User-based skill assessment techniques for operational hydrodynamic forecast systems

Abstract

NOAA’s National Ocean Service (NOS) applies hydrodynamic models for the development, transition and implementation of operational forecast systems in US estuaries, ports, lakes and the coastal ocean. These models and systems have applications in the support of safe and efficient marine navigation and emergency response, as well as marine geospatial and ecological applications. As part of the forecast system development, skill assessment is required to evaluate each model system. A forecast model has to be ensured that it has been developed and implemented in a scientifically sound and operationally robust way, and its performance should meet or exceed target benchmarks before its products are approved for release to the public. This paper discusses a new approach for the development of standards and procedures for skill assessment of operational forecast systems, which are designed primarily to support marine transportation and emergency response. Accordingly, this skill assessment focuses on variables, statistics and targets that are of the most use to these communities. The components of skill assessment include: data analysis techniques, definition of model run scenarios and statistical variables, and NOS target values and acceptance criteria. In addition, a large suite of skill assessment software has been developed to compute the new skill assessment scores using data files containing modelled variables and observed data. This software ingests model forecasts and observational data, fills in gaps, creates time series, extracts skill-specific parameters, and computes the skill scores. The software has been tested and applied to all NOS operational forecast systems. © 2010 2010 Taylor and Francis Group LLC.