Abstract:
This paper presents the estimation of geographically centred and probabilistically correct areal
reduction factors (ARFs) from daily rainfall data to explain the unique relationship between
average design point rainfall and average areal design rainfall estimates at a catchment level
in the C5 secondary drainage region in South Africa as a pilot case study. The methodology
adopted is based on a modified version of Bell’s geographically centred approach. The
sample ARF values estimated varied with catchment area, storm duration and return period,
hence confirming the probabilistic nature. The derived algorithms also provided improved
probabilistic ARF estimates in comparison to the geographically and storm-centred methods
currently used in South Africa. At a national level, it is envisaged that the implementation and
expansion of the methodology will ultimately contribute towards improved ARF estimations at
a catchment level in South Africa. Consequently, the improved ARF estimations will also result in
improved design flood estimations.
