Score-based Filtering for Land Data Assimilation

A Unified Framework for Bayesian Data Assimilation with Generative Models and Observation Interpolants

Bayesian data assimilation combines model forecasts with noisy observations, but sampling high-dimensional, non-Gaussian posteriors remains challenging. We introduce an observation-interpolant framework that turns pretrained stochastic interpolant, flow matching, and diffusion models into posterior samplers without ret...

Accounting for model error in ensemble variational data assimilation through learned generative dynamics

Many approaches have recently been proposed to combine data assimilation with machine learning for improved forecasts with geophysical models. Here, we focus on combining advanced sequential data assimilation schemes with a generative forecasting model which accounts for complex model errors. This generative model prov...

A localized weighted ensemble adjustment Kalman filter for high-dimensional nonlinear data assimilation

Particle filters provide a nonlinear Bayesian framework for data assimilation, but global weights often degenerate in high-dimensional geophysical systems. Ensemble Kalman filters are stable with practical ensemble sizes, yet their Gaussian and locally linear updates can be restrictive for nonlinear observations and...

EnJoi: Ensemble Joint Score Filter for Generative Data Assimilation

Data Assimilation (DA) aims to recover the full state of a dynamical system that is only partially observed. A solution is to use Score-based models to generate physically consistent trajectories that agree with the observations. These Autoregressive Diffusion models are trained by conditioning on the previous state; h...

Improving Radar Rainfall Estimate with Kalman Filter Based Bias Adjustment Methods

This study evaluates three real-time Kalman filter (KF) approaches—Mean Field (MKF), Site-Specific (SKF), and Lagrangian (LKF)—for radar rainfall bias correction in northern China. While SKF addresses spatial heterogeneity using independent gauge-location filters, LKF advects a gridded log-bias correction field with st...

Improvement Forecasts for Asynchronous Observations in Data-Driven Wildfire Spread Prediction