DAWIS: Data Assimilation with Windowed Inverse Sampling via Multitask Interpolants
Abstract
Flow- and diffusion-based generative models have recently emerged as flexible and highly efficient forecasting models for dynamical systems. When combined with inference-time guidance, they offer a promising route to high-dimensional non-Gaussian data assimilation (DA), the problem of combining forecasts with observations to estimate latent system states. Existing filters, however, condition on a fixed history and assimilate only the most recent observation, leaving them unable to revise past states when new observations arrive. Estimates then stay tethered to a history that later observations may contradict, and errors accumulate over the assimilation run. To this end, we introduce **DAWIS**, a unified DA method covering filtering, fixed-lag smoothing, and block smoothing within a single framework. DAWIS replaces the single flow time of a state-level prior with a multitask stochastic interpolant over a window of consecutive states, assigning a separate flow time to each. An assimilation cycle inverts the window to a vector of per-state turning points and regenerates it under observation guidance, with the turning points controlling how strongly each state is held fixed, revised, or generated from scratch. The same construction can also absorb the forecast into the assimilation cycle, removing the need for a separate forecasting model. Experiments on challenging nonlinear systems show that DAWIS improves on both filtering and smoothing baselines under sparse, noisy, and nonlinear observations. The code for DAWIS is available at https://github.com/Erik-Wikingsson/DAWIS
Related
- Xaurora: Generative Weather Forecasting with Denoising Stochastic Interpolants from a Foundation Model Prior
- Windowed Envelope Statistics for Time-Domain Significant Wave Height Estimation From HF Radar
- Latent Autoencoder Ensemble Kalman Filter for Nonlinear Data Assimilation
- Data assimilation with the 2D Navier–Stokes equations: Optimal Gaussian asymptotics for the posterior measure
- IEnSF: Iterative ensemble score filter for reducing error in posterior score estimation in nonlinear data assimilation
BibTeX
@misc{wikingsson2026dawis,
title = {{DAWIS: Data Assimilation with Windowed Inverse Sampling via Multitask Interpolants}},
author = {Erik Wikingsson and Martin Andrae and Tomas Landelius and Fredrik Lindsten},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.03314},
eprint = {2610.03314},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.03314},
}
RIS
TY - UNPB
TI - DAWIS: Data Assimilation with Windowed Inverse Sampling via Multitask Interpolants
AU - Erik Wikingsson
AU - Martin Andrae
AU - Tomas Landelius
AU - Fredrik Lindsten
PY - 2026
DA - 2026/10/02
PB - arXiv
DO - 10.48550/arxiv.2610.03314
UR - https://arxiv.org/abs/2610.03314
AB - Flow- and diffusion-based generative models have recently emerged as flexible and highly efficient forecasting models for dynamical systems. When combined with inference-time guidance, they offer a promising route to high-dimensional non-Gaussian data assimilation (DA), the problem of combining forecasts with observations to estimate latent system states. Existing filters, however, condition on a fixed history and assimilate only the most recent observation, leaving them unable to revise past states when new observations arrive. Estimates then stay tethered to a history that later observations may contradict, and errors accumulate over the assimilation run. To this end, we introduce **DAWIS**, a unified DA method covering filtering, fixed-lag smoothing, and block smoothing within a single framework. DAWIS replaces the single flow time of a state-level prior with a multitask stochastic interpolant over a window of consecutive states, assigning a separate flow time to each. An assimilation cycle inverts the window to a vector of per-state turning points and regenerates it under observation guidance, with the turning points controlling how strongly each state is held fixed, revised, or generated from scratch. The same construction can also absorb the forecast into the assimilation cycle, removing the need for a separate forecasting model. Experiments on challenging nonlinear systems show that DAWIS improves on both filtering and smoothing baselines under sparse, noisy, and nonlinear observations. The code for DAWIS is available at https://github.com/Erik-Wikingsson/DAWIS
ER -