Continental Shelf Research 232 (2022) 104629
6
allows us to assume that a likely overestimation of the Chl-a values in the
very coastal area does not significantly impact our conclusions.
3.3. Cross-shore and seasonal variability of Chl-a
Chl-a cross-shore sections (Fig. 6b) indicate that the lowest Chl-a
concentrations are observed from May to October at all locations, both
in coastal and offshore areas. The highest Chl-a concentrations, representing the productive season, are observed from November to April. In
the coastal area, the maximum Chl-a can exceed 2 mg m
-3
in winter
(Figs. 7a) and 0.5 mg m
-3
in summer (Fig. 7a). However, in the offshore
area, beyond 10 km from the coast, the average minimum reach 0.5 mg
m
-3
in winter and 0.2 mg m
-3
in summer (Figs. 6b and 7b). Beyond 10
km, the Chl-a concentration becomes stable during all seasons (Figs. 6b
and 7a). Indeed, we chose the distance of 10 km from the coast as the
shortest distance at which Chl-a seasonality becomes weak and stops
increasing towards the offshore (Figs. 6b and 7b).
The cross-shore gradient is well pronounced throughout the year and
increases exponentially towards the coast (Fig. 6b). The intensity of this
gradient is maximum during the productive season and is always
maximal at the coast and regularly decreases with increasing distance
from the coast (Fig. 6b), ranging from 0.2 to 2 mg m
-3
onshore (Figs. 6b)
and 0.1–0.5 mg m
-3
offshore (Fig. 6b).
3.4. Meridian variability and seasonality of Chl-a biomass index
The integrated cross-shore biomass index (Section 2.6) was
computed from each coastal point northwards to integrate the Chl-a
concentration up to a variable distance (Fig. 8b) where Chl-a drops
below the predefined value of 0.5 mg m
−3
(see methods). The aim was to
explore and quantify the integrated coastal Chl-a biomass along the
Algerian coastline (Fig. 8d). The value was carefully chosen as the best
threshold β = 0.5 mg m
−3
that intersects the different average crossshore climatological profiles over the year (Fig. 7b). The red line in
Fig. 8a represents the maximum distance effectively reached during the
productive season (we considered Chl-a <0.5 mg m
−3
as oligotrophic
and does not represent an enrichment). The resulting distance (Fig. 8b)
shows that the productive area varies over time from a few kilometres in
summer (cyan area in Fig. 8b) with a minimum of 1 km (when only one
sea pixel is considered) to a maximum of 80 km in winter, mainly
reached near the Alboran Sea in the west. The longitudinal gradient
along the coastline shows an apparent decrease of this distance
eastward, which is well in line with the decreasing influence of the
enrichment of Atlantic waters entering the Alboran Sea and moving
eastwards.
The average Chl-a along the same transect (Fig. 8c) shows extreme
variability between regions, from values <0.5 mg m
−3
in summer (this is
possible when a minimum of one sea pixel is considered) to values >2
mg m
−3
between April and November, i.e., during low-biomass season.
The resulting biomass index I B (Fig. 8d) is defined as the product of the
previous distance by the average Chl-a concentration along the same
transect (Eq. (6)). This index represents the spatial integration of the
most elevated Chl-a values along the cross-shore transect, while the
vertical dimension is partially considered by the attenuation depth of the
satellite measurement. Nevertheless, this depth represents a variable
fraction of the euphotic layer, according to the shape of the vertical Chla profile.
The Chl-a (Fig. 8c) represents a proxy of the average phytoplankton
biomass over the cross-shore profile, while the spatially integrated index
(I B , Fig. 8d) is predominantly determined by the length of the profile
(Fig. 8b). The I B index is expressed in g m
-2
and varies between 20 and
50 g m
-2
during the productive season (Fig. 8d) with a regular eastward
decrease. Several peaks in phytoplanktonic biomass are observed (I B is
>45 g m
-2
; between 2.2
◦
W and 0.5
◦
E from January to March, between
5.1
◦
E to 5.5
◦
E in January and February, and at 7.7
◦
E from January to
March–April). These peaks correspond to pronounced extensions of
Dist0.5 mg (Fig. 8b) as near the Alboran region or mostly to higher Chl-a
values in the central and eastern parts of the AB (Fig. 8c). I B is <10 g m
-2
everywhere during the low-biomass season, except in the Algiers and
Annaba bays (Fig. 8d). I B is, in fact, practically zero from June to
September in many locations where the Chl-a concentration barely exceeds 0.5 mg m
-3
(Fig. 8c).
4. . Discussion
The construction of a data set at 1-km spatial resolution allows
investigating and extracting the specific enrichments of coastal origin,
distinguishable from the large-scale seasonal cycle. Previous descriptions of the climatological cycle in the AB were based on monthly
averages (instead of fortnightly in this study) and at a much higher
spatial resolution. Therefore, our description of the cross-shore gradient
of Chl-a and its variability along a longitudinal gradient becomes much
more realistic and highlights the high importance of the coastal domain
(<10 km) in the whole basin.
Fig. 5. Results of the elimination of outlier pixels from the original uncorrected MODIS Level-2 data (left column), respectively for a fortnightly average (example of
1-15/01/2014) (a and b), and for a monthly climatological average (example of January 2003–2018) (c and d).
R. Harid et al.
6
allows us to assume that a likely overestimation of the Chl-a values in the
very coastal area does not significantly impact our conclusions.
3.3. Cross-shore and seasonal variability of Chl-a
Chl-a cross-shore sections (Fig. 6b) indicate that the lowest Chl-a
concentrations are observed from May to October at all locations, both
in coastal and offshore areas. The highest Chl-a concentrations, representing the productive season, are observed from November to April. In
the coastal area, the maximum Chl-a can exceed 2 mg m
-3
in winter
(Figs. 7a) and 0.5 mg m
-3
in summer (Fig. 7a). However, in the offshore
area, beyond 10 km from the coast, the average minimum reach 0.5 mg
m
-3
in winter and 0.2 mg m
-3
in summer (Figs. 6b and 7b). Beyond 10
km, the Chl-a concentration becomes stable during all seasons (Figs. 6b
and 7a). Indeed, we chose the distance of 10 km from the coast as the
shortest distance at which Chl-a seasonality becomes weak and stops
increasing towards the offshore (Figs. 6b and 7b).
The cross-shore gradient is well pronounced throughout the year and
increases exponentially towards the coast (Fig. 6b). The intensity of this
gradient is maximum during the productive season and is always
maximal at the coast and regularly decreases with increasing distance
from the coast (Fig. 6b), ranging from 0.2 to 2 mg m
-3
onshore (Figs. 6b)
and 0.1–0.5 mg m
-3
offshore (Fig. 6b).
3.4. Meridian variability and seasonality of Chl-a biomass index
The integrated cross-shore biomass index (Section 2.6) was
computed from each coastal point northwards to integrate the Chl-a
concentration up to a variable distance (Fig. 8b) where Chl-a drops
below the predefined value of 0.5 mg m
−3
(see methods). The aim was to
explore and quantify the integrated coastal Chl-a biomass along the
Algerian coastline (Fig. 8d). The value was carefully chosen as the best
threshold β = 0.5 mg m
−3
that intersects the different average crossshore climatological profiles over the year (Fig. 7b). The red line in
Fig. 8a represents the maximum distance effectively reached during the
productive season (we considered Chl-a <0.5 mg m
−3
as oligotrophic
and does not represent an enrichment). The resulting distance (Fig. 8b)
shows that the productive area varies over time from a few kilometres in
summer (cyan area in Fig. 8b) with a minimum of 1 km (when only one
sea pixel is considered) to a maximum of 80 km in winter, mainly
reached near the Alboran Sea in the west. The longitudinal gradient
along the coastline shows an apparent decrease of this distance
eastward, which is well in line with the decreasing influence of the
enrichment of Atlantic waters entering the Alboran Sea and moving
eastwards.
The average Chl-a along the same transect (Fig. 8c) shows extreme
variability between regions, from values <0.5 mg m
−3
in summer (this is
possible when a minimum of one sea pixel is considered) to values >2
mg m
−3
between April and November, i.e., during low-biomass season.
The resulting biomass index I B (Fig. 8d) is defined as the product of the
previous distance by the average Chl-a concentration along the same
transect (Eq. (6)). This index represents the spatial integration of the
most elevated Chl-a values along the cross-shore transect, while the
vertical dimension is partially considered by the attenuation depth of the
satellite measurement. Nevertheless, this depth represents a variable
fraction of the euphotic layer, according to the shape of the vertical Chla profile.
The Chl-a (Fig. 8c) represents a proxy of the average phytoplankton
biomass over the cross-shore profile, while the spatially integrated index
(I B , Fig. 8d) is predominantly determined by the length of the profile
(Fig. 8b). The I B index is expressed in g m
-2
and varies between 20 and
50 g m
-2
during the productive season (Fig. 8d) with a regular eastward
decrease. Several peaks in phytoplanktonic biomass are observed (I B is
>45 g m
-2
; between 2.2
◦
W and 0.5
◦
E from January to March, between
5.1
◦
E to 5.5
◦
E in January and February, and at 7.7
◦
E from January to
March–April). These peaks correspond to pronounced extensions of
Dist0.5 mg (Fig. 8b) as near the Alboran region or mostly to higher Chl-a
values in the central and eastern parts of the AB (Fig. 8c). I B is <10 g m
-2
everywhere during the low-biomass season, except in the Algiers and
Annaba bays (Fig. 8d). I B is, in fact, practically zero from June to
September in many locations where the Chl-a concentration barely exceeds 0.5 mg m
-3
(Fig. 8c).
4. . Discussion
The construction of a data set at 1-km spatial resolution allows
investigating and extracting the specific enrichments of coastal origin,
distinguishable from the large-scale seasonal cycle. Previous descriptions of the climatological cycle in the AB were based on monthly
averages (instead of fortnightly in this study) and at a much higher
spatial resolution. Therefore, our description of the cross-shore gradient
of Chl-a and its variability along a longitudinal gradient becomes much
more realistic and highlights the high importance of the coastal domain
(<10 km) in the whole basin.
Fig. 5. Results of the elimination of outlier pixels from the original uncorrected MODIS Level-2 data (left column), respectively for a fortnightly average (example of
1-15/01/2014) (a and b), and for a monthly climatological average (example of January 2003–2018) (c and d).
R. Harid et al.