A Methodology to Reduce Uncertainties in the High-Flow Portion of a Rating Curve

Abstract. Streamflow monitoring typically relies on the establishment of a rating curve that describes the relationship between stage and streamflow. This rating curve is based on direct and indirect measurements of discharge at various stages and application of open-channel hydraulics principles. Potential measurement errors increase with out-of-bank flow conditions because of difficulties in measuring flow velocity and flow depth during increased stages. With increasing frequency of large rain events attributed to climate change, it becomes critical to have an accurate rating curve for flood flows. This article presents techniques that, when used concurrently, can help identify errors and reduce uncertainty in the rating curve. They include commonly used techniques such as comparison of mean annual discharge per unit drainage area with those from nearby watersheds having similar characteristics. In addition, new techniques are proposed based on the hydrologic budget of a storm event and the analysis of the change in flow relative to time and stage. The methodology was tested using flow data from the 72 km² Goodwater Creek Experimental Watershed in northeast Missouri. Results showed that the original rating curve overestimated flow values by 30% for out-of-bank conditions. A revised rating curve was proposed based on channel conveyance. This revised rating curve satisfied the requirements of water balance and change in flow relative to time and stage. The resulting annual flows were reduced by up to 30%, better matching those of nearby watersheds. These techniques are useful for retrospective analysis of flow data series and for identifying rating curve inconsistencies.