e. 
f. 
=. 
30 June 1972 (S view). 
715 June 1972 (S view). 
37 
measured by a Moll-Gorezynski pyranometer. In 
principle, the determination of the net radiation 
by measurements of its four basic components is 
superior to the integrated measurements using a 
single all-wave net radiometer. This is par- 
ticularly so if the aim is to describe the radiation 
fluxes in terms of cloud, surface type, tempera- 
ture, or other meteorological parameters. 
Unfortunately, the Eppley precision long- 
wave pyrgeometers, which had become available 
only relatively recently, proved to be subject to 
large errors. In spite of extensive calibrations, 
both in the field by comparison with a Barnes 
thermal radiometer and in the laboratory with a 
black body device, we could not obtain consist- 
ent results. The errors are partly due to long- 
wave emission by the KRS-5 dome, supposedly 
semitransparent to long-wave radiation and hav- 
ing a low emissivity (Eppley Laboratory, 1972). 
After some time in use, it appears that the 
domes acquire an emissivity that is far from 
negligible. Furthermore, probably because of 
convection effects induced by considerable solar 
heating of the dark colored domes, the calibra- 
tion factors are not the same when the instru- 
ments point upward and downward. Although it 
is reported that the Eppley pyrgeometers may 
be used with some success if, among other 
things, the temperature of the KRS-5 dome is 
monitored, our experience with them shows that 
they are unsuitable as field instruments in their 
present design. We are not using the results of 
the pyrgeometer measurements in the present 
discussion. 
The Eppley short-wave pyranometers used in 
the present investigation were compared a few 
times in the field for internal consistency. 
Before and after the field season in 1972 and 
1973, two of the pyranometers used were cali- 
brated against a Linke-Feussner actinometer. 
The results of these calibrations gave factors 
within 1-2% of those recommended by the 
manufacturer. 
For the summer temperature and humidity 
measurements, we used recording thermo- 
hygrographs in standard, white-painted instru- 
ment shelters. Calibrations were carried out with 
an Assmann psychrometer in connection with 
changes of the paper record. 
The observations were made at sites adjacent 
to the Gas Arctic Research Station in 1972 and 
