An improved method for estimating surface humidity from daily minimum temperature

Minimum daily air temperature (T-n) is often used as a surrogate for mean daily dew point (T-d,T-day) to estimate near-surface humidity. This method is unreliable under arid conditions where nightly minimum temperatures may remain well above the dew point. Daily meteorological data from 52 weather stations within the continental US and Alaska were evaluated. Daily differences between T-d,T-day and T-n were large in arid climates, with corresponding vapor pressure differences averaging between 0.8 and 1.2 kPa on an annual basis. Sites with semi-arid characteristics showed a large degree of seasonality in the results with average vapor pressure differences ranging from 0.1 to 0.6 kPa between winter and summer months, Sites in other climate regimes generally showed small differences between T-n and T-d,T-day over the entire year, corresponding to average vapor pressure differences of less than 0.3 kPa. An empirical model was developed to improve the accuracy of T-n based humidity estimates using daily air temperature, annual precipitation and estimated daily potential evapotranspiration. The model reduced humidity estimation errors by up to 80% at semi-arid and arid sites and had minimal effects when T-n based humidity estimates were relatively accurate. The ratio of annual precipitation to estimated annual evapotranspiration was useful for distinguishing sites where T-n based humidity estimates were relatively accurate from sites where estimation errors were large. The results of this investigation provide a simple, more accurate method than T-n for estimating humidity in arid and semi-arid regions using general weather station data. (C) 1997 Elsevier Science B.V.