Date of this Version
JOURNAL OF ATMOSPHERIC AND OCEANIC TECHNOLOGY VOLUME 18 1470-1484
Air temperature measurement has inherent biases associated with the particular radiation shield and sensor deployed. The replacement of the Cotton Region Shelter (CRS) with the Maximum–Minimum Temperature System (MMTS) and the introduction of Automated Surface Observing System (ASOS) air temperature observing systems during the NWS modernization introduced bias shifts in federal networks that required quantification. In rapidly developing nonfederal networks, the Gill shield temperature systems are widely used. All of these systems house an air temperature sensor in a radiation shield to prevent radiation loading on the sensors; a side effect is that the air temperature entering a shield is modified by interior solar radiation, infrared radiation, airspeed, and heat conduction to or from the sensor so that the shield forms its own interior microclimate. The objectives of this study are to develop an energy balance model to evaluate the microclimate inside the ASOS, MMTS, Gill, and CRS shields, including the interior solar radiation, infrared radiation, and airspeed effects on air (sensor) temperature under day and night conditions. For all radiation shields, the model air temperature for shield effects was in good agreement between shields while the uncorrected ‘‘normal operating’’ temperatures were more variable from shield to shield. The solar radiation loading ratio was dramatically increased with a corresponding increase in the solar elevation angle for all shields except the ASOS shield, and are ranked as Gill MMTS CRS ASOS. The daytime infrared radiation effects on air temperature were ranked as ASOS Gill MMTS CRS, but the nighttime infrared radiation effects were not so large and were uniformly distributed among negative and positive effects on air temperatures. For the nonaspirated radiation shields (MMTS, Gill, and CRS), increasing ambient wind speed improved the accuracy of air temperatures, but it was impossible to reach the accuracy claimed by manufacturers when the in situ measurements were taken under lower ambient wind speed (4 5 m s1).