Long-Term Typhoon Trajectory Prediction: A Physics-Conditioned Approach Without Reanalysis Data

TL;DR

The paper tackles long-term typhoon trajectory prediction without relying on reanalysis data by introducing LT3P, a physics-conditioned, real-time data-driven model that fuses ERA5-derived physics with Unified Model forecasts via cross-attention and a bias-correction mechanism. It employs a two-phase training regimen: Phase 1 pre-trains a physics-conditioned encoder on ERA5 to capture geopotential height and wind fields, while Phase 2 uses UM inputs and a bias-corrector to align UM-derived features with ERA5-like representations, enabling accurate trajectories up to 72 hours ahead. Across Western North Pacific typhoons, LT3P achieves state-of-the-art performance, outperforming both NWP ensembles and data-driven baselines, and it demonstrates clear benefits from physics-guided representations and real-time data. The PHYSICS TRACK dataset release further supports reproducibility and accelerates research in real-time, physics-informed climate forecasting and AI for disaster resilience.

Abstract

In the face of escalating climate changes, typhoon intensities and their ensuing damage have surged. Accurate trajectory prediction is crucial for effective damage control. Traditional physics-based models, while comprehensive, are computationally intensive and rely heavily on the expertise of forecasters. Contemporary data-driven methods often rely on reanalysis data, which can be considered to be the closest to the true representation of weather conditions. However, reanalysis data is not produced in real-time and requires time for adjustment because prediction models are calibrated with observational data. This reanalysis data, such as ERA5, falls short in challenging real-world situations. Optimal preparedness necessitates predictions at least 72 hours in advance, beyond the capabilities of standard physics models. In response to these constraints, we present an approach that harnesses real-time Unified Model (UM) data, sidestepping the limitations of reanalysis data. Our model provides predictions at 6-hour intervals for up to 72 hours in advance and outperforms both state-of-the-art data-driven methods and numerical weather prediction models. In line with our efforts to mitigate adversities inflicted by \rthree{typhoons}, we release our preprocessed \textit{PHYSICS TRACK} dataset, which includes ERA5 reanalysis data, typhoon best-track, and UM forecast data.

Figures (4)

• Figure 1: Visualization and comparison of the UM and ERA5 data at pressure levels of 250, 500, and 700 hPa, along with a difference map of two datasets. The UM forecast data is at a lead time of +72 hours, and the ERA5 data corresponds to that forecasted time. This analysis covers the Western North Pacific basin, with latitudes ranging from 0 to 59.75°N and longitudes from 100°E to 179.75°E.
• Figure 2: Overview of LT3P. In phase 1, the physics-conditioned model $R(\cdot)$ is trained for the weather forecasting task to encode information from $\Phi$, $\vec{W}_{u}$, and $\vec{W}_{v}$. Subsequently, in phase 2, using UM as the input, the typhoon trajectory prediction model $D(\cdot)$ is trained in conjunction with the bias-corrector $B(\cdot)$ to make accurate predictions.
• Figure 3: Qualitative analysis between LT3P and other trajectory prediction baselines. Note that the typhoon trajectory provided by JMA exists at 24-hour intervals, so the three coordinates have been linearly connected. Additionally, the error (in km) of the final coordinate is also indicated. The probability map is visualized using kernel density estimation (KDE).
• Figure 4: Zonal wind bias of the UM before and after bias correction with respect to ERA5.