Measurements of natural airflow within a Stevenson screen, and its influence on air temperature and humidity records

Climate science depends upon accurate measurements of air temperature and humidity, the majority of which are still derived from sensors exposed within passively-ventilated louvred Stevenson-type thermometer screens. It is well-documented that, under certain circumstances, air temperatures measured within such screens can differ significantly from ‘true’ air temperatures measured by other methods, such as aspirated sensors. Passively-ventilated screens depend upon wind motion to provide ventilation within the screen, and thus airflow over the sensors contained therein. Consequently, instances of anomalous temperatures occur most often during light winds when airflow through the screen is weakest, particularly when in combination with strong or low-angle incident solar radiation. Adequate ventilation is essential for reliable and consistent measurements of both air temperature and humidity, yet very few systematic comparisons to quantify relationships between external wind speed and airflow within a thermometer screen have been made. This paper addresses that gap by summarising the results of a three month field experiment in which airflow within a UK-standard Stevenson screen was measured using a sensitive sonic anemometer, and comparisons made with simultaneous wind speed and direction records from the same site. The mean in-screen ventilation rate was found to be 0.2 m s-1 (median 0.18 m s-1), well below the 1 m s-1 minimum assumed in meteorological and design standard references, and only about 7% of the scalar mean wind speed at 10 m. The implications of low in-screen ventilation on the uncertainty of air temperature and humidity measurements from Stevenson-type thermometer screens are discussed, particularly those due to the differing response times of dry- and wet-bulb temperature sensors, and ambiguity in the value of the psychrometric coefficient