Seasonal and spatio-temporal dynamics of marine heatwaves in the Senegalo&ndash;Mauritanian Upwelling System and chlorophyll-a responses

Dione, Waly; Camara, Ibrahima; Lazar, Alban; Ruiz-Pino, Diana; Gaye, Amadou Thierno

doi:10.5194/egusphere-2026-2571

Preprints

https://doi.org/10.5194/egusphere-2026-2571

Preprints

21 May 2026

| 21 May 2026

Status: this preprint is open for discussion and under review for Ocean Science (OS).

Seasonal and spatio-temporal dynamics of marine heatwaves in the Senegalo–Mauritanian Upwelling System and chlorophyll-a responses

Waly Dione, Ibrahima Camara, Alban Lazar, Diana Ruiz-Pino, and Amadou Thierno Gaye

Abstract. Marine heatwaves (MHWs) represent an increasing threat to marine ecosystems, yet their dynamics remain poorly documented in tropical eastern boundary upwelling systems, where upwelling reaches its maximum in boreal winter rather than in summer. This study analyses the spatio-temporal variability of MHWs, their physical drivers, and biological impacts in the Senegalo–Mauritanian Upwelling System (SMUS) over 1982–2024, using satellite-derived SST, chlorophyll-a, and atmospheric reanalysis data.

A strong spatial heterogeneity is observed, with the highest cumulative number of MHW days (up to 50 days per year) occurring in shallow regions weakly influenced by upwelling, as well as in offshore transition zones. In contrast, lower occurrences (around 25 days per year) are observed in persistent upwelling areas, particularly at Cap Blanc and along the Petite Côte, which act as relative thermal refuges. At the seasonal scale, MHWs are more frequent and more persistent during the CP, particularly south of 18°N, while their intensity is modulated by the seasonal migration of the upwelling front.

Over the 42-year period, the Grande Côte region emerges as the most vulnerable coastal zone, showing the strongest increase in MHW occurrence, with significant trends (p < 0.05) in total MHW days (+10.25 days per decade) and duration (+2.22 days per decade). Along the Petite Côte, trends are approximately four times stronger during the CP (+5.16 days per decade) than during the warming period (+1.34 days per decade). No significant long-term trend is detected for maximum intensity at the annual scale. However, a significant seasonal increase (+0.16 °C per decade) is identified during the warming period in the Grande Côte region. Cap Blanc shows remarkable stability across all indicators, highlighting the buffering capacity of permanent upwelling.

Composite analysis of 283 MHW events identifies wind relaxation as the primary driver, reducing Ekman pumping and weakening upwelling circulation. However, the oceanic response differs strongly between regions. In the northern SMUS (Cap Blanc), where upwelling is quasi-permanent, MHWs are moderate and adjust rapidly, reflecting a thermodynamically dominated regime. In contrast, in southern regions with seasonal upwelling, wind relaxation leads to a near-complete collapse of upwelling, resulting in stronger and more persistent heat accumulation, characteristic of a dynamically controlled regime.

These physical contrasts translate into distinct biological responses. In the southern SMUS, MHWs lead to a systematic decrease in chlorophyll-a (up to −2 mg m⁻³). In contrast, Cap Blanc exhibits a season-dependent response, with an increase in biomass (up to +3 mg m⁻³) during the CP, sustained by residual nutrient supply and moderate warming, and a decrease in biomass (up to −2.7 mg m⁻³) during the warming period due to enhanced stratification.

These results highlight the crucial role of local upwelling dynamics in modulating the characteristics of MHWs and their ecological impacts in this marine ecosystem of major socio-economic importance.

Received: 04 May 2026 – Discussion started: 21 May 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Waly Dione, Ibrahima Camara, Alban Lazar, Diana Ruiz-Pino, and Amadou Thierno Gaye

Status: open (until 13 Oct 2026)

Post a comment Subscribe to comment alert

RC1:
'Comment on egusphere-2026-2571', Anonymous Referee #1, 10 Sep 2026 reply
Note: I wrote these comments in a markdown editor and pasted them here; some markdown remains in places.

General comments
This study addresses a worthwhile question, which is, how do marine heatwaves differ among the permanent and seasonal upwelling regions of the Senegal–Mauritania coast, and what chl-a anomalies accompany them? The long SST record, regional comparisons, and separation of cooling and warming periods provide a useful basis for answering this. The seasonal reversal in chl-a anomalies at Cap Blanc is interesting. The subject fits Ocean Science, and the regional analysis could contribute if the results and their interpretation are better substantiated.

The main challenge is that the manuscript goes too quickly from associations among SST, winds, upwelling indices, and chl-a to quite specific explanations of circulation, nutrient supply, and phytoplankton growth. Several of those explanations are not established by the measurements. Most consequentially, negative anomalies in the thermal upwelling index are treated as evidence that upwelling has "collapsed". The index is itself partly determined by the coastal SST used to identify heatwaves, and an anomaly below zero does not mean that the underlying index is below zero. The distinction between a northern regime dominated by surface heat exchange and a southern regime dominated by ocean dynamics also needs stronger support. These interpretations underpin much of the abstract and conclusion.

The manuscript also has further problems with event selection, composite uncertainty, seasonal comparisons, and reproducibility. Some statements contradict the reported results, including the assertion that Cap Blanc is stable across all indicators. The presentation is generally understandable, but repeated mechanistic explanations and an extended conclusion obscure the distinction between observations and proposed explanations.

I recommend major revision. The spatial and seasonal descriptions are potentially valuable, but the central mechanistic claims need either additional analysis or substantial qualification before we can accept them. My five comments below point to the main work required; several could change the conclusions, so they should be resolved before rewriting the abstract and discussion.

Specific comments
1. l.195–204, 224–234, 399–447, 465–497, and 526–549: How do we know that upwelling collapses?
Figures 7e and 8e show CUI anomalies. However, negative values are repeatedly interpreted as a breakdown or near-complete cessation of upwelling. These are different statements. Appendix A reports positive cooling-season CUI values of approximately 1.5–2.0 °C in the southern regions. An anomaly of −0.5 °C could therefore leave a substantial positive coastal–offshore temperature difference. Please show the absolute index and the corresponding event-date climatology alongside its anomaly, then reassess the descriptions of persistent, weakened, and collapsed upwelling. Even an absolute thermal-index reversal would require care when interpreted as a circulation reversal.

The calculation also depends on the choice of reference. From Eq. (3), CUI′ = SST′offshore − SST′coast, so selecting events with unusually high coastal SST can produce negative CUI anomalies without independently demonstrating a reduction in vertical transport. Please separate the coastal and offshore SST contributions and support the circulation interpretation with an independent diagnostic. An alongshore wind-stress/offshore Ekman-transport index would be useful here. Wind-stress curl measures Ekman pumping, but it does not by itself represent all coastal upwelling driven by offshore transport at a coastal boundary. Likewise, a wind-speed anomaly does not fully describe changes in the upwelling-favourable wind component.

The heat-flux analysis needs the same distinction between an indicative association and an established/quantified mechanism. At l.391–398, the approximate heat-budget calculation suggests that ocean processes dominate warming in all three regions. The abstract and conclusion nevertheless describe Cap Blanc as thermodynamically dominated. Please quantify the surface-flux contribution separately during development and decay, integrate the flux over explicitly defined intervals, and report the assumed mixed-layer depth and its sensitivity. For scale, 20 W m⁻² sustained for five days would warm a 20 m layer by about 0.11 °C, using seawater density 1025 kg m⁻³ and heat capacity 3990 J kg⁻¹ K⁻¹. This shows why the integration interval and mixed-layer depth must be considered. Differences in composite peak height or recovery time do not indicate/reveal which heat-budget term dominates.

The acknowledgement at l.228–234 that this is not a full mixed-layer heat budget is appropriate, but then the subsequent claims need to respect that limitation. An acceptable revision could keep the observed associations and describe the mechanisms as hypotheses. If you retain the strong attribution, you also need stronger heat-budget evidence. [Sala et al. (2025), *Leading dynamical processes of global marine heatwaves in an ocean state estimate*](https://os.copernicus.org/articles/21/2463/2025/) (and cited refs.), provide a relevant Ocean Science example of quantifying contributions during MHW onset/decline; however, this is not a requirement to reproduce that particular modelling approach, and I am showing this FYI.

2. l.150–170, 209–239, and Figures 7–9: Make the event table and composite construction reproducible.
How were the regional events identified? Detecting events in a spatially averaged SST series is not equivalent to detecting them at individual grid cells (or contiguous regions of grid cells) and then averaging their properties. Table 1 contains 92, 102, and 89 regional events, which sum correctly to 283; the seasonal totals also agree with the Figure 8 caption. However, the methods do not explain precisely how they were constructed. Please specify the spatial averaging and weighting, the event-selection rule, and whether you count simultaneous events in different regions separately. For the spatial chl-a composites, state whether each grid cell uses its own event dates or a regional pool.

The baseline description is inconsistent. Section 2.3 specifies 1982–2011 for MHW detection, whereas l.237–238 say that the full 1998–2024 period was used to calculate the seasonal chl-a climatology and detect events for the biological analysis. Was the original dataset restricted to the chl-a era, or were MHWs redetected against a different SST baseline? Please distinguish the chl-a climatology more clearly from the SST detection climatology. Unless there is a specific reason to change event definitions, use the same SST baseline and compare the physical and biological composites over the same event subset. Otherwise, differences between their samples could be mistaken for physical–biological relationships.

Please clarify the filtering at l.224–225. Does the 30-day low-pass filter act on the climatological seasonal cycle, the anomaly time series, or both? Was SST filtered before detection? Report the filter type, window, phase treatment, and how you handled record boundaries. The abrupt 2–3-day warming described at l.378–379 and the short fluctuations in Figures 7–9 need to be reconciled with this statement. If you applied a centred filter to anomalies, it can spread changes into the period before onset and affect inferred biological lead times. An unsmoothed or minimally smoothed sensitivity comparison would resolve this.

The composite curves and maps also need uncertainty estimates. Show the number of contributing events or valid observations at each lag, and explain the handling of overlapping windows, nearby events, events longer than 30 days, and incomplete windows. Use an uncertainty procedure that preserves dependence within events and accounts for temporal clustering; resampling individual daily values would not do this. Compare with seasonally matched non-event windows, or another justified reference, to establish how unusual the composite anomalies are. Where long-term changes could explain persistent offsets, test sensitivity to those changes while retaining the fixed-baseline analysis for historical MHW trends. Please also define the preconditioning, onset, and offset intervals used in Figure A3. A window centred on onset does not align with event termination, so its post-onset mean is not automatically a common decay trajectory.

3. l.133–148, 235–239, and 449–505; Table 3: Reveal the chl-a result before assigning nutrient, light, and temperature mechanisms.
The Cap Blanc seasonal contrast merits attention, and Table 3 usefully shows that individual events differ. Please develop this evidence. Give uncertainty intervals for the event-level responses and positive-response proportions, test the seasonal contrast with appropriate allowance for temporal dependence, and show whether the mean is sensitive to a few large blooms. The 22/34 positive cooling-period responses at Cap Blanc and 6/26 warming-period responses are more informative than a claim based only on the largest composite anomaly. The southern response is predominantly negative, but not always: Table 3 reports positive responses in 14% of Grande Côte events and 30% of Petite Côte events across the year. The abstract's description of a systematic decrease needs qualification.

The chl-a product must be identified precisely, including its product and dataset identifiers, processing version, and whether daily values are observed Level 3 fields or temporally and spatially filled Level 4 fields. This difference is important for event timing. The cited [Garnesson et al. (2019) product paper](https://os.copernicus.org/articles/15/819/2019/) describes both daily and daily-filled products, with the latter incorporating spatial and temporal interpolation. Please report quality flags, cloud-related missingness, minimum event coverage, and the number of retained observations. Explain the averaging of this skewed variable and add relative or log-scale anomalies to the absolute mg m⁻³ differences. Larger cooling-period losses in absolute units may partly reflect a larger seasonal chl-a baseline.

Surface chl-a is a useful proxy, but it does not directly measure phytoplankton carbon, primary production, nutrient limitation, or growth rate. Nutrients, light, mixed-layer depth, grazing, and phytoplankton physiology are not measured in this study. Changes in pigment content, vertical redistribution, transport, and losses could contribute to the observed signal. So, the detailed N–L–T explanation at l.465–497 remains a proposed interpretation. Please define this framework before using it and reduce the claims that its separate terms have been assessed/identified. In particular, the study cannot establish enhanced photosynthetic efficiency or a measured reduction in primary production from chl-a alone.

The negative southern chl-a anomalies already present at day −30 deserve explicit consideration. They may indicate shared environmental preconditioning rather than a response caused by the subsequent threshold-defined MHW. This is another reason to examine matched reference windows and the filtering. The warming-period observations in Table 3 remain valid observations even where the proposed upwelling explanation is inapplicable. Please discuss that limitation instead of labelling the responses themselves as not physically interpretable. Also, keep fisheries and spawning implications as motivation and possible consequences, since the analysis contains no direct measurements of those outcomes.

4. l.209–215, 288–312, 328–369, and 517–525; Tables 1–2: Reassess the seasonal and trend claims against the actual metrics.
Cooling and warming periods contain seven and five months, respectively. Comparing their raw event counts or cumulative days does not demonstrate a higher occurrence rate in the longer season. Table 1 is especially useful... Grande Côte has 47 cooling-period events and 55 warming-period events, despite the longer cooling period. Its cooling-period events last longer on average, but they are not more numerous. At Cap Blanc, mean duration is almost identical between periods. Please distinguish event frequency, total MHW days, and event duration throughout. Keep the seasonal totals if they represent the intended measure of exposure, but also provide an occurrence rate or proportion of eligible days when comparing seasons. Explain how events crossing a season or year boundary contribute to each metric, particularly because the current assignment uses the date of maximum intensity.

The statement at l.16 that Cap Blanc is stable across all indicators conflicts with its significant annual MHW-day trend of +4.35 days decade⁻¹ in Table 2 and Figure 6. A narrower near-coast area in Figure 4 could behave differently from the regional average, but that distinction must be explicit. Moreover, a non-significant trend is not evidence of no change. The smaller Cap Blanc trend may support relative moderation of thermal exposure, but it does not show (or even hint at) complete stability or protection of an ecosystem.

The claim that Grande Côte trends are strongest particularly during the cooling period also needs precision. Table 2 gives MHW-day slopes of +6.75 and +6.00 days decade⁻¹ for cooling and warming periods, while duration slopes are +2.35 and +2.64 days decade⁻¹, respectively. These do not support a general statement that both trends are much stronger in the cooling period. Report uncertainty and test contrasts directly where differences among regions or seasons are central to the argument. Significance in one series and non-significance in another do not demonstrate a significant difference between them. Separately estimated Sen slopes need not add across seasons, so non-additivity alone would not establish an error.

Please state the annual or seasonal observations used for the Mann–Kendall tests and Sen slopes, including how years without events are handled. Zero event counts are meaningful, but a year without events has no event duration or intensity to average. Address serial dependence and explain how you treated the many grid-cell tests. Confidence intervals for regional slopes, along with an appropriate assessment of spatial significance, would make the main comparisons more useful than significance stars alone. The exceptional recent years also require a sensitivity check on the inference about long-term change.

Lastly, define how you aggregate maximum intensity. Is it the mean of individual event maxima, the largest value in each year, or a maximum over the full record? These are different quantities, and the distinction is important when comparing Table 1 with Figures 2, 3, and 6. Figure A2 also requires correction or explanation... its month-specific MHW-day values reach hundreds, although the caption describes days per year. State the accumulation period and denominator explicitly.

5. l.107–148, 179–204, 385–387, 554–565, and 698–705: Verify the spatial calculations and provide the analysis workflow.
Please provide the exact regional masks, their areas, and the number of wet grid cells; the bathymetric dataset; and an explanation of how you selected cells shoreward of the 200 m isobath. Cap Blanc is defined as 20.8–22° N in the methods and Figure 1, but 20–22° N in Tables 1–2 and Appendix A. This affects whether part of the excluded Banc d'Arguin sector is included in the analysis. Specify the coastal and offshore sampling bands for CUI, including how the 5° westward displacement follows the coastline and how missing or land-adjacent cells are treated.

While reviewing Figure 1, I noticed a discrepancy between the spelling of place names in the Figure and in the text. Ensure they are correctly spelt (or a consistent spelling applied) both in the text and anywhere in Figures/Tables. Further, ensure that all place names in the text also appear in Figure 1.

A 0.25° grid represents roughly 25–28 km here and may resolve some shelf sectors with very few cells. Bilinear interpolation of 4 km chl-a onto that grid does not calculate a mean over each coarse cell. Please test whether the principal coastal and regional chl-a results persist with area-based aggregation of valid ocean pixels, or averaging within common regional masks at native resolution. Likewise, matching nominal grid spacing does not establish matching effective resolution of SST and reanalysis winds. Discuss what these products can resolve about the coastal fronts and wind structure used in the interpretation.

The EPV units and magnitudes need to be checked directly against the calculation. Figure 7d is labelled in 10⁻⁵ m s⁻¹, while its caption specifies 10⁻⁶ m s⁻¹; ditto for Figure 8. Appendix A reports 40–47 × 10⁻⁵ m s⁻¹ during winter at Petite Côte, equivalent to about 35–41 m day⁻¹. The event anomalies of 0.2–0.4 × 10⁻⁵ m s⁻¹ quoted at l.386 would be approximately 0.4–1% of those winter values, rather than 10%. These comparisons do not identify the source of the discrepancy, but they do show a problem somewhere. Please verify derivative distances in metres, longitude scaling with latitude, array orientation, coastal differences, and the powers of ten used in plotting. Equation (2) is the constant-f curl approximation; explain its use across this latitudinal domain or assess the effect of using curl(τ/f)/ρ instead.

The public source datasets and an unmodified detection toolbox do not reproduce the complete study. Please archive the scripts for preprocessing, masks, indices, event selection, seasonal allocation, composites, statistical analysis, and figures, together with a dated software version and the derived event catalogue. Identify the exact functions used for trend estimation. If the toolbox's `mean_and_trend.m` was used, its [currently available source](https://github.com/ZijieZhaoMMHW/m_mhw1.0/blob/master/mean_and_trend.m) uses MATLAB regression for trend estimates; it does not by itself implement the Mann–Kendall/Sen procedure described here. I am more concerned that you identify the study's implementation than that you provide evidence that this particular routine was used. I also think the journal's [data policy](https://www.ocean-science.net/policies/data_policy.html) requires access to the data needed to replicate figures.

More specific comments
l.31–34 and references at l.660–662: Please check both the numerical claim about increasing MHW frequency and its citation. The reference list gives Eric Oliver et al.'s eastern Tasmania paper, whereas the global trend discussion appears to require [Oliver et al. (2018), *Longer and more frequent marine heatwaves over the past century*](https://www.nature.com/articles/s41467-018-03732-9). Confirm that the stated percentage, metric, and time interval match the source. Also verify the claim that EBUS contain 20% of global marine biomass on 1% of the ocean surface. Biomass, primary production, and fish catches are different quantities.

l.55–67 and 84–96: Define the new contribution more precisely relative to the regional work already cited. [Imbol Koungue et al. (2025)](https://consensus.app/papers/interannual-variability-of-net-primary-productivity-in-koungue-prigent/65d49839e5b75e6cb23c577caec70998/?utm_source=chatgpt) examined low primary-productivity events and Dakar Niños, including local wind forcing and remote coastal-trapped waves. [Zhan et al. (2024)](https://consensus.app/papers/reduced-and-smaller-phytoplankton-during-marine-zhan-feng/cc9f955266935ae2b8924addc6a70e62/?utm_source=chatgpt) assessed phytoplankton changes in EBUS, with their Canary domain extending from [20 to 40° N](https://www.nature.com/articles/s43247-024-01805-w). Your study extends substantially farther south and resolves contrasting seasonal regimes, which is a plausible contribution. Explain that contribution without implying the broader physical–biological question has not been studied. Neither cited study is missing from the bibliography, but the manuscript lacks more direct/precise engagement with their findings and scope.

l.126–132: Cite the version-specific OISST paper, [Huang et al., *Improvements of the Daily Optimum Interpolation Sea Surface Temperature (DOISST) Version 2.1*](https://doi.org/10.1175/JCLI-D-20-0166.1), alongside any historical product reference. The description of satellite inputs should also reflect the actual period used. [NOAA's product documentation](https://www.ncei.noaa.gov/products/optimum-interpolation-sst) records changes in satellite inputs, including the use of AVHRR and VIIRS in later production. An AVHRR-only description is incomplete for the full record.

l.153–162: Please report the smoothing of the daily climatology and threshold, in addition to the 11-day sampling window. The linked [detection function](https://github.com/ZijieZhaoMMHW/m_mhw1.0/blob/master/detect.m) includes a default 31-day smoothing window; identify the settings actually used. The claim that 1982–2011 captures a complete AMO cycle is not supported, as it spans parts of both a negative and a positive phase. The cited [Knight et al. (2005)](https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2005GL024233) discusses an observed period of approximately 65 years. A fixed 30-year baseline is defensible, but describe what it measures and test sensitivity to the baseline rather than calling a later baseline intrinsically biased or the earlier anomalies the true amplitude.

l.253–280: The coldness of upwelled water alone does not explain fewer percentile-defined MHWs, because the threshold is local and seasonally varying. Discuss warming relative to the local baseline, variability, and persistence. The explanations involving heat advection, frontal gradients, and stratification also need to be separated: a horizontal SST gradient does not directly measure vertical stratification or advective heat convergence.

l.362–367: An annual total approaching 180 MHW days does not establish that one event lasted 180 days. Describe 2023 as an exceptional year unless an individual event duration is shown. AMO and ENSO contributions remain hypotheses here because the authors have not analysed their relationships with the event series. Consider remote forcing when discussing possible drivers, rather than assigning local wind relaxation as the trigger for every event.

l.406–413 and Figure 7a; Figure A3: Please check the quoted Cap Blanc recovery value against the curve. Figure 7a appears to end near +0.6 °C at day +30, rather than +0.2 °C. Figure A3 also shows a Cap Blanc longwave anomaly near +8 W m⁻² at onset, which is not obviously negligible alongside the reported net-flux anomalies. Define the averaging intervals and match these descriptions with the plotted values.

l.669–670: The Schlegel, Oliver, and Chen (2021) reference does not match the identifiable paper by those authors. Please verify the intended source. Their published article is [*Drivers of Marine Heatwaves in the Northwest Atlantic: The Role of Air–Sea Interaction During Onset and Decline*, Frontiers in Marine Science, 8, 627970](https://www.frontiersin.org/journals/marine-science/articles/10.3389/fmars.2021.627970/full). Correct the bibliography and check whether each associated claim is supported by that paper.

l.508–552: Shorten the conclusion to the established spatial pattern, the qualified seasonal and trend comparisons, and the chl-a associations. Add a concise limitations paragraph in the discussion. It should cover the surface-only observations, resolution, incomplete mechanism attribution, event/composite uncertainty, and shorter chl-a record. Thermal exposure does not on its own establish ecological vulnerability or fisheries impacts, so qualify those terms.

Technical corrections
l.4-5: Be upfront re the identity of the data products used here.

l.10 and abstract: Define CP at first use, and distinguish the SST/physical record from the shorter chl-a record. If the record contains all of 1982–2024, it contains 43 calendar years, not 42.

l.100, 118, and 122: The study-area, SST-threshold, and SST-trend references to Figure A1 appear to require Figure 1. Check all figure references.

l.89–90, 93–95, and 348–349: Rewrite the incomplete or awkward sentences, including “Then, are examined”, “allows to disentangle”, and “spatially averages in each three sub-regions”.

Section 3.2.1 heading: Remove “inof”. **l.365–367:** Correct “begining”, “21th”, and “inprint”, and repair the sentence punctuation.

l.396: Correct “anti-symetrical”.

Figure 3: The warming-period panel labels repeat a1, b1, and c1; change them to a2, b2, and c2.

Figure 8 caption: Cooling and warming periods occupy the upper and lower groups of panels, not left and right panels.

Table 2 and Figure A5a3: Reconcile the Petite Côte warming-period MHW-day slope: +1.34 in the table and abstract, but +1.38 in the figure.

Figure A5 caption: The region-specific bars do not use the blue/orange/yellow regional colour scheme described in the caption.

Figure A1b2: Correct its internal panel label, which repeats a2.

Figures 2c and A6b: Supply units for the SST-gradient contours.

Figure 9: Add a reference magnitude for wind vectors and use a clearly identified zero-centred scale for chl-a anomalies.

Figure A3: Number panels consistently and define the abbreviated phase labels.

Table 3, p.26: Repair the caption, whose beginning is clipped in the supplied PDF.

l.563–565: Use the toolbox's name, `m_mhw`, consistently and replace the incomplete `/blob/master/` link with a working repository or archived-version link.

Throughout: Standardise the study-region name, Grande Côte/Petite Côte spelling, chl-a notation, and regional bounds. Distinguish raw indices from their anomalies in labels and prose. Correct the EPV units only after checking the calculations described above.

Reply
Citation: https://doi.org/10.5194/egusphere-2026-2571-RC1

Waly Dione, Ibrahima Camara, Alban Lazar, Diana Ruiz-Pino, and Amadou Thierno Gaye

Viewed

Total article views: 267 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
161	84	22	267	14	13

HTML: 161
PDF: 84
XML: 22
Total: 267
BibTeX: 14
EndNote: 13

Views and downloads (calculated since 21 May 2026)

Month	HTML	PDF	XML	Total
May 2026	86	28	8	122
Jun 2026	16	7	1	24
Jul 2026	25	21	5	51
Aug 2026	13	11	2	26
Sep 2026	21	17	6	44

Cumulative views and downloads (calculated since 21 May 2026)

Month	HTML	PDF	XML	Total
May 2026	86	28	8	122
Jun 2026	16	7	1	24
Jul 2026	25	21	5	51
Aug 2026	13	11	2	26
Sep 2026	21	17	6	44

Viewed (geographical distribution)

Total article views: 251 (including HTML, PDF, and XML) Thereof 251 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 14 Sep 2026

Short summary

Marine heatwaves (1982–2024) in the senegalo-mauritanian upwelling show strong spatial heterogeneity. Shallow zones face up to 50 heat days/year, while persistent upwelling areas like Cap Blanc act as refuges. Wind relaxation drives events by weakening upwelling. The Grande Coast shows increasing vulnerability. Biologically, heatwaves reduce chlorophyll in the south, whereas Cap Blanc shows seasonal biomass increases during cold periods due to residual nutrients.


Total:	0
HTML:	0
PDF:	0
XML:	0