In March 2017, tropical cyclone Debbie brought widespread rainfall and flooding to the far north coast of New South Wales (NSW) including the Tweed, Lismore, Byron, Richmond Valley, Kyogle and Ballina local government areas. Five hundred SES flood rescues were carried out and in NSW six people lost their lives as a result of the flooding. In support of NSW emergency services, Manly Hydraulics Laboratory (MHL) deployed a field team to the flood affected area to obtain flood status checks for stations owned by the NSW Office of Environment and Heritage (OEH) during the event. This was followed up after the event to determine post flood status checks and debris line surveys.

There is often a high demand for water level and rainfall data during and immediately following a major flood event. This raw data is used for emergency planning, preparation and response; and it is often confirmed in field as the event unfolds by emergency personnel. The limitations of data which has not been quality controlled should be understood before being applied to other purposes such as modelling.

This paper identifies examples from the March 2017 Cyclone Debbie where the raw data was corrected for issues associated with an extreme flooding event. The magnitude of level adjustments can be significant, in the order of ±0.100 metres to more than 3 metres.

MHL applies a process of quality control and quality coding to its flood, estuary, ocean tide and rainfall monitoring stations that expresses the factors including uncertainty of the collected data. Understanding the differences between quality controlled and raw data during extreme events is critical to inform users of water data. The cost of getting data wrong can be substantial.