Kipp&Zonen BSRN Scientific Solar Monitoring System - Manual (Page 78)

Forgan, B. W ., 1996: A new m ethod for calibrating reference and field pyranometers, Journal of
21
Atm ospheric and O ceanic Technology, 13 638 - 645.
66
(7) An averaging period where the difference between the reference instrum ent and the field
instrum ents is greater than 1%, and is greater than 3 standard deviations away from the m ean
difference should be discarded.
(8) A m inim um of 25 acceptable series m ust be com pleted for each com parison.
(9) All changes in the ratio between the reference and the field instrum ent(s) m ust be recorded.
Changes of less than 0.1%/year (norm alized) need not be reported to the archive.
Changes greater than 0.5%/year (norm alized) indicate significant drift in one or both of the
instrum ents and rem edial action should be taken im m ediately. If a third cavity radiom eter
is available, the com parison should be repeated in an attem pt to isolate the changes. Once
the problem instrum ent is isolated, it should be returned to the m anufacturer to identify the
cause of the change in responsivity.
(i) If the field cavity radiom eter is faulty, the reference instrum ent should be used as a
replacem ent until the field instrum ent is returned. At that tim e, the reference instrum ent
should be com pared against the W RR.
(ii) If one of the field pyrheliom eters is found to be in need of service, another pyrheliom eter
of the sam e m anufacture and m odel should be substituted while it is sent for service.
(iii) If the reference instrum ent has apparently changed its responsivity, the instrum ent
should be sent to the m anufacturer to determ ine the reason(s) for the change and then
com pared against the W RR before a new com parison with the field instrum ents is
perform ed.
The archive should be inform ed of the problem and its solution once obtained. Large changes
of this nature m ay require re-evaluation of previously obtained data. If expertise is not available
to analyse the effects such a change in responsivity m ay have on previous determ inations of the
responsivity of other instrum ents or on the data, contact the BSRN archive for further advice.
8.3 Pyranometer Calibration
The standard procedure for the calibration of pyranom eters adopted by the BSRN is that of Forgan
(1996) . This m ethod recognizes that the best calibration accounts for the clim atic regim e in which
21
the pyranom eter is located. The following steps provide a brief outline of the procedure.
(1) Two (or four for redundancy) pyranom eters are required with approxim ately equal sensitivity.
The original sensitivities can be determ ined by the sun/shade technique against a standard
radiom eter.
(2) One instrum ent is used as the global instrum ent, while the second is installed on site as the
shaded radiom eter. Care m ust be taken to ensure that the area blocked by the diffuse disk
m atches the field of view of the cavity radiom eter (or working pyrheliom eter) being used for
the m easurem ent of direct beam radiation.
(3) At about the tim e of the sum m er solstice, the two pyranom eters are switched during a period
of sunny weather.
(4) Using a series of sim ultaneous equations the sensitivity of the two pyranom eters can then
be calculated from the data obtained during the several days before and after the instrum ent
swap.
This procedure has several significant advantages over the sun/shade m ethod of pyranom eter calibration.
Firstly, the procedure does not require the instrum ent to be rem oved from service during the calibration
procedure with the exception of changing its location, which can be accom plished during darkness.