Open AccessImproved fine‐scale transport model performance using AUV and HSI feedback in a tidally dominated system
Author(s) -
Hibler L. F.,
Maxwell A. R.,
Miller L. M.,
Kohn N. P.,
Woodruff D. L.,
Montes M. J.,
Bowles J. H.,
Moline M. A.
Publication year - 2008
Publication title -
journal of geophysical research: oceans
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.67
H-Index - 298
eISSN - 2156-2202
pISSN - 0148-0227
DOI - 10.1029/2008jc004739
Subject(s) - hindcast , bathymetry , environmental science , marine engineering , plume , underwater , scale (ratio) , remote sensing , downscaling , computer science , data collection , meteorology , geology , oceanography , engineering , cartography , precipitation , physics , statistics , mathematics , machine learning , geography
One of the challenges for model prediction and validation is providing them with data of appropriate spatial and temporal resolution. The maturation and increased application of autonomous underwater vehicles (AUVs) in aquatic environments allows systematic data collection on these model‐relevant scales. The goal of this study was to apply a fine‐scale circulation and transport model (Delft3D) to improve AUV mission planning and use data collected by the AUV to evaluate and improve model performance. A dye release was conducted in a tidally dominated embayment, and a planning phase model based on the best available data was used as a baseline for evaluation and for AUV mission planning (forecast). The planning phase model correctly predicted the general shape and direction of the dye plume and allowed for successful mission planning. Subsequently, bathymetry data collected by the AUV was incorporated into the model (hindcast), with temperature and salinity collected before the experiment. Comparisons with fluorometer measurements from the AUV indicate that the model effectively predicted the edges of the plume and centerline location. The location was also confirmed by remote sensing from an aircraft. Thermal stratification was found to be an important fate mechanism in the final model, and the results demonstrate the integration of observational data sets for small, short‐duration surface‐contaminant releases. This study highlights the strength of a phased, iterative approach with observation platforms and may serve as a guide toward improving the performance and evaluation of future coastal hydrodynamic and transport modeling efforts.