Evaluating the Joint Effect of Tropical and Extratropical Pacific Initial Errors on Two Types of El Nino Prediction Using Particle Filter Approach

The accuracy of different types of El Nino-Southern Oscillation (ENSO) predictions is sensitive to initial errors in different key areas of the Pacific Ocean. To improve the accuracy of the forecast, assimilation techniques can be utilized to eliminate these initial errors. However, limited studies have measured the extent to which assimilating ocean temperature data from different key regions in the Pacific Ocean can enhance two types of ENSO predictions. In previous research, three critical regions were identified as having initial errors in ocean temperature most interfering with two types of El Nino predictions, namely the North Pacific for Victoria Mode-like initial errors, the South Pacific for South Pacific Meridional Mode-like initial errors, and the subsurface layer of the western equatorial Pacific. Based on these initial error patterns, we quantified the effect of assimilating ocean temperature observation datasets in these three key regions using the particle filter method. The result indicates that ocean temperature initial accuracy in the tropical western area near the thermocline region is important for improving the prediction skill of CP-El Nino compared with the other two sensitive areas. However, three key areas are all important for EP-El Nino predictions. The most critical area varies among different models. Assimilating observations from the north and south Pacific proves to be the most effective for improving both types of El Nino predictions compared to the other two areas' choices. This suggests that the initial accuracy of ocean temperature in these two regions is less dependent on each other for enhancing El Nino predictions. Additionally, assimilating observations from all three sensitive areas has the best results. In conclusion, to enhance the accuracy of two types of El Nino predictions, we need to ensure the initial accuracy of ocean temperature in both tropical and extratropical regions simultaneously.

Automated summary

The accuracy of ENSO predictions is influenced by initial errors in different key areas of the Pacific Ocean. Assimilation techniques can be used to improve the accuracy by eliminating these errors. However, there have been limited studies on the impact of assimilating ocean temperature data from different regions on ENSO predictions.