Accession Number:

ADA550135

Title:

Implementation of the Automated Numerical Model Performance Metrics System

Descriptive Note:

Memorandum rept.

Corporate Author:

NAVAL RESEARCH LAB STENNIS DETACHMENT STENNIS SPACE CENTER MS OCEANOGRAPHY DIV

Personal Author(s):

Report Date:

2011-09-26

Pagination or Media Count:

31.0

Abstract:

Known as AutoMetrics, this set of software modules was developed to provide model-observation and model-model comparison matchups and statistics to help modellers and forecasters assess the performance of ocean circulation models, primarily the Navy Coastal Ocean Model NCOM and the Hybrid Coordinate Ocean Model HYCOM. This system is fully automatic and gives users much flexibility in what model runs to be evaluated by this system, in real-time operations or for long-term historical model runs. The AutoMetrics system was the result of a project under the Rapid Transition Program to transition to NAVOCEANO software that provided information to the oceanographers so that they can assess the performance of numerical ocean prediction models run at NAVOCEANO. The Navy Coastal Ocean Model NCOM Rowley et al., 2002 Martin, 2000 was initially the primary target, but also the Hybrid Coordinate Ocean Model HYCOM Wallcraft, 2007 can be used in this system as well. In fact, any ocean circulation model output in the format described below can be used in AutoMetrics. Primarily ocean profiles of observed temperature and salinity are collected and matched with profiles of modeled same parameters corresponding to same location and time. This also applies to surface-only observations. The relationship between observations and model output consists of one-to-many, because from a given model domain and its several runs there exist multiple forecasts for a given time and place. To further complicate matters, the availability time of both observations and model output are considered arbitrary and the system must continually check for both to provide the most up-to-date matches in a timely manner. Ultimately a data base is built with a set of matches that can be analysed statistically in a variety of ways.

Subject Categories:

  • Physical and Dynamic Oceanography
  • Computer Programming and Software

Distribution Statement:

APPROVED FOR PUBLIC RELEASE