Kipp&Zonen BSRN Scientific Solar Monitoring System - Manual (Page 84)

72
= pyrgeom eter body tem perature (K)
= pyrgeom eter dom e tem perature (K)
= the electrical output from the therm opile
= a correction factor for infrared irradiance on unshaded dom es.
Details are given in Philipona et al. (1995).
The therm istor tem peratures are calculated using the Steinhart and Hart equation with the standard
coefficients provided by the m anufacturer:
T = a + b(lnR) + c(lnR) -1 3
where T = tem perature (K)
a,b,c = the standard coefficients provided by the m anufacturer
R = the resistance in ohm s.
Newer pyrgeom eters are now being tested, but have not yet becom e com m on in the BSRN network.
The equations provided by the m anufacturer of these instrum ents should be applied to obtain the infrared
flux.
9.3 Quality Assurance Techniques
9.3.1 General testing procedures
9.3.1.1 Redundancy
Having m ore than one instrum ent m easure the sam e signal is useful. This provides a m eans of flagging
a signal as problem atic when the redundant m easurem ents differ and gives backup observations during
tim es when routine m aintenance is being done or an instrum ent is m alfunctioning. It is recognized
that having m ultiple instrum ents is not always feasible and this is therefore not m andated within the
BSRN fram ework.
9.3.1.2 Visual inspection
The m ost rapid m eans of determ ining gross problem s with the incom ing data are visual. It is highly
recom m ended that the DAS provide near-real-tim e (m inutes) graphical displays of the data, whether
converted to engineering units or sim ply transducer signals. A prelim inary conversion provides the
technician a better appreciation for the data, but large changes such as infinite resistance or zero signal
can be determ ined easily even from transducer signals. W hile the data being stored at a one m inute
interval provide significant inform ation for later quality assurance testing, the initial displays need only
be the m ean values. The m ore frequent the processed signal is output, the better chance the observer
has of observing unusual phenom ena.
On clear days, rapid sam pling of the data can provide a graphical m eans of ascertaining whether individual
instrum ents are level, or if any biases are in the solar-tracking instrum ents being used. Such changes
will be obvious through the asym m etry of the data.
Grouping of incom ing values is also beneficial. For exam ple, placing the tem perature signals of all
of the pyranom eters on one graphical display provides a rapid m eans of determ ining if one instrum ent
(or its ventilator) is m alfunctioning by showing large tem perature departures from the other instrum ents.
9.3.1.3 Lim it Checking
Autom atic lim it checks can be program m ed into m any DAS’s. These lim its can be such that flags are
autom atically inserted into the data stream to warn the operator of potential problem s. A key exam ple
is the use of lim its to test the resistance of instrum ents frequently (hourly, daily) and provide a warning
if the resistance is above or below a norm al set of lim its. Resistance lim its of this type can also be
effectively set out with respect to therm istor m easurem ents.
Sim ilar checks can be set up with respect to voltage signals. If instrum ents have known ranges, lim its
can be set to warn the operator if the instrum ents exceed the range. Two types of ranges m ust be