A Note on NCOM Temperature Forecast Error Calibration Using the Ensemble Transform
Abstract
During the MREA07 trial, off the NW coast of Italy in the late spring and summer of 2007, Navy Coastal Ocean Modeling (NCOM) multiple nests free-run ensembles were made available in real-time for the LASIE07 and BP07 events and a fairly complete set of observations were collected inside the inner nests domains. This note addresses the problem of predicting NCOM local unbiased 0-24 h forecast errors by perturbing a limited number of possible error sources through Monte-Carlo simulations, without local data assimilation. It discusses preliminary results using the Ensemble Transform [Bishop. C.H., and Toth, Z., 1999: Ensemble transformation and adaptive observations. Journal of the Atmospheric Sciences. 56.1748-1765) to calibrate the ensemble spread by adjusting its characteristics (spread-skill relationship and magnitude) to an observed or pre-estimated error field. A small (10 members) ensemble of free runs was used for water column temperature forecast Root Mean Square (RMS) error prediction. After being post-processed they were compared with observed errors and those estimated using time variability as an error proxy. The ensemble runs were generated through atmospheric forcing perturbations using the space-time deformation method as proposed by [Hong. H.X., Bishop, C. 2007. Ensemble and probabilistic forecasting. IUGG XXIV General Assembly 2007. Perugia, Italy, 2-13 July], keeping independent initial conditions. Because at the starting time all runs shared the same IC the ensemble was run for roughly two weeks for spinning up and then used during the following 10 days for data comparisons, during which the ensemble spread did not diverge and was consistent with the observed dynamics.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2009
- Accession Number
- ADA511322
Entities
People
- Clark D. Rowley
- Elaina Rodriguez
- Emanuel Coelho
- Germana Peggion
- Gregg A. Jacobs
- Richard Allard
Organizations
- United States Naval Research Laboratory