A study of residential areas done by Hoyt (1939) in the North American context concluded that the land use pattern was not a random distribution, nor sharply defined rectangular areas or concentric circles, but rather sectors. Thus, the effect of direction and time was added to the effect of distance.
(click on image to read more)
The model assumes a relationship between the socio-economic status (mainly income) of households and the distance from the CBD. The further from the CBD, the better the quality of housing, but the longer the commuting time. Thus, accessing better housing is done at the expense of longer commuting times (and costs).
Counterurbanisation, pattern of land use, sustainable living - recap these and other information from the links below
Settlements were built in sheltered areas such as valleys or in bays on coastal locations. Aspect (the direction in which a settlement faces) was also an important consideration. Settlements would be more ideally located on south-facing slopes in the Northern Hemisphere and on north-facing slopes in the Southern Hemisphere.