Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Data Assimilation using ERA5, ASOS, and the U-STN model for Weather Forecasting over the UK

Wenqi Wang, Jacob Bieker, Rossella Arcucci, César Quilodrán-Casas

TL;DR

This work investigates data assimilation for UK weather forecasting by adapting a global U-STN12 ML model to the UK climate and evaluating SPEnKF-based DA with ERA5 T850 and ASOS T2m data. It tests four DA scenarios, including Gaussian-noised T850, synthetic T850 observations from the ML model, ASOS T2m observations, and ERA5 T2m as a surrogate, using 24-hour assimilation intervals and RMSE as the key metric. The findings indicate that assimilating atmospheric variables improves forecast accuracy, while directly assimilating surface temperature data can degrade performance due to data sparsity and interpolation errors; model-based synthetic observations offer the most robust gains for longer lead times. The study highlights the need for improved interpolation, region-specific hyperparameter tuning, and multi-layer atmospheric representations to effectively leverage ground observations for DA in regional forecasting.

Abstract

In recent years, the convergence of data-driven machine learning models with Data Assimilation (DA) offers a promising avenue for enhancing weather forecasting. This study delves into this emerging trend, presenting our methodologies and outcomes. We harnessed the UK's local ERA5 850 hPa temperature data and refined the U-STN12 global weather forecasting model, tailoring its predictions to the UK's climate nuances. From the ASOS network, we sourced T2m data, representing ground observations across the UK. We employed the advanced kriging method with a polynomial drift term for consistent spatial resolution. Furthermore, Gaussian noise was superimposed on the ERA5 T850 data, setting the stage for ensuing multi-time step synthetic observations. Probing into the assimilation impacts, the ASOS T2m data was integrated with the ERA5 T850 dataset. Our insights reveal that while global forecast models can adapt to specific regions, incorporating atmospheric data in DA significantly bolsters model accuracy. Conversely, the direct assimilation of surface temperature data tends to mitigate this enhancement, tempering the model's predictive prowess.

Data Assimilation using ERA5, ASOS, and the U-STN model for Weather Forecasting over the UK

TL;DR

This work investigates data assimilation for UK weather forecasting by adapting a global U-STN12 ML model to the UK climate and evaluating SPEnKF-based DA with ERA5 T850 and ASOS T2m data. It tests four DA scenarios, including Gaussian-noised T850, synthetic T850 observations from the ML model, ASOS T2m observations, and ERA5 T2m as a surrogate, using 24-hour assimilation intervals and RMSE as the key metric. The findings indicate that assimilating atmospheric variables improves forecast accuracy, while directly assimilating surface temperature data can degrade performance due to data sparsity and interpolation errors; model-based synthetic observations offer the most robust gains for longer lead times. The study highlights the need for improved interpolation, region-specific hyperparameter tuning, and multi-layer atmospheric representations to effectively leverage ground observations for DA in regional forecasting.

Abstract

In recent years, the convergence of data-driven machine learning models with Data Assimilation (DA) offers a promising avenue for enhancing weather forecasting. This study delves into this emerging trend, presenting our methodologies and outcomes. We harnessed the UK's local ERA5 850 hPa temperature data and refined the U-STN12 global weather forecasting model, tailoring its predictions to the UK's climate nuances. From the ASOS network, we sourced T2m data, representing ground observations across the UK. We employed the advanced kriging method with a polynomial drift term for consistent spatial resolution. Furthermore, Gaussian noise was superimposed on the ERA5 T850 data, setting the stage for ensuing multi-time step synthetic observations. Probing into the assimilation impacts, the ASOS T2m data was integrated with the ERA5 T850 dataset. Our insights reveal that while global forecast models can adapt to specific regions, incorporating atmospheric data in DA significantly bolsters model accuracy. Conversely, the direct assimilation of surface temperature data tends to mitigate this enhancement, tempering the model's predictive prowess.
Paper Structure (9 sections, 1 equation, 2 figures)

This paper contains 9 sections, 1 equation, 2 figures.

Table of Contents

  1. Introduction
  2. Methods
  3. Datasets
  4. ERA5
  5. ASOS Dataset
  6. Model
  7. Data assimilation
  8. Results
  9. Conclusion

Figures (2)

  • Figure 1: (a) Geographical distribution of ASOS surface observation stations and annual average T2m values between latitudes 50° to 57.75° and longitudes -6° to 1.875°(2022). (b) Annual average T850 temperature pattern from ERA5 within the same latitudinal and longitudinal bounds (2022).
  • Figure 2: Root Mean Square Error (RMSE) Comparison for the U-STN12-SPEnKF Model Across Different DA Sources.