A Rating Curve Model Accounting for Cyclic Stage‐Discharge Shifts due to Seasonal Aquatic Vegetation. Issue 3 (1st March 2021)

• Record Type: Journal Article
• Authors: Perret, E.; Renard, B.; Le Coz, J.
• Is Part Of: Water resources research. Volume 57:Issue 3(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-03-01
• Subjects: aquatic vegetation -- Bayesian modeling -- hydrometry -- rating curve -- streamflow -- uncertainty; Hydrology -- Periodicals; 333.91
• DOI: 10.1029/2020WR027745
• Languages: English
• ISSNs: 0043-1397

Abstract

Abstract: Managing stage‐discharge relations at hydrometric stations affected by seasonal aquatic vegetation is challenging. Their ratings vary continuously in time so that establishing the streamflow time series is difficult. To the best of our knowledge, no rating curve model exists to deal with a vegetated channel control, only manual adjustment methods. To address this issue, a temporal rating curve model accounting for transient changes due to vegetation is developed. The model is built on basic concepts from open‐channel hydraulics and plant physiology. In the model, the flow resistance varies through time as a function of the plant development (growth and decay) and of the plant ability to reconfigure. Model parameters and uncertainty are estimated through Bayesian inference combining prior knowledge on the hydraulic controls at the station and observational data. In addition to the traditionally used stage‐discharge gaugings, the model allows the use of qualitative observations on the plant development state (e.g., no plant, growth, and decay). The model is tested and validated for a French hydrometric station where frequent gaugings (twice a month) and written comments about vegetation are available over more than 20 years. Relative errors between the simulated discharges and the observed discharges mostly range between ±20% for the temporal rating curve. In contrast, they are higher than ±50% with a standard model with no vegetation module. This case study highlights the importance of using observations about plant development to predict water discharge. We finally discuss some possible improvements and extensions of the model and recommend methods for data collection. Key Points: A rating curve model accounting for aquatic vegetation was developed and estimated through Bayesian inference Information on the vegetation development state should be used to estimate the rating curve Discharge prediction in presence of aquatic plants was improved using the new model (relative errors decrease from ±50% to ±20%)