NICOP - Forecasting from the Deep Ocean to the Coast: Predictability of shelf circulation impacted by a Western Boundary Current
Abstract
This project will contribute to the ONR goal of understanding ocean processes at high spatial and temporal resolution as they impact naval operations through the development of high-resolution ocean modelling and prediction capability in a region of complex ocean" dynamics.While there is an increasing push towards providing forecasts that resolve finer scale features in coastal regions, the p"redictability of the coastal ocean is not yet well understood. Operational ocean forecasts available at present typically resolve the slowly evolving mesoscale circulation (with model resolutions of ~10km) and predictive skill has continued to increase through imp"rovements in modeling and data assimilation techniques. With improved model resolution and physics, model error increases more rapid"ly. This so-called ~curse of resolution~ stems from the fact that finer grid resolution and numerics increase the model error by resolving finer-scales of short period turbulence that have faster error growth.This project will aid our understanding of the predictability of fine-scale processes in a dynamic continental shelf and slope region impacted by an intense western boundary current (The East Australian Current). The results of this project are relevant to analogous WBCs e.g the Kuroshio in the South China Sea and the Gulf Stream adjacent to the US east coast. Building on a strong foundation of previous ROMS modelling work we will develop a 750m resolution shelf model to understand predictability of coastal dynamics. This study will address its objectives by performing the following tasks in the nested Coastal domain:1. Develop a high resolution data assimilative model: We will develop a comprehensive and carefully tuned 4D-Var data assimilative configuration for the coastal ocean. The Coastal model will be nested inside our existin"g Regional model 2-year reanalysis. Data streams that will be assimilated include satellite-derived SSH and SST, surface radial curr""ents from a HF radar array, and data from Argo floats, glider missions and moorings. The system will be carefully tuned to ensure ap"propriate estimates of the background and observation error covariances to ensure a high degree of optimality in the analysis. Data will be assimilated over a 2-year period with forecasts made out to 5 days following each assimilation window.2. Assess predictive skill: We will assess the predictive skill of the Coastal model compared to available observations for forecast periods of 5 days over the 2 years. The predictive skill will be compared to that of the Regional model in the coastal region.3. Identify model forcin"g errors: The 4D-Var assimilation technique determines increment adjustments to the initial state, boundary conditions, and atmosphe"ric forcing to produce an improved ocean state-estimate that better fits the observations. We will analyse these increments to provi"de insight into the forecast error sources that may originate from the initial conditions, boundaries and surface forcings.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Nov 03, 2017
- Source ID
- N629091712141
Entities
People
- Moninya Roughan
Organizations
- Office of Naval Research
- United States Navy
- University of New South Wales