Reproduce “DAWIS: Data Assimilation with Windowed Inverse Sampling via Multitask Interpolants” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Erik Wikingsson, Martin Andrae, Tomas Landelius, Fredrik Lindsten
DAWIS extends data assimilation to windows of states, allowing observations to revise earlier estimates while updating the present. Per-state turning points govern how much each state can change during guided regeneration, letting one learned prior support filtering, fixed-lag smoothing, and block smoothing. On Saturating SQG, its block smoother achieves CRPS 0.68 ± 0.06 versus 0.86 ± 0.09 for Block Gibbs; evaluations use synthetic observations, not operational networks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.