The NSW Government's Flood Policy provides for:

• a framework to ensure the sustainable use of floodplain environments,
• solutions to flooding problems,
• a means of ensuring new development is compatible with the flood hazard.

Implementation of the Policy requires a four stage approach, the first of which is preparation of a Flood Study to determine the nature and extent of the flood problem.

The South Creek Flood Study was initiated as a result of flooding of roads and residential areas, most recently in March 2003 and April 1998. This report has been prepared by Webb, McKeown & Associates for Warringah Council.

The specific aims of the South Creek Flood Study are to:

• define flood behaviour in the South Creek catchment,
• prepare flood hazard and flood extent mapping,
• prepare suitable models of the catchment and floodplain for use in subsequent Floodplain Risk Management Studies and Creek Rehabilitation Studies and Plans.

Description of Creek Systems: South Creek has a catchment area of approximately 7.3 square kilometres and lies entirely within the boundaries of Warringah Council. It drains into Narrabeen Lagoon adjacent to Cromer Golf Course.

A large portion of the catchment has been developed for residential or commercial/light industrial purposes. The only area of natural bushland remaining includes the majority of the Wheeler Creek sub-catchment and land to the west of Cromer Golf Course.

There are seven road crossings in the study area:

• Willandra Road - Upper (South Creek),
• Mclntosh Road (South Creek),
• Alkira Circuit (South Creek),
• Willandra Road - Lower (South Creek),
• Carcoola Road (South Creek),
• Toronto Avenue (South Creek),
• Little Willandra Road (Wheeler Creek).

Cromer Golf Course is situated at the downstream end of the study area, adjacent to Narrabeen Lagoon. Further upstream (after Toronto Avenue) the floodplain becomes more restricted and fringed by residential developments. For the majority of its length the floodplain is largely in areas of open space, however some residential properties have been inundated by floodwaters in the past. A key feature of the creek near and downstream of the confluence with Wheeler Creek is the significant amount of vegetation in the channel.

Upstream of Little Willandra Road the creek becomes confined to a narrow floodplain fringed by open space. There are also several cascades and waterfalls.

The key phases of the South Creek Flood Study that have been undertaken are summarised below:

Review all available data, namely: * reports, photographs, Council records, * questionnaire survey of residents, interviews, * rainfall data from the Bureau of Meteorology and Manly Laboratory, * survey data - a comprehensive field survey was undertaken in 2002 and 2004, * obtain survey of historical flood levels.

Determine Approach: A rainfall-runoff approach was adopted due to the absence of long term historical flood data. This approach involved the setting up of two computer models - a hydrologic model to convert rainfall to runoff and a hydraulic model to convert the runoff to flows, flood levels and velocities.

Calibration to Historical Flood Levels: Following establishment of both models they were then used to simulate the historical floods of April 1998 and March 2003. This was achieved by ensuring the flood levels from the hydraulic model matched those actually recorded.

Determination of Design Flood Levels: Following calibration of the models design rainfall data were obtained from Australian Rainfall and Runoff (1987). These data were then input to the hydrologic/hydraulic models to determine design flood behaviour including the calculation of design flood levels. The lower parts of the creek are influenced by a combination of flows entering from the South Creek catchment and elevated water levels in Narrabeen Lagoon. The design analysis assumed that all culverts were blocked by debris. This approach is consistent with current best management practice following the floods in North Wollongong in August 1998. Sensitivity analyses were also undertaken. Due to the limited amount of calibration data available and in view of the sensitivity analyses, it is estimated that the order of accuracy is up to ±0.3 m for the reach downstream of Toronto Avenue and ±0.5 m for the reaches upstream. These orders of accuracy are typical of such studies and can only be improved upon with additional observed flood data to refine the model calibration.

Flood Problem Areas: The study has indicated that at all of the road crossings floodwaters will cross the road in the 1% AEP event (assuming that the culverts under the roads are blocked by debris). The yards of many properties adjoining the creek will also be inundated.

Outcomes: The main outcomes of this study are as follows:

• full documentation of the methodology and results,
• preparation of flood contour/hazard and extent maps for the South Creek catchment,
• the basis for a subsequent Floodplain Risk Management and Creek