Air temperature measurement errors in naturally ventilated radiation shields

Two sources of systematic errors are considered for estimating air temperature. The first source is ambiguity of the definition of the standardized measurement height over vegetated surfaces of varying heights. Without such a standardization, evaluation of the horizontal air temperature gradient is contaminated by the vertical variation of air temperature. This error is generally small in daytime unstable conditions, but increases with increasing stability at night. In an attempt to reduce such error, the use of the zero-plane displacement height for standardizing the measurement height is proposed. The second source of systematic errors is radiative forcing on the sensor-shield systems. A series of experiments is performed over a grass field to investigate the radiatively induced error in the air temperature estimate by the Onset HOBO Pro thermistor in a naturally ventilated multiplate shield. The magnitude of this error is estimated by comparing air temperature measurements by a platinum resistance temperature detector (RTD) sensor in a mechanically aspirated shield. In contrast to the errors resulting from the first source, the radiatively induced error increases with increasing instability. An empirical model is developed for correcting the radiatively induced temperature error using information on wind speed and net or shortwave radiation. The robustness of the model is examined with independent data.