Generating observation guided ensembles for data assimilation with denoising diffusion probabilistic model
This addresses data assimilation challenges in fields like weather forecasting, but it is incremental as it builds on existing ensemble and diffusion model techniques.
The paper tackles the problem of data assimilation with imperfect simulation models by generating pseudo-ensembles using a denoising diffusion probabilistic model trained on noisy and sparse observation data, resulting in better performance than established ensemble methods.
This paper presents an ensemble data assimilation method using the pseudo ensembles generated by denoising diffusion probabilistic model. Since the model is trained against noisy and sparse observation data, this model can produce divergent ensembles close to observations. Thanks to the variance in generated ensembles, our proposed method displays better performance than the well-established ensemble data assimilation method when the simulation model is imperfect.