Establishing a firm baseline on the magnitude of carbon sequestered by the biological pump will allow changes due to anthropogenic forcing to be detected and predicted. Quantifying the flux of carbon into the ocean interior requires high-quality, in-situ observations sustained over many years. Ocean time-series are critical for providing sustained data in real and delayed-mode: they provide a contemporary picture of conditions in locations of interest, as well as calibration and validation data for forward modelling of the future state across all carbon pools. The Southern Ocean Time Series (SOTS) is the longest biogeochemical moored time-series in the Southern Ocean, a region undergoing relentless anthropogenic forcing and becoming fresher, warmer, less oxygenated and more acidic. Tracking the local air-sea CO2 exchange, biological production, and carbon export that drive the OBCP in the subantarctic Southern Ocean provides a level of detail not discoverable by remote sensing e.g the timing and extent of surface coccolithophorid blooms can be observed from space, but the analysis of morpho-species shifts that can influence calcification rates and carbon export requires powerful microscopy on high-frequency samples captured by autonomous samplers, and sediment traps at depth. Using observations from two deep-water moorings at SOTS as a case-study, we discuss how in-situ observations can reveal and quantify variability across daily, seasonal, and annual cycles for many OBCP processes. The combination of deep sediment traps, water column sensor arrays and autonomous samplers is designed to capture as many of the key carbon pathways and processes as possible, complimenting the global coverage afforded by remote sensing.

Data about air-sea pCO2 fluxes are both spatial and temporal limited in many oceanic areas. However, novel approaches like the combination of in-situ and remote sensing data allow overcoming these limitations, expanding the time frame and spatial coverage of the in situ observations. In this study, we aimed to assess the air-sea pCO2 fluxes in a high oligotrophic area, like the Northeastern Levantine basin (East Mediterranean Sea). The main purposes were: i) to determine the spatial and temporal variability of the fluxes, ii) to assess the annual CO2 budget to understand if the area acts as a sink or a source of CO2, and iii) to identify multiannual trends. We developed an algorithm based on the empirical relationship between chlorophyll (Chl), sea surface temperature (SST), and partial pressure of CO2 (pCO2) in surface seawater. Data of sea surface temperature, chlorophyll, and pCO2 were collected during 7 seasonal cruises between 2018 and 2020 with a un underway system equipped with pCO2 sensor HydroC-CONTROS. The empirical algorithm has been then used to determine past data of pCO2 (starting from 2003) based on remote sensing data of SST (AVHRR ) and Chl (ESA product). Air-Sea pCO2 fluxes between 2003 and 2018 were calculated based on the Climate Data Store products for atmospheric XCO2 (SCIAMACHY/ENVISAT, TANSO-FTS/GOSAT, OCO-2) and wind speed (NASA ASCAT). The multiple linear regression applied to the pCO2, Chl-a and SST presented a very high r2 value, indicating that 96% of the variability found in pCO2 was dependent on SST and Chlorophyll concentration. The area displayed a clear spatial variability with Rhodes Gyre showing a distinct feature of low pCO2 concentration compared with the rest of the study area. The region between the coasts of Cyprus and Turkey and further extending west was marked by high pCO2 concentrations. High concentrations were also observed in Antalya Bay and further along the coast towards Crete. These patterns were detected in all seasons, even though the concentrations changed significantly between seasons. Air-sea pCO2 fluxes calculated between 2003 and 2020 displayed a clear seasonality with a net outflux (sea to air) in summer and a net sink (air to sea) in winter. Transitional seasons (Spring and Fall) presented almost a neutral flux. The time series also revealed strong interannual differences in the intensity of the air-sea CO2 flux, particularly in winter. On an annual average, the Northeastern Levantine basin behaved as a small source of CO2 for the atmosphere (0.29-0.35 mmol/m2/d). An analysis of the 17 years of data showed that there was a slight decreasing trend of the total yearly air-sea CO2 flux of -0.06 mmol/m2/d. This decrease mirrored the increase in atmospheric CO2 over the same time frame suggesting a response of the basin to the increase of atmospheric CO2.

Substantial uncertainties due to internal climate variability obscure accurate quantification of the fate of anthropogenic carbon in the Earth system. Yet such information is necessary to verify the effectiveness of fossil fuel emission reduction measures in support of policy transition. Initialized near-term prediction systems based on Earth system models have proven successful in constraining effects of climate variability on the global carbon cycle. Predictions of the ocean and land carbon sinks and variations of atmospheric CO2 concentrations are now possible by extending decadal carbon predictions systems with the ocean and land carbon cycle components. New studies demonstrate establishing predictive skill of the global carbon cycle, with emerging casual explanatory capacity of ESMs for skillful predictions of near-term variations of carbon sinks and atmospheric CO2, as well as contributing to overall understanding of decadal variability of the carbon cycle. A predictive skill for the global ocean carbon sink of up to 6 years is found for some models. Longer regional predictability horizons are found across single models. On land, a predictive skill of up to 2 years is primarily maintained in the tropics and extra-tropics enabled by the initialization of the physical climate. Furthermore, anomalies of atmospheric CO2 growth rate inferred from natural variations of the land and ocean carbon sinks are predictable at lead time of 2 years and the skill is limited by the land carbon sink predictability horizon. However, such predictions of the global carbon cycle still remain a cutting-edge activity of only a few modeling groups. Furthermore, as of now, only physical-state variables are assimilated into decadal prediction systems. A potential advantage of assimilating carbon cycle observational products remains poorly explored. I will present new outcomes and discuss future opportunities for predictions of the global carbon cycle and the planet’s breath maintained by variations of atmospheric CO2.

The ocean has played a major role in mitigating climate change by taking up roughly a quarter of annually emitted CO2. However, this has resulted in the acidification of the surface ocean with potential impacts on marine ecosystems. The impact of acidification may be felt sooner than anticipated due to the occurrence of extreme events. Here we investigate those extreme events. We use a global surface ocean pH product to investigate the character of pH extremes and how these coincide with marine heatwaves to form compound extreme events (CEX). The pH product used in this study is the output of a machine learning approach that uses satellite data to fill the sparse, ship-based surface ocean CO2 measurements. We combine satellite estimates of pCO2 and total alkalinity to derive global monthly by 1° ⨉ 1° resolution estimates of pH, dissolved inorganic carbon (DIC), and other parameters of the marine carbonate system (OceanSODA-ETHZ; Gregor and Gruber, 2020). We find that while CEXs are rare, they occur more frequently in the low latitude regions – up to 50% of maximum possible occurrence of compound extremes. The high coincidence is due to the dominance of temperature as a driver of pH between 10° – 40° N/S. Further, we find that El Niño/La Nina events are strong drivers of large-scale CEXs in the low latitude bands of the Pacific ocean. Compound extreme events are very rare in the high latitudes (< 10% of possible occurrence at > 40° N/S), where seasonal changes in pH driven by DIC are larger than those driven by temperature. Finally, we find that the Northeastern Pacific is a hotspot for long-lasting (>12 months) and intense CEXs. The characterisation of CEX events is a step towards understanding oceanic extremes, yet much work still needs to be done to understand their in-situ impact.

Satellite data provide many of the environmental variables necessary to estimate regional and global air-sea fluxes of carbon dioxide by a “bulk formula” method, with the clear exception of the upper ocean concentration of dissolved carbon dioxide. Empirical estimates require a combination of satellite data - often available as averages for cells defined in latitude, longitude, and time – with in situ data, in most cases along irregular ship tracks. The ideal workflow for this combination is not obvious and existing studies have taken divergent paths in a number of steps that substantially affect the calculated flux. The most prominent disparity results from different approaches to the inevitable mismatch between temperature measurements. One set of temperatures accompany measurements of carbonate parameters from ships or other in situ platforms, while satellites now provide refined sea surface temperature (SST) products that are usually regarded as a “gold standard” for this important climate variable. The most common practice is to estimate the fugacity outside of the measuring platform and combine this with a solubility calculated from a gridded product of SST satellite-measured temperature, disregarding discrepancies between in situ and satellite temperatures. In effect, it is assumed that the fugacity is correct for the “cell”, although the temperature is wrong and the carbonate system is very sensitive to temperature. An alternative method uses measurements at the equilibrator (for a shipboard measurement) and assumes that concentration of dissolved inorganic carbon is the conserved quantity in calculating dissolved carbon dioxide for the grid cell. These pathways lead to divergent results including a substantially larger net global ocean sink following the latter method. Presently, any debate is theoretical and argues over “the least bad option” among these pathways. However, empirically-driven progress is possible by diving deeper into the temperature data. Typically, there will be four distinct temperatures that can be readily associated with a shipboard equilibrator-based measurement of carbon dioxide. (1) Equilibrator temperature, (2) estimated temperature beyond the ship’s hull, (3) satellite measurements of SST “close” to the ship, and (4) the “cell” value of SST. The four data streams enable a separation of discrepancies into components. Estimating (2) from (1) can be more or less robust depending on the platform (e.g. an equilibrator cannot always be placed ideally on a ship of opportunity). A substantial bias between (2) and (4) is strongly associated with the contradiction in fluxes by alternative methods. Introducing (3) should provide insight into the major cause of the discrepancy between (2) and (4). Where the estimated bias between (1) and (2) is primarily at fault, this is evident in a bias between (2) and (3). Those occurrences are important since conservation of dissolved inorganic carbon is almost certain within the “plumbing” of a shipboard system, but ocean fugacity may be miscalculated if the warming from ocean to equilibrator is poorly estimated. All methods are weak if measurements on the ship track are a poor representation of the cell as a whole and this may be apparent by comparing (3) and (4).

.

Some regional algorithms for estimating size-fractionated primary production and phytoplankton carbon from the optical properties were developed in South China Sea (SCS), based on the in situ dataset collected in both the open and coastal waters. At first, a significant log-linear relationship was found between primary production (PP) and the product of the phytoplankton absorption coefficient aph and PAR (R2 =0.64). This relationship may be influenced by variations in dominant algal species. For example, Prochlorococcus-dominant samples ( average PAR of 0.79 mol m-2 h-1 ) might show the photoinhibition, and their PP values showed a decreasing trend with the increasing production of aph and PAR. However, For Diatoms-dominant and Haptophytes-dominant groups, their PP values showed an increasing trend as the increasing product of aph and PAR. Meanwhile, in situ measured hyperspectral absorption and irradiance data were also used to estimate the size-fractionated PP. Both the Size-fractionated quantum yield of photosynthesis and phytoplankton absorption coefficient were derived from the non-water absorption coefficient. Results demonstrated that the estimated size-fractionated PP showed good agreement with in-situ measurement. R2s are 0.41, 0.69, and 0.83, and median absolute percentage differences are 23.33%, 25.86%, and 31.05% for pico-, nano-, and micro-PP, respectively. For estimation of pico-phytoplankton carbon，the empirical algorithm were built to estimate the cell abundances of picophytoplankton, including Prochlorococcus (Pro), Synechococcus (Syn), and autotrophic picoeukaryotes (PE) from phytoplankton absorption coefficient at 443 nm. Then the contribution proportions of Pro, Syn and PE carbon contents of the picophytoplankton carbon biomass were calculated in the SCS. The algorithm for Syn seemed to be sensitive to variations in nutrient levels and herbivory pressure in coastal waters where nano- or microphytoplankton dominated. Based on the algorithm for estimating the size-fractionated phytoplankton carbon by Roy et al.(2017), we re-parameterized this algorithm and applied it to vertical measurements of the non-water absorption coefficient by ACS in SCS. The validation results showed the mean absolute errors and the biases between estimated and field picophytoplankton carbon were

The increase of atmospheric CO₂ levels by as much as about 10% since the beginning of 21st century and its impact on the Earth’s climate and the biosphere represent a major concern. A compilation of in-situ data over the global coastal ocean indicates that the world’s coastal shelves absorb about 17% of oceanic CO₂ influx, although these areas represent only 7% of the oceanic surface area. However, large uncertainties in coastal carbon fluxes in the coastal margins exists due to the under sampling of the coastal ocean in both space and time. Indeed, Satellite remote sensing, in conjunction with in-situ data, allow the collection of various physical and biological parameters at regional and global scales at different temporal resolutions not accessible from other in-situ observation methods. The main objective of the CO2COAST project (ANR funding) is to estimate the surface-ocean CO₂ partial pressure, pCO₂w, CO₂ flux, and associated uncertainties, from satellite remote sensing over the global coastal waters at high spatial resolution (1kmx1km). Based on these estimations, the respective contribution of the estuaries vs. continental shelves to the CO₂ fluxes will be evaluated over the global coastal waters. The global coastal database used to accomplish this aim is constituted of 11,36.10⁶ in situ data points of pCO₂, (SOCAT database) for which 580.10³ satellite match-up database has been built (first at 4 km, and later at 1 km). This match-up database gathers in-situ pCO₂w, and satellite measurements of remote sensing reflectance, Rrs, chlorophyll concentration, Chl, absorption by colored dissolved organic matter, acdom, sea surface salinity, SSS, and temperature, SST. Such a multidimensional dataset requires "intelligent investigation” using machine learning methods (ML) to exploit spatial and temporal complex structures, find patterns, and fuse heterogeneous sources of information efficiently. In a preliminary study, a ML algorithm has been applied to test two approaches, global and class-based regression, to estimate pCO₂w . The results were advantaging using the class-based approach. This result highlights that the relationship between physical and biogeochemical factors differ from a region to another while explaining the variability of pCO₂w. In this same study, the input parameters were Rrs, SST, SSS, and the coordinates. However, in a second stage, testing other configurations involving different input parameters (time, Chl, acdom, etc) is essential to evaluate the contribution of these parameters to accurately estimate the pCO₂w. For that, another ML algorithm is applied; 2S-SOM, a variant algorithm of the Self-organizing maps algorithm (SOM). Through an unsupervised learning and neural network classification, the dataset is finely clustered while evaluating the weights of each parameter in each cluster. These weights are automatically assigned while minimizing the intra-class variance of each cluster. The application of such ML algorithm will allow to better understand the drivers of the pCO₂w variability and to estimate this latter while using biotic and abiotic parameters such as Rrs, SST, SSS, acdom and Chl. The methodology under development is instrumental to build a coherent and robust satellite database of coastal pCO₂w at high spatio-temporal resolution (Daily, 1-4 km product, from 1997-2021).

Phytoplankton primary production (PP) plays an important role in the ocean carbon cycle. With the development of remote sensing, phytoplankton absorption coefficient (aph) has been considered a reliable bio-optical proxy for estimating marine PP. A PP model derived from aph and PAR was built based on a dataset collected during 2019 in coastal and open environment in the South China Sea (SCS). The model was validated by an independent dataset collected in 2018, and both the K-fold cross-validation and a sensitivity analysis were also conducted. The results indicated that there was a significant log-linear relationship between PP and the production of aph (443) and PAR with Adj.R^2 being 0.64, and compared to aph (443), the model was more sensitive to changes in photosynthetically active radiation (PAR). In addition, the influence of different dominant phytoplankton groups on the model performance was also discussed. Diatoms-dominant and Haptophytes-dominant groups have similar performance as the model, and their PP values showed an increasing trend as the increasing produce of aph (443) and PAR. However, Prochlorococcus-dominant samples with average PAR of 0.79 mol m^(-2) h^(-1) might show the photoinhibition, and the PP values showed a decreasing trend with the increasing production of aph (443) and PAR.

Coccolithophores occupy a major role in the marine carbon cycle through production and export of calcite plates (coccoliths), majorly contributing to particulate inorganic carbon (PIC) content in the open ocean. When shed, they cause strong light scattering in comparison with other phytoplankton groups. Coccolithophores can form massive blooms covering thousands of square kilometers, which appear bright turquoise in contrast to the usually dark blue ocean and are therefore reasonably well detectable with ocean color imagery. The most abundant Coccolithophore species globally and specifically in the North Sea, is Emiliana huxleyi, with three different morphotypes of varying calcite production rates. Even though E. huxleyi is tolerant to a wide range of environmental conditions, models predict a general decrease of that species under future climate change conditions, including increased temperatures and ocean acidification. Preliminary analysis of the Continuous Plankton Recorder time-series (1993-2018) in the North Sea suggests a significant increase in Coccolithophore abundance over time in the entire North Sea, and particularly in the northern regions. A 20-fold increase in cell number per liter was observed between the 1990s and 2010s. Our results suggest that Coccolithophores bloom every year in the North Sea with highest abundances in summer. A shift in bloom timing from June to July since the 2010s seems to have occurred in the central and southern North Sea. In addition, we plan to undertake field cruises to collect samples using “traditional” methods, which will be analyzed with electron microscopy, and to combine time-series with data from Sentinel-3 OLCI and autonomous FerryBoxes. This could help to assess long and short-term bloom dynamics, recognize possible morphotype changes, identify physical and biogeochemical drivers and observe small and large-scale distribution patterns in high resolution. Our results indicate trends, which can be used to describe the current and foreseen influence of pelagic species like Coccolithophores on the carbon cycle in a shallow shelf sea. The high spatial and temporal resolution of our observations improves the understanding of coastal system dynamics and could help predict future blooms of this important phytoplankton group.

To compensate for the shortcomings of passive water color remote sensing, which is difficult to work at night and has poor spatial coverage in polar region, this paper proposes a method to quantify total particulate organic carbon (POC) based on spaceborne lidar CALIPSO data. Firstly, the influence of the transient response of detector is eliminated, and then the crosstalk caused by the polarization splitter is corrected for the real signal of water body. Then, with the corrected signal, the total column-integrated depolarization ratio and the column-integrated below-surface depolarization ratio were calculated. After the influence of the water surface was removed by using the variance of wave slope based on wind speed, the global distribution results of subsurface backscatter are obtained. Finally, POC was quantified from subsurface backscatter. The comparison with MODIS POC shows that there is a good agreement between them. The POC product at night and in polar regions can compensate for passive water color remote sensing and study temporal and spatial changes of global primary productivity. The results provide a basis for the study of global primary productivity and carbon storage using spaceborne lidar data.

Satellite-derived inherent optical properties (IOPs; spectral absorption and scattering coefficients) are used as inputs to models that estimate marine parameters such as: primary productivity, different components of the oceanic carbon pool, and phytoplankton community composition. Thus, the quality of satellite derived IOPs is critical for deriving data products that inform us of biogeochemical processes relevant to the ocean carbon cycle. Accordingly, the continual improvement of algorithms that derive marine IOPs is critical. NASA’s Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission, scheduled to launch in early 2024, will carry the first dedicated hyperspectral ocean color radiometer, the Ocean Color Instrument (OCI), as well as two multi-angle polarimeters (SPEXOne and HARP2). For the PACE era, NASA will extend beyond legacy IOP algorithms, such as the Generalized Inherent Optical Properties algorithm framework (GIOP), and consider innovative methodologies that take advantage of PACE’s advanced capabilities. To achieve this, NASA has competitively selected the PACE Science and Application Team (SAT) to investigate state-of-the-art methodologies. The PACE Project Science Office then has a formal process of evaluating SAT innovations and considering their merit for potential adoption as post-launch algorithms. Key to this process is: (i) close liaison with PACE SAT members, (ii) the ability to build a software framework for testing multiple approaches, and (iii) synergising multiple PACE SAT innovations into a standardized/general approach. Within this presentation, we will discuss the current evolution of the NASA IOP algorithms for the PACE era and outline our strategy for delivering high quality IOPs as well their associated standard uncertainties.

Typhoon events have large impacts on marginal seas’ environmental conditions with implications for carbon cycling. However, little is known about the responses of chlorophyll-a (Chl-a) and particulate organic carbon (POC) to typhoon events in the Yellow-Bohai Sea (YBS). In this study, we utilized satellite-derived datasets of Chl-a and POC, together with key physical parameters, to analyze their responses to the typhoon events during 2003-2020. Over the past 18 years, there were 6 years experiencing one typhoon event and two typhoon events in summer of each year, with reduced wind intensity (lower than category 3) and heavy rainfall. Our result showed an increase in Chl-a in summer of typhoon years in the western Bohai Sea (BS) and western/central South Yellow Sea (SYS), especially in the upper layer of Yellow Sea Cold Water Mass, but a decrease in most parts of the North Yellow Sea (NYS). POC revealed an increase in western/central SYS (by ~50%) and a decrease in the BS and most parts of the NYS. There was a decrease of POC:Chl-a ratio in the BS and central SYS in typhoon years, indicating that elevated POC was largely resulted from enhanced phytoplankton growth. Some coastal regions in the YBS showed an increase in POC:Chl-a ratio, suggesting that there were other POC sources rather than biological contribution, such as sediment resuspension and terrestrial runoff. Our study highlights the complex impacts of typhoon events on the carbon cycle in marginal seas.

Long Island Sound - one of the largest estuaries and most important natural resources of North America - is a biogeochemical transformer of autochthonous and allochthonous dissolved organic carbon (DOC) that shapes estuarine ecological functioning and biological diversity. Here, we merged almost two decades of satellite data from Envisat/MERIS and Sentinel-3/OLCI, to examine, for the first time, spatial patterns, seasonal cycles, and decadal variability in DOC across this important and complex ecosystem from space. Five atmospheric correction (AC) approaches (BAC, C2RCC, Acolite, MUMM, and POLYMER) were first evaluated for OLCI using in situ radiometric data we collected across the estuary and AERONET-OC time series data collected at the Long Island Sound Coastal Observatory (LISCO).  We found that POLYMER was the optimal AC method, with mean APD of 11.86%, RMSE of 0.00061 sr-1 and bias of -5.63%. A multiple-linear regression (MLR) algorithm we previously designed for complex estuarine systems (Cao et al., 2018) was optimized for the LIS estuary using more than four years of bio-optical data we collected in this system, and applied to generate a long-term DOC data record from MERIS (2002-2012) and OLCI (2016-2021). Higher DOC concentrations were consistently observed in Western LIS, strongly influenced by human activities, while sharp gradients and distinct DOC plumes were captured near major river mouths, including the Housatonic and Connecticut Rivers, consistent with freshwater riverine export and tidal marsh DOC outwelling. Empirical Orthogonal Function (EOF) showed that seasonality was the most important determinant influencing the temporal variability in DOC distributions, while specific environmental factors, including wind speed/direction and precipitation, were also important. This study offered the first comprehensive description of DOC dynamics across the Sound from satellite observations, and an analysis of the key factors driving biogeochemical variability at seasonal and interannual scales.

The diffuse attenuation coefficient (Kd) determines the penetration rate of solar radiation in the ocean. Due to the wavelength limitation of traditional ocean color satellites, while remote sensing agencies distributed Kd(490) as a standard, there are no Kd products in the ultraviolet bands (Kd(UV)), which limit the research of the impact of UV radiation on ecosystem. In this study, six algorithms (both empirical and semi-analytical) developed for the estimation of Kd(UV) (at the near-blue UV bands, specifically 360, 380, and 400 nm) were assessed from a dataset of 317 points collected globally. In particular, the semi-analytical algorithm used remote sensing reflectance (Rrs) in the near-blue UV bands estimated from a recently developed deep learning system (UVISRdl) as the input. For Kd(380) in a range of 0.018 – 2.34 m-1, it is found that the semi-analytical algorithm has greater accuracy, where the average absolute relative difference (MARD) is 0.19, and the coefficient of determination (R2) is 0.94. For the empirical algorithms, the MARD values are 0.23 – 0.90, with R2 as 0.70 – 0.92, for this evaluation dataset. For a VIIRS and in situ matchup data covering oceanic and coastal waters (N = 62), the MARD of Kd(380) is 0.21, R2 is 0.94 by the semi-analytical algorithm. These results indicate that deep learning system combined with semi-analytical algorithms can provide reliable Kd(UV) from traditional satellite ocean color measurements. Such an approach offers a route to fill the gap of near-blue UV data in the global ocean for historical ocean color satellites and provides a method for evaluate and understand UV related processes in the ocean.

Seasonality of primary production is a reliable indicator for monitoring the state of an ecosystem and is controlled by the biogeochemical cycles of nutrients in the ocean; however, the linkage between the primary production seasonality and nutrient supply processes is not sufficiently understood. Here, we show satellite-derived geographic distributions of primary production-related seasonality parameters in the subarctic North Pacific coupled with seasonal variations in in situ nutrient concentrations. The seasonal patterns were classified into 15 clusters. The difference in the primary productions between observed and that predicted from the nitrate consumption (March–August) in each cluster clearly distinguished the area containing HNLC waters from the areas with continuous supplies of nutrients and iron that persist even after the spring bloom. Combining the patterns with existing ship-based observational data indicates that the iron and nutrient supplied from intermediate water by ocean mixing possibly helps sustain biological production and causes east-west gradient in seasonality of primary production in the subarctic North Pacific.

Mangroves are highly productive wetland plants with notable capability for atmospheric carbon sequestration and long-term carbon storage. Therefore, their existence is essential for climate change mitigation. This study was conducted in Perancak Estuary, Bali, Indonesia that experienced substantial mangrove loss due to intensive aquaculture development in the 1990s. Six monitoring stations were selected for studying the production and the decomposition of mangrove litter, comprised of three intact forests and three restored forests with plantation age ≥ 14 years. Monthly production of mangrove litterfall was assessed over the year (February 2020 – January 2021), which was separated into three main categories: leaf, reproductive and wood parts. Leaf litter decomposition experiment was also performed to inspect the interspecific and forest stages variation in organic matter formation, which four major mangrove species were selected: Rhizophora apiculata, Bruguiera gymnorhiza, Avicennia marina and Sonneratia alba. The results showed that annual litterfall production from intact and restored mangroves in Perancak Estuary were 10.18 and 13.96 Mg ha-1 year-1, which translated to approx. 4581 and 6282 Kg C ha-1 year-1 of annual litterfall carbon stock, respectively. Although restored mangroves had significantly higher plant litterfall production than intact mangroves, no significant difference detected for leaf litter decomposition between these forest types.

In their latest report, the Intergovernmental Panel on Climate Change expressed low confidence in satellite-based estimates of trends in marine primary production, citing the insufficient length of the time series as well as the lack of independent validation methods. Independent validation of basin-scale primary production estimates is compromised since all available in situ data from photosynthesis-irradiance measurements and all remotely-sensed data on chlorophyll concentration and available light are used for the modelling of primary production. Independent, concurrent, in situ, daily, water-column primary production measurements are not sufficient in numbers or in geographic distribution, for a global validation. Moreover, indirect methods of validation, such as the comparison with bulk property estimates, are compromised by incompatibility of time scales and representation of different components of primary production (gross vs. net production or new production or export production). In this study, we address the uncertainty in satellite-based primary production estimates by assessing the errors inherent to the calculation, in which each element of the calculation is considered separately. This method closely follows the validation approach described in the Guide to the expression of Uncertainty in Measurement (GUM). We assess the error in each of the input quantities to the primary production model (biomass, photosynthetic parameters and light) and propagate the errors through the model to obtain the uncertainty in primary production. By doing this on a pixel-by-pixel basis, we can address the uncertainties in primary production at regional scales and pinpoint regions where more in situ and remote-sensing data is needed to improve the confidence in satellite-based estimates of trends in marine primary production.

Particulate Inorganic Carbon (PIC) plays an important role in the ocean carbon cycle. In this study, we evaluated several satellite-based approaches to estimate global PIC concentrations from the Ocean Colour Climate Change Initiative (OC-CCI) products, including empirical algorithms from previous studies and a newly developed Random Forest machine learning approach. A large set of in situ and satellite match-up PIC data was used to validate the existing candidate PIC algorithms and the machine learning approach. Our results showed that the Random Forest method can retrieve the PIC concentration relatively well. In addition, we also assessed the application of the empirical algorithms using a different set of input variables provided from the OC-CCI products (remote sensing reflectance, chlorophyll-a) to improve the estimation of global PIC concentrations.

So far the oceans have helped to slow the full impact of climate change and this carbon absorption is fundamentally changing their chemical composition and altering their ecosystems. Only satellites in space, orbiting at more than 400 km above the oceans, can provide the large scale observations needed to efficiently, and economically, monitor their health. This talk will discuss recent advances in how and why satellites are being used to quantify marine and global carbon budgets, and why ocean carbon data are already helping to close the global assessments. The talk will identify new opportunities, like the ability to measure atmospheric CO2 over the ocean from space, and the potential of nano satellite constellations. It will also discuss issues that have so far been largely ignored; those of the high carbon emissions resulting from our own research, and the potential impact that all space activities may be having on our global atmosphere.

The exchange of carbon between the land and ocean sinks is poorly quantified with existing in situ measurements and it is also not clear how seasonality and long term variability impact this flux of carbon. Satellite Earth observations have excellent spatial and temporal coverage and can now provide high accuracy observation-based carbonate system monitoring in large river outflows. Here we demonstrate how satellite observations can be used to create decadal timeseries of the inorganic carbonate system conditions in the Amazon and Congo River outflows. A matchup database of carbonate system variables is used to assess the accuracy and precision of empirical algorithms driven by Earth observation data for deriving total alkalinity (TA) and dissolved inorganic carbon (DIC) within these river outflows. This then allows the calculation of the complete carbonate system using PyCO2SYS. Total combined uncertainties for TA and DIC are propagated through the carbonate system calculation allowing uncertainty estimates for all carbonate system variables. Combining these carbonate data with river flow gauging data, and a computer vision approach along with published depth profile information allows the total riverine land to ocean inorganic carbon outflow and variability to be quantified for the Amazon. We identify an annual outflow of 41 to 44 ± 3.0 Tg C yr-1 with a standard deviation of 4.33 Tg C yr-1 and coefficient of variation of 0.10 to 0.11. We will discuss this result for the Amazon, its relevance to global land to ocean carbon flow, and describe how this methodology could be extended and applied to other rivers.

Particulate inorganic carbon (PIC) is one of the four major pools of carbon found in marine environments. The pelagic PIC fraction consists of calcium carbonate from phytoplankton (specifically coccolithophores) and animals (e.g. foraminifera, pteropods). PIC is a dense biomineral ubiquitous in all marine systems, playing a key role in two fundamental oceanic carbon pathways: the biological carbon pump and the alkalinity pump. Both of these pathways affect the partial pressure of CO2 in the sea, but in opposing ways. Because of the importance of PIC production to the ocean’s carbon cycle, there is a profound need to understand the factors that control PIC concentration in space and time. Remote sensing of PIC focuses primarily on the detection of coccolithophores. Coccolithophores, and their detached calcite plates called coccoliths, are optically active, and can significantly contribute to the backscattering of light in the ocean. The current NASA algorithms for estimating PIC (both the Standard Product of Balch et. al (2005) and the Developmental Product of Mitchell et. al (2017)) make the assumption that the backscattering cross-section, or mass-specific backscattering, of PIC (b*b PIC) is taken as the average value for a single species of coccolithophore, Emiliania huxleyi. While Emiliania huxleyi is globally ubiquitous, it is not the most abundant species at tropical and sub-tropical latitudes. Here, we present a new b*b PIC that varies with PIC concentration, similar to variability of the mass-specific phytoplankton absorption coefficient due to the package effect. The new b*b PIC was derived on global field measurements, representing a wide range of coccolithophore species, and thus providing a more spatially robust estimation of b*b PIC. We show the new b*b PIC reduces the error on the current PIC remote sensing algorithms.

The ocean has absorbed ~40% of man-made carbon dioxide (CO2) emissionssince the beginning of the industrial era. This so-called "oceancarbon sink", which primarily sequesters emissions in the form ofdissolved inorganic carbon (DIC), plays a key role in regulatingclimate and mitigating global warming. However, we still lack amechanistic understanding of how physical, chemical, and biologicalprocesses impact the ocean DIC reservoir in both space and time, andhence how the storage rates of emissions may change in the future.Here we use a global-ocean biogeochemistry model (ECCO-Darwin), whichingests both physical and biogeochemical observations to improve itsaccuracy, to map how ocean circulation, air-sea CO2 exchange, andmarine ecosystems have modulated the ocean DIC budget for 1995-2018.We find that in the upper ocean, circulation provides the largestsupply of DIC while biological processes are the largest sink.Year-to-year changes in the ocean carbon sink are dominated by ElNiño-Southern Oscillation (ENSO) events in the equatorial PacificOcean, which then affect DIC globally through far-reachingocean-atmosphere teleconnections. In summary, our data-constrained,global-ocean DIC budget constitutes a significant step forward towardsunderstanding climate-related changes to the ocean DIC reservoir andthe potential consequences for marine ecosystems.

The Southern Ocean is the windiest place in the world, with frequent intense storms. The winds in these storms don't just deepen the mixed layer, but push water away from Antarctica, creating enormous upwelling, as well as driving large fluxes of carbon and heat between the ocean and the atmosphere. Unfortunately, these fluxes cannot be observed directly from space; we rely on vector
wind measurements and in situ ship and float-based observations to determine them. Our space-based observing network, however, only captures the vector winds over the Southern Ocean twice per day at best. We need more frequent vector winds to ensure that the wind fields in the climate reanalyses we use are accurate. Our estimates of the Southern Ocean air-sea carbon fluxes, based on these reanalysis winds, are uncertain and about 50% of the global uncertainty in air-sea carbon exchange is associated with the Southern Ocean. We show that higher winds are consistent with increased outgassing and reduced net uptake of atmospheric carbon by the Southern Ocean. We describe our observing system design experiment to determine the best additional scatterometer to add to the wind-observing constellation to capture more of the high winds and reduce the uncertainty in the global carbon budget.

Particulate Organic Carbon (POC) is in high dynamic with the highest turnover rate of any organic carbon pool on the planet, playing a key role in the biological carbon pump in shelf seas. Compared with the open ocean waters, the turbid coastal oceans in China are affected by large world-class rivers’ inputs, complex hydrodynamic and biogeochemical processes along the coast, resulting in optical complex particle types including inorganic mineral particles, algal particles, and detritus, which contribute to POC differently. Therefore, it is essential to develop optical identification algorithms for complex particle types and compositions, and subsequently approach POC. Here an extensive optical and biogeochemical dataset was collected from 6 cruises in turbid coastal oceans of China from different seasons in 2014, 2015 from Wei et al. (2019) and 2 cruises in 2020. Particle types (inorganic mineral particles, detritus and phytoplankton) were firstly distinguished according to their inherent optical properties. A semi-analytical algorithm for particle types and their proportions was proposed to estimate the organic particle mass fraction f-OSM (R2 = 0.864, RMSE = 6.793), using particle backscattering coefficients (bbp) and chlorophyll-a (Chla) instead of backscattering ratios, providing an effective inversion of the particle composition in turbid coastal oceans. Optical proxies such as bbp, aph, were subsequently selected for the different particle types. A POC retrieval model was developed based on the various fractional contributions of particle types from their proxies. The retrieval model was subsequently applied to Ocean Colour Climate Change Initiative dataset from 1998 to 2017, analysing the POC spatial-temporal pattern in the East China Sea, and furthermore, to investigate the potential sources of POC.

The Sentinel-3 OLCI Neural Network Swarm (ONNS) algorithm is designed to provide seamless information on biogeo-optical and -chemical properties of all natural water bodies [Hieronymi et al., 2017]. In order to capture the large optical diversity from inland waters to oceans, and therefore being able to observe land-sea exchange processes, ONNS utilizes an optical water type (OWT) classification scheme and applies specialized neural network algorithms for these classes. The latest developments include an atmospheric correction that is explicitly valid for all water types. Primal products of ONNS are several inherent optical properties (IOPs) of water constituents, but also parameters that describe the underwater light availability (Kd490) and concentrations of total chlorophyll-a and inorganic suspended matter (Chl & ISM). First comparisons with other ocean colour algorithms, for example regarding the absorption coefficient of coloured dissolved organic matter (acdom440), show advantages of the ONNS concept, e.g. through a proper separation of absorption fractions from CDOM and suspended particulate matter [Juhls et al., 2019]. Several additional water quality parameters can be estimated based on the retrieved IOPs, e.g. dissolved and particulate organic carbon concentrations (DOC & POC). Other novel products like the whitecap fraction of the sea surface in conjunction with phytoplankton biomass in water are useful for the study of air-sea exchange processes. An uncertainty estimate is provided for most of the ocean colour products.

Particulate Organic Carbon (POC) pool plays a vital role in the ocean carbon cycle. To understand the dynamics of the POC pool in the ocean, it is essential to capture consistent long-term time series data with adequate spatial resolution suitable for climate studies. The Ocean Colour Climate Change Initiative (OC-CCI) products (chlorophyll-a concentration, remote-sensing reflectance, and inherent optical properties) provide over two decades of consistent, error characterised, multi-sensor merged satellite data. In this study, we evaluated eight candidate POC algorithms applied to the OC-CCI version 5 data (1997-2020). The tested algorithms included those that had shown relatively good performance in earlier inter-comparison studies, as well as new algorithms that have emerged since then. All candidate algorithms were carefully validated using statistical metrics suggested by the original developers and the largest collection of in situ and satellite match-up POC data that we have been able to assemble. We also analysed the relationship between the POC and chlorophyll-a concentration to further assess the performance of the POC algorithms. The total average standing pools of POC were then estimated using the monthly OC-CCI version 5 data and then compared with the estimations from previous studies.

Understanding and monitoring the processes associated with the carbon cycle is of relevant importance, due to its implications on the climate system and the ocean, being the second largest carbon reservoir on planet Earth. Particulate Organic Carbon (POC) is a key biogeochemical parameter, as it helps us to quantify the primary production and export processes that occur in the ocean. Using different sensors, we can estimate POC, through the measurement of other bio-optical parameters such as the Particulate Backscattering Coefficient or Bbp. These sensors can be on board satellites or in situ, such as the buoys developed by the Biogeochemical Argo (BGC-Argo) program. The fusion of data from these two types of sensors has already been studied in previous works, demonstrating the possibilities of inferring the vertical distribution of the bio-optical properties of a water body up to 1000 meters depth. This study compares the results of the Bbp estimation in several depth layers using both in situ and OLCI data; and using only the data provided by the satellite sensor. That means that a higher spectral resolution than that used in previous works has been tested for the Remote sensing reflectance (Rrs), together with the inclusion of the Inherent Optical Properties (IOP), all derived from the C2RCC algorithm. The machine learning models applied show good prediction results, varying the contribution of the satellite products according to the depth of the estimated profile and the oceanic region observed (North Atlantic and Subtropical Gyres).

This study determined the spatial, and temporal distributions of Chlorophyll-a (Chl-a) concerning environmental factors in the Bay of Bengal (BoB) during 2003-2020. Three regions of interest (A, B and C) was selected by considering 2.5° × 2.5° grid for each in the BoB. Chl-a data had retrieved from MODIS-Aqua and the environmental factors (SST, POC, NPP, SSHA, Wind, Wind Vector, Current) data had been retrieved from satellites, whereas the in-situ Chl-a and nutrients data were retrieved from WOD, and the river discharge data were collected from BWDB for this study. The study showed that Chl-a ranged from 1.34 ± 2.23 mg/m3 to 0.12 ± 0.02 mg/m3 where as the coastal area comprised higher Chl-a irrespective of the month, season, and year. The post-monsoon period showed the highest concentrations (1.19 ± 2.12 mg/m3) of Chl-a while the pre-monsoon period showed the lowest concentrations (0.66 ± 1.12 mg/m3) of that in the coastal area. Moreover, during the northeast, and post-monsoon period Chl-a was observed as 0.16 ± 0.02 mg/m3, and 0.14 ± 0.01 mg/m3 respectively, whereas during the pre-, and southwest monsoon period the concentrations were lowest (0.13 ± 0.03 mg/m3) in the offshore area. Chl-a showed a rising, and falling trend during the study period for the study area, however, BoB showed a rising trend of Chl-a at a rate of 0.02 mg/m3 per decade. Chl-a and SST showed a significantly (p

Primary production by marine phytoplankton is one of the largest fluxes of carbon on our planet. In the past few decades, considerable progress has been made in estimating global primary production at high spatial and temporal scales by combining in situ measurements of photosynthesis-irradiance (P-I) parameters with remote-sensing observations of phytoplankton biomass. One of the major challenges in this approach lies in the assignment of the appropriate values for these model parameters that define the photosynthetic response of phytoplankton to the light field. In the present study, a global database of in situ measurements of P-I parameters and a 23-year record of climate-quality satellite observations were used to assess global primary production and its variability with seasons and locations as well as between years. In addition, the sensitivity of the computed primary production to potential changes in the photosynthetic response of phytoplankton cells under changing environmental conditions was investigated. Global annual primary production varied from 48.7 to 52.5 Gt C/yr over the period of 1998-2020. Inter-annual changes in global primary production did not follow a linear trend and regional differences in the magnitude and direction of change in primary production were observed. Trends in primary production followed directly from changes in chlorophyll-a and were related to changes in the physico-chemical conditions of the water column due to inter-annual and multi-decadal climate oscillations. Moreover, the sensitivity analysis in which P-I parameters were adjusted by ±1 standard deviation showed the importance of accurately assigning photosynthetic parameters in global and regional calculations of primary production. The light-saturation parameters of the P-I curve showed strong relationships with environmental variables such as temperature and had a practically one-to-one relationship with the magnitude of change in primary production. In the future, such empirical relationships could potentially be used for a more dynamic assignment of photosynthetic rates in the estimation of global primary production. Relationships between the initial slope of the P-I curve and environmental co-variables were more elusive.

The Arctic Ocean is considered to be a strong sink for atmospheric carbon dioxide (CO2). Air-sea CO2 flux is generally estimated by the bulk method using a parameterised gas transfer velocity and upper ocean CO2 fugacity (fCO2w) measurements. The fCO2w is often taken from a ship's seawater inlet at typically ∼5 m depth (fCO2w_bulk) by assuming that the upper ocean seawater is well-mixed. However, in the summertime Arctic, sea-ice melt results in shallow stratification (top ∼10 m), which can bias bulk CO2 flux estimates when the fCO2w measured at ~5 m depth is used. The micrometeorological eddy covariance flux technique is not affected by stratification. Here for the first time, we employ eddy covariance air-sea CO2 flux measurements during two Arctic cruises to assess the impact of sea-ice melt on Arctic Ocean CO2 uptake estimates. The sea surface CO2 fugacity (fCO2w_surface) is inferred from eddy covariance air-sea CO2 flux measurements. In sea-ice melt regions, fCO2w_surface values are consistently lower than fCO2w_bulk by an average of 39 μatm. Lower fCO2w_surface can be partially accounted for by fresher (≥27%) and colder (17%) meltwaters. A back-of-the-envelope calculation shows that neglecting the summertime sea-ice melt could lead to a ∼10% underestimate of the annual Arctic Ocean CO2 uptake.

Among the many calcifying marine organisms, coccolithophores are the major producer of particulate inorganic carbon (PIC). Calcium carbonate plates covering coccolithophores, called coccoliths, are responsible for a large increase of the water reflectance. Exploiting this high scattering property, several studies investigated the evolution and distribution of coccolithophore blooms through remote sensing techniques (e.g., Balch et al., 2018; Brown & Yoder, 1994; Cokacar et al., 2001; Holligan & Balch, 1991; Hopkins et al., 2015; Kopelevich et al., 2014; Mitchell et al., 2017; Moore et al., 2012). On the other hand, very few in situ reflectance spectra related to coccolithophore blooms are documented in the literature (Garcia et al., 2011; Iida et al., 2002; Smyth et al., 2002). The Black Sea is characterized by an almost regular yearly occurrence of coccolithophore blooms during spring. This study relies on the radiometric data of Ocean Color sites of the Aerosol Robotic Network (AERONET-OC) in the Western Black Sea to investigate the spectral features of the remote sensing reflectance RRS(λ) of marine waters during coccolithophore blooms. These autonomous multispectral measurements allow an unprecedented description of the evolution of these blooms. The radiometric products from recent Ocean Color satellite missions are also assessed in correspondence with coccolithophores blooms. Finally, implications on the use of RRS(λ) for the calculation of PIC concentration in coastal regions are discussed. The analysis shows not only elevated RRS(λ) in the blue-green spectral region in the presence of coccoliths, but it also confirms a shift toward the blue of the RRS(λ) spectra as the bloom declines and detached coccoliths accumulate at the surface. The assessment of satellite radiometric products, evaluated through AERONET-OC data, indicates that their accuracy is not significantly impacted by the extreme conditions created by the presence of coccolithophores blooms, respect to non-blooms conditions.

Internal waves are ubiquitous phenomenon in the ocean. However, their effect on primary production has seldom been studied. From the perspective of primary production, internal waves act to move the phytoplankton through a vertical light gradient and therefore influence the light intensities that phytoplankton experience, which subsequently effects productivity. Of special interest are internal waves around islands: a peculiar case of coastal trapped internal waves which occurs around closed shorelines. In such a case, wave energy is confined on a closed path, with waves having an integral number of wavelengths reinforcing themselves, with the potential of creating a resonant system. If the island geometry and stratification coincide with external forcing, pronounced internal island trapped waves (ITWs) can be excited. For the ITW with a diurnal period the key element is the interplay between the phase shift and the diurnal cycle of surface irradiance. Given that the wavelength of the internal wave equals the circumference of the island, full range of phase shifts occurs around the island. Therefore, amplification is bound to occur at some point around the island. Here we present the results of measurements undertaken around the Lastovo island in the Adriatic during the stratification period in 2021, where a set of thermistors was deployed at 9 equidistant depths between 5 and 45 m at opposite sides of the island, on rocky cliffs facing north and south, and at the control station at the island of Korčula. Bottom-moored ADCP was also deployed at the south of the island with current measurements available every 4 m between 11 and 83 m. Temperature oscillations at opposite sides of the island were out of phase, indicating the clockwise propagation of the internal waves around this northern hemisphere island. Several episodes with large diurnal thermocline oscillations, with range surpassing 20 m, were observed. Ship-based survey was also organized in July 2021, at the southern side of the island. CTD measurements were performed twice a day, around 06:00 (UTC+2) when deep thermocline was expected at the location, and around 18:00 (UTC+2) when shallow thermocline was expected. Deep chlorophyll maxima in the 40–60 m depth layer were following thermocline oscillations. CTD temperature measurements showed that the ITW range was decaying away from the coast. Complementary to the CTD measurements, discreet water nutrient and plankton (bacterio-, phyto- and zooplankton) samples were taken for further analysis, to test the hypothesis of ITWs as a main driver of primary production in the Lastovo island archipelago.

Oceanic particulate organic carbon (POC) is a relatively small (~4 Pg C) but dynamic component of the global carbon. Biogeochemical models historically focused on reproducing the sinking flux of POC driven by large fast-sinking particles (LPOC). However, suspended and slow-sinking particles (SPOC) dominate the total POC (TPOC) stock, support a large fraction of microbial respiration, and can make sizable contributions to vertical fluxes. Recent developments in the parameterization of POC reactivity in the PISCES model (PISCESv2_RC) have greatly improved its ability to capture POC dynamics. Here we evaluated this model by matching 3D and 1D simulations with BGC-Argo and satellite observations in globally representative ocean biomes. These comparisons rely on (1) a refined scheme for converting particulate backscattering at 700 nm (bbp700) to POC; (2) a novel approach for matching annual time series of BGC-Argo vertical profiles to PISCES 1D simulations; and (3) a critical evaluation of the correspondence between model tracers, SPOC and LPOC estimated from bbp700 profiles, and directly measured POC fractions. We show that PISCES captures the major features of SPOC and LPOC as seen by BGC-Argo floats across a range of spatiotemporal scales, with overall better model-observations agreement in the epipelagic (0–200 m) than in the mesopelagic (200–1000 m) layer. We also identified characteristic patterns of model-observations misfits in subpolar and subtropical gyres, which point to the need to better resolve the interplay between POC sinking, remineralization and fragmentation in PISCES. Beyond model evaluation results, our analysis indicates that a widely used satellite algorithm overestimates POC at high latitudes during the winter. Our approach can help constrain POC stocks, and ultimately budgets, in the epipelagic and mesopelagic ocean.

La variable spatiale et saisonnière de la profondeur de la couche de mélange (MLD) a été étudiée à l'aide des données hydrologiques de plusieurs bases de données sur la période d'octobre 1973 à mars 2017 à 10 °W entre les latitudes 2 ° N et 10 °S dans le golfe de Guinée. La méthode du seuil de densité avec le critère de 0,03 kg m -3 aété utilisée pour calculer la MLD. Dans la bande équatoriale, la MLD moyenne saisonnière est de 20 m quelle que soit la saison. A 6 °S et 10 °S, la MLD est relativement plus élevée pendant la saison froide. Le MLD varie entre 21 et 37 à 6 °S, et les moyennes saisonnières MLD sont respectivement de 21 et 40 m pendant les saisons chaudes et froides. A 10 °S, pendant la saison chaude, la MLD varie entre 28 et 52 m, et la moyenne saisonnière est de 39,5 m. Pendant la saison froide, la MLD varie entre 45 et 55 m avec une moyenne saisonnière de 49 m. Mots clés : Océan Atlantique tropical , Golfe de Guinée, Couche mixte, Variabilité spatiale, Variabilité saisonnière

Culture studies have repeatedly demonstrated that the parameters describing the photosynthetic response of marine phytoplankton can vary widely under different growth conditions (light, nutrients and temperature) and between species. Yet several remote sensing estimates of marine primary production either assign a single set of parameters for a given region/season or use global empirical relationships (e.g. the maximum photosynthetic rate as a function of sea-surface temperature). Our inability to develop a more mechanistic approach to parameter assignment is due to both an uneven distribution of experimental observations in both space and time and a lack of information on phytoplankton community structure and/or environmental conditions at the time the experiments were made. One of the aims of the ESA BICEP project is to expand existing global datasets of the photosynthesis-irradiance (PE) parameters.  This data mining effort has dramatically improved both the spatial and the temporal coverage of these parameters that are critical to convert maps of surface chlorophyll to estimates of water-column primary production.  We have used the > 10,000 experimental measurements and metadata assembled as part of the BICEP project to explore how changes in environmental forcing and the taxonomic structure of phytoplankton communities are related to variability in the PE parameters.  Here we focus on ‘regions of interest’ that cover the four ocean biomes defined by Longhurst.  These ocean biomes (Coastal, Polar, Trades and Westerlies) represent the primary unit of biogeographic division of the global ocean and provide a useful way of examining variability caused by large-scale changes in environmental forcing.  Our dataset reveals biome-specific differences in the relationship between taxonomic composition and phytoplankton photophysiology. By combining flow cytometric counts and HPLC pigment data in the Trades Biome, we show how variation in photoacclimatory status (intracellular pigment concentration and relative concentration of photoprotective pigments) is strongly related to photosynthetic performance.  The patterns of variability observed in this study can be used to improve our assignment of PE parameters for satellite-based studies of ocean primary production.

Oceans have been an integral part of human society. In the recent past, oceans are gaining increasing attention as an ecosystem that helps community well-being in eco-social aspects. Contribution of ocean to act as a biologically driven carbon flux and a store that sequester more than 80 % of global carbon is not well-documented in many countries. The potential to provide nature-based solutions by these massive ecosystem has been underestimated and overlooked. The blue carbon resources in the coast provide effective measures to combat climate change which shed light to manage ocean for a sustainable future. This study focuses on estimating blue carbon in the sea grass meadows in the southern coastal belt of Sri Lanka during the last decade. Sri Lanka, being a tropical island is a home to more than ten different species of sea grasses. In this investigation, we provide an estimation of extent of sea grasses in the coast including the lagoons using GIS and RS. This study highlights as to how, Sri Lanka can adopt Nationally Determined Contribution (NDC) mitigation actions as per the Paris Agreement, by ensuring ecological health of sea grasses. There is a dire need for an attitudinal change of coastal communities and the policy makers to recognize the ecosystem services provided by sea grass beds as many coastal ecosystems are under threat of destruction due to anthropogenic pressure. In this context, this study also provides a framework for community based sea grass conservation programme.

Phytoplankton in the sunlit layer of the ocean act as the base of the marine food web fueling fisheries, and also regulate key biogeochemical processes. Phytoplankton composition structure varies in ocean biomes and different phytoplankton groups drive differently the marine ecosystem and biogeochemical processes. Because of this, variations in phytoplankton composition influence the entire ocean environment, specifically the ocean energy transfer and the export of organic carbon to the deep ocean. As one of the algorithms deriving phytoplankton composition from space borne data, within the framework of the EU Copernicus Marine Service (CMEMS), EOF-PFT algorithm was developed using multi-spectral satellite data collocated to an extensive in-situ PFT data set based on HPLC pigments and sea surface temperature data (Xi et al. 2020, 2021; https://marine.copernicus.eu/). By using multi-sensor merged products and Sentinel-3 OLCI data, the algorithm provides global chlorophyll a data with per-pixel uncertainty for diatoms, haptophytes, dinoflagellates, chlorophytes and prokaryotic phytoplankton spanning the period from 2002 until today. Due to different lifespans and radiometric characteristics of the ocean color sensors, the consistency of the PFTs is evaluated to provide quality-assured data for a consistent long-term monitoring of the phytoplankton community structure. As current commonly used phytoplankton carbon estimation methods rely mostly on the backscattering property of phytoplankton, which could vary dramatically for different phytoplankton taxa, as a perspective of this study, phytoplankton carbon may be better estimated in a way that accounts for phytoplankton taxonomy.

Increased nutrients concentration in freshwaters are responsible for the increased frequency of harmful cyanobacterial blooms. Toxics blooms require rigorous monitoring to minimise any risks to organisms in natural ecosystems and human population. Primary production estimates from satellites and models are likely to exhibit large uncertainties in the presence of communities consisting of both algae and cyanobacteria. In situ instrumentation is needed to generate appropriate ground truth to develop and validate these methods. New fluorescence methods, that deploy a combination of LEDs, enable reproducible, rapid, and accurate measurements that can discriminate different types of phytoplankton. This is possible by tuning the fluorometer to detect the light-harvesting pigments in the photosynthetic apparatus which differ between phytoplanktonic types and by excitation wavelengths. Nevertheless, fluorometers able to record proxies of primary productivity need to be designed either for consistent sensitivity to all phytoplankton groups or capture their distinct responses. The new LabSTAF (Chelsea Technologies Group) fluorometer using inductions protocols was tested from April to October 2020, on a weekly basis on samples taken from Roadford lake (Southwest of England) where cyanobacteria occur annually. Different protocols targeting cyanobacteria and algae were applied to estimate their biomass and assess their physiology. These data, in combination with other measurements (absorbance, chlorophyll a concentration, turbidity, phytoplanktonic identification and counting and nutrients concentration) were used, in conjunction with protocol features of the instrument, to assess the potential of this fluorometer to discriminate cyanobacteria in natural samples. The LabSTAF is able to discriminate photosynthetic trends of both algal and cyanobacterial composition, over time, in the lake. Nevertheless, in the protocol targeting cyanobacteria, the LEDs were not able to fully saturate the sample which leads to an underestimation of cyanobacterial biomass. Moreover, this protocol also tended to include a signal from the algae which caused an overestimation of the cyanobacterial biomass. Brighter LEDs should be included in the LabSTAF to be able to fully saturate cyanobacteria. Moreover, an emission filter centred at 660 nm should also be considered to principally target cyanobacteria response and ameliorate the cyanobacterial biomass estimation.

Nearshore coastal waters often contain the highest levels of biodiversity and phytoplankton biomass. Yet, owing to difficulties in sampling in this dynamic region, from satellite and in situ, gaps remain in our understanding of the seasonality of phytoplankton in the nearshore, and their impact on pools and fluxes of carbon. Here, we analyse a unique annual dataset of chlorophyll-a concentration, a measure of phytoplankton biomass, and sea surface temperature (SST), collected by a surfer at a beach in Plymouth, UK on a near weekly basis between 2017 and 2018. By comparing this dataset with a separate in situ dataset collected 7 km offshore from the coastline (11 km from Bovisand Beach), and guided by satellite observations of light availability, we investigated differences in phytoplankton seasonal cycles between nearshore and offshore coastal waters. We observed significant differences in phytoplankton biomass between sites during the summer months of July and August. Offshore chlorophyll-a concentrations crashed during this period whereas chlorophyll-a concentrations in the nearshore remained high. Statistical comparisons between chlorophyll-a concentration and physical (SST and light) and chemical variables (nutrients) suggest that during the summer, the offshore site becomes nutrient limited at the surface, in contrast to the nearshore, where chlorophyll-a concentrations remain high. Our findings have implications for understanding how the spatial distribution of phytoplankton phenology within the coastal seas studied may be impacted by climate change. Additionally, the study emphasises the potential for using marine recreation as a platform for acquiring environmental data in otherwise challenging regions of the ocean to sample, for understanding phytoplankton carbon pools and fluxes, and for satellite validation of high resolution ocean colour products.

The ESA Earth Explorer Wind Mission ADM-Aeolus (Atmospheric Dynamics Mission, ESA, 2008), successfully launched on 22 August 2018, has the aim to provide global observations of wind profiles, demonstrating the impact of wind profile data on operational weather forecasting and on climate research. Within the Aeolus+ Innovation program, ESA has launched an Invitation To Tender (ITT, ESA AO/1-9544/20/I/NS) to carry out studies aimed at exploring, developing and validating innovative products and applications and exploiting the novel nature of Aeolus data. Although Aeolus’s mission primary objectives and subsequent instrumental and sampling characteristics are not ideal for monitoring ocean sub-surface properties, the unprecedented type of measurements from this mission are expected to contain important and original information in terms of optical properties of the sensed ocean volume. Being the first HSRL (High Spectral Resolution Lidar) launched in space, ALADIN (Atmospheric LAser Doppler Instrument) of ADM-Aeolus gives an unprecedented new opportunity to investigate the information content of the 355 nm signal backscattered by the ocean sub-surface components. Based on the above considerations, COLOR (CDOM-proxy retrieval from aeOLus ObseRvations), a selected Aeolus+ Innovation ITT project, aims to evaluate and document the feasibility of deriving an in-water AEOLUS prototype product from the analysis of the ocean sub-surface backscattered component of the 355 nm signal acquired by the ALADIN. An overview of the project and some preliminary results are presented.

Satellite remote sensing is currently the only observational tool that allows for global synoptic view of oceanic pools of carbon. Thanks to the large portfolio of available remote sensing algorithms, oceanic pools of particulate organic, inorganic, and phytoplankton carbon can be estimated. These crucial parameters then feed into subsequent, modeled estimates of the current and future oceanic and global carbon cycle. Despite their widespread use, only the remote sensing community remains aware of the shortcomings of these global estimates. Uncertainties are large and often unreported, with most originating from assumptions embedded into empirical algorithmic approaches, as well as a lack of high-quality validation data. In preparation for the upcoming NASA Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission, a suite of community supported protocols, data assemblage, match-ups and validation activities, and uncertainties estimates have been set in place, with the goal of improving estimates of global carbon pools (and the resulting fluxes). This groundwork will support legacy algorithms and it would hopefully inspire the development of new approaches that use both the hyperspectral (the Ocean Color Instrument) and multi-angle polarization (SPEXOne and HARP2) capabilities and potential of the PACE mission.

With the advent of in situ autonomous Argo profilers equipped with radiometric sensors (i.e., Biogeochemical-Argo), the knowledge on the spatial and temporal variability of the underwater light field has dramatically increased even in remote regions and during unfavourable conditions for ship-based sampling. Since 2012, new autonomous observations at four bands (380, 412/443, and 490 nm plus PAR – the photosynthetically available radiation) have allowed the characterization of the bio-optical behaviour of most open oceans (from ultra-oligotrophic gyres to the most productive regions) and marginal seas. The almost global distribution of 0–250 m radiometric profiles has provided an unprecedented global picture of the relationships between coloured dissolved organic carbon and optical properties of phytoplankton communities and how these vary with time. These profiles also gave inputs to unravel processes such as phytoplankton phenology, carbon production and export, both in synergy with biogeochemical models and space-based observations. The synergy between in situ BGC-Argo radiometry and Earth Observation has also been beneficial to validate Ocean Colour satellite products, by largely increasing the number of possible match-ups from a platform with no shadow. Here, we review recent advances on the understanding of the spatial and temporal variability of the underwater light field and related ocean carbon products, as obtained by BGC-Argo autonomous observations. In this context, we also identify the current observational and scientific limitations of BGC-Argo. Finally, we provide perspectives for future development that will enhance the utility of the BGC-Argo fleet. In particular, we focus on the recent deployments of BGC-Argo floats equipped with hyperspectral radiometers. These floats, by acquiring 140 bands in the ultraviolet and visible light, have recorded the underwater light field in the upper 400 m of the water column during bloom to post-bloom conditions and the formation of the deep chlorophyll maxima (DCM). The observed spectral variability of the underwater light captured in two different environments during high-chlorophyll DCM events (Mediterranean and Baltic Seas) have highlighted how such detailed spectral information allows monitoring phytoplankton diversity from the surface to the ocean interior. Developing synergies with current and future satellite missions (multi- and hyperspectral) will further improve such knowledge in space and time.

Accurate estimates of the oceanic particulate organic carbon concentration (POC) from optical measurements have remained challenging because interactions between light and natural assemblages of marine particles are complex, depending on particle concentration as well as properties such as particle composition and size distribution. In particular, the applicability of a single relationship between POC and particulate backscattering coefficient bbp across diverse oceanic environments is subject to high uncertainties because of the variable nature of particulate assemblages. Nevertheless, these relationships have been widely used for estimating oceanic POC, for example using in situ measurements of bbp from Biogeochemical (BGC) Argo floats. Despite these challenges, such an in situ-based approach to estimate POC remains scientifically attractive, especially in view of the expanding global-scale observations with the BGC-Argo array of profiling floats equipped with optical sensors such as the backscattering and chlorophyll-a fluorescence sensors. In the current study, we describe an improved approach to estimate POC which takes advantage of simultaneous measurements of bbp and chlorophyll-a fluorescence to better account for the effects of variable particulate composition on the relationship between POC and bbp. We formulated a multivariable regression model using a global dataset of field measurements of POC, bbp, and chlorophyll-a concentration (Chla), including surface and subsurface water samples from the Atlantic, Pacific, Arctic, and Southern Oceans. The analysis of a dataset of diverse seawater samples demonstrated that the use of bbp and an additional independent variable related to particle composition involving both bbp and Chla leads to notable improvements in POC estimations compared with a univariate model based on bbp alone. We expect this multivariable model to be particularly useful for estimating POC with measurements from autonomous BGC-Argo floats operating in various oceanic environments.

The particle size distribution (PSD) is an important parameter for ocean optics and biogeochemistry. The KSM09 ocean color algorithm (Kostadinov et al., 2009) uses Mie modeling of a single homogeneous spheres particle population to retrieve the PSD from the spectral shape and magnitude of the particulate backscattering coefficient (bbp). Here, we present a public, long-term (1997-2020) merged global ocean color data set based on a significantly updated PSD algorithm and using the merged, multi-platform 4 km monthly OC-CCI v5.0 (Sathyendranath et al., 2019) remote-sensing reflectance data and the Loisel and Stramski (2000) bbp algorithm. The new PSD algorithm uses a 2-component model in which phytoplankton cells and non-algal particles (NAP) are modeled as two distinct particle populations. Phytoplankton are modeled as coated spheres, based on the Equivalent Algal Populations (EAP) framework (e.g. Robertson-Lain and Bernard, 2018), and NAP are modeled as homogeneous spheres. Key variables of the data set are the power-law PSD parameters, and absolute and fractional size-partitioned phytoplankton carbon, estimated using existing allometric coefficients (Roy et al., 2017). Partial product uncertainties are estimated using a Monte Carlo approach and analytical error propagation. Validation effort results and the need for an empirical tuning of the PSD magnitude scaling parameter are presented. A version of this abstract/content is also planned to be presented at the upcoming Ocean Sciences Meetng 2022.

We formulated empirical POC (particulate organic carbon) algorithms for several satellite sensors to support generation of a global multi-decadal data record from multiple satellite ocean color missions. For the algorithm development we assembled a dataset of surface POC and spectral remote-sensing reflectance, Rrs, measurements collected in all major ocean basins encompassing tropical, subtropical, temperate as well as the northern and southern polar latitudes. Criteria for the inclusion or exclusion of data were based on documented consistency of measurement protocols as well as bio-optical characteristics of seawater which are consistent with open-ocean pelagic environments. We evaluated over 70 mathematical formulations that represent 7 distinctly different algorithm categories in terms of the definition of the independent variable involving Rrs. This analysis resulted in formulation of the potentially next generation version of global POC algorithms for SeaWiFS, MODIS, VIIRS, MERIS, and OLCI sensors. These algorithms are referred to as hybrid algorithms and combine the MBR (Maximum Band Ratio)-OC4 and BRDI (Band Ratio Difference Index) functions. Compared with the predecessor algorithms, the MBR-OC4 is intended to improve POC estimates mainly at high POC >200 mg m-3 and the BRDI at very low values

Ecosystem fragility is an often-used term in oceanography yet to this day it lacks a precise and widely accepted definition. Defining and subsequently quantifying fragility would be of great value, for such measures could be used to objectively ascertain the level of risk marine ecosystems face. Risk assessments could further be used to define the level of protection a given ocean region requires from economic activity, such as fisheries. With this aim the concepts of marginal production and fragility are defined for marine photosynthesis, the base of the oceanic food web. It is demonstrated that marine photosynthesis is always fragile with respect to light, implying variability in surface irradiance acts unfavourably on biomass, whereas it can be both fragile and antifragile with respect to mixed-layer depth, implying variability in mixed-layer depth can act both favourably and unfavourably on biomass. Quantification of marginal production and fragility is presented on data from two open ocean stations: Hawaii Ocean Time Series and Bermuda Atlantic Time Series. The calculations are then placed in a dynamical context using a model with stochastic forcing. Seasonal cycle of biomass is modelled and the effect of asymmetries in the response of primary production to mixed-layer depth variability is analysed. A new tipping point for marine phytoplankton is discussed. Using the new definitions presented here a rich archive of data, thus far untapped in this context, can be used to quantify primary production fragility. The presented definitions can also be used to predict when primary production enters the fragile state during the seasonal cycle.

Remote estimation of phytoplankton primary production has long been recognized as an important method for investigating the feedback of aquatic ecosystems on global climate change. One of the earlier proposed models, the Theory-based primary Production Model (TPM) has the potential to be applicable to a variety of water bodies due to its semi-analytical scheme. Its accuracy is greatly dependent on whether the photophysiological response of phytoplankton is adequately parameterized by a suite of photosynthetic parameters, two of which are the assimilation number (PBmax) and light saturation parameter (Ek). Yet, the remote assignment of PBmax and Ek is acknowledged as a challenging task and has been made limited progress, hampering the utilization of the TPM. In this study, we developed a machine-learning algorithm, the Enhanced Random Forest Regression (ERFR), to retrieve PBmax and Ek from satellite observations, which was then incorporated with the TPM (together termed as TPMERFR) to estimate daily, depth-integrated primary production (IPP). The ERFR was trained and validated using extensive in situ datasets from worldwide representative water areas. Our evaluations with independent in situ data and matchup data show that the ERFR outperformed conventional empirical and semi-analytical algorithms in both clear and turbid waters, and could better capture the variability of and Ek than look-up-table methods. Consequently, satellite-based IPP estimates from the TPMERFR were more accurate as indicated by the lower root mean square differences (RMSDs) of 0.27 and 0.24 for clear and turbid waters, respectively. By contrast, the selected benchmark models generally yielded IPP estimates with RMSDs of 0.27–0.53 and 0.29–0.62 for clear and turbid waters, respectively. Furthermore, the TPMERFR was implemented to climatological satellite products (2010–2019) to reassess global IPP, where reasonable distributions were preliminarily demonstrated, especially in coastal and inland water areas.

Phytoplankton size structure is an emergent property of marine ecosystems that determines the magnitude and fate of net primary production in the open ocean. Remote sensing reflectance measurements taken by ocean color satellites allow the inversion of the particle backscattering spectra and the estimation of the particle size-abundance distribution. Taking advantage of these advances, we developed a model based on the theoretical framework of the metabolic theory of ecology to estimate marine net primary productivity from space.The model scales basic biochemical reactions and physiological processes from individual cells to the entire phytoplankton community, integrating the nonlinear scaling of cell growth with size and the impact of temperature and nutrient limitation on metabolic rates and on the pace of ecological processes. The model is also vertically resolved and accounts for the impact of light limitation and photoacclimation. The inclusion of the metabolic scaling of phytoplankton growth leads to deviations of up to 30% in productivity estimates for a given standing biomass. Combined with the effect of temperature and nutrient limitation, the model predicts consistent large scale productivity gradients and a global net primary production of 60 Pg of C per year. However, the size-structured model reveals a larger contribution of highly diverse, small-sized phytoplankton classes with respect to previous assessments, and suggest a greater importance of oligotrophic regions to carbon cycling in the ocean. Our results highlight how the metabolic theory of ecology provides a framework to bridge the growing capabilities of ocean satellites to characterize phytoplankton trait diversity into remote sensing estimates of marine primary production.

The ocean primary production is a key element of the carbon pool, making up more than 50% of its global value. In order to quantify on global scale primary production, models are used which mostly rely on satellite input on the phytoplankton biomass, the light in the ocean and the sea surface temperature, besides considering the characterization of the response of the photosynthesis rate to the light intensity obtained from measurements with data sampled in the field. Past studies on global primary production modelling have demonstrated that significant improvements in its quantification can be achieved when the composition of the phytoplankton and the spectral characterization of the underwater light field, respectively, is taken into account. Reliable time series data globally distributed for those two components are still lacking. We present recently developed global and coastal satellite retrievals from multispectral ocean color (including OLCI on Sentinel-3), and hyperspectral atmospheric TROPOMI (Seninel-5P), and terrestrial DESIS (on ISS) data on the phytoplankton composition and spectral radiation attenuation. They offer by themselves only an incomplete description of the composition of phytoplankton or the spectral light and limited spatial and temporal resolution. Therefore, the combination of the different retrievals from these different sensors will enable time series data set applicable for improving for global marine primary production estimates. Strategies for their combination and testing the benefits also for coupled ocean ecosystem modelling will be discussed as well. Furthermore, the spatial and short-term temporal distinction between phytoplankton groups in highly productive areas such as upwelling areas would be valuable to a better understanding of the dynamics of phytoplankton communities in these areas and its rule in shaping trophic chains.

Assessments of the global ocean carbon sink require spatially and temporally complete fields of pCO2 (sw) as one of the key datasets. These fields are produced by extrapolating in situ pCO2 (sw) measurements with satellite observations of parameters that account for the variability of, and control on, pCO2 (sw). Both biological and physical ocean processes control pCO2 (sw) but currently, chlorophyll-a (Chl a) is often used as a proxy for the biological control. However, the full biological contribution to pCO2 (sw) is determined by the balance between photosynthesis (primary production; PP) and respiration, known as the net community production (NCP) and NCP can now be quantified from satellite observations. In this study, we evaluate the impact of including either satellite derived NCP, PP, Chl a or no biological parameters within a neural network scheme to produce complete fields of pCO2 (sw) for the South Atlantic Ocean. Estimates of pCO2 (sw) using NCP, PP or Chl a showed similar broad scale accuracy, although including NCP produced more accurate pCO2 (sw) fields in regions with strong biological signals, such as the Benguela and Equatorial upwellings and the Amazon Plume. These improvements to complete fields of pCO2 (sw) converted the South Atlantic Ocean from a net CO2 sink (as determined using no biological parameters), to a CO2 source (when using NCP). The future improvements in pCO2 (sw) estimates that could be achieved by reducing the uncertainties in satellite NCP, PP or Chl a were assessed. We show that using NCP would enable the greatest achievable reduction in pCO2 (sw) uncertainties, a situation that was confirmed by using in situ observations. Using satellite derived NCP within pCO2 (sw) extrapolation schemes and reducing satellite NCP uncertainties will ultimately improve our understanding and confidence in quantifying the global ocean as a CO2 sink.

The vertical distribution of primary production (PP) varies in time and space and cannot be readily estimated from space. We show here (using own and archived data) a universal relationship between the depth of the deep chlorophyll maximum (DCM) and nutricline depth (DNO3, nitrate = 1 µmol kg−1), which can be explained with a simple model including light and nutrient limitation. Total PP is shown to be a function of DNO3. In addition, the fraction of total water column PP occurring in the upper 10 m shows a significant dependence on nutricline depth (varying from about 11 % in oligotrophic regions to 80% or more (average 31%) in regions where DNO3 is located at < 20 m). This PP-DNO3 relationship was applied in a global PP-model based on calculated PP in the surface layer from satellite data (assuming that satellite observations are a good proxy for conditions in the upper 10 m) and climatological distributions of DNO3. The global PP estimates indicate that ~25% of ocean PP occurs in the upper 10 m. Comparisons of global PP estimates derived using this PP-DNO3 relationship indicate oligotrophic regions of the ocean may be more productive than usually assumed. The fit between the two approaches is better where DNO3 is closer to the surface in, e.g., the northern North Atlantic. We argue, however, that estimating PP using DNO3 depth may improve estimates of PP compared to models where the vertical distribution of PP is primarily estimated from surface chlorophyll distributions. To compare the two approaches, we applied the maximum photosynthesis rate (PBmax) as configured in previous models (VGPM). We note, however, that PBmax estimates determined in situ in the North Atlantic were lower and did not exhibit the temperature dependence assumed for PBmax. Thus, existing parameterisations may overestimate PP in this region. References Richardson, K., and Bendtsen, J. (2021). Distinct seasonal primary production patterns in the sub-polar gyre and surrounding seas. Front. Mar. Sci. 8:785685. doi: 10.3389/fmars.2021.785685 Richardson, K., and Bendtsen, J. (2019). Vertical distribution of phytoplankton and primary production in relation to nutricline depth in the open ocean, Mar. Ecol. Prog. Ser., 620, 33–46. Richardson, K., Bendtsen, J., Kragh, T., Mousing, E.A. (2016). Constraining the distribution of photosynthetic parameters in the global ocean, Frontiers in Marine Science, 3, 269, 10.3389/fmars.2016.00269.

The formation of Deep Chlorophyll Maxima (DCM) during the thermal stratification period is a common feature in oligotrophic marine ecosystems, among which the Eastern Mediterranean Sea (EMS). DCM usually represent photo-acclimation processes resulting in enhanced Chlorophyll a (Chl a) concentration, but this is not necessarily reflected in enhanced primary production (PP) rates, as is the case in the EMS, where PP maxima are usually located in the near surface layers (20-30 m). There are cases, however, when the combination of sufficient light and nutrients diffused from below the euphotic zone may allow PP maxima in deeper layers, corresponding to or overlaying DCM. In this work we investigate the occurrence of such deep PP maxima in the EMS during the onset of thermal stratification (i.e. late spring), as well as the underlying physical and biogeochemical features regulating the coupling between Chl a and PP vertical distributions. Our data show that during late spring in the EMS, the bulk of PP may occur in the deeper layers of the euphotic zone. In the reported case studies, PP maxima were located between 50 and 75 m and DCM between 50 and 100 m. However, in most cases the assimilation number (Chl a normalized PP) did not follow PP maxima and was consistently higher within the near surface layers. Overall, our data show that the contribution of DCM layer to total water-column integrated primary production may be substantial for these oligotrophic waters, ranging from 22 to 85%, while this contribution is negatively correlated with the DCM depth. Finally, the in-situ water-column integrated PP data are compared to current satellite-based model estimates of PP and the implications are discussed considering the importance of the DCM in the EMS.

For the past three decades, polar-orbiting ocean color satellites have afforded the means to obtain local, regional to global scale estimates of oceanic net primary production that have greatly aided studies on ocean carbon cycling, food web dynamics and climate change. Despite considerable progress, the ability of ocean color satellites to provide accurate estimates of daily ocean productivity has not been realized because polar-orbiting satellites are unable to account for variations in carbon fixation rates caused by changing incident irradiance levels and changes in phytoplankton physiology over the course of the day. Here for the first time, we have attempted to exploit the unique short-temporal measurements provided by the Korean Geostationary Ocean Color Imager (GOCI), to obtain hourly and daily measurements of surface- and euphotic-column integrated Net Primary Productivity (NPP). These estimates are based on the Absorption Based Productivity Model (AbPM), which is supplemented by a bio-optical database comprised of measurements made at different times of the day during the Korea-US Ocean Color (KORUS-OC) cruise in May-June 2016. Photo-physiological rate estimates varied across different water types encountered around the Korean Peninsula. Because of their limited number, these estimates were regionally scaled through the use of dynamic optical-biogeochemical (O-BGC) provinces. Daily and weekly integrated NPP satellite estimates of NPP derived from GOCI data, when compared against in-situ measurements, clearly underscore the superiority of geostationary over polar-orbiting ocean color satellites, which stems from the ability of geostationary satellites to account for short term changes in phytoplankton light absorption and incident irradiance fields that fluctuate greatly over the course of the day.

We have re-analyzed primary production and optical data from the Joint Global Ocean Flux Study, the North Atlantic Bloom Experiment (1989), specifically the RV Atlantis-II cruise from 24 April-7 May. The daily observations and experiments occur on a water column of decreasing mixed layer depth, and increasing phytoplankton biomass (as indicated by chlorophyll-a concentrations). The data allow us to examine several parameters that describe, or contribute to, the depth of the productive layer (euphotic zone): (1) the depth of 1% of surface (PAR) irradiance, (2) the depth of 1% of Ed(488,0-), (3) the compensation depth (where gross primary production is balanced by autotrophic respiration), (4) the depth of the base of the fluorescence maximum, (5) the mixed layer depth, and (6) the depth where integral gross primary production balances integral autrophic respiration (a proxy for the critical depth). We find that these parameters are in reasonable agreement, and the depth of the productive layer accords with the 1% depth horizon of surface irradiance. However, we note the difference between these results and those elsewhere. Therefore, we suggest that a biological definition of the euphotic zone (compensation depth, base of fluorescence maximum), rather than an optical definition, is more useful, especially with regard to remote sensing.

Studies of regional and global trends in the marine ecosystem require long-term data records of oceanic carbon stocks, and thus the consistency of carbon data products derived from different satellite missions is critical. Particulate organic carbon (POC) is a relatively small but dynamic reservoir of carbon in the ocean, and satellite-based ocean color observations can be used to provide information about variability in the magnitudes and seasonal phenology of surface POC concentration and to potentially examine long-term trends at regional, basin, and global scales. Using a recently proposed suite of POC algorithms formulated for different ocean color sensors (Stramski et al., 2021; doi: 10.1016/j.rse.2921.112776), we generated time-series of surface POC concentration in the global ocean from satellite missions of SeaWiFS, MODIS-Aqua, and VIIRS-SNPP which span a 24-year period. These data are used to conduct direct intercomparisons of the POC time-series provided by each mission with an emphasis on periods of mission overlap. An intercomparison analysis was performed for different geographically or biogeochemically-defined ocean regions to examine the degree of consistency in the POC data product generated by different sensors, and to identify any potential biases among mission-specific POC products. This analysis provides a basis for the creation of a consistent long-term data record of ocean surface POC concentration from multiple satellite missions, enabling more advanced analyses of long-term variations in the standing stock of POC within the upper ocean.

The relative contributions of dissolved and particulate organic matter, as well as particulate inorganic matter, to the total carbon export affect the magnitude and efficiency of the biological pump. Traditional methods for quantifying carbon export rely on snapshot, ship-based observations that often do not distinguish between dissolved and particulate organic matter, or their various export pathways, and that do not consider particulate inorganic carbon. In this study, we present a novel approach that leverages observations from one 5-sensor BGC float in the Northeast Pacific Ocean to partition carbon export potential into four distinct carbon pools: particulate inorganic carbon, particulate organic carbon, sinking particulate organic carbon, and dissolved organic carbon. Year-round observations reveal more complex carbon cycle dynamics among the carbon pools than previously anticipated. This helps to resolve the regional conundrum of a persistent particle sinking flux observed by sediment traps during a season that is known to be heterotrophic. By combining float-based net primary production (NPP) estimates with in situ particulate organic carbon (POC) sinking flux estimates, we capture two high export ratio (sinking- POC / NPP) periods with different underlying mechanisms. The proposed methods herein could be applied broadly as the biogeochemical Argo array expands, leading to greatly improved understanding of the biological carbon pump at global scales while generating a field dataset suitable for training new satellite algorithms that capture various components of biogenic carbon cycling.

The gravitational sinking of biogenic carbon particles in the ocean are an essential component of the ocean carbon cycle. Biogenic carbon particles are generated by phytoplankton in the sunlit surface ocean and comprise Particulate Organic and Inorganic Carbon (POC and PIC). Here we compare and contrast the downward transfer of carbon particles associated with blooms of calcifying and non-calcifying phytoplankton, corresponding to high PIC and high POC concentrations, respectively. We merged ocean colour satellite observations of POC and PIC in the near-surface ocean with water column observations of sinking particles obtained from BioGeoChemical-Argo floats (BGC-Argo) operating in the productive North and South Atlantic Ocean. Our results indicate that calcifying phytoplankton blooms had a pronounced impact on the downward transfer of sinking particles compared to blooms of non-calcifying phytoplankton in both regions. In the South Atlantic where floats operated at high temporal resolution, we show that particles associated with a calcifying phytoplankton bloom sank twice as fast. In the North Atlantic, where floats operated at lower temporal resolution, the transfer of particles was deeper and more efficient than for blooms of non-calcifying phytoplankton whose particle flux was strongly attenuated in the upper mesopelagic ocean.

Quantifying the parameters of the carbonate system through derivations of other variables is a key tool for the reconstruction of data in regions where it is scarce or non-existent. Satellite data and sparse in situ observation can be used to reconstruct the carbonate system variables such as dissolved inorganic carbon (DIC), whose concentration changes due to the cumulative effect of respiration, carbonate dissolution/precipitation, and the ocean-atmosphere interaction. The study area covers 22 to 50° N of the Northeast Pacific coast (California Current System, CCS). The aim was to reconstruct basin-scale sea surface DIC concentration from surface temperature and chlorophyll data from the AQUA-Modis sensor, with daily imagery of 4x4 km resolution. Algorithms proposed in the literature were used to determine DIC only in spring (eighteen-year climatology). The results showed spatial variability of DIC in the CCS, with the highest values in the southern portion off Baja California, Mexico, while the lower values were off the California coast region, USA. The use of algorithms for the reconstruction of DIC is an essential tool for understanding the Spatio-temporal variability of DIC and its controlling mechanism.

The Ocean Color Instrument on NASA’s Plankton, Aerosol, Cloud, ocean Ecosystem (PACE) mission will have increased spectral resolution and expansion into the ultraviolet spectrum, improving estimates of ocean properties. One of NASA’s PACE mission core objectives is to understand Earth’s Ocean ecosystem through improved estimates of phytoplankton concentration and community composition. Knowledge of community composition is essential for modeling carbon export from the surface ocean and ecological responses to climate variability. Some critical gaps exist in how we define, measure, and derive phytoplankton community composition (PCC) from ocean color. In this presentation we will discuss the following gaps. First, we must define the term PCC before we can align the data products for both algorithm development and validation. Second, less than a handful of satellite algorithms currently exist that derive any level of PCC from hyperspectral data and go beyond size class paradigm. Third, phytoplankton taxonomy data have been largely under-utilized in PCC algorithm development and validation. Advanced technologies like in-flow-imaging instruments have paved the way for the collection of higher spatial and temporal resolution of taxonomy information. When paired with bio-optical measurements, phytoplankton imagery will prove to be powerful tool for algorithm development and validation. Fourth, the availability of flow cytometry data that are necessary to quantify abundances of small phytoplankton, such as cyanobacteria and picoeukaryotes, is lacking. Lastly, no hyperspectral algorithms exist that partition PCC based on carbon estimates, which would fill a knowledge gap in the ocean biological carbon pump. Moreover, readily available standardized conversion tables to convert phytoplankton biovolume to carbon do not exist and would be necessary to support algorithm development from taxonomy data.

Primary production (PP) is highly sensitive to changes in the ecosystem and can be used as an early warning indicator for disturbance in the marine environment. Previous indicators of good environmental status of the North Atlantic, North and Celtic Seas suggested that daily PP should not exceed 2-3 gC m-2 d-1 during phytoplankton blooms and that annual PP should be

The EPIC/DSCOVR observations of the Earth’s surface lit by the Sun made from the first Lagrange point several times during the day are used to estimate daily averaged downward planar and scalar irradiance and average cosine for total light just below the ice-free ocean surface in the EPIC spectral bands centered on 317.5, 325, 340, 388, 443, 551, and 680 nm, and integrated values over the PAR and UV-A spectral ranges. A budget approach is used to estimate the planar fluxes just above and just below the surface. The sub-surface scalar fluxes and average cosines are derived from the above-surface fluxes using LUTs of clear sky and overcast situations and the estimated cloud factor (the ratio of actual and clear sky irradiance reaching the surface). Algorithm uncertainty is assigned to each estimate as a function of above-surface clear sky daily mean flux and cloud factor. Evaluation is performed using in situ above-surface planar flux measurements at existing open-ocean sites (fixed platforms and moored buoys) after transformation to below surface quantities  via radiation transfer simulations with best information about the controlling parameters. Time series at selected oceanic locations demonstrate the algorithm ability to capture variability for investigating ocean response to changes in available light over a wide range of temporal scales. The suite of EPIC daily averaged sub-surface radiation products, spectral and wavelength-integrated over key spectral ranges, is useful to address science questions pertaining to biogeochemical cycling of carbon, nutrients, and oxygen, as well as mixed-layer dynamics and upper ocean circulation. The methodology can be easily adapted to other satellite missions, in particular the future PACE and GLIMR.

Accounting for phytoplankton size class specific-variations in photophysiological parameters has shown promise in improving performance of satellite-based methods for estimation of primary production. Here, we utilize a large dataset of phytoplankton photosynthetic and optical properties and associated environmental conditions from the northern Gulf of Mexico to evaluate satellite approaches for estimating size class-specific variations in primary production. While no significant differences were found in the photosynthetic parameters PBmax (maximum rate of photosynthesis normalized to chlorophyll a) and αB (initial light-limited slope of photosynthesis-irradiance relationship normalized chlorophyll a) between different geographic regions, maximum quantum yield of carbon fixation in photosynthesis (Φcmax) differed significantly between regions and was higher for diatom-dominated communities. Multiple linear regression models, specific for the different phytoplankton communities, using a combination of environmental and bio-optical proxies as predictor variables showed considerable promise for estimation of the photophysiological parameters on a regional scale. Predictor values of phytoplankton size class-specific absorption indices along with other variables derived from satellite ocean color imagery (MODIS Aqua) were used to refine estimates of P-E parameters on synoptic scales, and subsequently applied to a bio-optical algorithm for quantifying phytoplankton group-specific estimates of primary production. Accounting for such variations in phytoplankton community composition should in principle allow for improved modeling of primary production for the river-influenced continental margin of the northern Gulf of Mexico as well as other systems in which there are substantial, spatial gradients in phytoplankton community composition.

The objective of the project is to track the status of coastal tidal wetlands on the U.S. Atlantic Coast at 30-m spatial resolution from 1986 to 2020. Coastal tidal wetlands are strong carbon sinks, with significant potential for climate change mitigation, through improved management. They are highly altered ecosystems, however, at substantial risk due to widespread and frequent land-use change, coupled with sea-level rise, leading to disrupted hydrologic and ecologic functions and ultimately, significant reduction in climate resiliency. Knowing where, when and how the changes have occurred, and the nature of those changes and their associated risks, is paramount to coastal communities and natural resource management. Large-scale mapping of the coastal tidal wetland changes is extremely difficult due to their inherent dynamic nature. To bridge this gap, we developed an automated algorithm for DEtection and Characterization of cOastal tiDal wEtlands change (DECODE) using dense Landsat time series. DECODE assembles all available Landsat observations and introduces a water level regressor for each pixel to flag the spectral breaks (characterized to condition change and cover change) and estimate harmonic time-series models for the divided temporal segments (classified to vegetated wetlands, unvegetated wetlands and open water). Besides, the water level regressor measures the impact of variations in water level on Landsat observations. Consequently, DECODE can export an additional annual product that presents the degree of tidal fluctuation. The case study of northeastern United States indicates overall accuracies of approximately 95.8% (cover classification) and 99.8% (change characterization), respectively. The condition change accounted for most of the coastal wetland, in which 84.3% of the changes only modified the condition of tidal vegetation, yet 15.7% of the changes led to the cover conversion. An interactive display of annual/accumulated change maps is available at: https://gers.users.earthengine.app/view/decodechange, and annual land cover maps are available at https://gers.users.earthengine.app/view/decodecover.

In the past decades the Arctic has experienced stronger temperature increases than any other region globally. Shifts in hydrological regimes and accelerated permafrost thawing have been observed and are likely to mobilize organic matter through rivers into the Arctic Ocean. Consequently, Arctic rivers such as the Lena River in Siberia need to be monitored closely to measure changes and better understand the Arctic carbon cycle. At the Research Station Samoylov Island in the Lena River Delta water samples are taken at a high frequency since 2018. From these samples, biogeophysical parameters, including concentrations of dissolved organic carbon (DOC) and colored dissolved organic matter (CDOM), are retrieved. In this study we collocated these samples with overpasses of the Ocean and Land Colour Instrument (OLCI) for the ice-free periods (May – October). The observations were atmospherically corrected using Polymer and used to derive absorption of CDOM (aCDOM). From aCDOM we deduced the amount of DOC using known relationships. The dataset includes over 2000 scenes for the five ice-free months of each year. First results reveal that simple band ratios provide a good estimate for aCDOM. These leave some residuals for further investigation. The best performing retrieval will then be used to create a DOC algorithm. Preliminary remotely sensed DOC agrees well with the in situ DOC data (r²=0.92, n=120) and captures the seasonal variability of DOC. We are thus able to simultaneously assess and refine the quality of satellite retrievals for the Lena River. We aim to improve existing algorithms for DOC monitoring and develop new algorithms, using a bootstrapping approach. These advances would extend existing in situ monitoring capabilities at the Lena River. Furthermore, they are beneficial for the pan-Arctic monitoring of DOC fluxes and broaden our understanding of the Arctic and global carbon cycle in a changing climate.

Chlorophyll concentration is used as a proxy for primary productivity. Satellite information provides continuous time series on this variable. For this reason, we obtained satellite images of SST, components of wind effort, surface currents, and chlorophyll-a concentrations from 2002 to 2020 from MODIS-Aqua. Also, the daily images of surface currents and wind are products of the GEKCO (LEGOS France) program. This information was used to study interannual variability, seasonal and regional trends in the Gulf of California. Chlorophyll concentration differences were obtained from the averages of two periods, from 2016/01 to 2020/8 and 2002/8 to 2007/2 to detect changes over time of chlorophyll concentrations. The difference in chlorophyll concentrations between two periods shows regions with different trends in Chl-a time, some regions being characterized by their negative or positive differences. Positive values ​​indicate an increase in chlorophyll concentration and negative values ​​indicate a decrease over a period of almost 20 years in the Gulf of California. It should be noted that the intermediate season, 2007/03 to 2015/12, was not considered as El Niño and La Niña events were swarming and masking the natural trends of chlorophyll concentrations. Based on the differences, geographic polygons were delimited and the time series of any variable by region were obtained, in the period from 2002 to 2020. Time series were constructed from the monthly series of satellite images of all the mentioned variables to correlate chlorophyll concentrations with the other variables. This analysis of regional trends is the basis for the proposal of regionalization in a climate change scenario. Additionally, the analysis of the time series of all the satellite variables shows a change of regime starting in 2012 and to date, the satellite variables maintain this change.

Droughts and climate-change-driven warming are leading to more frequent and intense wildfires, arguably contributing to the severe 2019–2020 Australian wildfires. The environmental and ecological impacts of the fires include loss of habitats and the emission of substantial amounts of atmospheric aerosols. Aerosol emissions from wildfires can lead to the atmospheric transport of macronutrients and bio-essential trace metals such as nitrogen and iron, respectively. It has been suggested that the oceanic deposition of wildfire aerosols can relieve nutrient limitations and, consequently, enhance marine productivity, but direct observations are lacking. Here we use satellite and autonomous biogeochemical Argo float data to evaluate the effect of 2019–2020 Australian wildfire aerosol deposition on phytoplankton productivity. We find anomalously widespread phytoplankton blooms from December 2019 to March 2020 in the Southern Ocean downwind of Australia. Aerosol samples originating from the Australian wildfires contained a high iron content and atmospheric trajectories show that these aerosols were likely to be transported to the bloom regions, suggesting that the blooms resulted from the fertilization of the iron-limited waters of the Southern Ocean.

Marine heatwaves (MHWs) have occurred in all of Earth’s ocean basins over the past few decades, with severe negative impacts on marine organisms and ecosystems. Of particular concern are compound ocean extremes events, i.e., multiple extreme events that occur simultaneously or in close sequence, as their individual effects may interact synergistically. Yet, the location and likelihood of these compound events, their underlying processes and their evolution under climate change are currently unknow. By combining the satellite-derived sea surface temperature data with a novel satellite-based ocean acidity product, we show that globally 1.8 in 100 months (or one out of five present-day marine heatwaves) are compound marine heatwave-ocean acidity extreme (OAX) events. This almost twice as many as expected from 90 percentile extreme event exceedances if MHWs and OAX events were statistically independent. Compound MHW-OAX events are most likely in the subtropics and less likely in the equatorial Pacific and the mid-to-high latitudes. The compound event likelihood results from opposing effects of temperature and dissolved inorganic carbon on [H+]. More compound events occur where the positive effect on [H+] from increased temperatures during MHWs is larger than the negative effect on [H+] from co-occurring decreases in dissolved inorganic carbon. Our results suggest that some of the observed high impact MHWs were also compound MHW-OAX events, in particular in the low-to-mid latitudes. The reported impacts of the low-to-mid latitude MHWs on marine organisms and ecosystems may therefore be also connected to additional stress from high acidity extremes.

Since 2013, marine heatwaves have become recurrent throughout the equatorial and northeastern Pacific Ocean and are expected to increase in intensity relative to historic norms. Ecological ramifications associated with these high temperature anomalies include reduced surface nutrient supply and chlorophyll levels, affecting the abundance and productivity of higher trophic organisms such as marine mammals and seabirds. The community composition of phytoplankton is a key mediator of the efficiency by which organic carbon is transferred to higher trophic levels and exported to the deep ocean. However, in situ observations of phytoplankton functional types (PFTs) in the open ocean are difficult to obtain, and thus, little is known about the impact that marine heatwaves can have on the assemblage of the phytoplankton community. We assimilate remote sensing ocean color data into an ocean biogeochemistry general circulation model to assess changes in the surface composition of PFTs in the Pacific Ocean over the ocean color satellite record (2002-2020). The largest changes in the partitioning of PFTs are observed in the Gulf of Alaska (GOA) and the ENSO 3.4 region, with opposite repercussions on total chlorophyll concentrations. At GOA, a low surface supply of nitrate during the heat anomaly in 2014 caused a decrease in (nutrient-rich-adapted) diatoms and an increase in (nutrient-depleted-adapted) dinoflagellates, leading to a positive anomaly in surface chlorophyll. In the ENSO 3.4 region, a decline of 40 % in the mean surface chlorophyll concentration during the 2016 El Niño event is associated with a nearly total collapse in diatoms, which are mostly displaced by chlorophytes, and to a lesser degree, by cyanobacteria. The combination of higher-temperature and lower-chlorophyll anomalies has become more common over the last decade, and our results provide insight on the phytoplankton types expected to dominate in these new environmental conditions.

Extreme events and disturbances (e.g., tropical cyclones, extreme precipitation and flooding often associated with land falling storms) affect aquatic carbon cycling at multiple spatiotemporal scales, and are most impactful in the coastal ocean with highly connected ecosystems that include rivers, estuaries, wetlands and the continental shelf. Although many field studies have reported on the strong carbon signatures associated with these extreme events, observational constraints have prevented a better assessment of their contribution to carbon cycling in the coastal ocean.  With projected increase in intensity, frequency and precipitation associated with these extreme events on widely different coastal regions (e.g., river-, tidal-dominated estuaries and wetlands), there is a critical need to improve mechanistic understanding of their impacts on carbon cycling. Although ocean color remote sensing has been widely used to assess phytoplankton response to the passage of tropical cyclones in the oceanic and shelf waters, studies related to carbon cycling especially in coastal and estuarine waters have been limited. In this presentation, we describe recent studies that examined carbon dynamics and fluxes in the more optically complex coastal environments. In a tidally-influenced wetland estuarine system, high spatial resolution satellite data from multiple sensors revealed the strong impact of extreme weather events on dissolved organic matter dynamics. In two river-dominated estuaries, multiple satellite sensor data revealed the contrasting impacts of two tropical cyclones on the particulate and dissolved organic matter dynamics and fluxes; a slow moving cyclone with extreme precipitation contributed a large estuarine-shelf carbon pulse, while a fast moving and strong cyclone resulted in a much smaller carbon flux to coastal waters. These studies demonstrate that the use of multiple satellite data with different resolutions and improved ocean color algorithms will be critical to assess contribution of extreme events to carbon fluxes in the coastal ocean.

Due to the essential ecosystem functions that submerged aquatic vegetation (SAV) such as kelp forests, eelgrass meadows and rockweed beds provide, up-to-date knowledge about their abundance and growth dynamics is critical to protect these threatened communities. Mapping and monitoring of SAV helps to assess the impacts of management and conservation efforts, to observe marine water quality and not least to construct blue carbon budgets. However, the highly fragmented nature of marine plants and algae, with patches consisting of many different communities, makes monitoring and mapping challenging. Many applied survey methods have in common that they provide very detailed information, but with the disadvantage that they are time and labor intensive, which limits our ability to comprehensively understand the abundance and dynamics of these underwater coastal communities at large scale. Consequently, these methods alone are not suitable for regular nationwide SAV assessments, which are required, for example, for reporting obligations resulting from directives such as the EU Water Framework Directive. In this study, we present our innovative approach for nationwide mapping of coastal habitats at 10 m spatial resolution with Copernicus Sentinel-2 imagery. By applying a combination of Copernicus Sentinel-2 satellite data, novel machine learning techniques and advanced data processing we have created the first-ever nationwide overview of the distribution of shallow water vegetation in Denmark (https://marine-vegetation.satlas.dk/) and Sweden in 2018 and 2020, respectively. As part of these activities, the entire methodological workflow was wrapped into an easy-to-use web application for non-technical specialists. By using the cloud-based web application, the user can execute the entire mapping process, from the selection of the latest Sentinel-2 imagery to the final SAV classification, in just a few clicks. We will give a brief overview of the methods implemented in our SAV mapping approach, show the cloud-based application and present examples of coastal habitat maps inclusive quantitative evaluation of the mapping accuracy and limitations of the approach.

Remote detection of seagrass in turbid coastal waters is challenging due to the need for satellite passes coincident with low tide and low turbidity conditions in order to detect seagrass. In the past this has meant that targeted aerial imagery has been the preferred collection medium for monitoring seagrass distribution. Commercial high resolution satellites from the PlanetScope constellation now provide us with daily coverage of the coastal waters of the US, this collection frequency can help overcome previous issues with turbidity and tidal state, by providing multiple images per month. Here we use the high frequency of passes available from PlanetScope to retrieve distribution and density of seagrass in the coastal bays of Virginia for 2019, 2020 and 2021 . All passes in which seagrass were visible were processed to seagrass presence/absence using fixed training patches and support vector machine learning available in ArcGIS Pro. The frequency with which a pixel was classified as seagrass was found to be correlated with seagrass density previously estimated as percent cover from aerial imagery. The changing distribution and density from the spring through the summer was also detectable allowing us to highlight the reduction in seagrass density after the warm summer months. The evidence points towards using frequency presence as a measure of seagrass (or submerged aquatic vegetation) density throughout the Chesapeake Bay, an area that has been difficult to monitor from satellites in the past.

Blooms of pelagic macroalgae have been reported around the world, among which are Sargassum fluitans/natans in the Atlantic, Sargassum horneri in the East China Sea, and Ulva prolifera in the Yellow Sea. Long-term remote sensing observations suggest their increasing trends. In several recent papers, such macroalgae blooms have also been hypothesized to play potentially a major role in carbon fixation and sequestration. Here, from literature review and using published values on macroalgae coverage, biomass density, carbon/biomass ratio, primary productivity, and carbon sequestration efficiency, we provide first-order carbon estimates of pelagic macroalgae. It is found that, compared to phytoplankton, pelagic macroalgae may contribute significantly to carbon stock, carbon fixation, and carbon sequestration in the macroalgae “niche” area, but such a role may diminish when the area is enlarged at basin scales.

.

In ESA’s Biological Pump and Carbon Exchange Processes (BICEP) project, we have been investigating satellite methods to map marine primary production, phytoplankton carbon, particulate organic carbon, particulate inorganic matter and dissolved organic carbon. Time series of each of these products at 9 km, monthly resolution is being generated. The main input to the calculations is the ocean-colour fields generated by the Ocean Colour Climate Change Initiative (OC-CCI). These are supplemented by fields of photosynthetically available radiation at the surface of the ocean (from NASA), sea-surface temperature, and sea-surface salinity (from CCI). For most of the products, the time series extends from 1998 to 2020, unless limited by availability of input data. The primary production computations (Kulk et al. 2020, 2021) rely on an extensive in situ database of photosynthesis-irradiance parameters. The same parameter set is used, along with a photo-acclimation model, to compute phytoplankton carbon, ensuring that the allocation of resource (light) between production of carbon and chlorophyll is treated in an internally consistent manner (Sathyendranath et al. 2020). Various algorithms available for calculation of particulate organic carbon have been compared, before selecting one of the better-performing algorithms for generation of the time series. The algorithm for mapping dissolved organic matter is a novel one, that makes use of machine-learning tools. In situ data bases have been created for validation and comparison of products, and for generation of uncertainty estimates. Algorithms for estimation of biological export production have also been implemented. In this presentation, we provide our preliminary estimates of the biologically mediated pools and fluxes of carbon in the oceanic mixed layer. Reference: CEOS (2014) CEOS Strategy for Carbon Observations from Space. The Committee on Earth Observation Satellites (CEOS) Response to the Group on Earth Observations (GEO) Carbon Strategy. Issue date: September 30 2014. JAXA, Japan Kulk G, Platt T, Dingle J, Jackson T, Jönsson B, Bouman HA, Babin M, Doblin M, Estrada M, Figueiras FG, Furuya K, González N, Gudfinnsson HG, Gudmundsson K, Huang B, Isada T, Kovac Z, Lutz VA, Marañón E, Raman M, Richardson K, Rozema PD, Van de Poll WH, Segura V, Tilstone GH, Uitz J, van Dongen-Vogels V, Yoshikawa T, Sathyendranath S (2020). Primary production, an index of climate change in the ocean: Satellite-based estimates over two decades. Remote Sensing 12:826; doi:10.3390/rs12050826. Kulk G, Platt T, Dingle J, Jackson T, Jönsson B, Bouman HA, Babin M, Doblin M, Estrada M, Figueiras FG, Furuya K, González N, Gudfinnsson HG, Gudmundsson K, Huang B, Isada T, Kovac Z, Lutz VA, Marañón E, Raman M, Richardson K, Rozema PD, Van de Poll WH, Segura V, Tilstone GH, Uitz J, van Dongen-Vogels V, Yoshikawa T, Sathyendranath S (2021). Correction: Kulk et al. Primary Production, an Index of Climate Change in the Ocean: Satellite-Based Estimates over Two Decades. Remote Sensing 13:3462; doi:10.3390/rs13173462 Sathyendranath, S, Platt, T, Kovač, Ž, Dingle, J, Jackson, T, Brewin, R JW, Franks, P, Marañón, E, Kulk, G, and Bouman, HA (2020) Reconciling models of primary production and photoacclimation [Invited]. Applied Optics, 59: C100-C114. https://doi.org/10.1364/AO.386252

Marine phytoplankton rival terrestrial ecosystems in their productivity, fixing roughly 50 PgC yr-1. A portion of this organic carbon is exported out of the surface ocean by the various pathways of the ocean’s biological carbon pump, including gravitational settling of organic particles (the “gravitational pump”), physical mixing and advection of organic carbon (the “mixing pump”), and active transport by vertically migrating metazoans (the “migrant pump”). Carbon exported by these pathways can be sequestered as respired CO2 in the deep ocean for years to centuries, but the magnitude of each pathway’s contribution to carbon export and sequestration remains highly uncertain. Here, satellite observations and ocean biogeochemical data are assimilated in a numerical model of the biological pump to quantify carbon export and sequestration by each of these pathways both globally and regionally. We determine that global carbon export is 10.3 Pg C yr-1 and total sequestration is 1,300 Pg C. The gravitational pump exports 7.3 Pg C yr-1, 85% of which is zooplankton fecal pellets and 15% sinking phytoplankton aggregates, while the migrant pump exports 1.0 Pg C yr-1 and the mixing pump exports 1.9 Pg C yr-1. These pathways have different sequestration times, with an average of 140 years for the gravitational pump, 150 years for the migrant pump, and only 50 years for the mixing pump. Global maps reveal the greatest carbon sequestration, and longest sequestration times, in the northern high latitudes, and shortest sequestration times in the subtropical gyres. These results suggest that ocean carbon storage will weaken as the ocean warms and the subtropical gyres expand.