This paper describes the urban heat island in Quebec City on a typical clear winter night. This is a follow-up to a study made on a summer night in August 1980; the reader will find a detailed description of the region in Leduc et al. (1980). On February 25, 1980 at 21: 30 EST, nearly 200 measurements of temperature were taken at preselected points by 17 mobile observers; a mini-sonde was launched near the Duberger meteorological tower while upper air data were obtained from a sonde launched at Valcartier CFB. Additional data were available from the five regular climatological stations in greater Quebec City and from 2 other thermographs in order to evaluate cooling rates.On the day of the 25th, a developing low pressure area over the eastern coast brought cold air from the northwest toward Quebec City. On that day, the maximum temperature was -1 "C while the minimum on the following night dropped to -19 "C. At the time of the experiment, the sky was clear and the winds were westerly at 1.5 m s-' at the surface and northwesterly at 6 m s-l at 122 m. A strong inversion was present in the lowest 200 m while upper inversions were noted on the upper air sounding.As noted in the summer experiment, cooling rates, according to the Oke et al. (1972) criteria, reflected a rural behaviour at all stations equipped with a thermograph. It is to be noted that no data on cooling rates are available for the highly urbanized sectors.One notes a relatively warm zone in the downtown area where temperature reached -8 "C; there is also a cold zone, at the same location as during the summer experiment. The mean temperature gradient between the urban and rural sectors was 5 "C, which is comparable with that observed in August 1979 (6 "C). Thus, the urban heat island does not seem to be more intense in winter than in summer. Relatively large areas of uniform temperature were detected at places where the summer experiment indicated sharp spatial contrasts; this could be attributed to the presence of snow on the ground. The great number of measurements taken in the more urbanized sector revealed a complex spatial arrangement of smaller heat islands which are separated by boulevards or open areas (covered with snow), For the city dwellers, the temperature differences resulted in relatively sharp dontrasts in clothing requirements for comfort.Urban morphology was analyzed and a multiple regression model was formulated between temperature on the one hand and altitude, land use and distance to the warmest point, on the other hand. It shows that, for both winter and summer the most important factor in explaining the heat island is distance to the warmest point while land use comes second; distance explains the form and land use explains the line structure of the heat island.
