Pith. sign in

REVIEW 3 major objections 6 minor 188 references

Satellite-only models predict Pseudo-nitzschia bloom occurrence on the Portuguese coast with ROC-AUC up to 0.77 under strict year-and-cluster holdout.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-10 17:04 UTC pith:NXHSN6SQ

load-bearing objection Solid applied ML paper: honest moderate AUCs under a real year∪ cluster CV, with the main novelty in the Portuguese matchup and river-aware blocking rather than the model class. the 3 major comments →

arxiv 2607.07834 v1 pith:NXHSN6SQ submitted 2026-07-08 cs.LG

Predicting Pseudo-nitzschia harmful algal blooms along the Portuguese Coast using satellite-derived predictors

Ayman Bnoussaad , El Khalil Cherif , Ligia Pinto , Ramiro Neves , Alexandra D. Silva , Alexandre Bernardino This is my paper
classification cs.LG
keywords Pseudo-nitzschiaharmful algal bloomsremote sensingmachine learningspatio-temporal forecastingPortuguese Atlantic coastupwellingsatellite predictors
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper asks whether Pseudo-nitzschia harmful algal blooms along the northern Portuguese coast can be anticipated from satellites alone, without any local water samples as inputs, under evaluation rules that match real forecasting. Using a decade of monitoring matched to sea-surface temperature, an upwelling index, chlorophyll-a, and plankton functional types, the authors show moderate skill: tree ensembles reach ROC-AUC about 0.74 with physical predictors and about 0.77 when biological satellite products are added. The evaluation withholds entire years and entire river-influenced coastal clusters at once, so the numbers measure generalisation to new seasons and new shoreline segments rather than memorisation of nearby samples. Feature rankings put seasonal timing, location, and multi-week temperature and upwelling history first, with recent chlorophyll refining risk inside those windows. A sympathetic reader cares because shellfish closures carry real cost and cell counts alone do not always flag toxin risk; a remote early-warning layer could focus limited field sampling where it matters most.

Core claim

Under a leakage-resistant spatio-temporal cross-validation that simultaneously withholds calendar years and river-aware spatial clusters, Pseudo-nitzschia HAB occurrence (defined as cell concentration above 10,000 cells per litre) on the L1-L2 Portuguese hotspot is moderately predictable from exclusively satellite-derived predictors. Extra Trees with environmental plus biological features reach ROC-AUC 0.77 plus or minus 0.06; Random Forest with environmental features alone reaches 0.74 plus or minus 0.05. Seasonal harmonics, spatial coordinates, and lagged physical forcing dominate, while short-lag chlorophyll and plankton functional types refine likelihood within favourable windows.

What carries the argument

Union-of-year-and-cluster blocking combined with river-aware spatial clustering of sampling sites: each validation fold holds out every observation from one calendar year together with every observation from one freshwater-influence cluster, so lagged satellite predictors cannot leak information from the held-out regime.

Load-bearing premise

The claim treats a fixed cell-count threshold of 10,000 cells per litre as an adequate operational proxy for harmful risk, even though toxin and harvest-ban events can occur at lower abundances and toxicity itself is not the modelled target.

What would settle it

Re-run the same year-and-cluster cross-validation after re-labelling HAB events from matched shellfish domoic-acid measurements (or actual harvest-ban days) instead of the fixed cell threshold; if discrimination collapses well below the reported 0.7 range, the occurrence proxy does not support operational early warning.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The manuscript develops a satellite-only machine-learning framework to predict Pseudo-nitzschia HAB occurrence (PN > 10,000 cells L⁻¹) along the northern Portuguese coast (zones L1–L2, 2013–2023; 1,440 observations). Predictors include SST, a Bakun-style upwelling index, chlorophyll-a, plankton functional types, seasonal harmonics, spatial coordinates, and lags up to 45 days. Sampling sites are partitioned by a river-aware (ROFI) clustering scheme, and models are evaluated under a stringent year∪cluster cross-validation (66 folds) that withholds entire years and spatial clusters simultaneously. Across linear, margin-based, and tree ensembles, HAB occurrence is moderately predictable: Random Forest reaches ROC–AUC 0.74±0.05 with environmental features; Extra Trees reaches 0.77±0.06 when biological predictors are added. Feature importance attributes skill primarily to seasonality, location, and multi-week SST/UI lags, with short-lag Chl-a refining predictions. Confusion matrices at fixed FPRs and spatial probability maps illustrate operational threshold trade-offs and adaptive sampling use.

Significance. If the reported discrimination holds under the stated protocol, the work provides a carefully scoped, leakage-resistant demonstration that satellite-only predictors can support early-warning of Pseudo-nitzschia bloom occurrence on an eastern-boundary upwelling coast. Strengths include strict separation of IPMA labels from CMEMS predictors, multi-model and nested feature-set comparison, fold-wise mean±std over 66 year×cluster splits, open code for CMEMS matchup and ROFI clustering, and explicit operational framing (threshold trade-offs, adaptive sampling maps). The contribution is methodological and regional rather than a universal HAB solution, but it is a credible baseline for satellite-supported monitoring support where in situ coverage is sparse.

major comments (3)
  1. §2.4 and Table 1: The supervised target is a fixed threshold HAB = 1 if PN > 10,000 cells L⁻¹ (~14% positives). The paper repeatedly notes that DA toxicity and harvest bans can occur at low cell abundances and that toxicity is not the modelling target (Introduction; §2.4; Discussion 4.3; Conclusion limitations). That disclosure is appropriate, but the operational claim in the abstract and highlights (“operationally relevant skill for … early-warning systems”) still rests on occurrence as a risk proxy. A sensitivity analysis over alternative thresholds (e.g. 5×10³, 2×10⁴, or a percentile-based definition) and/or a secondary analysis against ban-day labels would show whether the reported AUC range is robust to the definitional choice that is load-bearing for uptake.
  2. §2.7 and Figure 10: The year∪cluster blocking is a genuine strength, but the manuscript does not quantify residual leakage risk from (i) multi-week lagged predictors that may still share large-scale forcing with held-out years/clusters, or (ii) the fact that a held-out cluster’s other years remain in training (and a held-out year’s other clusters remain in training). Reporting a pure leave-one-year-out and pure leave-one-cluster-out ablation alongside the union design would make the “realistic forecasting” claim more transparent and would show how much of the 0.74–0.77 AUC is carried by each blocking axis.
  3. §2.7.1 / Table 2: Hyperparameters for the winning RF (env-only) and Extra Trees (full) configurations are stated as selected within training folds, but the selection procedure (grid vs. default, search space, whether nested CV was used) is not fully specified, and only the final settings are given. Given class imbalance and 66 folds, a brief nested-CV or fixed-hyperparameter sensitivity check is needed to confirm that the 0.03 AUC gain from adding biological predictors is not partly an artefact of unequal tuning effort across feature sets.
minor comments (6)
  1. Appendix A vs. main text §2.1/Figure 7: Main text refers to “river-aware K-means” and K=6, while Appendix A primarily describes DBSCAN with an augmented distance and notes k-means as an alternative. Please state unambiguously which algorithm produced the six clusters used for CV blocking.
  2. Table 1: “Remote-sensed biological uncertainty estimates” list CHL and PFT uncertainty ranges but it is unclear whether these uncertainty fields enter any of the three model configurations or are only descriptive. Clarify in §2.3/§2.6.
  3. Figure 9 caption and §2.5: Toxicity-related ban periods are highlighted, but ban labels are not used as a modelling target. A short note in the figure caption that bans are shown for context only would avoid reader confusion.
  4. §2.3 / Appendix B: Coastline angle defaults (−32°) and zone-specific adjustments for L1/L2 are mentioned; give the numerical angles used for L1 and L2 so the UI is fully reproducible from the text.
  5. Typographical/consistency: “Pseudo-nitzschia” is occasionally concatenated without space or italics in headings; “Downwelling/Upwelling Index” units and sign convention should be stated once in the main text as well as Appendix B.
  6. References: Several recent HAB ML and Portuguese Pseudo-nitzschia monitoring papers are cited appropriately; ensure all arXiv/software DOIs (CMEMS_Data_Analysis, ROFI_Clustering) remain accessible at publication.

Circularity Check

0 steps flagged

No significant circularity: independent in-situ labels, external satellite predictors, and out-of-fold year×cluster evaluation; no claimed prediction reduces to its inputs by construction.

full rationale

This is a standard empirical machine-learning forecasting paper, not a first-principles derivation. The supervised HAB target is constructed solely from independent IPMA in-situ Pseudo-nitzschia cell counts via a fixed literature threshold (PN > 10 000 cells L⁻¹, citing Lane et al. 2009); satellite predictors (CMEMS SST, Bakun-style UI, CHL, PFTs and lags) are never used to define the label. Seasonal harmonics (sin/cos of DOY) and the upwelling index are conventional encodings, not tautologies of the target. Models are trained and scored under a leakage-resistant year∪ cluster cross-validation that withholds entire years and spatial blocks; reported ROC–AUC values (0.74–0.77) are therefore out-of-fold generalisation estimates, not in-sample fits renamed as predictions. Feature-importance rankings and confusion matrices are likewise computed on held-out folds. Self-citations point only to open-source code repositories that implement data pipelines and clustering; they supply no uniqueness theorems or load-bearing lemmas. Hyper-parameter selection inside the CV folds is ordinary ML practice and does not force the discrimination numbers. Consequently no step in the claimed chain reduces, by the paper’s own equations or by self-citation, to its own inputs. Score 0 is the correct, proportionate finding.

Axiom & Free-Parameter Ledger

8 free parameters · 6 axioms · 1 invented entities

The central claim is an empirical ML predictability result. It rests on a small set of domain thresholds and design choices (HAB cutoff, lag horizon, cluster count, matching tolerance, coastline angle) plus standard assumptions that satellite physical and bio-optical fields proxy bloom-conducive conditions. No new physical entities are postulated; the ROFI clustering and Bakun UI are methodological constructs built from prior hydrology and Ekman theory.

free parameters (8)
  • HAB cell-concentration threshold = 10000 cells L^-1
    Binary target HAB=1 iff PN > 10,000 cells L⁻¹; fixed from literature rather than learned or toxicity-linked.
  • Maximum lag depth = 45 days
    Environmental and biological predictors lagged 0–45 days; horizon chosen as ecologically plausible and used for all models.
  • Number of river-aware spatial clusters K = 6
    Sampling sites partitioned into six clusters that define CV spatial blocks.
  • River-influence weight alpha / ROFI decay L = L=30 km; alpha not numerically fixed in main text
    Augmented distance and river-influence feature use alpha and L=30 km decay scale for ROFI extent.
  • Spatial matchup tolerance = 0.05 degrees
    Nearest CMEMS grid cell accepted within 0.05° of each in situ sample.
  • Coastline orientation for upwelling index = -32 degrees (default)
    Default theta = -32° for western Iberian margin with zone-specific L1/L2 adjustments in Bakun UI.
  • Random Forest hyperparameters (env-only winner) = n_estimators=400, max_depth=8, min_samples_leaf=3
    Selected by CV for highest mean ROC–AUC: n_estimators, max_depth, min_samples_leaf, balanced class weights.
  • Extra Trees hyperparameters (full-feature winner) = n_estimators=200, min_samples_leaf=2
    Selected by CV for highest mean ROC–AUC on environmental+biological set.
axioms (6)
  • domain assumption Satellite-derived SST, wind-based upwelling index, CHL and PFTs (with lags) contain transferable information about conditions conducive to Pseudo-nitzschia bloom occurrence.
    Load-bearing premise of the entire satellite-only forecasting design (Introduction; Sections 2.3–2.6).
  • domain assumption A fixed PN concentration threshold of 10,000 cells L⁻¹ defines HAB occurrence in a way that is operationally meaningful for early warning.
    Target construction in Section 2.4; paper notes toxicity can occur at low abundances yet still uses this binary label.
  • ad hoc to paper Withholding the union of one calendar year and one spatial cluster yields a realistic estimate of operational generalisation without temporal or spatial leakage.
    Core evaluation design in Section 2.7; performance claims are interpreted under this protocol.
  • domain assumption Bakun/Ekman coastal upwelling index with local coastline angle is an adequate scalar of nutrient-injection forcing for this margin.
    UI derivation Appendix B and environmental feature section; classical oceanographic assumption.
  • standard math ROC–AUC (and fixed-FPR confusion matrices) are appropriate primary metrics under ~14% positive-class imbalance.
    Stated in Section 2.7.2; standard imbalanced classification practice.
  • domain assumption In optically complex ROFI waters, CMEMS CHL/PFT retrievals remain useful enough as community-state proxies despite acknowledged bio-optical uncertainty.
    Section 2.3 explicitly flags coastal retrieval uncertainty yet still uses these fields as Model 3 predictors.
invented entities (1)
  • River-aware ROFI spatial clustering of L1–L2 sampling sites independent evidence
    purpose: Define ecologically coherent spatial blocks for leakage-resistant cross-validation and coastal segmentation.
    Methodological construct combining haversine distance with river-mouth exposure; not a new physical object, but paper-specific blocking entity used for all skill claims.

pith-pipeline@v1.1.0-grok45 · 27435 in / 3861 out tokens · 46903 ms · 2026-07-10T17:04:12.486022+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Predicting Pseudo-nitzschia harmful algal blooms along the Portuguese Coast using satellite-derived predictors." pith.science (2026). https://pith.science/paper/NXHSN6SQ

@misc{pith2026260707834,
  author       = {Pith},
  title        = {Pith review of: Predicting Pseudo-nitzschia harmful algal blooms along the Portuguese Coast using satellite-derived predictors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NXHSN6SQ}},
  note         = {Machine review of arXiv:2607.07834}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Pseudo-nitzschia diatoms pose recurrent risks to coastal ecosystems and shellfish harvesting along the Portuguese Atlantic coast. Here we develop and evaluate a spatio-temporal machine-learning framework to predict harmful algal bloom (HAB) occurrence using exclusively satellite-derived predictors under realistic forecasting constraints. We characterised environmental and biological variability across shellfish production zones (L1-L9) using 5,882 observations, providing system-wide context. Predictive models were developed for zones L1-L2, a hotspot for Pseudo-nitzschia and domoic acid events, using a decade-long dataset (2013-2023; 1,440 observations; more than 1,000 satellite-based predictors including sea surface temperature, an upwelling index, chlorophyll-a, and plankton functional types). Sampling locations were partitioned into ecologically meaningful sub-regions using a river-aware spatial clustering scheme. A stringent spatio-temporal cross-validation strategy that simultaneously withholds entire years and spatial clusters prevents leakage and closely mimics real-world forecasting conditions. HAB occurrence proved moderately predictable across model classes and feature configurations. Ensemble tree-based methods achieved the strongest discrimination: Random Forest reached 0.74 +/- 0.05 with environmental predictors; Extra Trees reached 0.77 +/- 0.06 with biological variables added. Feature-importance analyses revealed that seasonal structure, spatial context, and lagged environmental conditions dominate model decisions, while biological indicators refine bloom likelihood within physically favourable periods. The framework demonstrates operationally relevant skill for satellite-supported HAB early-warning systems along eastern boundary upwelling coasts.

Figures

Figures reproduced from arXiv: 2607.07834 by Alexandra D. Silva, Alexandre Bernardino, Ayman Bnoussaad, El Khalil Cherif, Ligia Pinto, Ramiro Neves.

Figure 1
Figure 1. Figure 1: Cumulative shellfish harvesting ban days reported by IPMA for each production area between 2013 and [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Annual shellfish harvesting ban days by production area (2013–2023). Bar colours represent total ban days [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Spatial distribution of the shellfish production areas along the Portuguese Atlantic coast. Areas L1 (Litoral [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Spatial distribution of trace metals and metalloids along the Portuguese coast derived from EMODnet [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Mean coastal Downwelling/Upwelling Index by production area along the Portuguese margin (2020–2023). [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Zonal ranges of the coastal Downwelling/Upwelling Index along the Portuguese margin (2020–2023). Bar [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: River-aware spatial clustering of sampling locations along the northern Portuguese coast. River mouths [PITH_FULL_IMAGE:figures/full_fig_p009_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Seasonal and monthly distribution of sampling events (2013–2023) for the L1 and L2 production zones. [PITH_FULL_IMAGE:figures/full_fig_p010_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Time series (2013–2023) of (top) PN cell concentrations, with toxicity-related harvesting ban periods [PITH_FULL_IMAGE:figures/full_fig_p014_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Spatio-temporal cross-validation schematic. For each fold, all observations from one target year (entire [PITH_FULL_IMAGE:figures/full_fig_p016_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Schematic of predictor configurations and modelling flow. Three progressively enriched predictor sets [PITH_FULL_IMAGE:figures/full_fig_p018_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: ROC curves for the HAB target using Environmental-only features. Ensemble tree models (Random [PITH_FULL_IMAGE:figures/full_fig_p020_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: ROC curves for the HAB target using Environmental + Biological features. Incorporating CHL and PFTs [PITH_FULL_IMAGE:figures/full_fig_p021_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: Top features for the model using Environmental-only predictors (Random Forest) and Environmen [PITH_FULL_IMAGE:figures/full_fig_p022_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: Confusion matrices for the Random Forest model (Environmental-only) evaluated at three fixed false [PITH_FULL_IMAGE:figures/full_fig_p023_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: Confusion matrices for the Extra Trees model (Environmental+Biological) evaluated at three fixed false [PITH_FULL_IMAGE:figures/full_fig_p023_16.png] view at source ↗
Figure 17
Figure 17. Figure 17: Spatial illustration of HAB predictions and observations along the Portuguese coast on two contrasting [PITH_FULL_IMAGE:figures/full_fig_p024_17.png] view at source ↗

discussion (0)

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

Reference graph

Works this paper leans on

188 extracted references · 188 canonical work pages · 5 internal anchors

  1. [1]

    International Journal of Environmental Science and Technology , volume=

    Harmful algal blooms (red tide): a review of causes, impacts and approaches to monitoring and prediction , author=. International Journal of Environmental Science and Technology , volume=. 2019 , publisher=

    work page 2019

  2. [2]

    Harmful Algae , volume=

    Pseudo-nitzschia physiological ecology, phylogeny, toxicity, monitoring and impacts on ecosystem health , author=. Harmful Algae , volume=. 2012 , publisher=

    work page 2012

  3. [3]

    Harmful Algae , volume=

    A HAB warning system for shellfish harvesting in Portugal , author=. Harmful Algae , volume=. 2016 , publisher=

    work page 2016

  4. [4]

    Toxins , volume=

    Bivalve shellfish safety in Portugal: variability of faecal levels, metal contaminants and marine biotoxins during the last decade (2011--2020) , author=. Toxins , volume=. 2023 , publisher=

    work page 2011

  5. [5]

    Harmful Algae , volume=

    Can Pseudo-nitzschia blooms be modeled by coastal upwelling in Lisbon Bay? , author=. Harmful Algae , volume=. 2010 , publisher=

    work page 2010

  6. [6]

    Progress in Oceanography , volume=

    The physical oceanography of upwelling systems and the development of harmful algal blooms , author=. Progress in Oceanography , volume=. 2010 , publisher=

    work page 2010

  7. [7]

    Harmful Algae , volume=

    Initial skill assessment of the California harmful algae risk mapping (C-HARM) system , author=. Harmful Algae , volume=. 2016 , publisher=

    work page 2016

  8. [8]

    Frontiers in Marine Science , volume=

    Satellite observations of phytoplankton functional type spatial distributions, phenology, diversity, and ecotones , author=. Frontiers in Marine Science , volume=. 2017 , publisher=

    work page 2017

  9. [9]

    Proceedings of the National Academy of Sciences , volume=

    Molecular forecasting of domoic acid during a pervasive toxic diatom bloom , author=. Proceedings of the National Academy of Sciences , volume=. 2024 , publisher=

    work page 2024

  10. [10]

    Water , volume=

    Pseudo-nitzschia blooms in a coastal upwelling system: Remote sensing detection, toxicity and environmental variables , author=. Water , volume=. 2019 , publisher=

    work page 2019

  11. [11]

    Remote Sensing , volume=

    A new algorithm using support vector machines to detect and monitor bloom-forming Pseudo-nitzschia from OLCI data , author=. Remote Sensing , volume=. 2024 , publisher=

    work page 2024

  12. [12]

    Water , volume=

    Comparison of in-situ chlorophyll-a time series and sentinel-3 ocean and land color instrument data in Slovenian National Waters (gulf of trieste, Adriatic Sea) , author=. Water , volume=. 2021 , publisher=

    work page 2021

  13. [13]

    Regional Studies in Marine Science , volume=

    Monitoring Tangier (Morocco) coastal waters for As, Fe and P concentrations using ESA Sentinels-2 and 3 data: An exploratory study , author=. Regional Studies in Marine Science , volume=. 2019 , publisher=

    work page 2019

  14. [14]

    Remote Sensing , volume=

    Determination of bathing water quality using thermal images Landsat 8 on the west coast of Tangier: Preliminary results , author=. Remote Sensing , volume=. 2019 , publisher=

    work page 2019

  15. [15]

    ISPRS Journal of Photogrammetry and Remote Sensing , volume=

    Characterization of harmful algal blooms (HABs) in the Arabian Gulf and the Sea of Oman using MERIS fluorescence data , author=. ISPRS Journal of Photogrammetry and Remote Sensing , volume=. 2015 , publisher=

    work page 2015

  16. [16]

    Desalination and Water Treatment , volume=

    Reverse osmosis desalination system and algal blooms Part I: harmful algal blooms (HABs) species and toxicity , author=. Desalination and Water Treatment , volume=. 2016 , publisher=

    work page 2016

  17. [17]

    Harmful Algae , volume=

    Harmful algal blooms: A climate change co-stressor in marine and freshwater ecosystems , author=. Harmful Algae , volume=. 2020 , publisher=

    work page 2020

  18. [18]

    International Journal of Oceans and Oceanography , volume=

    Harmful algal blooms and their impacts in the middle and outer ROPME sea area , author=. International Journal of Oceans and Oceanography , volume=

    work page

  19. [19]

    Harmful Algae , volume=

    Harmful algal blooms and public health , author=. Harmful Algae , volume=. 2016 , publisher=

    work page 2016

  20. [20]

    Biological Reviews , volume=

    Emerging threats and persistent conservation challenges for freshwater biodiversity , author=. Biological Reviews , volume=. 2019 , publisher=

    work page 2019

  21. [21]

    Harmful Algae , volume=

    Harmful algal blooms and their effects in coastal seas of Northern Europe , author=. Harmful Algae , volume=. 2021 , publisher=

    work page 2021

  22. [22]

    Nature Climate Change , volume=

    Toxic algal bloom induced by ocean acidification disrupts the pelagic food web , author=. Nature Climate Change , volume=. 2018 , publisher=

    work page 2018

  23. [23]

    Ecology of Harmful Algae , pages=

    The economic effects of harmful algal blooms , author=. Ecology of Harmful Algae , pages=. 2006 , publisher=

    work page 2006

  24. [24]

    Freshwater Algae of North America , pages=

    Harmful algal blooms , author=. Freshwater Algae of North America , pages=. 2015 , publisher=

    work page 2015

  25. [25]

    Harmful Algae , volume=

    Economic and sociocultural impacts of fisheries closures in two fishing-dependent communities following the massive 2015 US West Coast harmful algal bloom , author=. Harmful Algae , volume=. 2018 , publisher=

    work page 2015

  26. [26]

    Harmful Algae , volume=

    Understanding the blob bloom: Warming increases toxicity and abundance of the harmful bloom diatom Pseudo-nitzschia in California coastal waters , author=. Harmful Algae , volume=. 2017 , publisher=

    work page 2017

  27. [27]

    Limnology and Oceanography , volume=

    Effects of changing pCO2 and phosphate availability on domoic acid production and physiology of the marine harmful bloom diatom Pseudo-nitzschia multiseries , author=. Limnology and Oceanography , volume=. 2011 , publisher=

    work page 2011

  28. [28]

    Natural Toxins , volume=

    Environmental stress and domoic acid production by Pseudo-nitzschia: a physiological perspective , author=. Natural Toxins , volume=. 1998 , publisher=

    work page 1998

  29. [29]

    Harmful Algae , volume=

    Variability patterns and phenology of harmful phytoplankton blooms off southern Portugal: Looking for region-specific environmental drivers and predictors , author=. Harmful Algae , volume=. 2022 , publisher=

    work page 2022

  30. [30]

    Phycologia , volume=

    Pseudo-nitzschia (Bacillariophyceae) species, domoic acid and amnesic shellfish poisoning: revisiting previous paradigms , author=. Phycologia , volume=. 2012 , publisher=

    work page 2012

  31. [31]

    Harmful Algae , volume=

    Climate change and harmful algal blooms: insights and perspective , author=. Harmful Algae , volume=. 2020 , publisher=

    work page 2020

  32. [32]

    2005 , publisher=

    Remote Sensing of Coastal Aquatic Environments: Technologies, Techniques and Applications , author=. 2005 , publisher=

    work page 2005

  33. [33]

    pelagic seed banks

    Turbulence, watermass stratification and harmful algal blooms: an alternative view and frontal zones as “pelagic seed banks” , author=. Harmful Algae , volume=. 2002 , publisher=

    work page 2002

  34. [34]

    Global Ecology and Oceanography of Harmful Algal Blooms , pages=

    The role of life cycle characteristics in harmful algal bloom dynamics , author=. Global Ecology and Oceanography of Harmful Algal Blooms , pages=. 2018 , publisher=

    work page 2018

  35. [35]

    Cyanobacterial harmful algal blooms: state of the science and research needs , pages=

    Global warming and cyanobacterial harmful algal blooms , author=. Cyanobacterial harmful algal blooms: state of the science and research needs , pages=. 2008 , publisher=

    work page 2008

  36. [36]

    Ecology of Harmful Algae , pages=

    Harmful algal bloom dynamics in relation to physical processes , author=. Ecology of Harmful Algae , pages=. 2006 , publisher=

    work page 2006

  37. [37]

    Manual on harmful marine microalgae , volume=

    Harmful algal blooms: a global overview , author=. Manual on harmful marine microalgae , volume=. 2003 , publisher=

    work page 2003

  38. [38]

    Journal of Phycology , volume=

    Ocean climate change, phytoplankton community responses, and harmful algal blooms: a formidable predictive challenge 1 , author=. Journal of Phycology , volume=. 2010 , publisher=

    work page 2010

  39. [39]

    Reviews in Aquaculture , volume=

    Assessing risks and mitigating impacts of harmful algal blooms on mariculture and marine fisheries , author=. Reviews in Aquaculture , volume=. 2020 , publisher=

    work page 2020

  40. [40]

    Ecological Indicators , volume=

    Predicting coastal algal blooms with environmental factors by machine learning methods , author=. Ecological Indicators , volume=. 2021 , publisher=

    work page 2021

  41. [41]

    , author=

    Harmful algal blooms. , author=. 2017 , publisher=

    work page 2017

  42. [42]

    Environmental Modeling & Assessment , volume=

    Multiple effects of environmental factors on algal growth and nutrient thresholds for harmful algal blooms: application of response surface methodology , author=. Environmental Modeling & Assessment , volume=. 2016 , publisher=

    work page 2016

  43. [43]

    Harmful Algae , volume=

    Monitoring, modeling and projection of harmful algal blooms in China , author=. Harmful Algae , volume=. 2022 , publisher=

    work page 2022

  44. [44]

    Journal of Marine Science and Engineering , volume=

    A review of recent machine learning advances for forecasting harmful algal blooms and shellfish contamination , author=. Journal of Marine Science and Engineering , volume=. 2021 , publisher=

    work page 2021

  45. [45]

    and Bryan, J

    Yang, X. and Bryan, J. and Okubo, K. and Jiang, C. and Clements, T. and Denolle, M. A. , title=. Geophysical Journal International , volume=. 2022 , month=

    work page 2022

  46. [46]

    and Huang, W

    Qi, Q. and Huang, W. and Zhang, D. and Han, L. , title=. Applied Sciences , volume=. 2023 , url=

    work page 2023

  47. [47]

    and Embury, O

    Merchant, C.J. and Embury, O. and Bulgin, C.E. and Block, T. and Corlett, G.K. and Fiedler, E. and Good, S.A. and Mittaz, J. and Rayner, N.A. and Berry, D. and Eastwood, S. and Taylor, M. and Tsushima, Y. and Waterfall, A. and Wilson, R. and Donlon, C. , title=. Scientific Data , volume=

    work page

  48. [48]

    Brand, L. E. and Sunda, W. G. and Guillard, R. R. L. , title=. Journal of Experimental Marine Biology and Ecology , volume=

    work page

  49. [49]

    Bruland, K. W. and Franks, R. P. and Knauer, G. A. and Martin, J. H. , title=. Earth and Planetary Science Letters , volume=

    work page

  50. [50]

    and Costa, M

    Caeiro, S. and Costa, M. H. and Ramos, T. B. and Fernandes, F. and Silveira, N. and Coimbra, A. and Medeiros, G. and Painho, M. , title=. Ecological Indicators , volume=

    work page

  51. [51]

    Cullen, J. T. and Sherrell, R. M. , title=. Marine Chemistry , volume=

    work page

  52. [52]

    Limnology and Oceanography , volume=

    Trace metal nutrition of marine phytoplankton: a review , author=. Limnology and Oceanography , volume=

    work page

  53. [53]

    Figueiras, F. G. and Labarta, U. and Fernández-Reiriz, M. J. , title=. ICES Journal of Marine Science , volume=

    work page

  54. [54]

    and others , title=

    Ferreira da Silva, E. and others , title=. Environmental Earth Sciences , volume=

    work page

  55. [55]

    and Pan, G

    Guo, M. and Pan, G. and Zhang, M. and Song, Y. , title=. Acta Ecologica Sinica , volume=

    work page

  56. [56]

    Environmental Pollution , volume=

    Lead toxicity and its impact on phytoplankton growth and diversity in freshwater ecosystems , author=. Environmental Pollution , volume=

    work page

  57. [57]

    Lane, T. W. and Saito, M. A. and George, G. N. and Pickering, I. J. and Prince, R. C. and Morel, F. M. M. , title=. Met Ions Life Sci. , volume=

    work page

  58. [58]

    Proceedings of the National Academy of Sciences , volume=

    Biogeochemical evidence for the importance of cadmium in marine phytoplankton , author=. Proceedings of the National Academy of Sciences , volume=. 2000 , publisher=

    work page 2000

  59. [59]

    and others , title=

    Lima, A. and others , title=. Journal of Geochemical Exploration , volume=

    work page

  60. [60]

    Maldonado, M. T. and Hughes, M. P. and Rue, E. L. and Wells, M. L. , title=. Proceedings of the National Academy of Sciences , volume=

    work page

  61. [61]

    Limnology and Oceanography , volume=

    Influence of dissolved and particulate iron on the growth of phytoplankton from the Southern Ocean , author=. Limnology and Oceanography , volume=

    work page

  62. [62]

    Science , volume=

    The biogeochemical cycles of trace metals in the oceans , author=. Science , volume=. 2003 , publisher=

    work page 2003

  63. [63]

    Treatise on Geochemistry , publisher=

    Marine bioinorganic chemistry: The role of trace metals in the oceanic cycles of major nutrients , author=. Treatise on Geochemistry , publisher=

    work page

  64. [64]

    and Cobelo-García, A

    Prego, R. and Cobelo-García, A. and Millward, G. E. and Nombela, M. A. , title=. Estuarine, Coastal and Shelf Science , volume=

    work page

  65. [65]

    and Prego, R

    Santos-Echeandía, J. and Prego, R. and Cobelo-García, A. and Millward, G. E. , title=. Estuarine, Coastal and Shelf Science , volume=

    work page

  66. [66]

    Sunda, W. G. , title=. Frontiers in Microbiology , volume=

    work page

  67. [67]

    Wells, M. L. and Trick, C. G. and Cochlan, W. P. and Hughes, M. P. and Trainer, V. L. , title=. Harmful Algae , volume=

    work page

  68. [68]

    from Monterey Bay, California , author=

    Nitrogenous preference of toxigenic Pseudo-nitzschia spp. from Monterey Bay, California , author=. Harmful Algae , volume=. 2007 , publisher=

    work page 2007

  69. [69]

    Pitcher, G. C. and Figueiras, F. G. and Hickey, B. M. , title=. Ecology of Harmful Algae , editor=. 2006 , publisher=

    work page 2006

  70. [70]

    Coastal upwelling indices, West Coast of North America, 1946–71 , author=

    work page 1946

  71. [71]

    Journal of Sea Research , volume=

    Composition and interannual variability of phytoplankton in a coastal upwelling region (Lisbon Bay, Portugal) , author=. Journal of Sea Research , volume=

    work page

  72. [72]

    Remote Sensing of Environment , volume=

    Unravelling region-specific environmental drivers of phytoplankton biomass and phenology in a coastal upwelling system , author=. Remote Sensing of Environment , volume=

    work page

  73. [73]

    Anderson, D. M. , title=. Estuaries , year=

    work page

  74. [74]

    Wells, M. L. and Trainer, V. L. and Smayda, T. J. and Karlson, B. S. and Trick, C. G. and Kudela, R. M. and Ishikawa, A. and Bernard, S. and Wulff, A. and Anderson, D. M. and Cochlan, W. P. , title=. Harmful Algae , year=

    work page

  75. [75]

    Trainer, V. L. and Hickey, B. M. and Horner, R. A. , title=. Harmful Algae , year=

    work page

  76. [76]

    McCabe, R. M. and Hickey, B. M. and Kudela, R. M. and Lefebvre, K. A. and Adams, N. G. and Bill, B. D. and Gulland, F. M.D. and Thomson, J. R. and Cochlan, W. P. and Trainer, V. L. , title=. Geophysical Research Letters , year=

    work page

  77. [77]

    Nature Communications , volume=

    Tracking the global reduction of marine traffic during the COVID-19 pandemic , author=. Nature Communications , volume=. 2021 , publisher=

    work page 2021

  78. [78]

    Remote Sensing Letters , volume=

    Effects of the Covid-19 pandemic on the oceans , author=. Remote Sensing Letters , volume=. 2021 , publisher=

    work page 2021

  79. [79]

    Remote Sensing of Coastal Aquatic Environments , pages=

    Remote sensing of harmful algal blooms , author=. Remote Sensing of Coastal Aquatic Environments , pages=. 2005 , publisher=

    work page 2005

  80. [80]

    Oceanography , volume=

    Global ecology and oceanography of harmful algal blooms , author=. Oceanography , volume=

    work page

Showing first 80 references.