Pith. sign in

REVIEW 3 major objections 3 minor 75 references

Improved Global Ocean Heat Content Estimation by Modeling Vertical Spatio-Temporal Dependence

T0 review · 3 major / 3 minor · reviewed 2026-07-14 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Jointly mapping two ocean layers with their correlation cuts global OHC anomaly uncertainty by up to 15%.

desk verdict Abstract-only methods paper: joint bivariate GP mapping of two OHC layers is a sensible fix for a real operational flaw, with a claimed ~15% uncertainty cut that we cannot verify yet. read the letter →

arxiv 2607.11832 v1 pith:6MLDZOW4 submitted 2026-07-13 stat.AP physics.ao-ph

Thea Sukianto , Donata Giglio , Mikael Kuusela This is my paper
classification stat.APphysics.ao-ph
keywords oceanheatcontentArgofloatsGaussianprocessesverticaldependenceuncertaintyquantificationspatio-temporalmappingclimatemonitoring
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper aims to show that ocean heat content (OHC) anomalies are better estimated when the vertical dependence between pressure layers is modeled rather than ignored. Using Argo float data from 2004–2022, the authors map two layers jointly with bivariate locally stationary Gaussian processes and conditional simulations, instead of mapping each layer separately and then summing. Because fewer observations exist at depth, earlier work split the water column and treated layers independently, which made total-OHC uncertainties hard to quantify correctly. Accounting for the correlation between layers improves the anomaly maps and reduces global OHC anomaly uncertainty by as much as 15 percent. Those tighter, properly calibrated uncertainties then allow clearer tests of whether regional and global heat-content changes are statistically significant, which the authors illustrate with climatological case studies.

What carries the argument

Bivariate locally stationary Gaussian processes used to represent the joint spatio-temporal field of the two pressure layers, combined with conditional simulations that propagate the estimated vertical dependence into total-OHC maps and uncertainties.

What would settle it

If independent validation against withheld deep Argo profiles or against independent full-depth hydrographic sections shows that the joint bivariate maps produce larger or poorly calibrated total-OHC errors than the separate-layer maps, the claimed 15 percent uncertainty reduction would be falsified.

Watch

Extended reading notes

Core claim

Modeling the correlation between two pressure layers with bivariate locally stationary Gaussian processes and conditional simulations improves OHC anomaly mapping and yields up to a 15 percent reduction of global OHC anomaly uncertainties compared with mapping the two layers separately without accounting for their dependence.

Load-bearing premise

The true vertical dependence of OHC anomalies between the two chosen pressure layers is adequately described by a bivariate locally stationary Gaussian process, so that the reported uncertainty reduction is not an artifact of model misspecification.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 3 minor

Summary. The manuscript proposes a joint mapping method for ocean heat content (OHC) anomalies in two pressure layers using Argo data (2004–2022). Prior practice maps layers independently, which complicates uncertainty quantification when layers are summed. The authors instead use bivariate locally stationary Gaussian processes with conditional simulations to map the layers jointly while accounting for vertical correlation. They report improved OHC anomaly maps and up to a 15% reduction in global OHC anomaly uncertainties relative to independent mapping, and illustrate the use of the resulting uncertainties for assessing statistical significance of regional and global OHC anomalies in climatological case studies.

Significance. If the reported uncertainty reduction and map improvements hold under proper diagnostics, the contribution is practically important: OHC is the primary reservoir of Earth’s energy imbalance, and reliable layer-summed uncertainties are needed for regional and global significance statements. Joint bivariate modeling of vertical dependence is a natural and methodologically coherent improvement over independent layer mapping. The use of conditional simulations for uncertainty quantification is a strength if implemented and validated carefully. The work is applied and empirical rather than theoretical; its value rests on whether the bivariate locally stationary GP is adequate and whether the 15% figure is robust.

major comments (3)
  1. [Abstract (methods and results not available)] Abstract-only review: the central claim of up to ~15% global OHC-anomaly uncertainty reduction and improved maps cannot be verified without methods, residual diagnostics, cross-validation, and sensitivity analyses. The load-bearing premise is that a bivariate locally stationary Gaussian process adequately captures vertical spatio-temporal dependence between the two chosen pressure layers. The full manuscript must report diagnostics that could falsify local stationarity or Gaussianity (e.g., residual correlation structure, QQ/score diagnostics, leave-region-out or temporal holdout skill) and show that the uncertainty reduction is not an artifact of misspecification or of the chosen covariance form.
  2. [Abstract (hyperparameters and layer definition)] The free parameters of the procedure—GP covariance hyperparameters (length scales, variances, cross-layer correlation), the local-stationarity window / nonstationarity structure, and the two pressure-layer boundaries—are estimated from the same Argo data used for mapping. The manuscript must clarify how hyperparameters are estimated, whether uncertainty in those hyperparameters is propagated into the reported OHC uncertainties, and how sensitive the 15% reduction and significance case studies are to layer boundaries and to the bivariate covariance specification. Without that, the headline uncertainty reduction remains incompletely supported.
  3. [Abstract (uncertainty comparison design)] The abstract states that independent layer mapping “complicates the estimation of uncertainties when the maps are summed.” The joint method’s advantage should be demonstrated not only via a global scalar uncertainty reduction but via explicit comparison of summed-layer variance (including the cross-layer covariance term) against the independent-mapping baseline on the same grid and period. If the full text only reports a global percentage without decomposing where the gain comes from (cross-covariance vs. better marginal fits), the claim that dependence modeling is the operative mechanism is under-supported.
minor comments (3)
  1. [Abstract] The abstract’s “up to a 15 percent” phrasing should be paired in the full text with the spatial/temporal distribution of the reduction (global mean, regional range, time dependence) so readers can judge whether the figure is typical or a peak value.
  2. [Methods (expected)] Clarify in the introduction/methods how the two pressure-layer boundaries are chosen and whether results are robust to alternative partitions used in the OHC literature.
  3. [Case studies (expected)] Case-study significance statements should specify the null, the spatial aggregation, and multiple-testing or spatial-dependence adjustments if regional maps are interpreted jointly.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: empirical joint-vs-separate mapping comparison on Argo data, not forced by definition or self-citation.

full rationale

Only the abstract is available. The central claim is an empirical comparison: bivariate locally stationary Gaussian process mapping of two pressure layers (with conditional simulations) versus separate univariate mapping, reporting improved OHC anomaly maps and up to ~15% reduction in global OHC anomaly uncertainties on Argo 2004–2022 data. That comparison is not equivalent to its inputs by construction; the bivariate model is a modeling choice whose adequacy is an assumption, not a tautology that renames a fitted target as a prediction. Hyperparameter estimation from the same data used for mapping is standard geostatistical practice and does not constitute the self-definitional or fitted-input-called-prediction circularity patterns that raise the score. No uniqueness theorem, load-bearing self-citation chain, or ansatz smuggled via prior author work is present in the abstract. The result is falsifiable against the separate-mapping baseline and against external OHC products; therefore the derivation chain as stated is self-contained against the available text. Score 0 is the honest finding under the abstract-only limit.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

Abstract-only: free parameters are the usual GP hyperparameters (length scales, variances, cross-layer correlation, local-stationarity windows) fitted to Argo fields; axioms are standard geostatistical modeling choices (Gaussian process residuals, local stationarity, two-layer partition). No new physical entities are introduced. Exact fitted values and full axiom list cannot be audited without the paper body.

free parameters (3)
  • GP covariance hyperparameters (length scales, variances, cross-layer correlation)
    Bivariate locally stationary GPs require estimated spatial/temporal scales, marginal variances, and a cross-covariance parameter; these are fitted to Argo OHC anomaly fields and drive both the maps and the reported uncertainty reduction.
  • Local-stationarity window / nonstationarity structure
    Locally stationary models introduce a scale or partition over which stationarity is assumed; choice affects dependence estimates between layers and is not fixed by first principles in the abstract.
  • Two pressure-layer boundaries
    The analysis is framed for a fixed two-layer partition of the water column; the depth cut is a modeling choice that defines the bivariate target and is not derived in the abstract.
assumptions (3)
  • domain assumption OHC anomaly fields in the two layers are well modeled as a bivariate Gaussian process with local stationarity.
    Joint mapping and conditional-simulation uncertainties rest on this probabilistic model; the abstract does not supply non-Gaussian or nonstationarity diagnostics.
  • domain assumption Argo profiling float data from 2004–2022 adequately sample the spatio-temporal field for the GP posterior to be meaningful globally and regionally.
    Sparse deep sampling is acknowledged as motivation; the method still assumes the available floats suffice for the claimed global uncertainty reduction and significance tests.
  • standard math Standard properties of Gaussian processes and conditional simulation for uncertainty quantification.
    Conditional draws from the bivariate GP posterior are used to propagate layer dependence into total OHC uncertainty; this is textbook GP methodology.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Improved Global Ocean Heat Content Estimation by Modeling Vertical Spatio-Temporal Dependence." pith.science (2026). https://pith.science/paper/6MLDZOW4

@misc{pith2026260711832,
  author       = {Pith},
  title        = {Pith review of: Improved Global Ocean Heat Content Estimation by Modeling Vertical Spatio-Temporal Dependence},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6MLDZOW4}},
  note         = {Machine review of arXiv:2607.11832}
}
read the original abstract

Estimating ocean heat content (OHC) with reliable uncertainties is critical for understanding and monitoring the evolution of Earth's climate, as the ocean has stored most of the energy accumulated in the climate system due to Earth Energy Imbalance. Here, we use Argo profiling float data from 2004-2022 to map OHC. As fewer Argo observations are available deeper in the water column, previous studies have partitioned the ocean into at least two pressure layers and mapped each separately, which complicates the estimation of uncertainties when the maps are summed to get the total OHC. In this work, we consider the case of two pressure layers and propose an improved mapping and uncertainty quantification method using bivariate locally stationary Gaussian processes and conditional simulations to map the two sections jointly while accounting for the correlation between them. We find that modeling this correlation results in improved OHC anomaly mapping and up to a 15 percent reduction of global OHC anomaly uncertainties in comparison to mapping the two layers separately without accounting for their dependence. These estimated uncertainties are essential to analyze the statistical significance of OHC anomalies on both regional and global scales, which we demonstrate using several climatological case studies.

Figures

Figures reproduced from arXiv: 2607.11832 by the authors.

Figure 1
Figure 1. Comparison of cross-validated monthly median absolute prediction errors between [PITH_FULL_IMAGE:figures/full_fig_p019_1.png] view at source ↗
Figure 2
Figure 2. Relative difference in predictive variance between the univariate and bivariate [PITH_FULL_IMAGE:figures/full_fig_p020_2.png] view at source ↗
Figure 3
Figure 3. Cross-validation for conditional spatio-temporal dependence. The solid and [PITH_FULL_IMAGE:figures/full_fig_p023_3.png] view at source ↗
Figures from the paper (10 more)
Figure 4
Figure 4. Figure 4: Total global OHC anomaly (15–1850 dbar) standard error comparison. The red [PITH_FULL_IMAGE:figures/full_fig_p024_4.png]
Figure 5
Figure 5. Figure 5: Comparison of statistical significance for total regional OHC anomalies (15- [PITH_FULL_IMAGE:figures/full_fig_p025_5.png]
Figure 6
Figure 6. Figure 6: Estimated total vertical cross-correlation between 15–975 dbar and 975–1850 dbar [PITH_FULL_IMAGE:figures/full_fig_p027_6.png]
Figure 7
Figure 7. Figure 7: Total OHC (15–1850 dbar) time series with 68% (dark gray) and 95% (light gray) [PITH_FULL_IMAGE:figures/full_fig_p028_7.png]
Figure 8
Figure 8. Figure 8: Ocean heat uptake (15–1850 dbar) with 95% uncertainties (gray = univariate [PITH_FULL_IMAGE:figures/full_fig_p029_8.png]
Figure 9
Figure 9. Figure 9: Cross-correlation of monthly global OHC anomalies (15–1850 dbar) with 68% [PITH_FULL_IMAGE:figures/full_fig_p030_9.png]
Figure 10
Figure 10. Figure 10: Estimated cross-correlation parameters 34 [PITH_FULL_IMAGE:figures/full_fig_p034_10.png]
Figure 11
Figure 11. Figure 11: Vertically integrated OHC residuals for the collection of 20 [PITH_FULL_IMAGE:figures/full_fig_p036_11.png]
Figure 12
Figure 12. Figure 12: Vertically integrated OHC residuals for the collection of 20 [PITH_FULL_IMAGE:figures/full_fig_p036_12.png]
Figure 13
Figure 13. Figure 13: Vertically integrated OHC residuals for the collection of 20 [PITH_FULL_IMAGE:figures/full_fig_p037_13.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

75 extracted references · 15 canonical work pages

  1. [1]

    2024 , issn =

    Neural likelihood surfaces for spatial processes with computationally intensive or intractable likelihoods , journal =. 2024 , issn =. doi:https://doi.org/10.1016/j.spasta.2024.100848 , author =

    doi:10.1016/j.spasta.2024.100848 2024

  2. [2]

    Journal of Computational and Graphical Statistics , volume =

    Matthew Sainsbury-Dale and Andrew Zammit-Mangion and Jordan Richards and Raphaël Huser , title =. Journal of Computational and Graphical Statistics , volume =. 2025 , publisher =

    2025

  3. [3]

    The Galerkin finite element method for elliptic problems

    Quarteroni, Alfio. The Galerkin finite element method for elliptic problems. Numerical Models for Differential Problems. 2014. doi:10.1007/978-88-470-5522-3_4

    work page doi:10.1007/978-88-470-5522-3_4 2014

  4. [4]

    2025 , howpublished=

    Neural Conditional Simulation for Complex Spatial Processes , author=. 2025 , howpublished=

    2025

  5. [5]

    The Graph Neural Network Model , year=

    Scarselli, Franco and Gori, Marco and Tsoi, Ah Chung and Hagenbuchner, Markus and Monfardini, Gabriele , journal=. The Graph Neural Network Model , year=

  6. [6]

    G. C. Johnson and others. Global Oceans. Bulletin of the American Meteorological Society. 2024. doi:10.1175/BAMS-D-24-0100.1

    work page doi:10.1175/bams-d-24-0100.1 2024

  7. [7]

    The Annals of Applied Statistics , number =

    Beomjo Park and Mikael Kuusela and Donata Giglio and Alison Gray , title =. The Annals of Applied Statistics , number =. 2023 , doi =

    2023

  8. [8]

    WIREs Computational Statistics , volume =

    Porcu, Emilio and Furrer, Reinhard and Nychka, Douglas , title =. WIREs Computational Statistics , volume =. doi:10.1002/wics.1512 , year =

    doi:10.1002/wics.1512

Show all 75 references
  1. [9]

    Proceedings of the 19th International Conference on Artificial Intelligence and Statistics , pages =

    Deep Kernel Learning , author =. Proceedings of the 19th International Conference on Artificial Intelligence and Statistics , pages =. 2016 , editor =

    2016

  2. [10]

    Space and Space-Time Modeling using Process Convolutions

    Higdon, Dave. Space and Space-Time Modeling using Process Convolutions. Quantitative Methods for Current Environmental Issues. 2002

    2002

  3. [11]

    Scandinavian Journal of Statistics , volume =

    Wallin, Jonas and Bolin, David , title =. Scandinavian Journal of Statistics , volume =. doi:https://doi.org/10.1111/sjos.12141 , year =

    doi

  4. [12]

    2022 , note =

    The SPDE approach for Gaussian and non-Gaussian fields: 10 years and still running , journal =. 2022 , note =. doi:https://doi.org/10.1016/j.spasta.2022.100599 , author =

    2022 doi

  5. [13]

    and others , Doi =

    Hakuba, Maria Z. and others , Doi =. Trends and Variability in. Surveys in Geophysics , Number =

  6. [14]

    , title=

    Kuusela, Mikael and Stein, Michael L. , title=. 2018 , journal=

    2018

  7. [15]

    Lyman and Gregory C

    John M. Lyman and Gregory C. Johnson. Global High-Resolution Random Forest Regression Maps of Ocean Heat Content Anomalies Using In Situ and Satellite Data. Journal of Atmospheric and Oceanic Technology. 2023. doi:10.1175/JTECH-D-22-0058.1

    2023 doi

  8. [16]

    , title =

    Rodrigues, Alexandre and Diggle, Peter J. , title =. Scandinavian Journal of Statistics , volume =. doi:10.1111/j.1467-9469.2009.00675.x , year =

    2009 doi

  9. [17]

    Journal of the American Statistical Association , volume =

    Tilmann Gneiting , title =. Journal of the American Statistical Association , volume =. 2002 , publisher =

    2002

  10. [18]

    doi:10.1017/9781009157896 , year =

    IPCC , title =. doi:10.1017/9781009157896 , year =

    doi

  11. [19]

    2000 , journal=

    Argo , title=. 2000 , journal=

    2000

  12. [20]

    Progress in Oceanography , volume =

    The 2004–2008 mean and annual cycle of temperature, salinity, and steric height in the global ocean from the. Progress in Oceanography , volume =. 2009 , issn =. doi:10.1016/j.pocean.2009.03.004 , author =

    2004 doi

  13. [21]

    2012 , issn =

    Nonstationary modeling for multivariate spatial processes , journal =. 2012 , issn =. doi:10.1016/j.jmva.2012.05.011 , author =

    2012 doi

  14. [22]

    2016 , publisher=

    Pattern Recognition and Machine Learning , author=. 2016 , publisher=

    2016

  15. [23]

    2010 , publisher=

    Handbook of Spatial Statistics , author=. 2010 , publisher=

    2010

  16. [24]

    Haas , title =

    Timothy C. Haas , title =. Journal of the American Statistical Association , volume =. 1990 , publisher =

    1990

  17. [25]

    Haas , title =

    Timothy C. Haas , title =. Journal of the American Statistical Association , volume =. 1995 , publisher =

    1995

  18. [26]

    2018 , note =

    Modeling and emulation of nonstationary Gaussian fields , journal =. 2018 , note =. doi:10.1016/j.spasta.2018.08.006 , author =

    2018 doi

  19. [27]

    Frontiers in Marine Science , VOLUME=

    Meyssignac, Benoit and others , TITLE=. Frontiers in Marine Science , VOLUME=. 2019 , DOI=

    2019

  20. [28]

    An imperative to monitor Earth's Energy Imbalance

    von Schuckmann, K and Palmer, M D and Trenberth, K E and Cazenave, A and Chambers, D and Champollion, N and Hansen, J and Josey, S A and Loeb, N and Mathieu, P-P and Meyssignac, B and Wild, M. An imperative to monitor Earth's Energy Imbalance. Nat. Clim. Chang

  21. [30]

    Global Oceans

    Johnson, Gregory C and others. Global Oceans. Bull. Am. Meteorol. Soc

  22. [31]

    Hurricane Harvey links to ocean heat content and climate change adaptation

    Trenberth, Kevin E and Cheng, Lijing and Jacobs, Peter and Zhang, Yongxin and Fasullo, John. Hurricane Harvey links to ocean heat content and climate change adaptation. Earth's Future

  23. [32]

    Mikael Kuusela and Donata Giglio , title =

  24. [33]

    F. N. Fritsch and R. E. Carlson , journal =. Monotone Piecewise Cubic Interpolation , urldate =

  25. [34]

    Journal of Atmospheric and Oceanic Technology , year = 2020, month = apr, volume =

    Two Interpolation Methods Using Multiply-Rotated Piecewise Cubic Hermite Interpolating Polynomials. Journal of Atmospheric and Oceanic Technology , year = 2020, month = apr, volume =. doi:10.1175/JTECH-D-19-0211.1 , adsnote =

    2020 doi

  26. [35]

    2020 , note=

    Statistics for Mapping Ocean Heat Content with Argo Floats: Modeling and Uncertainty Quantification , author =. 2020 , note=

    2020

  27. [36]

    Johnson and Stephen Riser and Russ Davis and John Gilson and W

    Dean Roemmich and Gregory C. Johnson and Stephen Riser and Russ Davis and John Gilson and W. Brechner Owens and Silvia L. Garzoli and Claudia Schmid and Mark Ignaszewski , journal =. The Argo Program: Observing the Global Ocean with Profiling Floats , urldate =

  28. [37]

    Fifteen years of ocean observations with the global

    Riser, Stephen C and Freeland, Howard J and Roemmich, Dean and Wijffels, Susan and Troisi, Ariel and Belb. Fifteen years of ocean observations with the global. Nature Climate Change , volume=. 2016 , publisher=

    2016

  29. [38]

    Bridges-2: A platform for rapidly-evolving and data intensive research

    Brown, Shawn T and Buitrago, Paola and Hanna, Edward and Sanielevici, Sergiu and Scibek, Robin and Nystrom, Nicholas A. Bridges-2: A platform for rapidly-evolving and data intensive research. Practice and Experience in Advanced Research Computing

  30. [39]

    2005 , issn =

    The evolution of El Niño, past and future , journal =. 2005 , issn =. doi:https://doi.org/10.1016/j.epsl.2004.12.003 , author =

    2005 doi

  31. [40]

    McPhaden and Stephen E

    Michael J. McPhaden and Stephen E. Zebiak and Michael H. Glantz , title =. Science , volume =. 2006 , doi =

    2006

  32. [41]

    Record-breaking warming and extreme drought in the Amazon rainforest during the course of El Ni \ n o 2015-2016

    Jim \'e nez-Mu \ n oz, Juan C and Mattar, Cristian and Barichivich, Jonathan and Santamar \' a-Artigas, Andr \'e s and Takahashi, Ken and Malhi, Yadvinder and Sobrino, Jos \'e A and Schrier, Gerard van der. Record-breaking warming and extreme drought in the Amazon rainforest d...

    2015

  33. [42]

    Observing and predicting the 2015/16 El Ni \ n o

    L'Heureux, Michelle L and Takahashi, Ken and Watkins, Andrew B and Barnston, Anthony G and Becker, Emily J and Di Liberto, Tom E and Gamble, Felicity and Gottschalck, Jon and Halpert, Michael S and Huang, Boyin and Mosquera-V \'a squez, Kobi and Wittenberg, Andrew T. Observing...

    2015

  34. [43]

    Impact of the 2015 El Nino event on winter air quality in China

    Chang, Luyu and Xu, Jianming and Tie, Xuexi and Wu, Jianbin. Impact of the 2015 El Nino event on winter air quality in China. Sci. Rep

    2015

  35. [44]

    , year =

    McDougall, Trevor and Barker, P.M. , year =. Getting started with TEOS-10 and the Gibbs Seawater (GSW) Oceanographic Toolbox , volume =

  36. [45]

    McDougall

    Trevor J. McDougall. Potential Enthalpy: A Conservative Oceanic Variable for Evaluating Heat Content and Heat Fluxes. Journal of Physical Oceanography. 2003

    2003

  37. [46]

    Descriptive Physical Oceanography (Sixth Edition) , publisher =

    Chapter 3 - Physical Properties of Seawater , editor =. Descriptive Physical Oceanography (Sixth Edition) , publisher =. 2011 , isbn =. doi:https://doi.org/10.1016/B978-0-7506-4552-2.10003-4 , author =

    2011 doi

  38. [47]

    Principles of geostatistics

    Matheron, Georges. Principles of geostatistics. Econ. Geol

  39. [48]

    Journal of Machine Learning Research , year =

    Nikola Kovachki and Zongyi Li and Burigede Liu and Kamyar Azizzadenesheli and Kaushik Bhattacharya and Andrew Stuart and Anima Anandkumar , title =. Journal of Machine Learning Research , year =

  40. [49]

    Bayesian Data Analysis

    Gelman, Andrew and Carlin, John B and Stern, Hal S and Dunson, David B and Vehtari, Aki and Rubin, Donald B. Bayesian Data Analysis

  41. [50]

    Loeb and David R

    Norman G. Loeb and David R. Doelling and Hailan Wang and Wenying Su and Cathy Nguyen and Joseph G. Corbett and Lusheng Liang and Cristian Mitrescu and Fred G. Rose and Seiji Kato. Clouds and the E arth’s Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) Top-of-At...

    2018 doi

  42. [51]

    Rose and David A

    Seiji Kato and Fred G. Rose and David A. Rutan and Tyler J. Thorsen and Norman G. Loeb and David R. Doelling and Xianglei Huang and William L. Smith and Wenying Su and Seung-Hee Ham. Surface Irradiances of Edition 4.0 Clouds and the Earth’s Radiant Energy System (CERES) Energy...

    2018 doi

  43. [52]

    Zenodo , version =

    Giglio, Donata and Sukianto, Thea and Kuusela, Mikael , title =. Zenodo , version =

  44. [53]

    1995 , issn =

    A multidimensional unfolding method based on Bayes' theorem , journal =. 1995 , issn =. doi:https://doi.org/10.1016/0168-9002(95)00274-X , author =

    1995 doi

  45. [54]

    doi:10.1088/1361-6420/aaa34d , year =

    Calvetti, Daniela and Dunlop, Matthew and Somersalo, Erkki and Stuart, Andrew , title =. doi:10.1088/1361-6420/aaa34d , year =

    doi

  46. [55]

    Learning Uncertainties the Frequentist Way: Calibration and Correlation in High Energy Physics , volume=

    Gambhir, Rikab and Nachman, Benjamin and Thaler, Jesse , year=. Learning Uncertainties the Frequentist Way: Calibration and Correlation in High Energy Physics , volume=. Physical Review Letters , publisher=. doi:10.1103/physrevlett.129.082001 , number=

    doi

  47. [56]

    Gramacy and David Higdon , title =

    Annie Sauer, Robert B. Gramacy and David Higdon , title =. Technometrics , volume =. 2023 , publisher =

    2023

  48. [57]

    , publisher =

    Trenberth, Kevin E. , publisher =. El Ni\. doi:https://doi.org/10.1002/9781119548164.ch2 , year =

    doi

  49. [58]

    CLIVAR Exchanges , volume=

    Argo: The global array of profiling floats , author=. CLIVAR Exchanges , volume=. 1999 , publisher=

    1999

  50. [59]

    The Annals of Applied Statistics , number =

    Drew Yarger and Stilian Stoev and Tailen Hsing , title =. The Annals of Applied Statistics , number =. 2022 , doi =

    2022

  51. [60]

    Mary Lai Salvana and Jian Cao and Mikyoung Jun , year=

  52. [61]

    Multivariate Type

    Bolin, David and Wallin, Jonas , year=. Multivariate Type. Journal of the Royal Statistical Society Series B: Statistical Methodology , publisher=. doi:10.1111/rssb.12351 , number=

    doi

  53. [62]

    Trenberth and John T

    Kevin E. Trenberth and John T. Fasullo and Karina von Schuckmann and Lijing Cheng. Insights into E arth's Energy Imbalance from Multiple Sources. Journal of Climate. 2016. doi:10.1175/JCLI-D-16-0339.1

    2016 doi

  54. [63]

    Nychka , title =

    Thea Sukianto and Mikael Kuusela and Donata Giglio and Anirban Mondal and Pulong Ma and Douglas W. Nychka , title =

  55. [64]

    , year =

    McDougall, Trevor and Barker, P.M. , year =. Getting started with

  56. [65]

    Bamston and Muthuvel Chelliah and Stanley B

    Anthony G. Bamston and Muthuvel Chelliah and Stanley B. Goldenberg , title =. Atmosphere-Ocean , volume =. 1997 , publisher =

    1997

  57. [66]

    Statistics for Spatial Data

    Cressie, Noel A C. Statistics for Spatial Data

  58. [67]

    Interpolation of Spatial Data

    Stein, Michael L. Interpolation of Spatial Data

  59. [68]

    Incorporating correlated nugget effects in multivariate spatial models: An application to Argo ocean data

    Saduakhas, Damilya and Bolin, David and Jin, Xiaotian and Simas, Alexandre B and Wallin, Jonas. Incorporating correlated nugget effects in multivariate spatial models: An application to Argo ocean data

  60. [69]

    The Annals of Applied Statistics , number =

    Saduakhas, Damilya and Bolin, David and Jin, Xiaotian and Simas, Alexandre B and Wallin, Jonas , title =. The Annals of Applied Statistics , number =. 2026 , doi =

    2026

  61. [70]

    2021 , author =

    Optimization Toolbox version: 9.1 (R2021a) , publisher =. 2021 , author =

    2021

  62. [71]

    and others , TITLE=

    Wong, Annie P. and others , TITLE=. Frontiers in Marine Science , VOLUME=. 2020 , DOI=

    2020

  63. [72]

    Johnson and Stephen Riser and Russ Davis and John Gilson and W

    Dean Roemmich and Gregory C. Johnson and Stephen Riser and Russ Davis and John Gilson and W. Brechner Owens and Silvia L. Garzoli and Claudia Schmid and Mark Ignaszewski , journal =. The

  64. [73]

    2023 , journal=

    Argo , title=. 2023 , journal=

    2023

  65. [74]

    Journal of the Royal Statistical Society Series C: Applied Statistics , volume =

    Korte-Stapff, Moritz and Yarger, Drew and Stoev, Stilian and Hsing, Tailen , title =. Journal of the Royal Statistical Society Series C: Applied Statistics , volume =. 2025 , month =

    2025

  66. [75]

    and others , TITLE =

    von Schuckmann, K. and others , TITLE =. Earth System Science Data , VOLUME =. 2023 , NUMBER =

    2023

  67. [76]

    Neural Methods for Amortized Inference

    Zammit-Mangion, Andrew and Sainsbury-Dale, Matthew and Huser, Raphaël. Neural Methods for Amortized Inference. Annual Review of Statistics and Its Application. 2025. doi:https://doi.org/10.1146/annurev-statistics-112723-034123

    2025 doi

Pith tools

Reviewed July 14, 2026 · model on record in the stance chip above.