Application of space-time rainfall patterns derived from atmospheric reanalysis data to define the joint probability of riverine catchment and local stormwater flooding.

Flood outcomes for some urban areas can be influenced by the joint probability of floods generated from relatively small local stormwater catchments and from a larger river system. Flooding is influenced by the Annual Exceedance Probabilities (AEPs) of rainfall on the stormwater and river system catchments and the space-time pattern of rainfall across the catchments, which influence the magnitude and timing of the floods. This was the situation in the Hutt River catchment (area of 636 km²), located near Wellington, New Zealand and the urban stormwater catchments of Lower Hutt and Petone, which drain to the Lower Hutt River. Gridded space-time rainfall data, derived from atmospheric reanalysis, provide a means of characterising the space-time patterns of rainfall, which can be applied to model the joint probability of riverine and stormwater catchment floods. The Bureau of Meteorology Regional Reanalysis for the Australian Region (BARRA-R) data set was used to identify space-time rainfall patterns of several large events, from around the central part of New Zealand. Events spanned burst durations between 3 and 48 hours. Space-time patterns from these events were translated to the study area and centred on either the Hutt River catchment or the Lower Hutt and Petone stormwater catchments. A RORB model was run, in a Monte Carlo joint probability framework, to calculate the AEPs of Hutt River floods that were coincident with stormwater flood events in Lower Hutt and Petone with defined AEP. The model was also run to calculate the AEPs of Lower Hutt and Petone stormwater floods that were coincident with riverine generated floods on the Hutt River with defined AEPs. The RORB model was calibrated to large gauged flood events and design floods from the RORB model were verified to flood frequency analysis at gauges on the Hutt River. [ABSTRACT FROM AUTHOR]