| Autor: |
Ioannis Samos, Helena Flocas, Petroula Louka |
| Jazyk: |
angličtina |
| Rok vydání: |
2023 |
| Předmět: |
|
| Zdroj: |
Environmental Sciences Proceedings, Vol 26, Iss 1, p 158 (2023) |
| Druh dokumentu: |
article |
| ISSN: |
2673-4931 |
| DOI: |
10.3390/environsciproc2023026158 |
| Popis: |
Data assimilation is a technique used to combine observational data with numerical weather analysis fields to produce input conditions for Numerical Weather Prediction (NWP) models aimed at more accurate forecasts. In this study, two different regional Background Error (BE) covariance statistics are evaluated for the implementation of a Weather Research and Forecast (WRF) model using Variational Data Assimilation (VAR) schemes. Two different WRF model forecasts, initialized at different times, are compared to calculate different regional BE statistics based on the National Meteorological Center (NMC) technique. These statistics are then used in the three-dimensional variational (3DVAR) data assimilation process to produce analysis fields for a 15 day period during September 2019. The study compares the forecasts produced using the two different ΒΕ statistics and presents the results of the analysis fields during the medicane “Ianos” formed in September 2020 in the eastern Mediterranean Sea. The study demonstrates the importance of BE statistics’ usage in data assimilation. The results suggest that different initialization times can lead to significant differences in weather evolution. The study also highlights the need for caution in the choice of BE statistics and the importance of best practices in data assimilation. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
