Innovative Approaches for Geometric Uncertainty Quantification in an Operational Oil Spill Modeling System

Abstract

Reliable and rapid real-time prediction of likely oil transport paths is critical for decision-making from emergency response managers and timely clean-up after a spill. As high-resolution hydrodynamic models are slow, operational oil spill systems generally rely on relatively coarse-grid models to provide quick estimates of the near-future surface-water velocities and oil transport paths. However, the coarse grid resolution introduces model structural errors, which have been called “geometric uncertainty”. Presently, emergency response managers do not have readily-available methods for estimating how geometric uncertainty might affect predictions. This research develops new methods to quantify geometric uncertainty using fine- and coarse-grid models within a lagoonal estuary along the coast of the northern Gulf of Mexico. Using measures of geometric uncertainty, we propose and test a new data-driven uncertainty model along with a multi-model integration approach to quantify this uncertainty in an operational context. The data-driven uncertainty model is developed from a machine learning algorithm that provides a priori assessment of the prediction’s confidence degree. The multi-model integration generates ensemble predictions through comparison with limited fine-grid predictions. The two approaches provide explicit information on the expected scale of modeling errors induced by geometric uncertainty in a manner suitable for operational modeling.

A bridge between invariant dynamical structures and uncertainty quantification

“…There is much interest in uncertainty quantification involving trajectories in ocean data sets. Recent efforts in this direction are [13,14,15,16]. Vieira et al [16] have developed a clustering method to partition the space of trajectory data sets into distinct flow regions.…”

“…This method contains free parameters and they use uncertainty quantification to assess the parameter dependence on the partitions they obtain. The work [15] discusses approaches for uncertainty quantification from a geometrical perspective, and uncertainty quantifiers based on distances are proposed, but no connections are proposed with invariant dynamical structures. Similarly in [14] uncertainty quantifiers based on a descriptive statistics of distances between modelled trajectories and observations are used, but no relations are presented between these and the invariant dynamical structures.…”

A survey on predicting oil spills by studying its causes using deep learning techniques

“…Diagnosis of offshore gas leak tested in this research. In [34] New methods for quantifying geometric uncertainty were developed using fine-and coarse-grid models in the lake's mouth along the north shore of the Gulf of Mexico. We proposed and tested a new data-driven ambiguity model alongside a multiple models integration approach to measure this ambiguity in a working sense by using geometric uncertainty procedures.…”

Numerical Study on the Influence of Model Uncertainties on the Transport of Underwater Spilled Oil

“…Li et al [ 15 ] investigated the influence of the error in ocean current on the three-dimensional transport of the underwater spilled oil and gave a possible interpretation. Feng et al [ 22 ] proposed a data-driven uncertainty model and a multi-model integration method to operationally quantify the effect of hydrodynamic model structural errors introduced by coarse grid solution on the prediction of oil transport paths. Yang et al [ 6 ] provided a review of the decision support tools for oil spill response, and pointed out that it was necessary to deeply investigate the modeling oil spill because of the uncertainties in the background environment field.…”