Highlights Satellite observations constrain scales with a ¼ power point of 220 km. A decorrelation scale of 36 km provides lowest trajectory errors. Filtering unconstrained features reduces trajectory errors by 20%.
Abstract Observation space-time resolution limits the scales at which ocean forecast systems provide skillful information. The ocean processes of concern are mesoscale instabilities for which an ocean forecast system requires regular corrections of initial conditions to maintain skillful forecasts, and the observations considered are the regular satellite and in situ. Predominantly, the satellite altimeter constellation is the main observing system for this problem. We define constrained scales as those in which the forecast system has skill. The constrained scales are determined by successively filtering small-scale variability from 1 km resolution assimilative model experiments to reach a minimum error relative to ground truth data. Independent observations are from the LAgrangian Submesoscale ExpeRiment (LASER) consisting of over 1000 surface drifters persisting for three months in the Gulf of Mexico. We also vary the decorrelation scale of the assimilation system to determine the decorrelation scale that produces the smallest forecast trajectory errors. In present ocean forecast systems using regular observations, the constrained scales are larger than defined by a Gaussian filter with e-folding scale of 58 km or ¼ power point of 220 km. The decorrelation scale of 36 km used in the assimilation second order auto-regressive correlation function provides lowest trajectory errors. Filtering unconstrained variability from the model solutions reduces trajectory errors by 20%.
Constrained scales in ocean forecasting
Advances in Space Research ; 68 , 2 ; 746-761
2019-09-07
16 pages
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
Capacity-constrained traffic forecasting model
Online Contents | 2011
|Capacity-constrained traffic forecasting model
Taylor & Francis Verlag | 2011
|The National Centre for Ocean Forecasting
British Library Conference Proceedings | 2006
|Ocean wave forecasting using recurrent neural networks
Tema Archiv | 2006
|