A novel oceanic index derived from sea surface height in the South China Sea

Abstract

In this thesis, we propose a novel oceanic index based on sea surface height anomaly (SSHA) to forecast rainfall in each region of Thailand. We first investigate lead-lag relationships between some of the well-known oceanic and atmospheric indices, commonly used for studying, and rainfall in each region. We next investigate spatial and temporal variations of SSHA variability in the South China Sea over the period of January 1993 to December 2022 using Empirical Orthogonal Function (EOF) analysis. Results of the EOF analysis show high and low oscillation patterns in the studied region along with their associated Principal Component (PC) time series. The 2nd PC time series shows significantly strong correlation with the rainfall in each region. However, the direct use of PC time series are computationally costly. Instead, we use SSHA information within the dominant regions, which show high variances observed in the 2nd mode of EOF analysis. Results of EOF lead a way to create a new oceanic index called the South China Sea Index (SCSI). We then explore the correlation between this new index and the rainfall in each region of Thailand. Results show that the SCSI has strong correlation with rainfall in each region. To forecast rainfall in each region in Thailand, we formulate models using the Seasonal Autoregressive Integrated Moving Averages with Exogenous Variables (SARIMAX) model by combining the oceanic and atmospheric indices used in studying climate change with a novel time series-index based on SSHA developed in this study. The results of this research will help predict monthly regional Thailand's rainfall, leading to a better understanding of rainfall variations in each region of Thailand.