<p>The study investigates the nonlinear dynamical behavior of meteorological variables of a coastal region and its nearby sea surface temperature (SST) from January 2010 to February 2024. The scheme use largest Lyapunov exponents (LLE), correlation dimensions, Cao’s Nearest Neighbors method, Recurrence quantification analysis (RQA) and Kolmogorov-Smirnov (K-S) normality test to study the complexity in temperature and SST time series. In addition, the multivariate phase space reconstruction (PSR) is utilized to develop short term forecasting models. Early Warning Signal (EWS) indicators such as variance, serial correlation, skewness, and kurtosis are calculated for each datasets and parameterized into PSR to enhance predictive capabilities. Additionally, a wind direction index representing monsoonal effects, particularly changes associated with the East India Coastal Current (EICC), is incorporated. This enhances understanding of heatwave dynamics and land-ocean heat transfer, especially during the southwest monsoon. Next the occurrence of positive Lyapunov exponents confirmed the presence of deterministic chaos in Daily Temperature and SST, and indicating short-term predictability. The proposed chaotic system demonstrates a robust capability to forecast extreme weather events, providing valuable insights for climate resilience and disaster preparedness in coastal regions. The comprehensive analysis underscores the significance of incorporating multi-dimensional nonlinear dynamics and EWS indicators in climate modeling, offering a nuanced approach in understanding and predicting the complex system interactions between atmospheric and oceanographic variables in coastal areas. This analysis underscores the significance of nonlinear dynamics and the application of early warning system indicators in climate modeling, emphasizing the intricate links among atmospheric and oceanographic variables in coastal regions.</p>