Abstract:
The design and testing of an all-thermistor (no thermopile) pyrgeometer (LWT) is descrbed. After calibration
by comparison with a Kipp & Zonen CG4, 9.2 months of data show mean differences of order 1-2 W m-2
with standard deviations of order 7-8 W m-2. Approximately half of the mean difference and the standard
deviation derived from 40 occasions when the LWT readings were anomalously high for periods of 2-l4
hours, principally at night. No reason has been found for the anomalous behavior. During the 9.2 months,
of data, there were also 11 periods of a few hours each when the Eppley PIR indicated noticeably higher
flux values than did the Kipp & Zonen CG4. The conclusion reached is that contemporary thermistors allow
temperature measurements of suffcient accuracy, and the thermopile can be eliminated from pyrgeometers.
The differences seen between the Kipp & Zonen and the Epply raise doubts about their absolute accuracies
on time scales of hours, although their long-term averages are quite comparable.
