion. Cold
isolate significant temperature difference attrib-
air drains were observed on some mornings when
utable to hamlets or vegetation in this winter
lapse occurred concurrently with sparse snow
study.
cover, and cold air drains occasionally produced
Our hypothesis that a river plane provides a
asystematic lesser temperatures.
temperature reference in winter, and that a mov-
Hill and ridge tops, adjacent slopes, and small
ing probe senses multiple surface temperatures
basins along slopes consistently exhibited the
that can be used to synthesize the vertical tem-
greatest morning air temperature differences stud-
perature structure in hilly terrain, has provided
ied. Greater snow depth, and its coincident greater
consistent and useful results in this case. This
uniformity of cover, increased these differences.
hypothesis seems worthy of additional examina-
The experiments and models of Magono et al.
tion in other terrain. The combination of moving
(1982) and Maki et al. (1986), in conjunction with
probe observation, with frequency analysis of tem-
the observations in a different geographic setting
perature difference, provides the geometric stan-
with more variable snow cover in this work, may
dard deviation of local temperature difference.
aid interpretation and extrapolation of meteoro-
The magnitude of the GSD appears related to air
logical records over snow-covered, data-sparse,
exchange over snow-covered ground.
and complex topography. The observations and
We propose that the principles of constructing
analyses of Parts I, II, and III agree with the
vertical temperature structure using a moving
Hokkaido model theory, showing slopes to be
probe, and the analysis of the frequency distribu-
relatively warmer, and basins relatively cooler on
tion of temperature differences as an estimate of
snow-covered winter mornings. The dynamic
local exchange, can be applied to prepare local
moving probe method of temperature observa-
dispersion estimates in data-sparse operational
tion showed slope, basin size, and snow cover to
settings.
be the dominant features determining winter
We showed through analysis of the frequency
morning air temperature in the vicinity of the
distribution and mean value of temperature dif-
Connecticut River. Temperature differences asso-
ferences in Parts I and II that a relatively repeat-
ciated with hamlets, barriers, and surface vegeta-
able pattern of temperature difference, with
tion were minor when compared. Small basins
respect to elevation and snow cover, is present
influenced local temperature and temperature
above the river surface and east of the Connecti-
structure even when strong cold advection oc-
cut River on winter mornings. Collections of ob-
curred over sparse snow cover. The larger basin
servations, obtained at more than one observation
point at the same (5-m) elevation, were used in
along the Connecticut River did not frequently
produce inversion structure in the lowest 100 m
preparation of these figures. Several of these col-
of air when less than 20 cm of snow covered the
lections grouped points from fields and woods,
35
Previous Page
