A quasi-geostrophic barotropic model, incorporating a planetary wave stabilization term, was integrated at the 500-mb level to obtain 12, 24, 36 and 48-hour height forecast charts. To achieve planetary wave stabilization, an approximation based on Phillips’ (1951) two-layer atmospheric model, was made to the horizontal divergence term. The integration was performed on the IBM 1410 computer located at McGill University. The calculations were done on a (49 x 38) rectangular, hemispheric grid network using initial 500-mb height data for February 15, 1959. The resulting forecast, for the particular day, showed a definite improvement over forecasts obtained with other barotropic models which did not include the stabilization term. The comparisons also demonstrated that the use of a large grid can result in an improved forecast.

The distributions of daily maximum and minimum temperatures recorded at McGill Observatory during the 88 year period from August 1874 to July 1962 have been analysed. The following main features emerge: 1) The frequency distributions (as judged by plots on arithmetic probability paper) fall into four well defined categories. A summer type corresponds to a normal distribution of temperatures. A winter type corresponds to a skew distribution, and transition types occur in the spring and the fall. 2) The interquartile range for both the maximum and minimum temperatures is highest in mid-winter and lowest in mid-summer, with secondary minima occurring in the spring and the fall.

Mean seasonal cross-sections along 80 degrees west were constructed from daily analysed data. By this procedure it was possible to formulate means and standard deviations with respect to various jet stream cores and selected temperatures as well as the usual geographical means. A set of such sections for the period June 1959 - May 1960 is presented and discussed.

The balance wind equation is derived, analysed and programmed for application in atmospheric models, and is compared and contrasted with other wind approximations. The mathematical concept of ellipticity of the equation is related to physical characteristics of the atmosphere. A review of three different numerical methods of solution is made; after simplification, one of these is adopted. The solution is performed on the Control Data G-20 computer located at the Central Analysis Office at Montreal Airport. The computations are carried out on a 1709 point octogonal grid covering the northern hemisphere from the Pole to 15°N latitude. The two 500 mb cases selected illustrate the efficiency of the balance equation in reducing the tendency for excessive anticyclogenesis frequently displayed in geostrophic barotropic forecasts, and show that it provides a good representation of the initial wind field.

The three-dimensional patterns of the large-scale divergence and vertical motion over North America are derived at twelve-hour intervals for a one-week period beginning May 21, 1962. The distribution of severe storms in the central and eastern United States during that period is examined and related to the patterns of vertical motion. The mean profile of large-scale vertical motion near early-morning storms is compared with that near late afternoon storms; the two profiles are found to be quite different. The morning storms correlate highly with the patterns of upward motion; the afternoon storms do not, presumably because other factors such as local heating are of comparable importance.

The simultaneous and continuous recording of two meteorological parameters, atmospheric pressure in a meso-scale network of microbarographs and precipitation by a 3-cm radar, permitted the study of two pressure phenomena (jump line and pressure dome) in relation to precipitation. The jumps were found to be of two basic types, one followed by a pressure dome and associated with severe weather, and the other showing as a step on the barograph traces and not associated with weather. The pressure domes were found to be closely associated with precipitation, some starting with, some developing towards and others intensifying with the radar echoes. In a preliminary study, it was shown how valuable information about pressure phenomena can be obtained from data (wind, pressure, precipitation) collected at one single station, rather than throughout a network.

Outgoing energy budgets for the sea surface are calculated for various circulation types over the Norwegian Sea, based on daily weather maps of January 1959-62. After a discussion of the various equations used, an analysis of the different energy terms is given, calculated from the available synoptic and upper air observations. The results are presented for each term of the energy budget and for the net outgoing energy flux, in the form of maps. They are then analysed for the main types along an air trajectory; a comparison between the types is made, using areal and latitudinal means of the individual energy budget terms and the net outgoing energy. In conclusion, the applicability of the present investigation and its implications are discussed.

Vertical profiles of the total precipitation in the area scanned by a radar have been obtained for the first time. Profiles for summer shower situations commonly show an accumulation of precipitation aloft, at a height which is observed to increase with the available energy of convection. This observation has been combined with calculations by ether researchers to give a three-fold relation between mean updraught speed, height of accumulation and energy of convection. Profiles for continuous rain are quite distinct from those of showers, generally showing the greatest amount of precipitation at the base of the storm. Profiles form a useful digest of the radar data for the synoptic meteorologist. The essential data can be reduced to a small array of digits, which are economically transmitted and readily interpreted to give insight into the precipitation pattern and the accompanying convection.

The utilization of radar as a method of observing clouds and cloud patterns has been most beneficial in the quest for an understanding of cloud physical processes, because radar is capable of "seeing" into the central parts of precipitating clouds, and can reveal the distribution and development of the precipitation patterns within them. One of the first extensive programs that used radar as a cloud physical research tool was the Thunderstorm Project (1949), which was carried out in Florida and Ohio in 1946 and 1947. Radar was used in this project to locate the levels of formation of first radar clouds or first echoes, and to study the vertical growth and horizontal and vertical extent of thunderstorms.

In an attempt to obtain direct evidence on the reflectivities and transmissions of thin high-level clouds in the infra-red, measurements of radiation at the surface are being made as part of a continuing study. The present work is a theoretical study of the results which one may expect to obtain for various typical atmospheric conditions. The conclusions reached are that the effects of carbon dioxide absorption and solar radiation are negligible, as are fluctuations in the ozone content. The effect of water vapour is important but can be determined from the routine radiosonde soundings. [...]