Analysis of the calibration and validation of hydrodynamic simulations with TELEMAC-3D model of the Patos Lagoon

Abstract

This work presents the analysis of the calibration and validation of the computational model for two configurations of the Jetties of the Rio Grande Barra for hydrodynamic studies of the Patos Lagoon and the continental shelf of southern Brazil. The model used was TELEMAC-3D, which solves the three-dimensional Navier-Stokes equations, considering the hydrostatic hypothesis, to describe the dynamics of free surface geophysical fluids. We developed two finite element triangular meshes, with about 75.000 elements and seven sigma levels. The numerical domain reaches depths up to 2.427 m, with fluvial and oceanic liquid boundaries. The horizontal turbulence model adopted was Smagorinsky and the vertical model of mixing length. Current velocity data obtained from December 2005 were used for calibration, varying the coefficient of wind friction, horizontal and vertical velocity diffusion, and the salinity tracer. The coefficient of wind influence was the factor that most influenced the model results. The calculated Relative Mean Absolute Error was 0.383 dn. for surface and 0.167 dn. for depth, rated good and excellent, respectively. For validation, we evaluated the model performance on reproducing the salinity behavior, related to the depth of the environment, in January 2017. The calculated Root Mean Square Error was 7.37 dn. and the Relative Mean Absolute Error was 0.228 dn., rating the model performance as good. These variances between metrics is uniformly acceptable for real models. Salt transport is a complex phenomenon and depends on both advective and diffusive transport. Thus, it is possible to conclude that he proposed computational model is able to reproduce a complex phenomenon reliably.