WeatherSyn: An Instruction Tuning MLLM For Weather Forecasting Report Generation

TL;DR

WeatherSyn introduces a new instruction-tuned multimodal large language model specifically designed for automated weather forecast report generation, outperforming existing models in accuracy and generalization.

Contribution

The paper presents the first instruction-tuning dataset and a specialized model for weather report generation, advancing automation in weather forecasting analysis.

Findings

WeatherSyn outperforms leading closed-source MLLMs in report accuracy.

The model demonstrates strong zero-shot transferability across regions.

The dataset covers 31 US cities and 8 weather aspects.

Abstract

Accurate weather forecast reporting enables individuals and communities to better plan daily activities and agricultural operations. However, the current reporting process primarily relies on manual analysis of multi-source data, which leads to information overload and reduced efficiency. With the development of multimodal large language models (MLLMs), leveraging data-driven models to analyze and generate reports in the weather forecasting domain remains largely underexplored. In this work, we propose the Weather Forecasting Report (WFR) task and construct the first instruction-tuning dataset for this task, named~\DatasetNameL, which covers 31 cities in America and 8 weather aspects. Based on this corpus, we develop the first model, \ModelNameL, specialized in generating weather forecast reports. Evaluation across multiple metrics on our dataset shows that \ModelNameL~ consistently outperforms leading closed-source MLLMs, particularly on structurally complex weather aspects. We further analyze its performance across diverse geographic regions and weather aspects. \ModelNameL~ demonstrates strong transferability across different regions, highlighting its zero-shot generalization capability. \ModelNameL~offers valuable insight for developing MLLMs specialized in weather report generation. .