Remote prediction of ground temperature in Australian soils and mapping its distribution

Abstract Earth-covered, or underground, housing is growing in popularity, particularly when designed on solar-passive principles. In Australia, no means of determining ground temperature data was available which could be used to assess site suitability for this type of building in feasibility studies of remote sites. Using long-term ground temperature records from 20 sample sites scattered throughout five states of Australia, an isothermal contour map is presented. This map shows the geographical distribution of a temperature differential which relates average ground temperature to average annual air temperature; in relation to average air temperature, the differential ranged from −1.5 to 5.0 K. Data from this map were tested against on-site ground temperature measurements taken throughout the five states from a further 73 bore-log readings of ground temperature at various depths. The results were sufficiently accurate to conclude that data from the map would be suitable for substitution of a particular function ( T m ) in an equation describing periodic ground temperature for the Australian continent. A computer program of this equation was used to extrapolate the ground temperature data to various depths relevant to earth-covered housing design. This would enable heat budget assessments to be made at the feasibility stage of the design process. The sampling data are interpolated spatially by computer program in terms of isothermal contours at 1°C intervals at depths of 1, 2, 3 and 500 cm for bare ground in full sun. The final results are presented as a set of maps. Various methods of assessment of ground temperature on an annual basis are described, utilising these maps in step-by-step procedures based upon the progressive testing of increments in soil roof-cover depth.