Accurate Estimation of Chlorophyll-a Concentration in the Coastal Areas of the Ebro Delta (NW Mediterranean) Using Sentinel-2 and Its Application in the Selection of Areas for Mussel Aquaculture

Abstract

Multispectral satellite remote sensing imagery, together with appropriate modeling, have been proven to provide chlorophyll-a maps that are useful to evaluate the suitability of coastal areas for carrying out shellfish aquaculture. However, current approaches used for chlorophyll-a estimation in very shallow coastal areas often fail in their accuracy. To overcome this limitation, an algorithm that provides an accurate estimation of chlorophyll-a concentration in the coastal areas of the Ebro delta (North Western Mediterranean) using atmospherically corrected Sentinel 2 (S2) remote sensing reflectances (Rrs) has been calibrated and validated. The derived chlorophyll-a maps created have been used in a dynamic carrying capacity model that covers areas from very rich waters inside the embayment to the more oligotrophic waters in the open sea. The use of carrying capacity models is recommended to evaluate the potential of marine coastal areas for bivalve mollusk aquaculture. In this context, the depletion of chlorophyll-a is an indicator of negative environmental impact and thus a continuous monitoring of chlorophyll-a is key. The proposed methodology allows estimation of chlorophyll-a concentration from Sentinel-2 with an accuracy higher than 70% in most cases. The carrying capacity and the suitability of the external areas of the Ebro delta have been determined. The results show that these areas can hold a significant mussel production. The methodology presented in this study aims to provide a tool to the shellfish aquaculture industry.

Assessing the Chlorophyll-a Retrieval Capabilities of Sentinel 3A OLCI Images for the Monitoring of Coastal Waters in Algoa and Francis Bays, South Africa

“…This observation is consistent with findings by other researchers [106][107][108][109] who observed good correlations between Sentinel-2-based Chl-a concentration measurements and in-situ data. This finding was further validated through ANOVA, which revealed statistically insignificant differences at seven of the eight stations that were used as sources of validation reference data (Table 5).…”

“…These and other factors may have reduced the ability of Sentinel images to capture representative signals that compared well with in-situ measurements. Basing analysis of the correlations between in-situ and the image-based measurements we obtained in this study (Table 4) on a two-tier high and low classification in which r values ≥ 0.538 show good correspondence [106], it is evident that at most stations (five out of eight) Sentinel images yielded good Chl-a estimations that can be reliably used to objectively inform the identification of environmentally friendly of the oceanic resources at our disposal. The results shown in Table 4 closely mimic what ANOVA revealed (Table 5).…”

Analysing Spatiotemporal Variability of Chlorophyll-a Concentration and Water Surface Temperature in Coastal Lagoons of the Ebro Delta (NW Mediterranean Sea, Spain)

“…The Ebro Delta extends 25 km from the mainland and forms two embayments, El Fangar and Els Alfacs (Figure 1). The former has an extension of 12 km 2 , a mean depth of 2 m, and a 1 km wide mouth, while the second has an extension of 50 km 2 , a mean depth of 4 m, and its mouth is 3 km wide [25]. Other coastal lagoons can be found in the delta, such as the El Garxal and La Tancada lagoons.…”

“…Likewise, the temperature pixels extracted were used to calibrate and validate a model able to relate estimated and measured temperature. Lastly, the performance of all the algorithms tested was evaluated using the R 2 and RMSE statistics [19,25].…”

“…So much so that their use has progressively increased in the last decades, yielding very promising results that prove their suitability for a wide range of environmental applications [17]. The previous include water quality monitoring in continental lakes and reservoirs [18][19][20][21], as well as coastal waters [22,23], evaluating the suitability of coastal areas for fisheries and aquaculture [24,25], characterisation and mapping of a variety of coastal habitats and their changes [26][27][28][29][30], or even the surveillance of harmful coastal algal blooms [31,32]. Furthermore, these techniques have also allowed assessing the impact of extreme storm events on water quality and flooding [14,33], measuring human impact on water transparency [34], or assessing the impact of meteorological freshwater episodes and winds on water temperature and Chl-a concentration in coastal lagoons [9], among many others.…”