1. Introduction
The net radiation is a basic parameter required for the investigation of the surface energy budget. The separate measurement of the incoming and the outgoing long- and shortwave radiation not only allows the net radiation to be determined more accurately than is possible with a single net radiometer but also enables the relationship between the four components to be fully characterized. To ensure the quality and reliability of the radiometer data and to facilitate the ease of deployment, an integrated four-component system was designed and implemented. The radiometer array was developed as part of the National Center for Atmospheric Research (NCAR)/Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) facility (Businger et al. 1990). Initially the four unmodified radiometers were mounted separately and their output sent separately to the main computer for processing. Further development involved modifications of the radiometers, their housing and mounting, and the ability to process the data with an in situ microprocessor.
The integrated radiometer array with the associated algorithm is a research tool and will be used to undertake intercomparisons with net radiometers, make sensitivity analyses, and investigate the radiative consequences of nonhomogenous vegetative surfaces.
We have used a value of cdc = 1.11 in our calculations. The value of the coefficient is subject to uncertainty and values from 2.5 to 3.0 were suggested by J. Deluisi in 1995. However, our calculation indicates that even if a high value of cdc = 3.0 is used rather than cdc = 1.11, the final value for net radiation differs only by approximately +7 W m−2 at night and −10 W m−2 in the daytime.
The pyranometers used are routinely calibrated by the Surface Radiation Research Branch, Air Resources Laboratory, National Oceanic and Atmospheric Administration (NOAA), and the pyrgeometers are routinely calibrated by Eppley Laboratories. An accuracy of approximately 2% for the calibrations of the pyranometers and the pyrgeometers is expected. However, given uncertainties in the correction terms, the leveling and the siting of the array the total accuracy of the system will not be as good.
2. Pyrgeometer modifications
In the standard Eppley pyrgeometer a single thermister embedded in the brass housing monitors the temperature of the case, which is temperature of the base of the thermopile and a single thermister mounted on the inside edge of the glass dome monitors the temperature of the protective hemisphere. Because of the mass and thermal conductivity of the case, the single point measurement is sufficient. However, because of the lower thermal conductivity of glass a single point measurement on the dome will not give a representative measurement of the dome temperature. To obtain a more representative measurement, the pyrgeometers were modified. Three 10-K ohm thermisters were attached 120° apart on the base of the dome by Eppley Laboratories. Circuitry to average the three temperatures was then assembled and housed in the cavity of the pyrgeometer. In the standard Eppley pyrgeometer, a battery- powered circuit is used to apply a bias to compensate for the longwave radiation emission by the top surface of the thermopile. This circuit was eliminated, and the output of the compensation thermister made available directly. A voltage-standard integrated circuit was installed to provide a reliable, fixed voltage to excite the thermisters. The internal wiring of the pyrgeometer was arranged so that the connector pin-out still corresponded to the original functions. Figure 1 shows the modification of the internal circuitry of the pyrgeometer.
3. Temperature calibration
To achieve the most accurate measurement of temperatures from the thermisters’ output, the integrated array was placed in a temperature chamber in the NCAR sensor calibration laboratory. The assemblage of thermisters, together with the excitation and amplification circuit, was calibrated to within 0.05°C, over the temperature range −20° to 50°C. Using the Steinhart–Hart equation (Steinhart and Hart 1968), a three-coefficient calibration was established for each thermister. These coefficients were loaded into an Electrically Erasable Programmable Read Only Memory (EEPROM) in the micrologger to allow appropriate calibration for the output serial message.
4. Radiometer ventilation
The radiometer domes are exposed to the atmosphere and are subject to the deposition of mineral dust, insect excrement, and vegetative debris. In addition, dew and frost often coat the domes in the morning hours and cause periods of measurement interruption. A ventilation system, blowing clean air over the surface of the domes, alleviates these problems. The degree to which the ventilation mitigates the local solar heating of the pyrgeometer dome could not be determined. A study of the influence of ventilation on pyranometers (Sanchez and Stoffel 1994) suggests that the effect is not significant.
The ventilator was designed using a Micronel, model D240, 12 VDC, axial fan (Micronel, Vista, California), drawing air through a volume packed with filter scrim to remove dust and cloud or fog droplets. The housing, which incorporates the standard Eppley radiation shield as its top cover, delivers a flow of clean air over the dome of the Eppley radiometer. Figure 2 is a schematic of the air flow in the device. Figure 3 shows the flow pattern of air around the dome when a smoke generator was used with the ventilator.
5. Leveling stand
The radiometers each view an entire hemisphere with a cosine-weighted response to incident radiation. As a first-order estimate, an error in the horizontal level will yield a corresponding cosine-dependent error in the measurement of both downwelling and upwelling radiative energy. Such a dependency would yield a measurement error of only 0.3% for a tilt error of 1°. However, as the tilt error increases to 3°, the measurement error would be expected to increase to 3%, and for 10°, the error would be an unacceptable 30%. This cosine relationship is an oversimplification because the downwelling shortwave radiation is predominately from the sun, the downwelling radiation from the whole cloud scattered sky, and both the upwelling short- and longwave radiation is from the underlaying vegetative surface with plants leaves at varying angles and different optical characteristics.
A leveling stand was designed and fabricated to accommodate the four Eppley radiometers in their ventilators. This allowed the whole system to be leveled as a single unit. The stand, which was based upon the design of optical table equipment, has independent roll and pitch adjustments that operate without cross-interference. An Applied Geometrics Model 900 Biaxial Clinometer electronic level (Applied Geometrics, Santa Cruz, California) with a remote digital readout allows the assemblage of radiometers to be leveled to within a 1/10°. Values for the north–south level Lns and the east–west level Lew are produced. Figure 4 shows the leveling stand mounted on the top beam of the radiometer stand. In the field this radiometer stand is deployed with the top beam up to 4 m above the surface. The ability to mount the integrated array high enough to see a large surface area is important.
6. Data logger
The analog voltages generated by the radiometers are quite small (∼10 mV dc) and previously the limitation of signal cable length restricted the choice of the most appropriate surface over which to site the radiometer array. The use of in situ data processing enables a robust serial data stream to be generated and the radiometer array can be located wherever the most representative vegetation occurs. Another advantage of a serial message is the freedom from possible confusion that often accompanies bundles of analog signals.
A dedicated micrologger, designed during the development of the PAMIII/Flux station (Militzer et al. 1995), is used to acquire the signals from the sensors. As the system developed, the number of input signals increased, and it was decided to use an integrated sensor concept (Pike et al. 1983) to improve overall performance and to offer a sensor adaptable to any host machine. The analog signals are amplified and digitized. Appropriate transfer functions are applied to the raw data to convert to engineering units. Intermediate and final products are then computed and passed on to the host machine as a serial message. For the application of the integrated system with the ASTER facility data is transferred at 1 Hz. The option also exists to send raw data to allow for cross-checking of the micrologger’s calculations.
7. Logger hardware
The logger was designed with flexibility in mind. To achieve this, a board-stacking approach is used with the digital circuit as the heart of the system. Different digital and/or analog boards can be stacked on the digital board. All available I/O lines are daisy-chained to each board. The digital section is built around a Motorola 68HC11 microprocessor with a 512-byte EEPROM, a 2-kbyte RAM, up to 24 I/O lines, and an 8-channel, 8-bit A/D (Motorola, Phoenix, Arizona). A WSI PSD301 programmable microcontroller peripheral device provides 32 kbyte of UVPROM. An external 32 kbyte of RAM are also available. Communication to a host machine is handled through an RS-232 or RS-485 interface. Additional circuitry includes an onboard temperature sensor and current monitoring, which is interfaced to the 68HC11 A/D microprocessor.
For the integrated surface radiation measurement system an A/D interface with signal conditioning is required. The A/D circuit is designed around the Crystal CS5505 16-bit A/D microprocessor (Crystal, Austin, Texas), having four analog input channels. One A/D board has the option for two CS5505 chips or a total of 8 A/D channels per board. The maximum data rate when all channels are active is five samples per second. The voltage input range can be either 0 to 2.5 volts or −2.5 to 2.5 volts. Front-end signal conditioning is provided for each channel. An excitation signal of 12 volts dc is also available to the sensors. Figure 5 schematically illustrates the configuration of the microprocessor.
8. Logger software
The software is designed around a simple task schedular driven by external interrupts. All software is written in c. An interactive interface allows the user to set a variety of parameters. Examples are the ability to set the channel sampling order, turning on or off channels in real time, and selecting the input voltage range per channel. Other features include a set of transfer functions that can be applied to a specific channel. For the integrated radiation system, specific transfer functions were written. Each transfer function can have up to four coefficients. All user-selected parameters are stored in EEPROM to allow for power cycling. The output data stream can be either calibrated information, raw binary, or both.
9. Data products
All the products calculated are available as a serial message from the integrated radiometer array. The products that are normally generated are listed below.
Ssw—calibrated response of the pyranometer thermopile
Slw—calibrated response of the pyrgeometer thermopile
Tp—temperature of the top of the thermopile
Td—temperature of the dome
Tc—temperature of the case housing the thermopile
uRsw—shortwave radiation measured by up-looking pyranometer
uRlw—longwave radiation measured by up-looking pyrgeometer
dRsw—shortwave radiation measured by down-looking pyranometer
dRlw—longwave radiation measured by down-looking pyrgeometer
Rn—net radiation
Lns—north–south level
Lew—east–west level
10. Results
The integrated array was set up during the MICROFRONT95 deployment 16 km north of Eldorado, Kansas, at a long-grass prairie site in spring 1995. The surface below the array was a uniform mat of dead grass. The measured and calculated components of the integrated surface radiation array, together with those of two Radiation Energy Balance Systems Q6 and Q7 net radiometers were monitored. Figures 6–9 show the response throughout the course of a single 24-h period. The time axis shows universal time coordinated. Although the data is acquired with 1-s resolution, the time resolution of the plots is 5 min. Figure 6 shows the thermopile responses and the compensation corrections for the two pyrgeometers. Note that the corrections uCp and dCp are, in fact, of greater magnitude than the value of the thermopile outputs uSlw and dSlw. Figure 7 shows the corrections caused by the heating of the pyrgeometers’ domes and the corrections required for the response of the sensors to shortwave radiation. These terms, uCdc, dCdc, uCυ, and dCυ, are considerably smaller in magnitude, and only the influence of shortwave radiation on the upward-looking pyrgeometer uCυ is truly significant. Figure 8 shows the four components of the radiation, iRsw, iRlw, oRsw, and oRlw, and the value of the calculated net radiation Rn. Figure 9 compares the value of the net radiation calculated from the four components with the values measured using two net radiometers.
REFERENCES
Alados-Arboledas, L., J. Vida, and J. L. Jimenez, 1988: Effects of solar radiation on the performance of pyrgeometers with silicon domes. J. Atmos. Oceanic Technol.,5, 666–670.
Albrecht, B., and S. K. Cox, 1977: Procedures for improving pyrgeometer performance. J. Appl. Meteor.,16, 188–197.
Businger, J. A., W. F. Dabberdt, A. C. Delany, T. W. Horst, C. L. Martin, S. P. Oncley, and S. R. Semmer, 1990: The NCAR Atmospheric-Surface Turbulence Exchange Research (ASTER) facility. Bull. Amer. Meteor. Soc.,71, 1006–1011.
Dickey, T. D., D. V. Manov, R. A. Weller, and D. A. Siegel, 1994: Determination of longwave heat flux at the air–sea interface using measurements from buoy platforms. J. Atmos. Oceanic Technol.,11, 1057–1078.
Militzer, J. M., M. C. Michealis, S. R. Semmer, K. S. Norris, T. W. Horst, S. P. Oncley, A. C. Delany, and F. M. Brock, 1995: Development of the prototype PAM III/flux-PAM surface meteorological station. Preprints, Ninth Symp. on Meteorological Observations and Instrumentation, Charlotte, NC, Amer. Meteor. Soc., 490–494.
Pike, J. M., F. V. Brock, and S. R. Semmer, 1983: Integrated sensors for PAM II. Preprints, Fifth Symp. on Meteorological Observations and Instrumentation, Toronto, ON, Canada, Amer. Meteor. Soc., 326–333.
Sanchez, J., and T. Stoffel, 1994: The influences of ventilators on pyranometer measurements. Proc. 1994 Annual Conf. of the American Solar Energy Society, San Jose, CA, American Solar Energy Society, 369–372.
Steinhart, J. S., and S. R. Hart, 1968: Calibration curves for thermisters. Deep-Sea Res.,15, 497.
The original and the modified internal circuitry of the Eppley pyrgeometer.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0046:AISRMS>2.0.CO;2
A schematic of the radiometer ventilator showing the airflow in the interior.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0046:AISRMS>2.0.CO;2
The airflow pattern generated by the ventilator.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0046:AISRMS>2.0.CO;2
The radiometers, housed within their ventilators and mounted on the leveling frame, attached to the beam of the radiation stand.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0046:AISRMS>2.0.CO;2
The data system hardware.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0046:AISRMS>2.0.CO;2
Thermopile responses and compensation corrections for the upward-looking and downward-looking pyrgeometers.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0046:AISRMS>2.0.CO;2
Dome–case corrections and corrections for the sensitivity to shortwave radiation for the upward-looking and downward-looking pyrgeometers.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0046:AISRMS>2.0.CO;2
The four components of the surface radiation budget and their sum.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0046:AISRMS>2.0.CO;2
A comparison of the four components of the surface radiation and the values from net radiometers.
Citation: Journal of Atmospheric and Oceanic Technology 15, 1; 10.1175/1520-0426(1998)015<0046:AISRMS>2.0.CO;2