Effective generation mechanisms of tropical instability waves as represented by high-resolution coupled atmosphere-ocean prediction experiments

Cusp-shaped fluctuations of the sea surface temperature (SST) front in the tropical Pacific, now known as tropical instability waves (TIWs), were discovered by remote sensing in the 1970s. Their discovery was followed by both theoretical and analytical studies, which, along with in situ observations, identified several possible generation mechanisms. Although modeling studies have shown that TIWs strongly influence the heat budget, their influence on local variations of realistically initialized predictions is not yet understood. We here evaluate a series of medium-range (up to ~ 10 days) coupled atmosphere-ocean predictions by a coupled model with different horizontal resolutions. Observational SST, surface wind stress, heat flux, and pressure data showed that representation of temporally and spatially local variations was improved by resolving fine-scale SST variations around the initialized coarse-scale SST front fluctuations of TIWs. Our study thus demonstrates the advantage of using high-resolution coupled models for medium-range predictions. In addition, analysis of TIW energetics showed two dominant sources of energy to anticyclonic eddies: barotropic instability between equatorial zonal currents and baroclinic instability due to intense density fronts. In turn, the eddy circulation strengthened both instabilities in the resolved simulations. This revealed feedback process refines our understanding of the generation mechanisms of TIWs.

Tropical instability wave modulation of chlorophyll-a in the Equatorial Pacific

The global daily gap-free chlorophyll-a (Chl-a) data derived using the data interpolating empirical orthogonal functions (DINEOF) technique from observations of the Visible Infrared Imaging Radiometer Suite (VIIRS) in 2020 and the in situ measurements at the Tropical Ocean Atmosphere (TAO) moorings are used to characterize and quantify the biological variability modulated by the tropical instability wave (TIW). Our study aims to understand how ocean physical processes are linked to biological variability. In this study, we use the TAO in situ measurements and the coincident VIIRS Chl-a data to identify the mechanism that drives ocean biological variability corresponding to the TIW. Satellite observations show that the TIW-driven Chl-a variability stretched from 90°W to 160°E in the region. The enhanced Chl-a pattern propagated westward and moderately matched the cooler sea surface temperature (SST) patterns in the Equatorial Pacific Ocean. In fact, the Chl-a variation driven by the TIW is about ± 30% of mean Chl-a values. Furthermore, the time series of Chl-a at 140°W along the equator was found to be in phase with sea surface salinity (SSS) at 140°W along the equator at the TAO mooring since late May 2020. The cross-correlation coefficients with the maximum magnitude between Chl-a and SST, Chl-a and SSS, and Chl-a and dynamic height were –0.46, + 0.74, and –0.58, respectively, with the corresponding time lags of about 7 days, 1 day, and 8 days, respectively. The different spatial patterns of the cooler SST and enhanced Chl-a are attributed to the phase difference in Chl-a and SST. Indeed, a Chl-a peak normally coincided with a SSS peak and vice versa. This could be attributed to the consistency in the change in nutrient concentration with respect to the change of SSS. The vertical distributions of the temperature and salinity at 140°W along the equator reveal that the TIW leads to changes in both salinity and nutrient concentrations in the sea surface, and consequently drives the Chl-a variability from late May until the end of the year 2020.

Purely satellite data-driven deep learning forecast of complicated tropical instability waves

Forecasting fields of oceanic phenomena has long been dependent on physical equation-based numerical models. The challenge is that many natural processes need to be considered for understanding complicated phenomena. In contrast, rules of the processes are already embedded in the time-series observation itself. Thus, inspired by largely available satellite remote sensing data and the advance of deep learning technology, we developed a purely satellite data-driven deep learning model for forecasting the sea surface temperature evolution associated with a typical phenomenon: a tropical instability wave. During the testing period of 9 years (2010-2019), our model accurately and efficiently forecasts the sea surface temperature field. This study demonstrates the strong potential of the satellite data-driven deep learning model as an alternative to traditional numerical models for forecasting oceanic phenomena.