The external and internal temperatures are calibrated using a 1000 ohm 0.05% resistor in place of the platinum resistance thermometers with 1000 ohms resistance at 0 C. Because the other resistances in the temperature circuit are known only to 1%, the temperature calibration will vary from one electronic unit to another. The correction factor determined from the calibration resistor is programmed into the read-only-memories for each unit. After the correction factors have been programmed into the AWS, a calibration box with 0.1% resistors is used in the field to check the temperature calibration.
The atmospheric pressure transducer is a Parascientific model 215 digiquartz pressure gauge. The transducer frequency changes from 40 kHz at zero pressure to about 36 kHz at 1000 hPa. The pressure resolution is about 0.05 hPa.
Paulin aneroid barometers calibrated against a mercury barometer of 10 mm bore are used to check the pressure gauge calibration. Comparisons are made between AWS units, the aneroid barometer at McMurdo Station, and with the mercury barometers at Scott Base, Antarctica. The calibrations should be within +/- 0.2 hPa. Two mercury barometers have been purchased for use at McMurdo, Antarctica but are not yet available.
The reference vacuum on the older pressure transducers can degrade with time with a maximum observed 4 hPa shift to lower pressure after five years. Thus, recalibration of each pressure transducer would be desirable every two to three years.
The Belfort model 123 aerovane measures wind direction and speed. The aerovane rotates a potentiometer wiper, and the fraction of full scale of the potentiometer is measured. The wind direction is checked by positioning the aerovane to the cardinal directions relative to the boom supporting the aerovane. North or the potentiometer zero is towards the antenna on the boom and has a dead zone of 5 deg. During the field installation the boom is usually aligned along the north-south line as determined from the sun's azimuth, longitude, and Greenwich Mean Time. In some cases the 180 deg end of the boom may point in a direction other than south. At Manuela site, the 180 deg end of the boom points up the glacier and a correction is added to the data during processing. At Byrd site the wind is usually out of the north so the boom was rotated 120 deg and the correction added during the data processing. The wind speed is determined from the aerovane tachometer voltage output as 0.0472 volt per meter per second. The aerovane tachometers are spun at 1800 rpm with a load of 1071.5 ohms and the output should be 9.20 +/-0.05 vdc.
Two additional wind sensors were used with AWS units for 1994. These were the Vaisala anemometer model WAA-15 and the R.M. Young wind monitor model 05103. The Vaisala WAA-15 was used as a backup sensor for measuring wind speed in the Adelie Coast area. It is a 3-cup opto-electronic anemometer. When rotating, the anemometer produces a pulsed output that is proportional to the wind speed. Rated accuracy is +/- 2% up to 75 m/s. The pulsed output was input into one of the digital counter channels for 5 seconds. This resulted in a calibration value of .293 m/s/bit.
The R.M. Young monitor 05103 also used a 1000 ohm potentiometer so that the wind direction was recorded identically wtih the Belfort/Bendix aerovanes. The wind speed was from the range of 0 to 1.0 volt full scale corresponding to 50 m/s. Thus the calibration for wind speed was a nominal .195 m/s/bit for the R.M. Young with +/- 1% up to 50 m/s.
The Vaisala HMP-35A humidity sensor output voltage varies linearly with relative humidity (U). The sensor is calibrated by placing it over saturated salt solutions with known relative humidities at room temperature: sodium chloride (U=75%), and lithium chloride (U=12%) are used. In addition, a dry inert gas, forced past the sensor, gives a 0% U, and the sensor output can be zeroed. Then, the gain setting can be set directly using a salt solution with a high relative humidity, such as sodium chloride. The resolution of the humidity sensor is about 1% and the drift is 2 to 3% per year in the field. The relative humidity data are not included on the summary pages but are included in the 3 hourly data sets.
Vertical Air Temperature Difference
Two junction thermocouples are used to measure the air temperature difference between 3 m and 0.5 m on the tower. The output is about 78 microvolts for 1.oC temperature difference between the junctions at 0.0 C, dropping to 60 microvolts at -80 C. Zero output is adjusted to 0.4 volts, so that 0 to 1 volt corresponds to a -6 C to +9 C range of air temperature differences between 3 m and 0.5 m. The resolution is 0.05 C. Calibration of the individual systems is done by applying known voltages to the amplifier input. The vertical temperature difference data are not included on the summary pages but are included in the 3 hourly data sets.