Calibration of HEC-RAS models for rating curve development in semi arid regions of Western Australia

Ross Doherty - WA Dept of Water , 22 October, 2010

The commencement of hydrographic monitoring in the north of WA commenced in the 1960s when Commonwealth Government funding was abundant. For gauging stations constructed on large alluvial river channels, cableway infrastructure was erected to provide the means to measure river flow for the development of rating curves. Initial gauging programs enabled rating curves to be developed but the extent of the work was inadequate with some facilities being used in only one or two years before deteriorating along with the gauging program. Consequently rating changes were not detected and data integrity suffered.

In the early 1990s, regional hydrographic units were able to take advantage of computer software which employed traditional open channel hydraulic theory to produce water surface profiles and theoretical rating curves. The acceptance of the concept of modelling river channels to develop rating curves has increased gradually. In recent years, LiDAR and RTK technology have made channel geometry information more available and has resulted in the development of ratings for many gauging stations, which were either long overdue for rating revision, or required initial rating development.

Hydrographic intuition indicates that modelling results are credible. However, verification of theoretical rating curves is considered necessary, even though the methodology may be difficult and subjective. This paper addresses some of the methods available for calibrating theoretical models using readily available hydrographic data. By
presenting several case studies the results of theoretically developed rating curves are shown to be reliable with an acceptable level of uncertainty.

The focus on open channel modelling for developing primary rating curves for stations in the semi arid areas of WA has changed the operational focus of hydrographers. Strategic wet season gauging strategies will still be maintained but the expectations will be more realistic with gauging success viewed as an opportunistic bonus. The primary focus must become the monitoring of physical changes in the river channel profile and identifying changes in channel roughness. As the use and acceptance of theoretical modelling increases, the long neglected practice of rating curve maintenance will again become routine, and the integrity of flow data may return to an acceptable standard.