Regional flood prediction equations are generally developed based on the recorded streamflow data in medium- to large-sized catchments, but these equations are often applied in practice to very small catchments. Since there is little/no recorded streamflow data available for very small catchments, the applicability of the developed regional flood prediction equations to these catchments cannot be verified directly. The empirical observations reveal that smaller catchments produce "steeper flood frequency curves" than larger catchments given all the flood generation factors remaining the same.
This paper uses data from 429 catchments from the states of Victoria, New South Wales and Queensland to examine the effects of catchment size on flood quantile estimates. An empirical scale correction factor is proposed that can partially account for the effects of decreasing catchment size on flood quantile estimates. Independent testing using 32 catchments shows that the proposed method provides reasonable results for the catchments as small as 2.3 km2, but its applicability to very small catchments (ie. smaller than 2.3 km2) cannot be verified due to the unavailability of recorded streamflow data.