Analog methods for tropical cyclone (TC) rainfall prediction compare current and past TCrelated information (e.g., atmospheric fields) to estimate rainfall. However, comparing spatially distributed data like atmospheric fields is challenging, as common metrics fail to capture multiple characteristics of the data simultaneously. For example, conventional cosine similarity excels at identifying changes in pattern similarity but falls short in capturing absolute magnitude similarities. To address this challenge, we provide a proof of concept for using a perceptual-based measure-the Structural Similarity Index (SSIM)-to identify analog TCs based on atmospheric field similarity. Our study demonstrates the effectiveness of SSIM in selecting analogous atmospheric fields for TC rainfall prediction in a basin, producing results comparable to existing studies. Compared with cosine similarity, SSIM proved more effective in selecting atmospheric field-based analogs, leading to improved basin rainfall prediction performance. This study not only underscores the effectiveness of perceptual similarity metrics in meteorological forecasting but also illustrates how image processing techniques can innovatively be applied in geosciences, treating spatial data as analogous to images.

There is a need to improve the prediction of tropical cyclone (TC) rainfall as climate change has led to increased TC rainfall rates. Enhanced reliability in predicting TC tracks has paved the way for statistical methodologies to utilize them in estimating current TC rainfall, achieved by identifying similar past TC tracks and obtaining their corresponding rainfall data. The widely used Fuzzy C Means (FCM) clustering algorithm, though popular, has limitations stemming from its clustering-centric design, hindering its ability to pinpoint the most appropriate similar TCs. Our study introduces the Sinkhorn Distance as a novel measure of TC similarity in rainfall prediction. Our findings indicate that the incorporation of Sinkhorn Distance significantly enhances the accuracy of TC rainfall predictions across WNP. When compared to the conventional approach using FCM, our Sinkhorn Distance-based methodology consistently yields better results, as demonstrated by metrics like RMSE and correlation coefficients. Collectively, the inclusion of Sinkhorn Distance stands as a valuable addition to our toolkit for discerning similar TC tracks, thus elevating the precision of TC rainfall predictions. With ongoing advancements in statistical and AI techniques, we anticipate even more refined approaches to further enhance our predictive capabilities. This study represents a leap forward in meeting the critical need for more accurate TC rainfall forecasts in the WNP Region.