DISCUSSION - CR97

Table 10. Number of observations of lesser temperatures.

ered in any way as a climatology. The

distances tabulated are relative to W,

Distance (km) <17.8C <20C <25C <30C <35C

Location

the point with most frequent subzero

days; note that the A is only 0.6 km

from c in basin I. There are two lapse

days when air of less than 17.8C

was uniquely observed at the apex

included in the tabulation. A smaller

set of observations along the north

south path including 24 of the days

<17.8C was observed at P and j. Sub-

zero air was rather uniformly ob-

served along the path on 21 of these

days, but no air temperature less than

25C was observed south of k during

19901993. There was one day that air

of 18C was observed along the river

Table 11. Frequency of winter morning temperatures at

south of u, but no air of less than 17.8C

Piermont bridge (P).

observed elsewhere.

<3 cm

310 cm

1020 cm

>20 cm

We have shown in Part II that fre-

quency of lapse and inversion structure

Number of observations

43

31

34

45

systematically varied with the depth

Number less than 17.8C

6

3

8

14

and continuity of snow cover. We also

showed that the frequency of greater

Number less than 20.0C

2

1

5

14

local temperature differences, and the

Number less than 25.0C

mean value of local temperature differ-

0

5

ences, increased as snow cover increased

Number less than 30.0C

0

1

in depth and uniformity. All of these

Median temperature, C

systematic differences were temperature

7.6

8.0

11.4

11.8

difference with respect to the tempera-

ture measured at P, the northernmost

river level measuring point along the observation

DISCUSSION

path. The temperatures observed at P are strati-

fied with respect to snow cover category in Table

The introduction to Part I showed that climatic

11. The entire set of temperatures observed at P

time scale differences of a factor of two in the

annual number of subzero (<17.8C) days occurs

are near normally distributed, but stratified seg-

ments depart from normality. We have used the

over distances of 30 km in northern New York

10- to 20-cm class described in Part 1, and com-

and New England. Part I of this work showed

bined the 10- to 20-cm class with the >30-cm class

that temperature differences along a 30- to 50-km

to provide four snow cover categories with ap-

homogeneous plane coincident with the bank of

the Connecticut River were on the order of 1C or

proximately equal numbers of observations. The

table indicates that during this series of observa-

less on a majority of 158 observing days over

tions lesser temperatures, and more frequent sub-

three winters, including 40 subzero days. Quasi-

zero (<17.8C) temperatures, accompanied

simultaneously, temperature differences of as

much as 18C were observed at points within 6

thicker and more uniform snow cover at P.

Considering these combined arguments indi-

km of the river. The number of mornings with

cates that P more often experiences lesser air tem-

subzero temperatures differed by one-third at two

peratures, with respect to its surroundings, as

points separated by less than 0.6 km in distance

snow cover thickens and becomes more uniform.

and 60 m in elevation.

This lesser temperature is associated with more

The most frequent occurrences of subzero

frequent formation of inversion structure along

morning temperature are at W, which is also the

the plane intersecting the river bank as snow thick-

point with the least mean and median morning

ness increases.

air temperature. The least frequent occurrences of

34