2006.07411
Sources and Sinks of Rare Trajectories in 2-Dimensional Velocity Fields Identified by Importance Sampling
Meagan Carney, Holger Kantz
incompletemedium confidence
- Category
- Not specified
- Journal tier
- Specialist/Solid
- Processed
- Sep 28, 2025, 12:55 AM
- arXiv Links
- Abstract ↗PDF ↗
Audit review
The paper correctly specifies the GPA incremental weights and qualitatively asserts an exponential tilt and increased tail probability with larger C, and it defines a backward-reconstruction-based statistic p_E. However, it omits the pathwise change-of-measure derivation (including the necessary initial pre-weighting to obtain a pure endpoint tilt by h(X_K)) and does not account for the Radon–Nikodym correction when using survivors to estimate probabilities under the original (untilted) dynamics. The model supplies the missing Feynman–Kac derivation, proves monotonicity rigorously, and provides a consistent inverse-weighted estimator for P(X_0∈E | X_K∈B), correcting the paper’s unweighted p_E definition.
Referee report (LaTeX)
\textbf{Recommendation:} major revisions
\textbf{Journal Tier:} specialist/solid
\textbf{Justification:}
This paper presents a novel and useful application of genealogical particle analysis to identify sources and sinks of rare trajectories in atmospheric flows. The methodology and empirical illustrations are clear and compelling. However, the theoretical claims (endpoint exponential tilt, dependence on the tilting parameter, and interpretation of the backward-reconstructed origins as probabilities under the original flow) are not supported by explicit change-of-measure derivations or stated assumptions, and the proposed p\_E statistic is biased for the target probability without inverse weighting. These issues can be addressed with relatively compact additions and clarifications, but they are essential for correctness.