Pith. sign in

REVIEW 4 major objections 6 minor 58 references

Probabilistic Deep Learning for Drought Forecasting: Role of Internal Climate Variability

T0 review · 4 major / 6 minor · reviewed 2026-08-04 · deepseek-v4-flash

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

Pith's one-line read Drought forecasts gain sharper risk bounds by treating internal climate variability as a learnable forecast quantity.

desk verdict A good idea and clean implementation, but the headline calibration claim is not established: the key assumption linking ensemble spread to real-world forecast uncertainty is never directly tested, and the paper's own long-run benchmark suggests the bound is systematically too wide. read the letter →

arxiv 2608.01864 v1 pith:3JORKNX5 submitted 2026-08-03 stat.AP cs.LGphysics.ao-ph

Henri Funk , Cornelia Gruber , Göran Kauermann , Helmut Küchenhoff , Magdalena Mittermeier This is my paper
classification stat.APcs.LGphysics.ao-ph
keywords droughtforecastinginternalclimatevariabilitylargeensembledeeplearningTemporalFusionTransformerSPEIuncertaintyquantificationEurope
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 argues that the internal climate variability that makes drought forecasts uncertain is not unstructured noise: it has spatial, seasonal, and temporal structure that can be learned from a 50-member climate model ensemble. The authors build a one-month-ahead European drought forecast with a Temporal Fusion Transformer trained on reanalysis data, then derive a lower uncertainty bound from the spread of forecasts across ensemble members after isolating internal variability through pairwise residual differencing. On the out-of-sample period 2020–2024, this ensemble-informed bound is better calibrated than the model's own reanalysis-trained quantile bound, especially under drought conditions where the reanalysis bound underestimates lower-tail risk. The central claim is that large ensembles can transfer physically plausible variability into machine-learning forecasts, yielding risk-aware drought bounds that a single historical record cannot provide.

What carries the argument

The key identity is the pairwise-differencing relation for forecast residuals. For ensemble members i and j, the residual epsilon_i = true water balance - forecast is decomposed into a common bias psi and an internal-variability term theta_i. The difference delta_ij = epsilon_i - epsilon_j cancels the bias, and under zero-mean, independent, equal-variance assumptions on the theta terms, E[delta_ij^2] = 2 sigma^2, where sigma is the spatio-temporal internal variability. The paper models the squared pairwise differences with a generalized additive model (GAM) with penalized splines in space and time, giving a smooth estimate sigma_hat, and constructs the lower bound as mu_hat - z_{1-alpha} * s

What would settle it

A falsifying observation would be a long or independent verification period in which the nominal 10% large-ensemble bound is exceeded by real observations at rates well above 10% in non-mountainous, non-marine regions, or a multi-model large ensemble that yields materially different sigma_hat than CRCM5-LE; either would indicate that the bias-cancellation or representativeness assumption does not hold.

Watch

Extended reading notes

Core claim

The paper's core discovery is that using climate-model ensemble simulations not as direct observations but as a training signal for the scale and spatio-temporal structure of internally generated forecast variability yields drought risk bounds that are better calibrated than bounds derived from a single reanalysis history. Concretely, applying the ERA5-trained forecasting model to each of 50 CRCM5 ensemble members produces a set of residual error fields; pairwise differences of these residuals cancel the common bias from applying the model across domains, leaving the unexplained internal variability. Modeling the expected squared pairwise differences as a smooth function of space and time gi

Load-bearing premise

The ensemble bound's validity rests on the assumption that the bias from applying an ERA5-trained forecasting model to CRCM5-LE simulations is identical across the 50 members (so pairwise differencing cancels it) and that CRCM5-LE's internal variability faithfully represents real-world forecast variability; if either fails, the bound is miscalibrated.

Editorial extensions

If this is right

  • During the 2020–2024 test period, the large-ensemble lower bound is better calibrated than the reanalysis-only bound in six of eight European regions and in every season at the pan-European level, and at the regional scale it captures all drought events for the Iberian Peninsula and Mediterranean.
  • The reanalysis-based bound fails to detect 74.2% of drought events and 90.3% of extreme drought events, while the ensemble bound misses about half of drought events and 71.4% of extreme events, showing that historical tail widths regress to average and cannot account for anomalous dry spells.
  • Internal variability should be treated as a forecast quantity in its own right rather than irreducible noise, so uncertainty bounds can adapt to shifting climate states instead of being fixed to historical variability.
  • The framework is transferable to other regions, drought indices, accumulation windows, and multistep lead times, and the trade-off between conservative lower-tail coverage and predictive sharpness can be controlled by the bound level alpha.
  • Tightening the bound level to 5% or 2.5% widens the lower tail and improves detection of extreme drought where the reanalysis bound collapses.

Reading between the lines

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

  • If internal variability is a learnable forecast quantity, the same pairwise-differencing scheme could be applied to extract variability-aware uncertainty bounds for other climate hazards (e.g., heatwaves, floods) from any single-model initial-condition large ensemble, provided the bias-cancellation assumption holds.
  • The paper's single-model ensemble leaves structural model uncertainty unaddressed; comparing sigma_hat across multiple large ensembles driven by different global models could reveal where CRCM5-LE's variability is systematically too narrow or too wide — the Alpine and British Isles failure modes may be early symptoms.
  • The normal-quantile construction of the lower bound is a convenience; a GEV or Pareto tail might improve calibration at extreme-drought thresholds, something the current data cannot confirm given the short test period.
  • A natural testable extension is to evaluate the same ensemble-informed bound on a longer hindcast or on a different reanalysis product to see whether the coverage improvement persists when the test period is not dominated by a record drying trend.
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

4 major / 6 minor

Summary. The paper develops a one-month-ahead probabilistic drought forecasting framework for Europe. A Temporal Fusion Transformer is trained on ERA5-Land water balance (1970–2019) to produce a central forecast, and two lower 10% bounds are compared: (i) a reanalysis-based bound taken directly from TFT quantile estimates, and (ii) an ensemble-based bound that subtracts a z-quantile times an estimated standard deviation obtained from pairwise differences of CRCM5-LE forecast residuals, with the squared differences modelled by a spatio-temporal GAM. Both bounds are transformed to SPEI-1 and evaluated on the unseen period 2020–2024 across eight PRUDENCE regions and at grid-cell level. The authors report that the ensemble-informed bound is better calibrated than the reanalysis bound in most regions and seasons, particularly for lower-tail drought risk, and conclude that internal climate variability should be treated as a forecastable quantity rather than unstructured noise.

Significance. If the result holds, the paper offers a transferable method for injecting large-ensemble information into machine-learning-based drought forecasts, with a clear practical benefit for risk-averse planning. The conceptual claim—that internal variability has learnable spatio-temporal structure that can improve lower-tail drought bounds—is interesting and goes beyond standard quantile-regression approaches. The paper also provides detailed hyperparameter tables, a full derivation of the pairwise-differencing estimator, and an honest discussion of failure modes (Alps, British Isles). However, the empirical evidence is currently too thin to support the headline claim: the evaluation is restricted to a single five-year dry period, the key equivalence between CRCM5-LE spread and real-world forecast-error variance is asserted rather than tested, and the reported full-period results in Table 6 are not reconciled with the calibration claim.

major comments (4)
  1. [Section 3.1, Table 2; Section 3.2, Table 3] The evaluation is based on one five-year test period (60 monthly values per region at regional level) and reports exceedance rates without confidence intervals or significance tests. At the nominal 10% level, with n=60, the exact binomial 95% CI is approximately 4.1%–19.5%, so a regional value of 23.3% (Alps, Large Ensemble) is outside that interval, while many other reported differences between Large Ensemble and Reanalysis (e.g., Eastern Europe 13.3% vs 10.0%) are within sampling noise. The seasonal grid-cell analysis in Table 3 has no uncertainty quantification at all, despite strong spatial and temporal dependence. Because the central claim is 'better calibrated across most regions and seasons,' the paper needs at least block-bootstrap confidence intervals, a paired test of exceedance rates, or a clearly reported spatio-temporal clustering procedure.
  2. [Section 2.3, Eqs. (5)–(7); Section A.1, Eqs. (8)–(10); Table 6] The ensemble bound assumes that the variance of CRCM5-LE residuals, after pairwise differencing, equals the variance of real-world forecast errors of the ERA5-trained model. This requires (a) CRCM5-LE internal variability to match real-world variability in amplitude and spatio-temporal structure, and (b) the forecast model's error behaviour on CRCM5 predictors to match that on ERA5 predictors. Neither assumption is tested. The paper's own Table 6 provides indirect evidence against correct sizing: over 1970–2024, the Large-Ensemble bound has below-bound rates of 0.26%–7.32% across regions, far below the nominal 10%, indicating that the bound is systematically too wide (overly conservative) during the calibration period. The 2020–2024 improvement is thus compatible with a mechanically wider bound capturing more observations during an anomalously dry period. A direct validation is needed, e
  3. [Section 2.3, Eq. (7); Section 4] The quantity sigma estimated from CRCM5-LE residual spread is labelled 'forecast variability,' but CRCM5-LE is a climate large ensemble, not an initialized forecast ensemble. The spread of uninitialized or forced climate trajectories does not, by itself, correspond to one-month-ahead forecast error variance under real-world initial conditions. The paper should either reframe the claim as using a perfect-model proxy (and then test that proxy against observed forecast residuals) or justify why ensemble spread of this type is the relevant quantity for short-lead forecast uncertainty. Without such a test, the sentence in the conclusion that 'internal variability is treated as a forecast quantity in its own right' is a conceptual overreach.
  4. [Section 2.3, Eq. (7); Section A.3] The lower bound in Eq. (7) uses a normal quantile z_{1-alpha} to define a 10% lower bound on water balance. The authors justify this by a central limit theorem argument, but no diagnostic is presented that monthly forecast residuals are approximately normal, and the subsequent SPEI-1 transformation is monotonic and will not repair a misspecified quantile. Heavy tails or skewness in water-balance residuals would directly bias the exceedance rates reported in Tables 2 and 3. The paper should provide a quantile–quantile plot or a formal normality test for the residuals used in Eq. (7), or replace the normal quantile with an empirical quantile from the residual distribution.
minor comments (6)
  1. [Abstract] The phrase 'better calibrated across most regions and seasons' should be quantified in the abstract (e.g., give pan-European exceedance rates for the two bounds).
  2. [Section 2.1] Please clarify the regridding: all variables are bilinearly interpolated to the CRCM5 grid, but for the ERA5-trained TFT, are the ERA5 predictors also regridded to 0.11°? This affects the comparability of the central forecast and the ensemble forecasts.
  3. [Table 4 caption] The caption says 'test loss' is evaluated on 'the independent holdout period 2020–2024.' Since hyperparameters were selected using a validation period, clarify the relationship between the validation period (presumably within 1970–2019) and the independent test period, and state whether the reported test loss uses the final selected hyperparameters only.
  4. [Figure 4] The hatching marks grid cells with exceedance between 0 and 20%, but the colour scale already maps 0–10% and 10–20% to very different colours. This makes the hatching redundant and potentially confusing. Please adjust the legend or hatching to represent the 'ideal calibration' region more clearly.
  5. [Section A.5 and Table 6] Table 6 is labelled 'full period 1970–2024' and includes the training period of the TFT. Since the forecast model was fit on 1970–2019, the full-period exceedance rates are not a valid out-of-sample calibration check. Please state explicitly that this table is descriptive/in-sample and should not be used to assess calibration.
  6. [Section 4] The statement that 'The years 2020–2024 include drought conditions that exceed the ERA5-Land record' is load-bearing for the interpretation of the test period but is not supported by a figure or quantitative comparison (e.g., against 1970–2019 SPEI minima). Please add supporting evidence or a reference.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the ensemble-based uncertainty estimate is derived from CRCM5 residuals and its calibration is assessed against independent ERA5 observations, so the central claim is not equivalent to its inputs by construction.

full rationale

The paper's derivation chain is transparent and non-circular. The uncertainty quantity σ is estimated from pairwise differences of CRCM5 forecast residuals (Eqs. 4–6, 8–10). This is a variance identity: E[(υ_i − υ_j)^2] = 2σ² under stated assumptions, not a quantity defined in terms of the target outcome. The lower bound in Eq. (7) then combines this CRCM5-derived σ with the ERA5-trained central forecast μ̂. The calibration claim is evaluated by comparing the fraction of observed ERA5-Land SPEI-1 values falling below the bound against the nominal 10% level (Tables 2, 3, 6; Figures 3, 4). These ERA5 observations are not used to fit σ or the GAM for σ; the evaluation is therefore external to the uncertainty-model estimation. The only self-citation of note is Gruber et al. (2026), cited for the pairwise-differencing technique. This is not load-bearing circularity because the present paper re-derives the mathematical identity in Section A.1 and the central empirical result—better calibration of the ensemble bound—is tested against an independent reanalysis record. The paper's weaknesses are assumptions about transferability (CRCM5 internal variability representing real-world variability, identical bias across members) and acknowledged limitations in complex orography (Section 4). These are validity concerns, not circular reductions: no equation equates the predicted calibration with the fitted input, and no parameter is fitted to the evaluation target. The derivation is self-contained and the evaluation is independent, so the paper does not exhibit circularity.

Assumptions & free parameters 4 free parameters · 6 assumptions · 0 invented entities

The central uncertainty estimate sigma_hat comes from a GAM fitted to CRCM5-LE residual differences, adding several fitted parameters (smooth terms, random intercepts, gamma dispersion). The TFT itself contributes a large number of learned weights. No new physical entities are introduced.

free parameters (4)
  • TFT weights and hyperparameters = per region, see Table 4
    The forecasting model is trained on ERA5-Land 1970-2019; its weights and architecture choices are fitted to historical data and affect both bounds.
  • GAM parameters for squared pairwise differences = beta_0, beta_1, smooth terms, random intercept
    Fitted to CRCM5-LE residual differences to produce sigma_hat(s,t); any misspecification propagates directly to the bound.
  • SPEI gamma distribution parameters = theta_hat(s) per grid cell
    Fitted to ERA5-Land water balance 1970-2019 via maximum likelihood (Eq. 12); used to transform forecasts and bounds to SPEI-1 units.
  • Nominal bound level alpha = 0.1
    Chosen by the authors; not fitted to data, but it is a design choice that affects the interpretation of calibration.
assumptions (6)
  • domain assumption Residual decomposition: eps_i = psi + theta_i, with the systematic bias psi identical across members
    A.1 Eq. 8; required for bias cancellation in pairwise differences. If bias differs across members, the estimator is contaminated.
  • domain assumption theta_i are independent, mean zero, and have equal variance across members
    A.1 Eq. 10; yields E[delta^2] = 2 sigma^2. Independence is plausible for initial-condition ensembles but unverified, and within-family perturbations may be correlated.
  • standard math mu_hat approximately normal by the central limit theorem
    Section 2.3; justifies using the normal quantile z_{1-alpha} in Eq. 7.
  • domain assumption CRCM5-LE internal variability is representative of real-world internal variability
    Central to transferring ensemble spread to ERA5-based forecasts; the authors acknowledge it fails in alpine orography (Section 4).
  • domain assumption Gamma distribution for squared pairwise differences in the GAM
    A.2; a plausible but unverified distributional choice for nonnegative squared differences.
  • standard math SPEI-1 gamma fit to ERA5 is an appropriate drought transform
    A.3; standard practice per Vicente-Serrano et al. (2010).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Probabilistic Deep Learning for Drought Forecasting: Role of Internal Climate Variability." pith.science (2026). https://pith.science/paper/3JORKNX5

@misc{pith2026260801864,
  author       = {Pith},
  title        = {Pith review of: Probabilistic Deep Learning for Drought Forecasting: Role of Internal Climate Variability},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3JORKNX5}},
  note         = {Machine review of arXiv:2608.01864}
}
read the original abstract

Predicting drought risk is essential for anticipating impacts on water resources, agriculture, ecosystems, and climate adaptation planning. Yet drought forecasts remain uncertain because variability can substantially alter regional precipitation and evaporative demand. Treating this variability as unstructured noise ignores the fact that internal variability has spatial, seasonal, and temporal structure and thus contains information that can be used to improve drought forecasting. We propose a deep-learning-based forecasting framework for European drought prediction and extend it with an uncertainty-aware drought bound that explicitly incorporates internal forecast variability from a large climate model ensemble. This bound represents a physically plausible lower-tail trajectory of future drought conditions and marks how severe drought could plausibly become under an unfavourable realisation of internal variability, giving adaptation planning a conservative, risk-averse reference. We compare the proposed bound with a lower bound derived from reanalysis data only and show that our proposed ensemble-informed bound is better calibrated across most regions and seasons. This is specifically true during anomalously dry conditions, when historical reanalysis alone underestimates lower-tail drought risk. Our results show that internal variability should be treated as a forecast quantity in its own right. More broadly, large ensembles provide a practical way to transfer physically plausible climate variability into machine-learning drought forecasts, yielding risk-aware bounds that are more informative for drought assessment under shifting climate conditions.

Figures

Figures reproduced from arXiv: 2608.01864 by the authors.

Figure 1
Figure 1. Seasonal SPEI-1 patterns over Europe during the evaluation period 2020–2024. prediction systems such as the North American Multimodel Ensemble (Kirtman, 2014) and ENSEMBLES (Weisheimer, 2009) estimate forecast uncertainty through the spread across physically consistent members (Lavaysse et al., 2015; Becker and Van Den Dool, 2016; Xu and Mo, 2018). Yet, the ex￾plicit integration of this perspective into AI-based dro… view at source ↗
Figure 2
Figure 2. Schematic overview of the proposed framework. ERA5- Land data from 1970–2019 are used to train the forecasting model ˆ𝑓 . Applying the trained model to ERA5-Land predictors for 2020– 2024 yields the central forecast 𝜇ˆ, while applying ˆ𝑓 to CRCM5-LE simulations over the same period provides an estimate of forecast variability 𝜎ˆ . The resulting forecasts are then transformed to SPEI-1 for evaluation. 2.1 Study Desig… view at source ↗
Figure 3
Figure 3. Prediction overview of the SPEI-1 for the entire domain of Europe and the eight European regions during the evaluation period 2020–2024 (solid colored lines). Observed ERA5-Land SPEI-1 (black line) is shown against forecasted values and 10% uncertainty bounds. Dotted bounds are reanalysis-based TFT quantiles; shaded bounds are CRCM5 ensemble-based bounds. Background colouring indicates the season. 5 [PITH_FULL_IMAG… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Fraction of observed seasonal SPEI-1 values falling below the 10% lower reanalysis bound (top row) and ensemble bound (bottom row), by season and grid cell, for the period 2020–2024. Values close to 10% (green) indicate a well-calibrated bound; values well above 10% in…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

58 extracted references · 4 canonical work pages

  1. [1]

    Hydrology and Earth System Sciences , volume=

    The 2018 northern European hydrological drought and its drivers in a historical perspective , author=. Hydrology and Earth System Sciences , volume=. 2020 , publisher=

    2018

  2. [2]

    Science advances , volume=

    Direct and seasonal legacy effects of the 2018 heat wave and drought on European ecosystem productivity , author=. Science advances , volume=. 2020 , publisher=

    2018

  3. [3]

    Journal of Climate , volume=

    Probabilistic seasonal forecasts in the North American Multimodel Ensemble: A baseline skill assessment , author=. Journal of Climate , volume=

  4. [4]

    Journal of Water and Land Development , volume=

    Drought forecasting using new machine learning methods , author=. Journal of Water and Land Development , volume=. 2013 , publisher=

    2013

  5. [5]

    Khalil and J

    Anteneh Belayneh and Jan Adamowski and B. Khalil and J. Quilty , year =. Coupling machine learning methods with wavelet transforms and the bootstrap and boosting ensemble approaches for drought prediction , journal =

  6. [6]

    International journal of climatology , volume=

    Standardized precipitation evapotranspiration index (SPEI) revisited: parameter fitting, evapotranspiration models, tools, datasets and drought monitoring , author=. International journal of climatology , volume=. 2014 , publisher=

    2014

  7. [7]

    Nature Geoscience , volume=

    Direct and lagged climate change effects intensified the 2022 European drought , author=. Nature Geoscience , volume=. 2024 , publisher=

    2022

  8. [8]

    Frontiers in Water , volume=

    Hot spots and climate trends of meteorological droughts in Europe--assessing the percent of normal index in a single-model initial-condition large ensemble , author=. Frontiers in Water , volume=. 2021 , publisher=

    2021

Show all 58 references
  1. [9]

    Climatic change , volume=

    A summary of the PRUDENCE model projections of changes in European climate by the end of this century , author=. Climatic change , volume=. 2007 , publisher=

    2007

  2. [10]

    A novel intelligent deep learning predictive model for meteorological drought forecasting , journal =

    Ali Danandeh Mehr and Amir Rikhtehgar Ghiasi and Zaher Mundher Yaseen and Ali Unal Sorman and Laith Abualigah , year =. A novel intelligent deep learning predictive model for meteorological drought forecasting , journal =

  3. [11]

    Journal of Open Research Software , volume=

    eofs: A library for EOF analysis of meteorological, oceanographic, and climate data , author=. Journal of Open Research Software , volume=

  4. [12]

    Journal of Climate , volume=

    The seasonal atmospheric response to projected Arctic sea ice loss in the late twenty-first century , author=. Journal of Climate , volume=

  5. [13]

    Nature Climate Change , author =

    Insights from. Nature Climate Change , author =. 2020 , note =. doi:10.1038/s41558-020-0731-2 , abstract =

    2020 doi

  6. [14]

    An improved

    Abhirup Dikshit and Biswajeet Pradhan and Alfredo Huete , year =. An improved. Journal of Environmental Management , volume =

  7. [15]

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

    Funk, Henri and Ludwig, Ralf and Küchenhoff, Helmut and Nagler, Thomas , title =. Journal of the Royal Statistical Society Series C: Applied Statistics , pages =. 2025 , month =. doi:10.1093/jrsssc/qlaf044 , url =

    2025 doi

  8. [16]

    Janga Reddy , year =

    Poulomi Ganguli and M. Janga Reddy , year =. Ensemble prediction of regional droughts using climate inputs and the SVM--copula approach , journal =

  9. [17]

    Predictability Desert

    Bridging the" Predictability Desert": A Probabilistic Bias Correction Framework for AI and Dynamical Subseasonal Forecasts , author=. Towards Trustworthy Predictions: Theory and Applications of Calibration for Modern AI , year=

  10. [18]

    Sources of

    Gruber, Cornelia and Schenk, Patrick Oliver and Schierholz, Malte and Kreuter, Frauke and Kauermann, Göran , month = jan, year =. Sources of. doi:10.48550/arXiv.2305.16703 , urldate =

    doi arXiv

  11. [19]

    Environmetrics , volume =

    Gruber, Cornelia and Funk, Henri and Mittermeier, Magdalena and Küchenhoff, Helmut and Kauermann, Göran , title =. Environmetrics , volume =. doi:https://doi.org/10.1002/env.70108 , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1002/env.70108 , year =

    doi

  12. [20]

    Applied Sciences , volume=

    Application of artificial intelligence to forecast drought index for the Mekong delta , author=. Applied Sciences , volume=. 2024 , publisher=

    2024

  13. [21]

    Statistical Science , author =

    Generalized. Statistical Science , author =. 1986 , note =. doi:10.1214/ss/1177013604 , number =

    1986

  14. [22]

    Bulletin of the American Meteorological Society , volume=

    The potential to narrow uncertainty in regional climate predictions , author=. Bulletin of the American Meteorological Society , volume=. 2009 , publisher=

    2009

  15. [23]

    Climate dynamics , volume=

    The potential to narrow uncertainty in projections of regional precipitation change , author=. Climate dynamics , volume=. 2011 , publisher=

    2011

  16. [24]

    Singh and Youlong Xia , year =

    Zengchao Hao and Vijay P. Singh and Youlong Xia , year =. Seasonal Drought Prediction: Advances, Challenges, and Future Prospects , journal =

  17. [25]

    Geophysical Research Letters , volume=

    European summer synoptic circulations and their observed 2022 and projected influence on hot extremes and dry spells , author=. Geophysical Research Letters , volume=. 2023 , publisher=

    2022

  18. [26]

    Kay, J. E. and Deser, C. and Phillips, A. and Mai, A. and Hannay, C. and Strand, G. and Arblaster, J. M. and Bates, S. C. and Danabasoglu, G. and Edwards, J. and Holland, M. and Kushner, P. and Lamarque, J.-F. and Lawrence, D. and Lindsay, K. and Middleton, A. and Munoz, E. an...

    2015 doi

  19. [27]

    B. P. Kirtman. The North American Multimodel Ensemble: Phase-1 seasonal to interannual prediction; phase-2 toward developing intraseasonal prediction. Bull. Amer. Meteor. Soc. 2014

    2014

  20. [28]

    Hydrology and Earth System Sciences , volume=

    Early warning of drought in Europe using the monthly ensemble system from ECMWF , author=. Hydrology and Earth System Sciences , volume=. 2015 , publisher=

    2015

  21. [29]

    Journal of Applied Meteorology and Climatology , volume=

    The ClimEx project: A 50-member ensemble of climate change projections at 12-km resolution over Europe and northeastern North America with the Canadian Regional Climate Model (CRCM5) , author=. Journal of Applied Meteorology and Climatology , volume=

  22. [30]

    Water Resources Research , volume =

    Jun Li and Zhaoli Wang and Xushu Wu and Chong. Water Resources Research , volume =. 2021 , title =

    2021

  23. [31]

    International journal of forecasting , volume=

    Temporal fusion transformers for interpretable multi-horizon time series forecasting , author=. International journal of forecasting , volume=. 2021 , publisher=

    2021

  24. [32]

    Journal of the Atmospheric Sciences , author =

    Deterministic. Journal of the Atmospheric Sciences , author =. 1963 , pages =. doi:10.1175/1520-0469(1963)020<0130:DNF>2.0.CO;2 , language =

    1963 doi

  25. [33]

    and Milinski, S

    Maher, N. and Milinski, S. and Ludwig, R. , TITLE =. Earth System Dynamics , VOLUME =. 2021 , NUMBER =

    2021

  26. [34]

    2024 , issn =

    Long-term drought prediction using deep neural networks based on geospatial weather data , journal =. 2024 , issn =. doi:https://doi.org/10.1016/j.envsoft.2024.106127 , url =

    2024

  27. [35]

    Drought forecasting using artificial neural networks and time series of drought indices , journal =

    Saeid Morid and Vladimir Smakhtin and Kaveh Bagherzadeh , year =. Drought forecasting using artificial neural networks and time series of drought indices , journal =

  28. [36]

    Earth system science data , volume=

    ERA5-Land: A state-of-the-art global reanalysis dataset for land applications , author=. Earth system science data , volume=. 2021 , publisher=

    2021

  29. [37]

    Reports on Progress in Physics , author =

    Predicting uncertainty in forecasts of weather and climate , volume =. Reports on Progress in Physics , author =. 2000 , pages =. doi:10.1088/0034-4885/63/2/201 , language =

    2000 doi

  30. [38]

    , author=

    Drought prediction based on SPI and SPEI with varying timescales using LSTM recurrent neural network. , author=. Soft Comput. , volume=

  31. [39]

    A review of machine learning methods for drought hazard monitoring and forecasting: Current research trends, challenges, and future research directions , journal =

    Foyez Ahmed Prodhan and Jiahua Zhang and Shaikh Shamim Hasan and Til Prasad. A review of machine learning methods for drought hazard monitoring and forecasting: Current research trends, challenges, and future research directions , journal =. 2022 , issn =. doi:https://doi.org/...

    2022

  32. [40]

    Eos , volume=

    Scientists must join forces to solve forecasting’s predictability desert , author=. Eos , volume=

  33. [41]

    Basic and Applied Ecology , volume=

    A first assessment of the impact of the extreme 2018 summer drought on Central European forests , author=. Basic and Applied Ecology , volume=. 2020 , publisher=

    2018

  34. [42]

    doi:10.5281/zenodo.10020800 , url =

    Schulzweida, Uwe , title =. doi:10.5281/zenodo.10020800 , url =

    doi

  35. [43]

    Thompson, David W. J. and Barnes, Elizabeth A. and Deser, Clara and Foust, William E. and Phillips, Adam S. , year = 2015, month = aug, journal =. Quantifying the. doi:10.1175/JCLI-D-14-00830.1 , urldate =

    2015 doi

  36. [44]

    Geographical review , volume=

    An approach toward a rational classification of climate , author=. Geographical review , volume=. 1948 , publisher=

    1948

  37. [45]

    2023 , author =

    Drought in Europe June 2023 , url =. 2023 , author =

    2023

  38. [46]

    Climate Dynamics , volume=

    Assessing natural variability in RCM signals: comparison of a multi model EURO-CORDEX ensemble with a 50-member single model large ensemble , author=. Climate Dynamics , volume=. 2019 , publisher=

    2019

  39. [47]

    Stochastic Environmental Research and Risk Assessment , volume =

    Tuong Quang Vo and Seon-Ho Kim and Duc Hai Nguyen and Deg-Hyo Bae , year =. Stochastic Environmental Research and Risk Assessment , volume =

  40. [48]

    Weisheimer

    A. Weisheimer. ENSEMBLES: A new multi-model ensemble for seasonal-to-annual predictions—Skill and progress beyond DEMETER in forecasting tropical Pacific SSTs. Geophys. Res. Lett. 2009

    2009

  41. [49]

    , month = may, year =

    Wood, Simon N. , month = may, year =. Generalized

  42. [50]

    Geophysical Research Letters , volume =

    Xu, Lei and Zhang, Xihao and Wu, Tingtao and Yu, Hongchu and Du, Wenying and Zhang, Chong and Chen, Nengcheng , title =. Geophysical Research Letters , volume =. doi:https://doi.org/10.1029/2024GL111134 , url =. https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1029/2024GL111...

    doi

  43. [51]

    Mo , title =

    Li Xu and Kingtse C. Mo , title =. 43rd NOAA Annual Climate Diagnostics and Prediction Workshop , year =

  44. [52]

    Quarterly Journal of the Royal Meteorological Society , author =

    The. Quarterly Journal of the Royal Meteorological Society , author =. 2020 , note =. doi:10.1002/qj.3803 , language =

    2020 doi

  45. [53]

    Climate dynamics , volume=

    Reanalysis-driven climate simulation over CORDEX North America domain using the Canadian Regional Climate Model, version 5: model performance evaluation , author=. Climate dynamics , volume=. 2013 , publisher=

    2013

  46. [54]

    Climate Dynamics , volume=

    Present climate and climate change over North America as simulated by the fifth-generation Canadian regional climate model , author=. Climate Dynamics , volume=. 2013 , publisher=

    2013

  47. [55]

    and Beguería, Santiago and López-Moreno, Juan I

    Vicente-Serrano, Sergio M. and Beguería, Santiago and López-Moreno, Juan I. , month = apr, year =. A. Journal of Climate , publisher =. doi:10.1175/2009JCLI2909.1 , language =

    doi

  48. [56]

    doi:10.2905/JRC.HW6V250 , urldate =

    doi

  49. [57]

    and Matsuura, Kenji , year = 2005, month = dec, journal =

    Willmott, Cort J. and Matsuura, Kenji , year = 2005, month = dec, journal =. Advantages of the Mean Absolute Error (. doi:10.3354/cr030079 , urldate =

    2005 doi

  50. [58]

    Atmosphere , VOLUME =

    West, Harry and Quinn, Nevil and Horswell, Michael , TITLE =. Atmosphere , VOLUME =. 2021 , NUMBER =

    2021

Pith tools

Reviewed August 4, 2026 · model on record in the stance chip above.