Local Off-Grid Weather Forecasting with Multi-Modal Earth Observation Data
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2024
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910164137803776 |
|---|---|
| author | Yang, Qidong Giezendanner, Jonathan Civitarese, Daniel Salles Jakubik, Johannes Schmitt, Eric Chandra, Anirban Vila, Jeremy Hohl, Detlef Hill, Chris Watson, Campbell Wang, Sherrie |
| author_facet | Yang, Qidong Giezendanner, Jonathan Civitarese, Daniel Salles Jakubik, Johannes Schmitt, Eric Chandra, Anirban Vila, Jeremy Hohl, Detlef Hill, Chris Watson, Campbell Wang, Sherrie |
| contents | Urgent applications like wildfire management and renewable energy generation require precise, localized weather forecasts near the Earth's surface. However, forecasts produced by machine learning models or numerical weather prediction systems are typically generated on large-scale regular grids, where direct downscaling fails to capture fine-grained, near-surface weather patterns. In this work, we propose a multi-modal transformer model trained end-to-end to downscale gridded forecasts to off-grid locations of interest. Our model directly combines local historical weather observations (e.g., wind, temperature, dewpoint) with gridded forecasts to produce locally accurate predictions at various lead times. Multiple data modalities are collected and concatenated at station-level locations, treated as a token at each station. Using self-attention, the token corresponding to the target location aggregates information from its neighboring tokens. Experiments using weather stations across the Northeastern United States show that our model outperforms a range of data-driven and non-data-driven off-grid forecasting methods. They also reveal that direct input of station data provides a phase shift in local weather forecasting accuracy, reducing the prediction error by up to 80% compared to pure gridded data based models. This approach demonstrates how to bridge the gap between large-scale weather models and locally accurate forecasts to support high-stakes, location-sensitive decision-making. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_12938 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Local Off-Grid Weather Forecasting with Multi-Modal Earth Observation Data Yang, Qidong Giezendanner, Jonathan Civitarese, Daniel Salles Jakubik, Johannes Schmitt, Eric Chandra, Anirban Vila, Jeremy Hohl, Detlef Hill, Chris Watson, Campbell Wang, Sherrie Machine Learning Atmospheric and Oceanic Physics Urgent applications like wildfire management and renewable energy generation require precise, localized weather forecasts near the Earth's surface. However, forecasts produced by machine learning models or numerical weather prediction systems are typically generated on large-scale regular grids, where direct downscaling fails to capture fine-grained, near-surface weather patterns. In this work, we propose a multi-modal transformer model trained end-to-end to downscale gridded forecasts to off-grid locations of interest. Our model directly combines local historical weather observations (e.g., wind, temperature, dewpoint) with gridded forecasts to produce locally accurate predictions at various lead times. Multiple data modalities are collected and concatenated at station-level locations, treated as a token at each station. Using self-attention, the token corresponding to the target location aggregates information from its neighboring tokens. Experiments using weather stations across the Northeastern United States show that our model outperforms a range of data-driven and non-data-driven off-grid forecasting methods. They also reveal that direct input of station data provides a phase shift in local weather forecasting accuracy, reducing the prediction error by up to 80% compared to pure gridded data based models. This approach demonstrates how to bridge the gap between large-scale weather models and locally accurate forecasts to support high-stakes, location-sensitive decision-making. |
| title | Local Off-Grid Weather Forecasting with Multi-Modal Earth Observation Data |
| topic | Machine Learning Atmospheric and Oceanic Physics |
| url | https://arxiv.org/abs/2410.12938 |