11.1 Introduction

Net radiation (R_n) is the algebraic sum of the downwelling broadband solar, the same as GHI, also called shortwave (S↓), the downwelling broadband infrared, also called longwave (L↓), the upwelling broadband shortwave (S↑), and the upwelling broadband longwave radiation (L↑), that is,

Net radiation is the energy that evaporates water or sublimates ice, produces sensible heat that raises the temperature of the atmosphere, and heats surface soil or water. A small amount of net energy is used to photosynthesize carbon dioxide and water into organic compounds, especially sugars, and oxygen in plants. Measuring the net energy with the highest accuracy is important because this is the energy that drives weather and climate.

Figure 11.1 plots all of the variables in Equation 11.1 for two of the National Oceanic and Atmospheric Administration Surface Radiation (SURFRAD) Network sites on the same day as a function of the local standard time of the site (http://www.srrb.noaa.gov/surfrad/index.xhtml). The downwelling components are in solid lines, and the upwelling components are dashed lines. Obviously, it was clear at both sites with global horizontal irradiance (GHI) reaching nearly the same maximum values. Since the desert has a higher albedo, the reflected GHI is higher than the vegetative surroundings at Goodwin Creek. Upwelling infrared is higher at both sites because the surface temperature and emissivity are greater than the atmosphere’s temperature and emissivity. The difference in upwelling and downwelling infrared at Goodwin Creek is smaller than the difference at Desert Rock because of the higher atmospheric water vapor content at Goodwin Creek (midday relative humidities were 10% and 40%, respectively). At both sites, the net energy is negative at night because the ground is warmer than the atmosphere and there is no solar radiation contribution to the energy budget. The daily integrated net energy is larger at Goodwin Creek than at Desert Rock with higher values during the day and night.

Four different types of commercial instruments are used to measure net radiation. The most basic instrument measures net radiation by determining the thermopile voltage created by connecting the hot and cold junctions of a thermopile to the top and bottom absorbing surfaces of the radiometer, respectively. The thermopile is covered with a dome that transmits most wavelengths of the shortwave and longwave spectrum, often polyethylene. A second type of net radiometer consists of two sensors that measure downwelling and upwelling radiation separately. The third type of net radiometer measures net shortwave and net longwave separately. The fourth method to measure net radiation is to construct net radiometers using four sensors that separately measure the four components of net radiation. These instruments and their accuracies will be discussed in the next three sections.

11.2 Single-Sensor (All-Wave) Net Radiometers

The simplicity of concept and construction of a single-unit-construction net radiometer would appear to be a good design. The hot junction of the thermopile is connected to the top of the detector, which receives downwelling shortwave and longwave radiation, and the cold junction is connected to the bottom of the detector, which receives upwelling shortwave and longwave radiation. The difference in temperature of these matched receivers is proportional to the net radiation. To reduce the effects of wind, dust, and precipitation a material that transmits both solar and infrared radiation protects the surfaces, usually polyethylene. Since thermopile responses are sensitive to temperature, it is best to include a measurement of the thermopile temperature needed to make this correction. A bubble level is used to ensure the proper horizontal orientation of the instrument, and a desiccant holder located within the housing is often included to eliminate internal condensation. The cost of this sensor tends to be much lower than the multisensor net radiometers. Examples of commercially available instruments of this type of net radiometer are the Radiation and Energy Balance Systems (REBS) Q–7. 1, the Middleton CN1-R, the EKO Instruments MF–11, and the Kipp & Zonen NR Lite2. Figure 11.2 is an example of a single-sensor (all-wave) instrument in its most common form.

The polyethylene domes transmit most of the radiation in the short and long wavelengths. The bubble levels allows for a proper (horizontal) orientation. Desiccated air flows through the support arm. Figure 11.3 is a novel design of the all-wave net radiometer using a rugged shield that does not require constant airflow to keep the protective shield inflated. The vertical shaft is to discourage birds from perching on this instrument.

There are design and operational issues when using these types of net radiometers. Many sensors of this type use a flexible polyethylene dome to protect the detector from dust, wind, and precipitation yet transmit most of the solar and infrared wavelengths. These domes use a continuous desiccated air or nitrogen stream to keep the domes inflated and free of condensation on the interior surfaces (Figure 11.2). Some manufacturers now use thicker polyethylene domes so that continuous inflation is no longer required, but the air inside continues to need desiccation. Manufacturers suggest that the domes be changed every 3 to 6 months, mainly because they can be easily scratched, and they become brittle and crack under freezing conditions. The domeless NR Lite2 (Figure 11.3), which has a black Teflon conical absorber covering the sensors, is more rugged than the polyethylene domes, but this instrument is the most sensitive all-wave net radiometer to wind speed. All of these sensors show some dependence on wind speed (Brotzge and Duchon, 2000; Smith et al., 1997), and manufacturers provide corrections for wind-speed dependency. Further, Brotzge and Duchon (2000) found that the effects of precipitation are pronounced in the NR Lite. Imperfect Lambertian responses (responses that are proportional to the cosine of the angle of incidence) are not thoroughly discussed in the literature, although Brotzge and Duchon pointed out some problems with the NR Lite, the predecessor to NR Lite2. Calibration is problematic because there is no accepted standard for this measurement. Calibrations are discussed at the end of the chapter.

11.3 Two-Sensor Net Radiometers

As mentioned in the introduction, two types of net radiometers employ two separate sensors; however, their mode of operation is fundamentally different. The Schulze-Däke net radiometer is composed of two separate pyrradiometers (2π steradian field of view instruments measuring both long-wavelength and short-wavelength irradiance) that measure all downwelling radiation with one and all upwelling radiation with the other. Polyethylene domes are used to shield the sensors’ surfaces from the weather but are thick enough to be self-supporting. Halldin and Lindroth (1992) found this instrument “showed superior performance, and its output was almost entirely within the accuracy of the reference measurements” (p. 762). This was a comparison to the net radiation instruments of the day, many of which still exist in some form. It appears that the Schulze-Däke net radiometer is no longer available. The Schenk 8111 pyrradiometer is similarly configured as the Schulze-Däke net radiometer and is available from Ph. Schenk GmbH Wien (see Figure 11.4).

The Kipp & Zonen model CNR 2 measures net shortwave and net longwave using separate sensors (Figure 11.5). The CNR 2 is a relatively new instrument. It uses glass domes over the shortwave detector and silicon windows over the long-wave detector. The infrared sensor field of view is 150°. The Hukseflux RA01 is similar in its configuration with the infrared measurements also limited to a 150° field of view. Blonquist, Tanner, and Bugbee (2009) recently compared the CNR 2 with other types of net radiometers. They concluded that the CNR 2 data were less accurate than four-component net radiometers but were better than the single-sensor net radiometers.

11.4 Four-Sensor Net Radiometers

The Kipp & Zonen CNR 1 is the oldest of this genre, but the manufacturer no longer offers it. Its replacement from Kipp & Zonen is the CNR 4. Hukseflux’s NR01 is the other four-component sensor available for net radiation measurements (see Figure 11.6). The NR01 infrared measurements have a field of view that is 150°, as was the case for the old Kipp & Zonen CNR 1. The 150° field of view instruments use a flat surface for better deposition of the interference filter for the infrared compared with the convex surfaces of the 180° instruments. These systems have the advantage that each sensor can be calibrated separately, and the downwelling shortwave sensor’s cosine response can be measured. The sensors provide details of the incident radiation for each of the components of net radiation. These four-component instruments do not, however, use the manufacturers’ top-of-the-line pyranometers or pyrgeometers because of the high cost of such a system.

In a very careful study using the current and best pyranometer and pyrgeometer calibration techniques, Michel, Philipona, Ruckstuhl, Vogt, and Vuilleumier (2008) found that a heated and ventilated CNR 1 could be field calibrated to yield net radiation uncertainties smaller than 10% as claimed by the manufacturer. However, using the manufacturer’s indoor calibrations led to significantly higher uncertainties. They found that even with field calibrations that an unheated and unventilated CNR 1 could not measure net radiation with uncertainties under 10%.

11.5 Accuracy of Net Radiometers

Blonquist et al. (2009) addressed the question of the accuracy for many of the currently manufactured net radiometers. They found that, in general, the accuracy increases with the cost of the net radiometer. The four-component systems were the best, followed by the two-component systems and then the all-wave net radiometers. However, one problem with this comparison is that there was no true reference. As a proxy for the reference, this study used the average of each separate sensor on the two Kipp & Zonen CNR 1’s and the three Hukseflux NR01’s. The most significant problem is the lack of a longwave, or broadband infrared, standard. The lack of a broadband infrared standard has been of concern for a long time and has been mentioned early and often in the literature (Blonquist et al., 2009; Brotzge and Duchon, 2000; Halldin and Lindroth, 1992; Ohmura et al., 1998). Recently, an interim standard for the measurement of infrared irradiance was developed at the World Radiation Center in Davos, Switzerland. The interim World Infrared Standard Group (WISG) (http:// www.pmodwrc.ch/pmod.php?topic=irc) consists of four pyrgeometers that have been calibrated using an absolute sky-scanning radiometer and black-body characterizations (Marty et al., 2003; Philipona et al., 2001). This infrared standard, along with the World Radiometric Reference (WRR) for solar radiation (Finsterle, 2006), presents an excellent opportunity to develop a true net radiation standard.

11.6 A Better Net Radiation Standard

With improvements in the measurement of the diffuse and direct components of solar radiation (Michalsky et al., 2007, 2011) and the establishment of the interim World Infrared Standard Group (WISG) at the World Radiation Center in Davos, Switzerland, for the measurement of infrared (http://www.pmodwrc.ch/pmod.php?topic=irc), there is the potential to develop an accurate working standard for net radiation.

The very best measurement of net radiation requires four first-class horizontal radiometers (two pyranometers for shortwave and two pyrgeometers for infrared) plus a first-class pyrheliometer and a solar tracker with shading for the two downwelling, horizontally mounted radiometers. These measurements are not widely made, but several Baseline Surface Radiation Network (BSRN) sites, which field first-class radiation sensors (Ohmura et al., 1998), are making these measurements. Due to the complexities associated with a 2n steradian field of view and other pyranometer design considerations, the most difficult measurement of shortwave irradiance is the downwelling component. Adding the independently measured direct normal component and the diffuse horizontal irradiance to compute the total horizontal shortwave irradiance minimizes the cosine response error common to current pyranometers. The thermal offset correction for properly calibrated pyranometers must also be made (Michalsky et al., 2007). Further, by shading the downwelling infrared from direct solar radiation, the solar correction to the infrared irradiance measurement is minimized.

A worthwhile study that should be undertaken in light of the recent improvements to solar and infrared radiometry is the characterization of commercial net radiometers using any one of the fully equipped BSRN sites.

11.7 Net Radiometer Sources

This list identifies known manufacturers that continue to produce net radiometers:

• EKO Instruments Co., Ltd.  (http://www.eko.co.jp)

• Hukseflux Thermal Sensors B.V.  (http://www.hukseflux.com/)

• Kipp & Zonen BV  (http://www.kippzonen.com/)

• Middleton Solar  (http://www.middletonsolar.com/)

• Ph. Schenk GmbH Wien  (http://www.schenk.co.at/schenk/)

• Radiation and Energy Balance Systems (no website; phone: (206) 624–7221)

Questions

1. What is considered the most accurate method for measuring net radiation at the surface?

2. What are the four different types of net radiometers?

3. Why is polyethylene a popular dome material for net radiometers?

4. Why is the fractional accuracy of net radiation lower than the individual radiation measurements?

5. Why is the measurement of net radiation important?