Abstract:
Design point rainfall estimates assume a uniform distribution of rainfall over a
catchment and hence are only representative of a limited area. For larger areas,
Areal Reduction Factors (ARFs) are used to convert design point rainfall depths or
intensities to an average areal design rainfall depth or intensity for a catchmentspecific
critical storm duration and catchment area. The overall purpose of this
study is to develop an enhanced methodology to express the spatial and temporal
rainfall variability at a Quaternary Catchment (QC) level by means of
geographically-centred and probabilistically correct ARFs. The ARF values
presented in this study are based on observed daily rainfall data as extracted from
223 rainfall stations situated in the C5 secondary drainage region. The
methodology adopted is based on a modified version of Bell’s (1976)
geographically-centred approach. Individual sets of ARF values were derived for
each of the 23 QCs present in the C5 secondary drainage region by considering
various storm durations (1, 8, 16, 24, 72 and 168 hours) and corresponding
recurrence intervals (2, 5, 10, 20, 50, 100 and 200 years). The climatological
variability in the two tertiary drainage regions (C51 and C52) of the C5 was also
recognised by conducting separate regression analyses in each region. The
statistical differences in the regional ARF values highlight the presence of dominant weather types in each region. The statistical differences also confirm
that ARFs are influenced by different rainfall-producing mechanisms while not
being constant for various storm durations and exceedance probabilities or
recurrence intervals such as geographically-centred probabilistically correct ARFs.
It is recommended that the findings from this study be expanded to other regions
in South Africa, ultimately to devise both improved design rainfall and flood
estimates.
