WINTER MORNING AIR-TEMPERATURE

Results of temperature measurements, which may be applied ro inference of winter temperatures in data-sparse areas. are presented. The morni ng air temperatures during three winters were measured at 80 places in a 10 km x 30 km area along the Connecticut River. NOAA climatologies show this region to have complex spatial variation in mean minimum tem perature. Frequency analysis techniques were applied to evaluate the d ifferences in daily local temperature. Temperature lapse or temperatur e inversion in the study area was inferred from the difference of surf ace temperature measurements 100 and 300 m above river level. The freq uency of inferred temperature lapse and the inferred lapse rate dimini shed as snow cover increased. The frequency of inferred temperature in version and inversion strength increased as snow cover increased. When more than 20 cm of snow covered the ground, an additional surface inv ersion was frequent in the layer less than 100 m above river level, an d two-thirds of river level temperatures less than -20 degrees C occur red concurrent with these conditions. The daily temperature difference s at the individual points, with respect to a defined point, were logn ormally distributed. The magnitude and geometric standard deviation of temperature differences throughout the study area were larger on morn ings when inversion was inferred. With respect to topography, temperat ure differences and the geometric standard deviation of temperature di fferences were smaller along flats or among basins than along or atop slopes on mornings when inversion was inferred. It is proposed that so me meteorologically prudent inferences of surface temperature and near -surface temperature lapse or temperature inversion can be made for si milar data-sparse areas.