Local atmospheric circulations and the mesoclimate of Durban.

Abstract

Rapid urban and industrial growth along the Natal coast has occurred with little concern for characteristics of local climate. In general, industrial growth has taken place without taking into account the potential for pollution transportation during the winter season; urban areas have developed without consideration for the vital need in sub-tropical latitudes to maintain adequate mixing of the lower atmosphere. The apparent inadequacy in planning for climate has partly been due to a lack of understanding of the nature and characteristics of local wind systems. The motivation for this study stems , therefore, from the need to evaluate the influence of land-sea and topographically-induced wind systems upon certain aspects of weather and climate on the Natal coast. During summer, moist and relatively cool air is advected almost daily over the Natal coast by the sea breeze . Observations of spatial and temporal variations of these winds were made in the period 1963-69 from recording stations in the Durban area and along a 40-mile transect approximately normal to the coast at Durban. Comparable observations have not as yet been made in South Africa nor have there been a ttempts to examine the influence of sea breezes upon selected climate and weather phenomenon. In this latter connection the role of the sea br eeze is examined in two separate fields in this study, namely urban climatology and thunderstorm development. Climatic discomfort caused by the combined effect of high temperatures and humidities on the Natal coast, is at a maximum in urban areas . Since these areas are also highly populated , a need exists to evaluate the ability of the sea breeze to reduce urban temperatures by advection of cooler air over the area, by eddy diffusion of heat or by displacement of urban temperatures aw~ from poorly ventilated areas. Thus climatic discomfort in these areas may also be reduced. Detailed observations of the influence of the sea breeze, in particular, and wind in general, upon the spatial variation of temperature , humidity and human comfort were, therefore, undertaken in the Durban area. It is believed that this i s the first time an attempt has been made to describe the spatial distribution of a comfort index in an urban area. The results should be of interest to the a rchitect and town planner. The sea breeze, strengthened by gradient and valley Vlinds, is shown to penetrate at least 40 miles inland. In accordance with this , it is suggested that the inland invasion of marine air takes place in response to lowered pressure in adjacent inland areas in southern Natal and provides the moisture required to feed thunderstorms which develop in this area. These storms subsequently move towards the coast following the retreating convergence zone between winds with an onshore and offshore component of motion and reach Durban after sunset. Land breezes are best developed in winter. Observational techniques were similar to those employed in the sea breeze study and provide the only detailed examination of spatial and temporal variations of land breezes in South Africa. Until recently the nature and characteristics of topographically-induced winds in Natal were also a relatively unexplored feature of local atmospheric circulations. However, Tyson (1967) has examined local winds in certain Natal valleys and his prediction that mountain- plain winds reach the coast during the night is also examined in this study. While the relatively weak. and shallow land breeze may be of secondary importance when compared with the sea breeze, it would be a mistake to underestimate the importance of these winds upon certain aspects of coastal climate. In p~rticular the tendency for land breezes to initiate cloud and precipitation at night is examined while the potential for the transportation of atmospheric pollution by these winds is also briefly discussed. A study of the diurnal variation of precipitation at Durban reveals a high frequency of low intensi~ precipitation at night. The nature of this precipitation differs in both frequency and" amount from high intensity rainfall which is the product of thunderstorm and frontal activity. It is suggested in this study that the land breeze plays a dominant role in providing the buoyancy necessary to cause nocturnal drizzle precipitation from shallow stratus cloud. The thesis is divided into four parts. Part I examines details of the physiography, weather, climate, observati ~nal methods and analysis techniques. Because of the variability of weather on the coast of Natal that section is dealt with in more detail than would have been deemed necessary for a similar study inland. Without knowledge of the characteristics of the atmospheric circulation, the behaviour of local wind systems in relation to large scale systems could not be adequately discussed. In Part II land and sea breezes and topographically-induced wind systems are examined in detail. Emphasis is placed not only on the observational characteristics of these winds such as onset, depth, velocity gradient, surging, relation to gradient winds and dissipation but also on their relation to theoretical models of the relevant wind system. The influence of local wind systems upon selected climatic elements is dealt with in Part Ill. Characteristics of the diurnal variation of precipitation are examined in Chapter 6. Rainfall frequencies and amounts are shown to be highest in the first half of the night and a model is developed to explain the influence of land breezes and mountain-plain winds upon low intensity but high frequency nocturnal rainfall. High intensity but low frequency rainfall produced by thunderstorms also occurs soon after sunset and a further model is advanced to explain this phenomenon in relation to the sea breeze. The effect of air movement upon the spatial variation of temperature, humidity and a discomfort index in both summer and winter is examined in Chapters 7 and 8. The relationship between these elements has permitted the development of an empirical model to predict values across the city of mean midday summer and winter temperatures as well as discomfort index values. In Part IV the most significant aspeots relating to the study as a whole are summarised. In conclusion the potential f or the transport of pollutants by land breezes and topographically induced winds is briefly discussed as an aid to planning and siaple spatial models are given showing generalised air movement and ventilation characteristics in the Durban area and along the Natal coast.