FY 1990

Uncertainties in tropical Pacific Ocean simulations: The seasonal and interannual sea level response to three analyses of the surface wind field

Busalacchi, A.J., M.J. McPhaden, J. Picaut, and S. Springer

In Proceedings of the Western Pacific International Meeting and Workshop on TOGA COARE, ORSTOM, Nouméa, New Caledonia, 24–30 May 1989, 367–377 (1989)

The purpose of this study is to characterize differences in time/space structure present among conventional descriptions of the tropical Pacific surface wind field, and in turn, to quantify the impact of these differences on our ability to model the dominant wind-forced variability of the tropical Pacific Ocean on seasonal and interannual time scales. A linear, multiple vertical mode ocean model is used as a transfer function to determine the influence of three distinct surface wind stress products for the period 1979–1983. This five-year period was chosen for study because it encompasses three years of a fairly regular seasonal cycle leading up to the 1982–83 El Niño for which there are several coincident oceanic and surface wind data sets. The three different wind analyses used are the Florida State University subjective analysis, the University of Hawaii subjective analysis, and the Fleet Numerical Oceanography Center objective analysis. We examine first the three mean seasonal cycle solutions prior to El Niño which then serve as self-consistent bases for analyzing the significant anomalies about the mean in 1982–83. The impact of uncertainties in the forcing functions is discussed relative to the dominant seasonal and interannual scales of variability for the wind-driven oceanic response. On seasonal time scales, critical differences in the wind stress products, of order 0.2–0.4 dynes cm⁻², were in wind regimes of surface convergence and significant gradients such as the ITCZ and SPCZ. These uncertainties in the wind fields were manifested in model sea level solutions as 6–12 cm discrepancies near the NECC Trough and east of New Guinea. On interannual time scales, the influence of greater sampling along the major ship tracks was evident. Away from the major shipping lanes rms differences in the wind stress anomalies (1979–1983) about the mean seasonal cycle (1979–1981) reached up to 0.5 dynes cm⁻². The combined effect of these differences in the wind products resulted in 8–20 cm rms differences in the model sea level simulations. The largest of these discrepancies tended to exist at the terminus of equatorial wave characteristics, e.g. in the east along the equator and in the west off the equator.