Waikanae: excessive flood damage in 1998?

Abstract

In October 1998 two flood events with estimated return periods of 28 and 16 years occurred one week apart on the Waikanae River. The October 1998 flood events are the largest flood events on record. Major damage resulting from the October 1998 events was disproportionate to the size of the events. The disproportionate level of damage raised questions about the ability of the river and its protection works to deal with the flood events. In this thesis there is analysis of flow and rainfall data records and the physical aspects of the October 1998 events are examined. Relationships between rainfall and flow in the catchment are established and analysed with past events to conclude whether flood behaviour has changed in the catchment. This information is used to review the current management and sustainability of flood protection. The nature of the weather front in the October 1998 events caused by the orographic effect of the north-westerly airflow resulted in the rainfall reaching the lower parts of the catchment first (Warwicks station). The lag time is short in the Waikanae catchment and similarities exist between the responses of the flow and rainfall for the two events. Since lag times and distance to the river gauge do not appear to be related, it can be concluded that the response time in the Waikanae catchment from any rainfall gauge will be approximately 3-5 hours. Results of past events indicate that the same shaped hydrograph can occur as a result of quite different conditions. It is possible that the size of the 1998 event is too low and should be closer to a 50-year return period. It is suggested that the October 1998 flood did not cause disproportionate damage in relation to the estimated return period/event size but that estimates of the event sizes were too conservative, due to the lack of accurate data of the flow records for the Waikanae River. It is also possible that some flood protection management decisions have been influenced by lack of accurate data. If estimates of the 1998 event sizes were increased as recommended, this would decrease the expectations of the capabilities of the current flood protection works.