Gradually Varied Flow modeling: How to choose a reach with maximum information content?

In any modeling exercise intended to imitate natural system, calibration process is considered to be an inevitable part as one cannot describe a natural system with complete certainty. The higher the number of tuning parameters, the more uncertain the model is. As such, should we be concern about the information content of data to be used for calibration and subsequent verification? The current research is devised to address this question in the context of Gradually Varied Flow (GVF) in open channel hydraulics. For this purpose, concurrent GVF and water depth sensitivity equations were solved numerically to identify the length of a reach with maximum information content invariably called influence distance. At first, the developed numerical scheme is validated by reproducing the previously limited analytical solutions available in the literature. Then, it was applied to a wide range of problems for which analytical solutions were not available. The results show how influence distance changes with different conditions of downstream and upstream flow and also different channel shapes. Next, the impact of dominant parameters was studied for different channels with regard to their practical implications. This could help practitioners to intelligently collect and calibrate natural channels in real world setting. Knowing where to collect the water level data would make computation of channel discharge more representative and accurate.