Effects of Climate Change on Moroccan Coastal Upwelling: Relationships between the NAO, Upwelling Index, and Sea Surface Temperature (1978-2024)

TL;DR

This study analyzes how climate change affects Moroccan Atlantic coastal upwelling by examining interrelationships among the NAO, Coastal Upwelling Index, and SST over 1978-2024 using ERA5 reanalysis data. It employs index calculations, seasonal correlations, Granger causality in a season-specific VAR framework, and breakpoint/trend analyses to quantify spatio-temporal dynamics and causality. Key findings include strong winter NAO–CUI coupling (DJF, r ≈ $+0.65$), a dominant local upwelling control in summer (JJA, r ≈ $-0.46$ with SST), a north–south gradient in upwelling intensity, widespread SST warming (~$+0.149^{\circ}$C per decade spatially), and a robust SST→NAO memory signal (lag 7–11 months). The results reveal a regional climate paradox where global warming coexists with localized cooling from intensified upwelling, with important implications for fisheries, aquaculture, and coastal tourism, and they advocate adaptive management through seasonal forecasting grounded in the identified lags and causal pathways.

Abstract

This study investigates climate change impacts on Moroccan Atlantic coastal upwelling (21$^{\circ}$--35$^{\circ}$N), analyzing interrelationships among the North Atlantic Oscillation (NAO), Coastal Upwelling Index (CUI), and Sea Surface Temperature (SST) from 1978--2024 using ERA5 reanalysis data. Methods include index calculations, seasonal correlations, Granger causality tests, linear trends, and breakpoint detection to evaluate spatio-temporal dynamics and causality. Results reveal seasonal variability: strong NAO--CUI coupling in winter (DJF; $r = +0.65$), local upwelling dominance in summer (JJA; $r = -0.46$ CUI--SST), and a north--south gradient with intense southern upwelling (mean Ekman transport $37.1 \pm 34.5$ m$^3$/s/100m). Trends show significant SST warming ($+0.0736$ units/decade; $+0.149^{\circ}$C/decade spatially), summer upwelling decline ($-0.0635$ units/decade in JJA), and weak NAO shifts. Granger causality indicates rapid local forcing (CUI $\to$ SST, lag 1--3 months), delayed teleconnections (NAO $\to$ SST, lag 2--4 months), and ocean feedback (SST $\to$ NAO, lag 7--11 months). Breakpoints (1992--1997) signal warming acceleration, highlighting a paradox of global SST rise amid localized upwelling-induced cooling per Bakun's hypothesis. Findings underscore implications for fisheries, aquaculture, and tourism, advocating adaptive management via seasonal forecasting. Keywords: Coastal upwelling, NAO, SST, climate change, Granger causality, Moroccan Atlantic, Ekman transport, spatio-temporal variability.

Figures (11)

• Figure 1: Geographic representation of the Atlantic coast of Morocco, indicating the main coastal cities studied along the coastline, from Dakhla to Tangier. The coastlines of neighboring countries are included to provide regional context, with Lisbon (Portugal) used as a reference point.
• Figure 2: Monthly and seasonal evolution of the NAO index (1978-2022). Colors represent seasons: DJF (blue), MAM (green), JJA (red), SON (orange).
• Figure 3: Seasonal evolution of normalized indices of NAO (blue), upwelling (green), and SST (red) along the Moroccan Atlantic coast between 1978 and 2024. Marked interannual variability is observed, particularly in winter (DJF) and summer (JJA), reflecting differentiated seasonal influence of NAO on upwelling intensity and sea surface temperature.
• Figure 4: Correlation matrices between NAO, Upwelling and SST indices for each season. Positive correlations (red) indicate direct relationships, while negative correlations (blue) indicate inverse relationships.
• Figure 5: Analysis of seasonal climate trends (1978-2024) for NAO, Upwelling and SST indices. Graphs show temporal evolution with linear trends (dotted lines) and corresponding regression equations. Trends are expressed in units per decade, with indication of statistical significance (* for $p < 0.05$).