Kipp&Zonen BSRN Scientific Solar Monitoring System - Manual (Page 77)

65
to guard against perform ance degradation between international com parisons. One m eans of m onitoring
perform ance is the use of the reference instrum ent in W MO Regional Pyrheliom eter Intercom parisons.
It should be cautioned that an instrum ent’s calibration coefficients should not be changed unless a
confirm ed shift in the instrum ent properties has occurred.
The reference instrum ent will be used to m onitor the responsivity of the field instrum ent on an ongoing
basis depending upon weather. This procedure should occur at least quarterly, if weather perm its.
W hile frequent com parisons between the reference instrum ent and the field instrum ent are beneficial,
it is not the goal of the procedure to com pare the instrum ents at every favourable opportunity. Therefore,
it is not necessary for the reference instrum ent to rem ain perm anently at the station. At sites where
a cavity radiom eter is being used continuously, norm al incidence pyrheliom eters should only be used
to fill in data gaps when the cavity radiom eter is in calibration m ode. The output of these therm opile
instrum ents should be correlated with the output of the cavity radiom eter obtained im m ediately before
and after the calibration cycle of the field cavity radiom eter.
In the special cases where two therm opile pyrheliom eters are used to m easure the direct beam radiation
at a station, only one cavity radiom eter is necessary. In this case, the single instrum ent will follow the
procedures set down by the reference instrum ent in the preceding paragraphs with the exception that
the norm al incidence pyrheliom eters be com pared against the reference instrum ent as frequently
as possible. This latter procedure, while m ore econom ical, will not provide the sam e overall quality
of data.
In all cases, the link between the W RR and the BSRN observatory instrum ents should be through a
cavity radiom eter.
8.2.2 Detailed procedures
As all solar radiation m easurem ents are linked to the output of the working cavity radiom eter or working
pyrheliom eters, great care m ust be taken to m aintain the highest standards in com paring the working
instrum ents with the reference radiom eter. This com parison is the prim e link between individual BSRN
stations and the W RR and thus all other BSRN stations. The following procedures are presented as
a m eans of ensuring that this prim e link is m aintained with the lowest uncertainty possible.
(1) The reference and field radiom eters should be co-located (within m etres). If possible, a
perm anent m ount should be constructed on the sam e optical axis as the field radiom eter
for the reference radiom eter. If this is not possible, extrem e care m ust be taken to ensure
the accuracy of the pointing of both instrum ents during the com parison.
(2) To reduce uncertainty, both instrum ents should be connected to the sam e data acquisition
system . All m easurem ents should be differential. Extrem e care m ust be taken to elim inate
unnecessary noise and ground loops. W hen a single digital m ultim eter (or equivalent) cannot
be used, a reference voltage and reference resistance should be m easured by the different
m ultim eters and any discrepancies corrected. The references used should be of the sam e
m agnitude as the signals being m easured.
(3) The am pling frequency should be the sam e as used during norm al operations.
(4) The averaging period should be a m inim um of 10 m inutes and a m axim um of 25 m inutes.
Following each averaging period those instrum ents operated in a passive m ode (e.g., Hickey-
Frieden) m ust be zeroed and calibrated. The m inim um tim e allotted for each of these functions
should be one m inute.
(5) All calibration activities m ust be conducted on days in which the cloud cover is less than 4/8's
and the cloud is greater than 15° distant from the solar disk. As a quantitative check of stability,
all averaging periods m ust have a standard deviation of less than 0.1% of the m ean solar
signal during the averaging period or 0.3 W m (whichever is greater).
-2
(6) The irradiance levels should be between 400 and 1100 W m during the com parison (the -2
m axim um value is dependent upon season, latitude and altitude).