The study examines the ability to predict extreme rainfall days (ERDs) in Myanmar, a country frequently affected by severe flooding due to intense rainfall. We used a physics-based model combined with real-world data to forecast ERDs based on the atmospheric ERA5 reanalysis data and records rainfall from 79 different stations in Myanmar. The findings reveal that ERDs are more common in specific regions of Myanmar during certain seasons and are associated with particular atmospheric conditions. In June and July, El Niño's rapid growth phase—which was linked to a large anticyclone anomaly over the Indian Ocean—caused an increase in ERDs in central Myanmar. The region experiences higher rainfall as a result of this anomaly's enhanced moisture transfer. Similar to this, a zonal SLP dipole in eastern Myanmar contributes to higher ERDs from April to September. Additionally, the study finds indicators of possible causation, such as SST anomalies and the SIO dipole index, that are precursors to higher ERDs. Using these physical indicators, prediction models with strong TCC values for monsoon core region and northern Himalayas foothill region of Myanmar at different times demonstrated notable accuracy. The study finds a strong correlation between summer ERDs and monsoon rainfall, indicating that predicting mean rainfall accurately is essential for forecasting ERDs. To comprehend and forecast future changes in these relationships, more research is necessary due to the non-stationarity of predictor-predictand interactions. The discovery clarifies important atmospheric variables that impact extreme rainfall episodes in Myanmar and advances our understanding of these occurrences.