Tải bản đầy đủ - 0trang
CHAPTER 5. THE CALIBRATION AND USE OF NET RADIOMETERS
SHERWOOD B. IDS0
and for scheduling irrigations (Franzoy and Tankersley, 1970; Heermann
and Jensen, 1970; Jensen, 1969, 1972; Jensen and Heermann, 1970; Jensen et ai., 1970, 1971). Thus, the study of net radiation has become a
subject of considerable importance for many soil scientists and
Over the years, several attempts have been made to develop techniques
€or estimating net radiation from other more easily measured parameters
(Davies, 1967; Fritschen, 1967; Gay, 1971; Idso et al., 1969b; Linacre,
1968, 1969; Monteith and Szeicz, 1961; Pasquill, 1949; Penman, 1948
Richardson, 1931; Scholte-Ubing, 1961; Stanhill et al., 1966). Owing to
effects of clouds, however, such procedures have not achieved a degree
of accuracy that is satisfactory for all-weather utilization. Indeed, even when
these procedures are applied only to clear sky conditions, Idso (1968,
1971a) and Nkemdirim (1972, 1973) have shown that they may not be
adequately generalized. Direct measurement thus remains the only reliable
method for obtaining precise information on net radiation under all the
varied conditions existing in the natural environment. Consequently, this
paper presents a summary of some of the basic aspects of net radiometers
and techniques for using them.
The primary component of all net radiometers is the thermopile. This
device usually takes the form of a flat black plate of some insulating material with a series of thermocouple junctions located on each side. The temperature difference across the plate is proportional to the net radiation to
which it is exposed; and by means of a calibration constant, the electrical
output of the thermopile is converted to yield this flux.
The basic design of the more common thermopiles in use today was
described by Funk (1959, 1962a) and Fritschen (1963, 1965). Other
variants were reported by Baumgartner (1952), Campbell et al. (1964),
Frttschen and Van Wijk (1959), and Hofmann (1952). Gates (1964),
Monteith (1959), and Monteith and Szeicz (1962) outlined the manufacture of simple versions. In addition, there is the economical net radiometer
transducer which utilizes a pair of thermometers. It has been described
by Suomi and Kuhn (1958) and Tanner et al. (1960, 1969).
There are two major types of net radiometers-ventilated and shielded.
In the first type, a forced stream of air is directed over the thermal transducer, or thermopile, to equalize as much as possible the effects of convection from its upper and lower surfaces. Examples of such instruments are
those developed by Gier and Dunkle (1951) and Suomi et al. (1954).
In the second type, transparent wind shields are used to reduce unequal
convection effects. Some of the materials that have been used to construct
CALIBRATION AND USE OF NET RADIOMETERS
FIG. 1. The thermopile of a Fritschen net radiometer showing the epoxy bobbin
around which are wound the several thermocouple junctions. The bobbin is encased
in a flat disk of similar epoxy and is now ready to be painted black.
these wind shields are rock salt coated with selenium (Aleksandrov and
Kurtener, 1941), sylvite coated with selenium (Khvoles, 1952), thallium
bromoiodide (Houghton and Brewer, 1954; Kreitz, 1953; Stern and
Schwartzmann, 1954), filters of germanium and silicon (Mills and Crawford, 1955), sheets of mica (Fritschen and Van Wijk, 1959), Saran plastic
film (Fritschen, 1960), and polystyrene (Fritschen, 1963). The material
used by most manufacturers today is polyethylene (Monteith, 1972).
Since many good reviews dealing with the various commercially available
net radiometers and their operating characteristics already exist (Gates,
1962, 1965a; Kondrat’yev, 1965; Monteith, 1972; Robinson, 1962; Sellers,
1965; Tanner, 1963), this information is not repeated, nor is the detailed
theory underlying their operation. Instead, this review is devoted to practical discussions of calibration techniques, basic considerations of data acquisition, and methods of modifying net radiometers for a variety of different
The most common method for calibrating net radiometers is the occulting technique (Collins and Rabich, 1971) or, more simply, the so-called
“shading technique.” Calibration is accomplished by positioning a net
radiometer alongside a pyranometer or solarimeter and alternately shading
SHERWOOD B. IDS0
and unshading both instruments simultaneously from the direct rays of the
sun (or some other light source) with identical small, opaque shields. In
the laboratory, a high-pressure xenon arc lamp is generally used as the
light source (Collins, 1970), although tungsten filament lamps have also
been used (Drummond, 1956; Fritschen, 1963; Hill et al., 1966).
If the net radiometer is positioned high enough above the ground so
that the shadow from the shield used to shade it from the direct beam of
the light source is insignificant in altering the radiation it receives from
below, the electromotive force (emf) change of the net radiometer’s output
FIG. 2. Calibrating a net radiometer by the common shading technique. The
lower left picture shows obscuration of the sun by the shield over the net
radiometer; and the lower right picture depicts the shading of the Eppley pyranometer
used as the standard.
CALIBRATION AND USE OF NET RADIOMETERS
between shaded and unshaded conditions can be equated to the known
energy flux change recorded by the standard pyranometer to yield the Calibration factor of the net radiometer. Important points to remember when
using this technique include the following: (a) the net radiometer must
not be positioned too close to the ground when calibrating; (b) the accuracy of the resulting calibration factor will be no better than that of the
solarimeter used to obtain it; and (c) this technique gives the calibration
factor for response to shortwave radiation (0 5 A _< 3 p m ) , which may or
may not be the same as the net radiometer’s response to long-wave radiation (A > 3 pm).
Chamber techniques are generally used to obtain long-wave radiation
calibration factors. Two major variations are presently in vogue, one due
primarily to Fritschen (1963) and the other to Funk (1959, 1962a,b).
FIG. 3. Back and front views of the calibration chamber constructed by L. J.
Fritschen for the calibration of net radiometers at the US. Water Conservation
Laboratory. It can be used for both short- and long-wave calibrations. In the
former instance, an artifical light source is used to alternately illuminate a net radiometer and a pyranometer located at the central chamber; in the latter instance,
the central chamber divider is removed, and the net radiometer is positioned there
and calibrated by calculation of the net long-wave radiation exchange between the
upper and lower chamber halves, which are kept at different temperatures by circulating water.
SHERWOOD B. IDS0
Fritschen’s chamber, developed from a prototype of Johnson (U.S.
Weather Bureau, unpublished manuscript, 1956), is composed of two cubical halves with their common side removed. The inner copper walls are
painted black and have copper tubing attached to their backsides. Two
water streams of different temperature flow through the tubes attached to
the upper and lower sets of five walls to create a net flux of thermal radiation. The net radiometer is positioned in the center of the chamber; and
its output is equated with the calculated net long-wave radiative flux; this
flux is derived from thermocouple-obtained wall temperatures and a knowledge of the wall emittance, which may be measured by a variety of techniques (Buettner and Kern, 1965; Conaway and Van Bavel, 1966; Fuchs
and Tanner, 1966, 1968; Idso and Jackson, 1969). The formula given by
Fritschen for calculating the net radiation in his chamber is not exact, however, and the correct expression derived by Idso (1970a) should be used
when employing this technique.
The long-wave calibration technique of Funk was developed from an
earlier version of MacDowall ( 1955 ) which makes use of a more complex
arrangement of two chambers. The net radiometer is positioned in the
center of one of the insulated chambers and allowed to view the second
(a blackbody cavity of different temperature) through a small aperture
€or only a short period of time. The blackbody cavity most used today
is an electrically heated cylindrical cavity of cast aluminum described by
Collins (1968), the emissivity of which is known to be better than 0.999.
Both Funk‘s and Collin’s analyses of the net radiation to which the net
radiometer is exposed, however, are slightly in error. Idso (1970a) provides a more exact evaluation of this flux.
Chamber techniques of the Funk and Fritschen type require no other
standard radiometer or solarimeter for comparison; and they can yield very
accurate long-wave calibration factors when properly utilized. The chambers are somewhat complex, however, and hardly worth the effort to construct for the calibration of just a few radiometers. Thus, they generally
are used only in major laboratories, such as the CSIRO Division of Atmospheric Physics, which calibrates some 400 instruments annually (Collins
and Rabich, 1971), and the U.S. Water Conservation Laboratory, which
at its peak production of Fritschen-type net radiometers calibrated about
A simpler technique for obtaining long-wave calibration factors of net
radiometers was developed by Idso ( 1970a,b). In this method a net radiometer is positioned anywhere from 5 to 20 cm above the floor of a constant-temperature room. Then, a blackened metallic plate about 40 cm
on a side (which was previously cooled about 10°C below the temperature
of the room) is brought into the room and set beneath the radiometer.
CALIBRATION AND U S E OF N E T RADIOMETERS
FIG.4. The setup of the flat-plate technique for obtaining long-wave calibration
factors of net radiometers.
The output of the radiometer is subsequently monitored, along with the
temperature of the flat plate (by means of 4 or 5 embedded thermocouples), as the cool plate gradually warms to ambient room temperature.
Since all radiative fluxes in the room are constant, except for the flux
emitted by the warming plate, the millivolt output of the net radiometer
can be plotted against the radiation received from the plate and its calibration factor determined as the slope of this line. The radiation from the plate
is specified by the plate’s emittance, temperature, and view factor with respect to the radiometer, which may be determined from a variety of standard heat transfer texts (Eckert and Drake, 1959; Jakob, 1957; Sparrow
and Cess, 1966). This technique has also been shown to be suitable for
calibrating soil heat flux plates (Idso, 1972b).
Both Fritschen and Funk concluded from their studies of short- and longwave calibration factors of polyethylene-shielded net radiometers that these
instruments were about 5% more sensitive to short-wave radiation than
to longwave radiation. A similar conclusion was also reached by Collins
and Kyle ( 1966). Thus, for years, practically all polyethylene-shielded net
radiometers have had a small strip of white paint applied to their thermopiles to reflect about 5 % of the incident solar radiation and thereby equate
their short- and long-wave calibration factors. Idso ( 1970a), however,
showed that the inaccuracies in the original derivations of the equations
representing the net long-wave radiation fluxes in the chambers of Fritschen
and Funk were such as to completely, and rather exactly, negate the need
for this paint. Experimentation with the flat plate technique also confirmed
this analysis. Thus, for the most exacting requirements of net radiation
measurement, polyethylene-shielded instruments without any white paint
on their transducers should be employed.
SHERWOOD B. IDS0
Utilizing the Basic N e t Radiometer
Most of the primary considerations involved in net radiometry are discussed by Fritschen (1963, 1965) and Funk (1959, 1962a). In addition,
there is an instruction manual by Fritschen and Mullins (1965) that is
whollly devoted to this subject. The methodology covers the physical installation of net radiometers, leveling of net radiometers, preventing of internal
and external condensation of water on radiometer domes, cleaning of
domes, and replacement of damaged domes. Only a few additional comments are required.
First, it has been my experience that radiometer calibration factors are
not adversely affected unless the domes are extremely dust laden. Thus,
cleaning the domes should be a rather conservative practice, since they
are so easily scratched. Second, although a standard placement height of
1 meter is often suggested for net radiometers, this topic is a subject in
itself. For instance, routine experimentation has been conducted with net
radiometers at heights of 0.1 m (Van Bavel and Fritschen, 1965), 0.3
m (Van Bavel, 1967), 0.5 m (Rosenberg, 1966; Stanhill et al., 1966),
0.75 m (Idso et al., 1969b), 1.0 m (Davies and Buttimor, 1969; Fritschen,
1967; Stanhill and Fuchs, 1968; Thompson and Boyce, 1967), 1.5 m
(Ekern, 1965; Fitzpatrick and Stern, 1966; Monteith and Szeicz, 1961),
2.0 m (Begg et al., 1964; Denmead, 1969; Morgan et al., 1970; Rosenberg,
1969), and 2.5 m (Stanhill et al., 1966). Of this group of investigators,
only the ones using the lowest (0.1 m) and highest (2.5 m) heights gave
any reasons for their choices-and these were only qualitative at that.
Idso and Cooley (1971, 1972) conducted a comprehensive study of net
radiometer height placement. They showed that if net radiation measurements are required over a surface such as a dry, bare soil that becomes
significantly warmer than ambient air during clear days, net radiometers
should be located at a height of only 20 or 25 cm. The reason for choosing
this level is that the intervening air between the net radiometer and the
ground interacts with the radiation from the soil and modifies it significantly
over even smaller distances than this. Thus, true net radiation is theoretically measured only at the surface, if surface and air temperatures are
different. However, as the surface is approached, the shadow of the net
radiometer begins to alter significantly the radiation emitted and reflected
by the surface. The level of compromise between these two effects was
found to be between 20 and 25 cm for radiometers of the Funk and Fritschen type (about 6 cm in diameter). If for some reason a different level
is selected, Kondo (1972) has developed procedures for estimating the
required corrections to surface conditions.
CALIBRATION AND USE OF NET RADIOMETERS
When surface and air temperatures do not differ greatly, such as over
an irrigated field or crop, considerations of homogeneity of field of view
tend to predominate in selecting the height of net radiometer placement.
View factors must then be calculated for various heights to determine the
best level for representative sampling of the total system or a specific portion of it, i.e., the bare soil between crop rows. Standard heat transfer
texts may be consulted for this exercise or the two more agriculturally
oriented papers of Reifsnyder and Lull (1965) and Reifsnyder (1967),
which contain some useful graphs that often obviate the need for an investigator to make the actual calculations. Other examples of view factor utilization in net radiometry have been published by Gates (1962, 1965a) and
Waggoner and Reifsnyder (1961).
Modifications for Different Applications
The basic Fritschen and Funk net radiometers are extremely versatile
instruments. Solar Radiation Instruments' (SRI) of Australia markets a
Funk type that is claimed to be suitable for use in nine different forms.
These are: (a) a regular net radiometer, (b) a short-wave balance meter,
(c) a solarimeter, (d) an albedometer, (e) a total hemisphrical radiometer,
( f ) a total albedo and long-wave emission radiometer, (g) an underwater
short-wave balance meter, ( h ) an underwater solimeter, and (i) an underwater albedometer.
Most of the procedures required to transform the basic net radiometer
into these other forms are rather simple and straightforward. By merely
removing the polyethylene domes and replacing them with glass, for instance, the net radiometer is transformed into a short-wave balance meter.
The SRI instrument may be purchased with interchangeable glass domes
specifically manufactured for this purpose. Working with the Fritschen net
radiometer, I have also made this transformation with the glass domes that
are normally used for the Kipp solarimeter, as they are of the proper size
to be held in place by the net radiometer's clamping rings.
By painting either the inner or outer surface of one of the glass domes
black to restrict light entry to only one side of the thermal transducer, the
shortwave balance meter may be transformed into either a solarimeter or an
albedometer, depending upon whether it is operated in an upright or an
inverted position. In both of those applications, however, additional steps
must be taken to ensure that the painted dome does not absorb an excessive
*Trade names and company names are included for the benefit of the reader
and imply no endorsement or preferential treatment of the product listed by the
U.S.Department of Agriculture.