Altimeter assimilation in the OCCAM global model Part II: TOPEX/POSEIDON and ERS-1 assimilation

• Published in: Journal of marine systems Vol. 26; no. 3; pp. 323 - 347
• Main Authors: Fox, A.D, Haines, K, de Cuevas, B.A, Webb, D.J
• Format: Journal Article
• Language: English
• Published: London Elsevier B.V 01.11.2000; Elsevier
• Subjects: altimetry assimilation; Earth, ocean, space; Exact sciences and technology; External geophysics; hydrographic data; Marine; model errors; Other topics; Physics of the oceans; WOCE; WOCE; model errors; altimetry assimilation; hydrographic data; Satellite observation; Sea level; Altimetry; Hydrography; World ocean; Data assimilation
• ISSN: 0924-7963; 1879-1573
• DOI: 10.1016/S0924-7963(00)00044-0

Maps of combined TOPEX/POSEIDON (T/P) and ERS-1 altimeter sea level anomalies for 1993 are assimilated every 10 days into a high resolution global ocean model, OCCAM. The assimilated mean sea level is based on the model simulation, but modified in the Kuroshio and Gulf Stream based on a dynamic sea level from climatological hydrographic data. Sea level updates are accompanied by vertical thermocline displacement and geostrophic current adjustments away from the equator. Sea level errors were used to weight the sea level updates, with the model forecast errors calculated at low spatial resolution using an adaptive method based on the work of Dee [Dee, D.P., 1995. On-line estimation of error covariance parameters for atmospheric data assimilation, Mon. Wea. Rev. 123, pp. 1128–1145]. The modified mean sea levels are very successful at deflecting the western boundary currents to more realistic separation points which results in better surface temperature fronts and diagnosed air–sea heat fluxes, based on temperature relaxation. The altimeter anomalies are successfully assimilated and the model is able to propagate these between assimilation times, although with some degradation and damping of the smaller scale features. Predictions of sea level changes over 10-, 20- and 40-day periods are better than the corresponding persistence forecasts except for the Antarctic Circumpolar Current area where there is little difference. AVHRR and subsurface hydrographic and ADCP current data from WOCE cruises were used for validation, and the model run with altimeter assimilation shows significantly better agreement with these independent data sets than the model simulation, for all areas away from the equator.