Some aspects of surfaces of discontinuity

The following are the chief points contained in the various sections of the paper:‐ (1) Some factors tending to produce sharp fronts at the earth's surface are discussed. (2) Examples of soundings of upper air temperature through rainy fronts are given. It is found that the surface of discontinuity is normally smoothed through a layer about a kilometre thick, inversions being rare, especially in deep depressions. A considerable thickness of nimbus cloud is often formed below the discontinuity, probably owing to turbulence in air made damp by rain. It is thought that some rain belts, with associated fronts resembling “occlusions,” are developed in polar air and are not strictly “occlusions” at all. (3) Further evidence is given showing that warm sectors are not surface phenomena, but are of fundamental importance in determining the upper air conditions over depressions. It is argued that the fall of pressure in the warm sector in a deepening depression must be due to the spreading over of air from higher latitudes in the upper part of the troposphere and in the stratosphere. The corresponding feature of a developing anticyclone is a spreading over of tropical air at high levels. (4) Subsidence is discussed in a rough quantitative manner. In a developing anticyclone the downward movement is probably of the order of a kilometre per day at the 3‐kilometre level. The development of inversions with dry air above them (comprising a very large percentage of all inversions in the troposphere above 500 metres), is considered to be due to subsidence combined with turbulence up to a definite limit. Emphasis is laid on the importance of both cloud particles and precipitation in preventing dynamical warming at a fixed level by subsidence. (5) A list of 16 striking wind discontinuities observed by pilot balloons in the British Isles in the last nine years is given, with some remarks on their relation with fronts and surfaces of subsidence. (6) Turbulence at sloping surfaces of discontinuity is discussed on the basis of a criterion due to L. F. Richardson. (7) The over‐running of warm air by cold air is referred to, and it is thought that except near the ground this takes the form of continuous rather than of discontinuous motion. It is shown that a vertical front of any appreciable magnitude must be very much smoothed out. An appendix is added dealing with the combination of rotary and translatory motions. The general line of argument is that the more important pressure changes are due mainly to large‐scale horizontal movements at high levels, considered in conjunction with movements at lower levels. When depressions grow deeper the resulting convergence in the lower levels influences the subsequent behaviour of already existing fronts and in certain cases forms new fronts. When anticyclones develop, the subsidence causes inversions to form which are entirely different from frontal surfaces, and in addition the divergence may sharpen up fronts at the boundaries of the anticyclones.