pith. sign in

arxiv: 2604.11691 · v1 · submitted 2026-04-13 · 🧮 math.PR

Asymptotic behavior of spatio-temporal point processes of exceedances

Carolin Forster , Marco Oesting This is my paper

Pith reviewed 2026-05-10 16:30 UTC · model grok-4.3

classification 🧮 math.PR
keywords spatio-temporal extremespoint processesregular variationexceedancesweak convergencelimit distributionmultivariate time series
0
0 comments X

The pith

The point process of spatio-temporal exceedances converges weakly to an explicit limiting distribution.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper establishes that for stationary regularly varying multivariate time series, the rescaled point process tracking occurrence times, spatial sites, and exceedance magnitudes converges weakly to a limit point process whose distribution can be written down explicitly. Exceedances are defined through site-dependent risk functionals rather than a single fixed threshold, allowing flexibility in how extremes are identified at each location. A reader would care because the result supplies a concrete mathematical object for describing how extremes cluster or spread across both space and time. This convergence merges scattered earlier findings into a single framework that supports approximation of joint tail probabilities without full simulation of the original series.

Core claim

Exploiting the framework of stationary regularly varying multivariate time series, the authors show weak convergence of the considered point processes of extremes and explicitly determine its limit distribution. The points of the process are given by the rescaled occurrence times, the sites and the rescaled values of exceedances, where the exceedances over a high threshold are flexibly defined via site-dependent risk functionals.

What carries the argument

The rescaled spatio-temporal point process of exceedances, whose points consist of occurrence times, spatial sites, and exceedance values under regular variation.

If this is right

  • Probabilities of joint extreme events at multiple sites and times can be approximated directly from the limit process.
  • The explicit limit supplies a basis for asymptotic inference on the frequency and dependence structure of spatio-temporal extremes.
  • Flexible site-dependent risk functionals allow the same convergence result to cover a wider range of practical definitions of what counts as an exceedance.
  • The merged framework recovers and extends earlier one-dimensional or purely temporal results as special cases.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The convergence result could be used to construct asymptotic confidence regions for the spatial extent of extreme clusters.
  • Testing whether real data satisfy the regular-variation assumption would be a direct empirical check of the paper's premise.
  • The limit process suggests a natural way to simulate realistic extreme scenarios for stress-testing in environmental or financial applications.

Load-bearing premise

The underlying process must belong to the class of stationary regularly varying multivariate time series.

What would settle it

Simulate many long realizations from a stationary regularly varying multivariate series, form the empirical point process of exceedances for increasing thresholds, and verify whether its distribution converges to the paper's explicit limit; persistent discrepancy would refute the claimed convergence.

read the original abstract

In this paper, we analyze the asymptotic behavior of the point process of exceedances in a spatio-temporal setting whose points are given by the rescaled occurrence times, the sites and the rescaled values of exceedances. Here, the exceedances over a high threshold are flexibly defined via site-dependent risk functionals. Exploiting the framework of stationary regularly varying multivariate time series, we merge and extend the results from the literature in order to show weak convergence of the considered point processes of extremes and to explicitly determine its limit distribution.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 2 minor

Summary. The manuscript claims that for stationary regularly varying multivariate time series, the spatio-temporal point process of exceedances—whose points consist of rescaled occurrence times, sites, and rescaled exceedance values defined via site-dependent risk functionals—converges weakly to an explicit limit distribution. This is obtained by merging and extending standard results from the literature on tail processes and vague convergence of point processes.

Significance. If the result holds, the work provides a unified treatment of extremes in spatio-temporal settings with flexible, site-specific exceedance definitions. This extends the usual regular-variation-plus-tail-process route to a more general context and supplies an explicit limiting measure, which is a clear strength for both theory and potential applications in risk modeling. The paper follows the standard vague-convergence arguments without non-standard mixing or scaling assumptions.

minor comments (2)
  1. The abstract states the convergence result but does not indicate the form of the limiting point process; adding one sentence on the structure of the limit would improve readability.
  2. In the introduction or setup section, the precise manner in which the site-dependent risk functionals modify the standard tail-process construction should be spelled out with a short display equation to avoid any ambiguity for readers familiar with the univariate or non-spatial case.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive assessment of our manuscript, the recognition of its unified treatment of spatio-temporal extremes, and the recommendation for minor revision. No major comments were provided in the report.

Circularity Check

0 steps flagged

No significant circularity; derivation merges external regular variation results

full rationale

The paper's central claim of weak convergence for the spatio-temporal point process of exceedances is obtained by merging and extending standard results on stationary regularly varying multivariate time series, with exceedances defined via site-dependent risk functionals. The derivation follows the usual route: regular variation implies a tail process, the point process is constructed via rescaled times/sites/marks, and convergence follows from vague convergence on the space of measures. No steps reduce by construction to fitted inputs, self-definitions, or load-bearing self-citations; the result rests on external theory rather than internal reparameterization or renaming.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Central claim rests on stationarity and regular variation of the multivariate time series; no free parameters or new entities introduced in the abstract.

axioms (1)
  • domain assumption The process is a stationary regularly varying multivariate time series
    Invoked to apply and extend existing convergence results for point processes of extremes.

pith-pipeline@v0.9.0 · 5369 in / 1032 out tokens · 40491 ms · 2026-05-10T16:30:17.925620+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

28 extracted references · 28 canonical work pages

  1. [1]

    Asadi, A

    P. Asadi, A. C. Davison, and S. Engelke. Extremes on river networks. Ann. Appl. Stat., 9 0 (4): 0 2023--2050, 2015

    work page 2023

  2. [2]

    Basrak and J

    B. Basrak and J. Segers. Regularly varying multivariate time series. Stochastic Process. Appl., 119 0 (4): 0 1055--1080, 2009

    work page 2009

  3. [3]

    G. P. Bopp, B. A. Shaby, and R. Huser. A hierarchical max-infinitely divisible spatial model for extreme precipitation. J. Amer. Statist. Assoc., 116 0 (533): 0 93--106, 2021

    work page 2021

  4. [4]

    T. A. Buishand, L. de Haan, and C. Zhou. On spatial extremes: with application to a rainfall problem. Ann. Appl. Stat., 2 0 (2): 0 624--642, 2008

    work page 2008

  5. [5]

    Castro-Camilo and R

    D. Castro-Camilo and R. Huser. Local likelihood estimation of complex tail dependence structures, applied to U . S . precipitation extremes. J. Amer. Statist. Assoc., 115 0 (531): 0 1037--1054, 2020

    work page 2020

  6. [6]

    Castro-Camilo, R

    D. Castro-Camilo, R. Huser, and H. Rue. A spliced gamma-generalized P areto model for short-term extreme wind speed probabilistic forecasting. J. Agric. Biol. Environ. Stat., 24 0 (3): 0 517--534, 2019

    work page 2019

  7. [7]

    Castro-Camilo, L

    D. Castro-Camilo, L. Mhalla, and T. Opitz. Bayesian space-time gap filling for inference on extreme hot-spots: an application to R ed S ea surface temperatures. Extremes, 24 0 (1): 0 105--128, 2021

    work page 2021

  8. [8]

    Cooley, D

    D. Cooley, D. Nychka, and P. Naveau. Bayesian spatial modeling of extreme precipitation return levels. J. Amer. Statist. Assoc., 102 0 (479): 0 824--840, 2007

    work page 2007

  9. [9]

    D. J. Daley and D. Vere-Jones. An introduction to the theory of point processes. V ol. II : General theory and structure . Springer, New York, 2008

    work page 2008

  10. [10]

    R. A. Davis and T. Hsing. Point process and partial sum convergence for weakly dependent random variables with infinite variance. Ann. Probab., 23 0 (2): 0 879--917, 1995

    work page 1995

  11. [11]

    R. A. Davis and T. Mikosch. The sample autocorrelations of heavy-tailed processes with applications to ARCH . Ann. Statist., 26 0 (5): 0 2049--2080, 1998

    work page 2049

  12. [12]

    E. F. Eastoe and J. A. Tawn. Modelling non-stationary extremes with application to surface level ozone. J. R. Stat. Soc. Ser. C. Appl. Stat., 58 0 (1): 0 25--45, 2009

    work page 2009

  13. [13]

    Embrechts, C

    P. Embrechts, C. Kl\" u ppelberg, and T. Mikosch. Modelling extremal events. Springer, Berlin, 1997

    work page 1997

  14. [14]

    Engelke, R

    S. Engelke, R. de Fondeville, and M. Oesting. Extremal behaviour of aggregated data with an application to downscaling. Biometrika, 106 0 (1): 0 127--144, 2019

    work page 2019

  15. [15]

    Forster and M

    C. Forster and M. Oesting. Non-stationary max-stable models with an application to heavy rainfall data. Extremes, 28 0 (3): 0 523--556, 2025

    work page 2025

  16. [16]

    Hazra, B

    A. Hazra, B. J. Reich, and A.-M. Staicu. A multivariate spatial skew- t process for joint modeling of extreme precipitation indexes. Environmetrics, 31 0 (3): 0 Paper No. e2602, 19, 2020

    work page 2020

  17. [17]

    Hsing, J

    T. Hsing, J. H \"u sler, and M. R. Leadbetter. On the exceedance point process for a stationary sequence. Probab. Theory Related Fields, 78 0 (1): 0 97--112, 1988

    work page 1988

  18. [18]

    R. W. Katz, M. B. Parlange, and P. Naveau. Statistics of extremes in hydrology. Adv. Water Resour., 25 0 (8-12): 0 1287--1304, 2002

    work page 2002

  19. [19]

    J. Koh, F. Pimont, J.-L. Dupuy, and T. Opitz. Spatiotemporal wildfire modeling through point processes with moderate and extreme marks. Ann. Appl. Stat., 17 0 (1): 0 560--582, 2023

    work page 2023

  20. [20]

    Kulik and P

    R. Kulik and P. Soulier. Heavy-tailed time series. Springer, New York, 2020

    work page 2020

  21. [21]

    M. R. Leadbetter. Weak convergence of high level exceedances by a stationary sequence. Z. Wahrscheinlichkeitstheorie und Verw. Gebiete, 34 0 (1): 0 11--15, 1976

    work page 1976

  22. [22]

    M. R. Leadbetter, G. Lindgren, and H. Rootz\'en. Extremes and related properties of random sequences and processes. Springer, New York-Berlin, 1983

    work page 1983

  23. [23]

    Oesting and R

    M. Oesting and R. Huser. Patterns in spatio-temporal extremes. J. Amer. Statist. Assoc., 2026+. To appear

    work page 2026

  24. [24]

    Oesting and A

    M. Oesting and A. Schnurr. Ordinal patterns in clusters of subsequent extremes of regularly varying time series. Extremes, 23 0 (4): 0 521--545, 2020

    work page 2020

  25. [25]

    Oesting, M

    M. Oesting, M. Schlather, and P. Friederichs. Statistical post-processing of forecasts for extremes using bivariate B rown- R esnick processes with an application to wind gusts. Extremes, 20 0 (2): 0 309--332, 2017

    work page 2017

  26. [26]

    B. J. Reich and B. A. Shaby. A hierarchical max-stable spatial model for extreme precipitation. Ann. Appl. Stat., 6 0 (4): 0 1430--1451, 2012

    work page 2012

  27. [27]

    S. I. Resnick. Extreme values, regular variation, and point processes. Springer, New York, 2008

    work page 2008

  28. [28]

    Vettori, R

    S. Vettori, R. Huser, and M. G. Genton. Bayesian modeling of air pollution extremes using nested multivariate max-stable processes. Biometrics, 75 0 (3): 0 831--841, 2019

    work page 2019