On the study of shelter effect with sheltered (screened) meteorological sensors

Abstract

Temperature and humidity measured by means of sensors placed in a Stevenson-type screen were compared with similar measurements made with sensors placed in the open, in wind-sheltered, and exposed plots of irrigated dry beans. The Stevenson-type screen introduced errors into the measurement of air temperatures and humidities. The errors were such as to underestimate the influence of wind shelters. Temperature in the screens was higher than in the open in either plot during the average day. Nocturnal temperatures were nearly identical in screened and open air in the exposed plot. In the wind-sheltered screen, however, nocturnal temperatures were lower than in the open. Vapor pressure was always greater throughout the day and night in the screens in either plot. “Apparent” shelter-effect (difference between screened sensors) and “real” shelter effect (difference between open sensors) were evaluated. Normally the daytime elevation of temperature in the wind-sheltered plot was underestimated by screened sensors. The nocturnal depression of temperatures in the wind-sheltered plot was exaggerated by the screens. “Apparent” shelter effect on vapor pressure and relative humidity was smaller than the “real” effect at all times. On calm days the screens may cause shelter effect on temperature to be underestimated and have little effect on humidity shelter effect. On windy days temperature shelter effect may be exaggerated slightly while humidity shelter effect is considerably underestimated by the screened sensors.