Pith. sign in

REVIEW 3 major objections 5 minor 12 references

Coupled Seasonal Data Assimilation of Sea Ice, Ocean, and Atmospheric Dynamics over the Last Millennium

T0 review · 3 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read A season-by-season data-assimilation scheme reconstructs last-millennium climate with higher verification skill than prior paleo-products while using roughly a quarter of the proxies.

desk verdict First seasonal last-millennium reanalysis with online DA and sea ice, but the pre-instrumental skill claim is thinner than it looks; still a solid contribution worth serious peer review. read the letter →

arxiv 2501.14130 v1 pith:BXH3CMIF submitted 2025-01-23 physics.ao-ph physics.data-an

classification physics.ao-phphysics.data-an
keywords paleoclimatedataassimilationseasonalresolutionlastmillenniumlinearinversemodelseaicereconstructionElNiñodiversityproxysystemonline
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

This paper introduces a data-assimilation reconstruction of the last millennium that resolves the seasonal cycle (March–May, June–August, September–November, December–February) for coupled atmosphere, ocean, and sea-ice fields. Its central claim is that an "online" assimilation scheme, in which a linear inverse model forecasts the climate state from one season to the next, uses proxy information more efficiently than earlier offline reconstructions: it attains the highest correlation skill in surface-temperature verification against instrumental records while assimilating far fewer proxies than competing products do, especially in boreal winter when proxy coverage is sparse. The paper further claims that reconstructed upper-ocean heat content, Arctic sea-ice concentration, and the seasonal evolution of El Niño events verify well against independent instrumental and satellite data, and that the reconstruction shows consistent skill in pre-instrumental epochs when checked against withheld proxies. If these claims hold, the result matters because it offers a seasonal-resolution, physically coupled view of the last millennium, and because the season-to-season update mechanism explains why summer-biased tree-ring proxies no longer dilute winter and annual reconstructions.

What carries the argument

The engine is a season-to-season Linear Inverse Model (LIM) whose state vector holds principal components of 2-meter temperature, sea surface temperature, upper-300-meter ocean heat content, and Northern-Hemisphere sea-ice concentration and thickness, trained on last-millennium simulations of two CMIP5 climate models. The LIM supplies the forecast prior for an ensemble square-root Kalman filter with 800 members and no localization; linear proxy system models, calibrated on modern instrumental temperatures in the EOF-truncated space with a diagonal error covariance, map proxies into that prior. The defining mechanism is the seasonal update strategy: a proxy with, say, June–August seasonality updates only the June–August ensemble, while annual-mean proxies are assimilated only once an annual mean can be formed from the seasonal forecasts, so the LIM's memory carries information from proxy-rich summer into proxy-poor winter.

What would settle it

A decisive test would be to run the bootstrap verification while withholding all proxies earlier than 1200 CE; if pre-1200 correlations for non-assimilated proxies collapse toward zero, the claim of consistent skill throughout the last millennium fails. A second, sharper test would compare the DJF temperature field over 1400–1700 against an independent winter-sensitive reconstruction excluded from the proxy database, requiring agreement within the stated ensemble spread.

Watch

Extended reading notes

Core claim

The core discovery is that cycling a linear inverse model forward season to season, and assimilating each proxy in the season it actually represents, produces a last-millennium reconstruction whose verified skill exceeds that of earlier paleo-data-assimilation products even though roughly one quarter as many proxies are used. Global-mean correlation against instrumental surface temperature is about 0.54 for annual means (versus 0.47–0.53 for three earlier reconstructions), and the largest edge comes in boreal winter (about 0.43 versus 0.37), a season with few tree-ring records. Reconstructed upper-ocean heat content correlates near 0.2 with an objective ocean analysis, Arctic sea-ice concentration correlates 0.11–0.21 with satellite data, and the Niño3.4 index correlates about 0.78 with a sea-surface-temperature analysis (coefficient of efficiency around 0.55). The paper attributes the winter advantage to the online scheme: summer proxy information persists through the seasonal forecast and updates the prior for data-poor seasons. It also shows that when seasonal proxies are restricted to updating annual means only, the reconstructed global-mean temperature difference between the Medieval Climate Anomaly and the Little Ice Age shrinks from 0.15 °C to 0.10 °C, arguing that the seasonal update is essential for recovering multicentennial variability.

Load-bearing premise

The reconstruction's pre-instrumental skill rests on the assumption that the linear proxy system models calibrated against twentieth-century instrumental temperatures stay valid with the same error statistics throughout 850–1850, so that the mapping from climate state to proxy and the assumed error covariances do not drift over time.

Editorial extensions

If this is right

  • If the skill claim is correct, future paleoclimate reanalyses can achieve equal or better verification with substantially smaller proxy networks, cutting the data burden for seasonal reconstructions.
  • Seasonal fields for ocean heat content and Arctic sea ice become available across the whole millennium, enabling direct study of seasonal sea-ice evolution and its links to temperature and orbital forcing.
  • The millennium-length ENSO reconstruction, verified against four El Niño onset classes in the twentieth century, provides a large sample for studying changes in El Niño diversity and seasonal evolution.
  • Because the season-to-season update yields a larger Medieval Climate Anomaly-to-Little Ice Age temperature difference than annual-mean updating, earlier annual reconstructions may have underestimated multicentennial variability by letting summer-biased proxies dilute winter and annual signals.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • An extension the authors do not pursue is adding atmospheric circulation or precipitation principal components to the LIM state; the paper's exclusion of high-frequency variables to protect forecast skill suggests this would require a careful balance between state dimension and memory.
  • Because the LIM is trained on two CMIP5 models, a natural test is to train the same system on a CMIP6 last-millennium simulation; the paper's claim that CMIP5-to-CMIP6 seasonal variability statistics are similar predicts verification skill would change little.
  • The winter-skill advantage is attributed to dynamical memory in the trained LIM, which implies a testable prediction: a model with weaker seasonal memory would show a smaller winter edge, so training the LIM on a model with poorly simulated ENSO persistence should degrade DJF skill more than JJA skill.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The manuscript introduces LMR Seasonal, a seasonal-resolution 'online' data assimilation reconstruction of the last millennium. A linear inverse model (LIM) trained on CMIP5 last-millennium simulations provides coupled forecasts of surface temperature, sea surface temperature, upper-ocean heat content, and Northern Hemisphere sea-ice concentration/thickness; an ensemble square-root Kalman filter assimilates PAGES2k v2 proxies with season-specific update timing. The paper claims that this reconstruction achieves the highest correlation skill against instrumental products while using fewer proxies, that verification against held-out proxies demonstrates robust pre-instrumental skill, and that the method captures ENSO evolution and the MCA-LIA contrast better than existing off-line reconstructions.

Significance. If the central claims hold, LMR Seasonal would be a significant advance: it is the first seasonal-resolution, coupled ocean-atmosphere-sea-ice reanalysis of the last millennium, and its online update strategy is a clear methodological improvement over static-prior off-line DA for seasonal-to-interannual memory. The manuscript has concrete strengths: it compares against three established paleo-DA products, verifies variables not directly assimilated (OHC300, sea-ice concentration) against independent observational datasets, and demonstrates sensible ENSO composites. However, the headline comparative claim rests on global-mean correlation differences of 0.01-0.07 without uncertainty estimates, and the pre-instrumental verification using withheld proxies is circular with respect to the stationarity of the proxy system models. These issues make the paper a strong candidate for major revision rather than acceptance in its current form.

major comments (3)
  1. [§3.1, Figs. 3-4] The abstract and Section 3.1 claim that LMR Seasonal 'achieves the highest correlation skill' compared to other paleo-DA products. The evidence is global-mean correlation differences of 0.01 (vs. LMRv2), 0.04 (vs. PHYDA), and 0.07 (vs. LMR Online) for annual-mean temperature (Fig. 3), and 0.06 (DJF) and 0.01 (JJA) vs. PHYDA (Fig. 4). No confidence intervals, significance tests, or bootstrap estimates are provided for these differences. Given that the verification period 1880-2000 substantially overlaps the PSM calibration period (GISTEMP/ERSST, also 20th-century), these small differences may reflect sampling variability or calibration artifacts. The authors should provide uncertainty bounds on the correlation differences and, where possible, a verification period that does not overlap calibration.
  2. [§3.2, Eq. (7), Fig. 11] The claim that 'verification against independent proxy records shows that reconstruction skill is robust throughout the last millennium' is not supported by the bootstrap experiment as designed. The 20% withheld proxies are evaluated by forward-modeling them from the reconstructed climate state using the same modern-calibrated PSM (Eq. 7). As the authors correctly note in Section 2.2, the PSMs are linear and calibrated on GISTEMP/ERSST. A multiplicative drift in the true proxy-climate sensitivity is invisible to this procedure: if the calibrated sensitivity is c times the true sensitivity and the assimilated proxies share this bias, the EnKF analysis (Eq. 11 in the low-observation-error limit) underestimates the temperature by roughly 1/c, and the forward model maps this underestimated temperature back to approximately the correct proxy value. Thus high holdout correlation and even high CE do not establish stationarity. The conclusion in Section 3.2 that 'the distribution of correlation values... is very similar, suggesting a robust PSM relationship' is therefore circular. A valid test would require, for example, calibration on one time period and validation on a non-overlapping period, pseudo-proxy experiments in which PSM parameters are perturbed or made time-varying, or comparison against independent non-PSM-based reconstructions.
  3. [§4, Fig. 13] The paper attributes the MCA-LIA difference (0.15°C) to the seasonal-update strategy based on a single experiment in which seasonal proxies update only the annual mean (0.10°C). No uncertainty estimate is given for either value, yet the conclusion 'our seasonal-update strategy appears to be essential to reconstructing the magnitude of the MCA-LIA difference' depends on the 0.05°C difference being statistically distinguishable from ensemble spread and from natural variability. The authors should report confidence intervals on the MCA-LIA difference for both the seasonal and annual-update experiments and test whether the difference between the two experimental outcomes is significant.
minor comments (5)
  1. [Fig. 3 caption] The word 'shwon' in the caption is a typo for 'shown'.
  2. [§2.2, references to supplementary figures] The text states that independent-proxy verification results are 'compare Fig. 11 with Supplementary Fig. S8', but Supplementary Fig. S8 is the Niño3.4 PHYDA comparison, not the proxy-verification figure. The intended cross-reference is likely Supplementary Fig. S9.
  3. [Throughout] The manuscript inconsistently uses 'El Niño', 'Nino3.4', and 'Niño3.4' (e.g., Figs. 8-10 and text). Please standardize the notation.
  4. [§2.3] The EnSRF update equations (Eqs. 10-12) are presented for the serial observation update but the text does not specify how observations are ordered or whether the ensemble is re-orthogonalized after each observation; a sentence clarifying the implementation would help reproducibility.
  5. [Data availability] The data availability statement says the reconstruction data and code 'will be released to the public once this manuscript has been accepted.' For a journal that values reproducibility, releasing code and data at the time of submission (or at least in a preprint repository) would strengthen the paper and allow readers to verify the claims.

Circularity Check

1 steps flagged · score 4.0 of 10

Pre-instrumental proxy verification reuses the modern-calibrated PSMs, so the key stationarity assumption is not independently tested; skill claims are otherwise substantially supported by independent OHC and sea-ice verification.

  1. fitted input called prediction [Abstract; Section 3.2 (Independent Proxy Verification), with Eq. (7) from Section 2.2]
    "Verification against independent proxy records shows that reconstruction skill is robust throughout the last millennium. ... For each epoch, the proxies are forward modeled from the reconstructed climate states using the PSM (7) for each proxy, yielding a direct comparison of the LMR Seasonal reconstruction to both the assimilated and independent proxy chronologies."

    The withheld-proxy verification is not independent of the PSM calibration. The PSMs (Eq. 7) are fitted to modern GISTEMP/ERSST data (Section 2.2), and the same H is used both in the EnKF update (Eqs. 8-11) and to forward-model the withheld proxies in Section 3.2. If the true pre-instrumental proxy-climate sensitivity differs from the calibrated sensitivity by a factor c, the analysis state is biased by roughly 1/c (low observation-error limit of Eq. 11), and the forward model maps that biased state back through the same biased H, yielding predicted proxy values close to the observed ones. The reported correlation and CE can remain near 1 while the reconstructed temperature is substantially biased.

full rationale

The paper's core machinery—LIM training on CMIP5 last-millennium simulations, EOF truncation, EnKF update, and the seasonal update strategy—is internally consistent and not circular: the LIM forecast skill is tested out-of-sample on the other model (Supplementary Figs. S1–S2), and the seasonal-update vs annual-update experiment (Section 4) is a controlled comparison, not a redefinition. The main circular burden is concentrated in the verification of pre-instrumental skill. Section 3.2 withholds 20% of proxies from assimilation but evaluates them through the same modern-calibrated linear PSMs (Eq. 7) used to produce the reconstruction, so the test cannot detect a common multiplicative bias in the proxy-climate relationship. Instrumental verification (Section 3.1) also overlaps the PSM calibration era and uses correlated instrumental products (GISTEMP/ERSST for calibration vs HadCRUT5/ERA-20C/HadISST for verification), further weakening the independence of the claimed skill. However, the paper provides genuinely independent support that is not circular: OHC300 and sea-ice concentration are not assimilated and yet show positive correlations with EN4 and satellite data (Figs. 5–6), and the ENSO composite comparisons use instrumental data that were not directly used in PSM calibration for those fields. These independent checks justify a moderate circularity score rather than a high one. Overall, the central claim is not reduced to its inputs by construction, but the key stationarity assumption is supported by a holdout test that is partially circular.

Assumptions & free parameters 8 free parameters · 7 assumptions · 0 invented entities

No new physical entities are invented. The free parameters are standard data-assimilation design choices (PC truncation, ensemble size, filter thresholds). The load-bearing assumptions are stationarity of the linear proxy system models calibrated on modern data and representativeness of the CMIP5-trained LIM; both receive partial support from out-of-sample tests.

free parameters (8)
  • EOF truncation rank for TAS, TOS, SIT, SIC = 15 PCs each (about 80% variance)
    State vector dimension in the LIM and EnKF; chosen to balance variance explained against overfitting.
  • EOF truncation rank for OHC300 = 30 PCs
    Follows Perkins and Hakim (2020) to capture extended upper-ocean memory.
  • PSM calibration correlation threshold = 0.05
    Proxies with weaker local temperature correlation are removed before assimilation.
  • PSM residual one-year lag autocorrelation threshold = 0.90
    Removes proxies with temporally correlated observation errors, which violate EnKF assumptions.
  • Ensemble size = 800
    Chosen to avoid localization and inflation methods in the EOF state space.
  • LIM integration time step = 6 hours
    Numerical discretization step in the stochastic LIM integration.
  • LIM lag for training covariance = 3 months
    Seasonal forecast step; C(tau) with tau = 3 months in Eq. (3).
  • Bootstrap holdout fraction and epochs = 20%, 50 epochs
    Design choices for independent proxy verification in Section 3.2.
assumptions (7)
  • domain assumption The coupled climate state follows a stable linear stochastic differential equation with stationary Gaussian noise.
    Eq. (1) in Section 2.1; the LIM forecast prior is valid only if linearity and stationarity hold on seasonal time scales.
  • domain assumption The truncated EOF basis (15 PCs for most variables, 30 for OHC300) captures enough variance for the reconstruction, and truncation error is absorbed by the PSM.
    Sections 2.1-2.2; the state space defines what the EnKF can represent and what the PSM maps from.
  • domain assumption Proxy system models are linear and calibrated on modern instrumental data remain valid over the last millennium.
    Section 2.2 Eq. (7) and Section 3.2; the pre-instrumental verification relies on this stationarity.
  • domain assumption CCSM4 and MPI-ESM-R last-millennium simulations provide statistics representative of real climate variability from 850 to 1850.
    Section 2.1; the LIM is trained on these models and never sees observations.
  • domain assumption Observation errors of different proxies are uncorrelated, so a diagonal R matrix is sufficient.
    Sections 2.2-2.3; off-diagonal covariances are not estimable from the calibration period.
  • domain assumption An 800-member ensemble has small enough sampling error to proceed without localization.
    Section 2.3; no localization or inflation is used, and the Southern Ocean degradation is attributed to teleconnection bias.
  • domain assumption Objective seasonality assignment identifies the season each proxy represents.
    Section 2.2; expert and objective seasonality give similar results, but any misassignment would bias the seasonal update.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Coupled Seasonal Data Assimilation of Sea Ice, Ocean, and Atmospheric Dynamics over the Last Millennium." pith.science (2026). https://pith.science/paper/BXH3CMIF

@misc{pith2026250114130,
  author       = {Pith},
  title        = {Pith review of: Coupled Seasonal Data Assimilation of Sea Ice, Ocean, and Atmospheric Dynamics over the Last Millennium},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/BXH3CMIF}},
  note         = {Machine review of arXiv:2501.14130}
}
read the original abstract

``Online" data assimilation (DA) is used to generate a new seasonal-resolution reanalysis dataset over the last millennium by combining forecasts from an ocean--atmosphere--sea-ice coupled linear inverse model with climate proxy records. Instrumental verification reveals that this reconstruction achieves the highest correlation skill, while using fewer proxies, in surface temperature reconstructions compared to other paleo-DA products, particularly during boreal winter when proxy data are scarce. Reconstructed ocean and sea-ice variables also have high correlation with instrumental and satellite datasets. Verification against independent proxy records shows that reconstruction skill is robust throughout the last millennium. Analysis of the results reveals that the method effectively captures the seasonal evolution and amplitude of El Ni\~{n}o events. Reconstructed seasonal temperature variations are consistent with trends in orbital forcing over the last millennium.

Figures

Figures reproduced from arXiv: 2501.14130 by the authors.

Figure 1
Figure 1. Proxies from PAGES2k V2 (PAGES2k Consortium and others, 2017). a. Locations and counts of proxy types after filtering by the specified standards indicated by the Subsection 2.2. b. Evolution of the number of proxies over time. c–d. Spatial distribution of PSM calibration correlations and 1-year lag residual (error) auto-correlations. perturbations x ′ 𝑖 , are update by x ′ 𝑎 = x ′ 𝑝 − " 1+ √︄ 𝑅𝑘 var(𝑦𝑒,𝑘) + 𝑅𝑘 # −1 … view at source ↗
Figure 2
Figure 2. LMR Seasonal update strategy. The light blue box represents the ensemble, the rose box the proxy, and the pink arrow the forecast step from the LIM. Curly brackets denote the update from the EnKF to integrate the proxy data into updating the prior ensemble. The text within the box indicates the seasonality of either the ensemble or the proxies. update mechanism. This novel update strategy has a significant impact on… view at source ↗
Figure 3
Figure 3. Annual mean surface temperature instrumental verification. a–d. Correlation between various DA reconstructions and HadCRUT5 (Morice et al., 2021) 2m air temperature during 1880-–2000. Results are shown for (a LMR Seasonal, b LMRv2, c PHYDA, d LMR Online) with the global-mean correlation and the number of used proxies given in the title for each subpanel. Correlation difference between LMR Seasonal (a) and other reco… view at source ↗
Figures from the paper (11 more)
Figure 4
Figure 4. Figure 4: Surface temperature seasonal instrumental verification a–d. The correlations between the LMR Seasonal and HadCRUT5 (Morice et al., 2021). For DJF (a), MAM (c), JJA (b), and SON (d) during 1880–2000. e–f. Correlations between PHYDA and HadCRUT5 are shown for DJF (e) and…
Figure 5
Figure 5. Figure 5: Ocean heat content from 300m to the surface (OHC300) instrumental verification. Correlation between LMR Seasonal OHC300 and HadleyEN4 OHC300 (Good et al., 2013) over the period 1940—2000 for the annual mean (a), DJF (b), JJA (c), MAM (d), and SON (e). Global-mean corre…
Figure 6
Figure 6. Figure 6: Northern Hemisphere sea-ice concentration (SIC) instrumental verification. Correlation between the LMR Seasonal SIC and satellite SIC data (Fetterer et al., 2017) during 1980–2000 are presented for the annual mean (a), DJF (b), JJA (c), MAM (d), and SON (e). Global-mea…
Figure 7
Figure 7. Figure 7: Global mean surface temperature (GMT) instrumental verification. Temporal verification of the ensemble-mean LMR Seasonal reconstructed GMT series (colored curves) against HadCRUT5 (Morice et al., 2021) GMT (black solid curve) in annual mean (a), DJF (b), MAM (c), JJA (…
Figure 8
Figure 8. Figure 8: Niño3.4 index instrumental verification. Temporal verification of the ensemble mean LMR Seasonal reconstructed Niño3.4 Index (colored curves) against HadISST (Rayner et al., 2003) (black solid curve) in all seasons (a), DJF (b), MAM (c), JJA (d), and SON (e). Dark shad…
Figure 9
Figure 9. Figure 9: Verification of the four classes of El Niño onset evolution of Wang et al. (2019) during 1900–2000. The left column displays composite analyses from HadISST (Rayner et al., 2003), and the right column shows the LMR Seasonal Reconstruction. Rows show composite averages …
Figure 10
Figure 10. Figure 10: Verification of four Strong Basin-Wide (Super) El Niño Cases’ onset evolution (1902, 1972, 1982, and 1997). The left column shows the evolution in HadISST (Rayner et al., 2003), and the middle column represents the LMR Seasonal Reconstruction. The right column depicts…
Figure 11
Figure 11. Figure 11: Verification of LMR Seasonal against assimilated and non-assimilated proxy data. The top row illustrates the distribution of correlation values between proxy values and LMR Seasonal estimates for assimilated (a) and non-assimilated (b) proxy data from 1880–2000 (blue)…
Figure 12
Figure 12. Figure 12 [PITH_FULL_IMAGE:figures/full_fig_p017_12.png]
Figure 13
Figure 13. Figure 13: Differences Between the Medieval Climate Anomaly (MCA, 950CE–1250CE) and Little Ice Age (LIA, 1400CE–1700CE) in four DA Reconstructions. a. Global Mean Surface Temperature (GMT) 20-year running mean in LMR Seasonal (red), LMRv2 (yellow), LMR Online (green), and PHYDA …
Figure 14
Figure 14. Figure 14: Time series of Arctic sea-ice area (a) and volume (b), and upper 300m ocean heat content anomaly (c) over the last millennium. The solid colored lines represent the ensemble mean, black solid lines denote the 30-year running means, dark shading the interquartile range…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

  1. [1]

    J., and Coauthors, 2017: Last millennium Northern Hemisphere summer temperatures from tree rings: Part II, spatially resolved reconstructions.Quaternary Science Reviews,163, 1–22

    Anchukaitis, K. J., and Coauthors, 2017: Last millennium Northern Hemisphere summer temperatures from tree rings: Part II, spatially resolved reconstructions.Quaternary Science Reviews,163, 1–22. Anderson, J. L., 2012: Localization and sampling error correction in ensemble Kalman filter data assimilation.Monthly Weather Review,140 (7), 2359–2371. Blanchar...

  2. [4]

    Magorian, 1993: Prediction of Niño 3 sea surface temperatures using linear inverse modeling

    Penland, C., and T. Magorian, 1993: Prediction of Niño 3 sea surface temperatures using linear inverse modeling. Journal of Climate,6 (6), 1067–1076. Penland, C., and L. Matrosova, 1994: A balance condition for stochastic numerical models with application to the El Niño-Southern Oscillation.Journal of climate,7 (9), 1352–1372. Penland, Cécile and Sardeshm...

  3. [11]

    Li, 2024: Why is the Pacific meridional mode most pronounced in boreal spring?Climate Dynamics, 62 (1), 459–471

    Meng, Z., and T. Li, 2024: Why is the Pacific meridional mode most pronounced in boreal spring?Climate Dynamics, 62 (1), 459–471. Meng, Z., F. Zhu, and G. J. Hakim, 2023: Sacpy–A Python Package for Statistical Analysis of Climate.AGU23. Morice, C. P., and Coauthors, 2021: An updated assessment of near-surface temperature change from 1850: The HadCRUT5 dat...

  4. [12]

    1.00 0.75 0.50 0.25 0.00 0.25 0.50 0.75 1.00 Correlation Figure S9: Same as Figure 11, but for the expert-seasonality based PSM. 32 Last Millennium Reanalysis (LMR) Seasonal ANN DJF MAM JJA SON 0.4 0.3 0.2 0.1 0.0 Mean Trend [ C/kyr] a MPI MPI-DA 95% - 5% ANN D J F M A M J J A S O N 0.4 0.2 0.0 Mean Trend [ C/kyr] b insolation trend 95% - 5% MMM Ann MMM 9...

  5. [36]

    Luo, Y.-M

    Wang, B., X. Luo, Y.-M. Yang, W. Sun, M. A. Cane, W. Cai, S.-W. Yeh, and J. Liu, 2019: Historical change of El Niño properties sheds light on future changes of extreme El Niño.Proceedings of the National Academy of Sciences, 116 (45), 22512–22517. Whitaker, J. S., and T. M. Hamill, 2002: Ensemble data assimilation without perturbed observations.Monthly we...

  6. [59]

    K., 2022:Reconstructing Arctic Sea Ice in the Instrumental Era

    Brennan, M. K., 2022:Reconstructing Arctic Sea Ice in the Instrumental Era. University of Washington. Brennan, M. K., and G. J. Hakim, 2022: Reconstructing Arctic sea ice over the Common Era using data assimilation. Journal of Climate,35 (4), 1231–1247. Briffa, K. R., P. D. Jones, and F. H. Schweingruber, 1992: Tree-ring density reconstructions of summer ...

  7. [261]

    Laskar, J., P

    Huang, B., and Coauthors, 2017: Extended reconstructed sea surface temperature, version 5 (ERSSTv5): upgrades, validations, and intercomparisons.Journal of Climate,30 (20), 8179–8205. Laskar, J., P. Robutel, F. Joutel, M. Gastineau, A. C. Correia, and B. Levrard, 2004: A long-term numerical solution for the insolation quantities of the Earth.Astronomy & A...

  8. [499]

    26 Last Millennium Reanalysis (LMR) Seasonal a LMR Seasonal (mean = 0.54, Nproxy =

    tree.TRW (n=275) tree.MXD (n=61) speleothem.d18O (n=3) marine.diatom (n=1) marine.d18O (n=1) marine.alkenone (n=4) marine.MgCa (n=3) lake.varve_thickness (n=5) lake.reflectance (n=2) lake.midge (n=1) lake.chironomid (n=1) lake.accumulation (n=1) lake.BSi (n=1) ice.melt (n=1) ice.dD (n=7) ice.d18O (n=28) hybrid (n=1) documents (n=3) coral.d18O (n=63) coral...

Show all 12 references
  1. [545]

    R" and“CE

    f LMR Seasonal - LMRv2 (mean = 0.02) g LMR Seasonal - PHYDA (mean = 0.03) h LMR Seasonal - LMR Online (mean = 0.06) 0.9 0.6 0.3 0.0 0.3 0.6 0.9 Correlation (vs. 20C Reanalysis) 0.45 0.30 0.15 0.00 0.15 0.30 0.45 Correlation Difference Figure S6: As in Figure 3, but for the cor...

  2. [624]

    Emile-Geay, G

    Zhu, F., J. Emile-Geay, G. J. Hakim, D. Guillot, D. Khider, R. Tardif, and W. A. Perkins, 2024: cfr (v2024. 1.26): a Python package for climate field reconstruction.Geoscientific Model Development,17 (8), 3409–3431. 23 Last Millennium Reanalysis (LMR) Seasonal Supporting Infor...

  3. [747]

    Emile-Geay, K

    Zhu, F., J. Emile-Geay, K. J. Anchukaitis, N. P. McKay, S. Stevenson, and Z. Meng, 2023: A pseudoproxy emulation of the PAGES 2k database using a hierarchy of proxy system models.Scientific Data,10 (1),

  4. [1880]

    e Non-assimilated proxies' mean (1880-2000) f Non-assimilated proxies' mean (before

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.