Estudio sobre técnicas de asimilación de observaciones en resultados de modelos meteorológicos
Loading...
Official URL
Full text at PDC
Publication date
2026
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
License

Citation
Abstract
RESUMEN: La asimilación de datos constituye una herramienta fundamental en la predicción meteorológica moderna, pues permite combinar de forma coherente observaciones experimentales y modelos numéricos para obtener estimaciones más precisas del estado atmosférico. En este trabajo se presenta una revisión bibliográfica de las principales técnicas de asimilación de datos utilizadas en meteorología, prestando especial atención a los métodos secuenciales y variacionales. En particular, se estudian el filtro de Kalman y las técnicas 3DVAR y 4DVAR, ampliamente empleadas en sistemas actuales de reanálisis atmosférico. Por otro lado, se implementa numéricamente un esquema 4DVAR sobre el modelo caótico de Lorenz-96 con el objetivo de reconstruir el estado inicial del sistema a partir de observaciones parciales y ruidosas. Los resultados obtenidos muestran la capacidad de esta técnica para reducir significativamente el error en la estimación del estado y ponen de manifiesto la importancia de la asimilación de datos en la modelización de sistemas atmosféricos complejos.
ABSTRAC: Data assimilation is a fundamental tool in modern weather forecasting, as it enables the coherent combination of observational data and numerical models in order to obtain more accurate estimations of the atmospheric state. This work presents a bibliographic review of the main data assimilation techniques used in meteorology, with special emphasis on sequential and variational methods. In particular, the Kalman filter and the 3DVAR and 4DVAR techniques, widely employed in modern atmospheric reanalysis systems, are studied. In addition, a 4DVAR scheme is numerically implemented on the chaotic Lorenz-96 model in order to reconstruct the initial state of the system from partial and noisy observations. The obtained results show the ability of this technique to significantly reduce the estimation error and highlight the importance of data assimilation in the modeling of complex atmospheric systems.
ABSTRAC: Data assimilation is a fundamental tool in modern weather forecasting, as it enables the coherent combination of observational data and numerical models in order to obtain more accurate estimations of the atmospheric state. This work presents a bibliographic review of the main data assimilation techniques used in meteorology, with special emphasis on sequential and variational methods. In particular, the Kalman filter and the 3DVAR and 4DVAR techniques, widely employed in modern atmospheric reanalysis systems, are studied. In addition, a 4DVAR scheme is numerically implemented on the chaotic Lorenz-96 model in order to reconstruct the initial state of the system from partial and noisy observations. The obtained results show the ability of this technique to significantly reduce the estimation error and highlight the importance of data assimilation in the modeling of complex atmospheric systems.











