Blooms : case sTudies in The norTh aTlanTic and arcTic WaTers

Blooms of a coccolithophore E. huxleyi are generally huge, occur annually and in the oceans of both Hemispheres. As a calcifying algal species, E. huxleyi is known to enhance the partial pressure of dissolved CO2 in the surface ocean, thus reducing its ability to absorb atmospheric CO2. Here we report on the results of our satellite study of CO2 enhancement in the atmospheric column over E. huxleyi blooms in the North, Greenland, Iceland and Barents seas. The study is based on OCO-2 and wind force and direction data, and E. huxleyi bloom masks developed by us earlier. Eight case studies are discussed herein relating to the time period 2015–2018. The results obtained are strongly indicative that, indeed, the phenomenon of E. huxleyi blooms noticeably affects the carbon fluxes between the atmosphere and the surface ocean: the quantified enhancement of CO2 content in the atmospheric column over the bloom area in five out of eight case studies proved to be in the range of 0.6–3.0 ppm. It is also shown that the magnitude of CO2 enhancement in the atmospheric column is significantly controlled by air advection in the boundary layer.

In addition to the production of particulate calcite, coccolithophores are capable of increasing dissolved CO 2 partial pressure within their blooming areas [Holligan et al., 1993;Kondrik et al., 2018].
Conjointly, these two mechanisms affect the carbon balance in surface ocean and tend to weaken marine carbon sinks, which has farreaching consequences in terms of planetary climate change [IPCC, 2014].
Within the coccolithophore group, E. huxleyi is the most widespread species in the world's oceans [Westbroek et al., 1985;Moore et al., 2012].It forms gigantic blooms with a surface of several thousand square kilometers [Kondrik et al., 2017], but sometimes exceeding one million square kilometers [Balch et al., 2014].
The aforementioned E. huxleyi bloom-driven enhancement of dissolved CO 2 partial pressure can reduce, nullify or even reverse the flux of CO 2 at the atmosphere-ocean interface.Indeed, Shutler et al. [2013] report on an average reduction in the monthly air-sea CO 2 flux by about 55 % across the marine tracts encompassing extensive E. huxleyi blooms in the North Atlantic, whereas the maximum reduction over the time period 1998-2007 was registered at 155 %.
Here we present our results on several case studies in the North, Iceland, Greenland and Barents seas.The study was designed to quantify the atmospheric columnar CO 2 over E. huxleyi blooms based on remote sensing data from the Orbiting Carbon Observatory OCO-2 that was put into orbit in 2014 to study CO 2 concentration and spatio-temporal distribution in the Earth's atmosphere [Crisp, 2015].The areas targeted in the above seas were identified in advance making use of E. huxleyi bloom masks developed on the basis of ocean color data from the ocean-colour climate-change initiative OC CCI data archive [Sathyendranath, Krasermann, 2014].methodology Previously, based on the developed bloom masking technology, i. e. the methodology of E. huxleyi bloom detection and contouring, the 1998-2018 time series of blooms of this alga were obtained for the Subarctic Atlantic and Arctic Seas [Kondrik et al., 2019;Selyuzhenok et al., 2019].For the revealed locations of E. huxleyi blooms, the 2015-2018 OCO-2 data were subjected to sieve analysis in order to ascertain the cases of OCO-2 footprint trajectory crossing both the bloom area and adjoining bloom-free waters.The identified situations were further analyzed as case studies in order to investigate on a quantitative basis if there was any impact of E. huxleyi bloom areas on XCO 2 registered by OCO-2.Thus, to assess the impact, XCO2 values registered along the OCO-2 footprint both over the bloom area and beyond it (either prior to reaching the bloom area or after leaving it) were compared.The resultant change in XCO 2 , i. e. ΔXCO 2 , was considered as a measure of the E. huxleyi bloom impact on the CO 2 exchange at the atmosphere-sea water interface, and hence, of the change in the CO 2 atmospheric columnar content.
All case studies also included the analysis of above water surface wind force and direction over the bloom area in order to clarify the issue of air mass advection across the satellite footprint trajectory.
In the case of the North Sea 2018, CCMP are unavailable, and in their stead ASCAT data, version 2.1 were exploited (http://www.remss.com/missions/ascat/).To better harmonize scatterometric and radiometric wind measurements, the ASCAT data were generated with the use of a new Geophysical Model Function, C-2015.
Thus, the wind vectors that are laid upon the maps illustrating our case studies represent 8-day wind force and direction averages specifically over the areas of E. huxleyi blooms.
atmospheric co 2 content.The column averaged dry air mole fraction, XCO 2 is defined as the ratio of "the altitude-dependent CO 2 number density integrated over the atmospheric column and the column abundance of dry air" [Crisp, 2015].
Having a 16-day ground-track repeat cycle, OCO-2 yields XCO 2 values with single-sounding random errors in the range of 0.5-1 ppm at solar zenith angles up to 70°, and at the spatial resolution of 3 km 2 in nadir, i. e. 1.25 km in width and ~2.4 km in length, which corresponds to a ~1.8 mrad instantaneous field of view and 3 Hz sampling.In 2018, the OCO-2 data processing algorithms were improved at NASA, and the current and retrospective products (L1B/L2 Version 8 and L2LiteFileVersion 9; October 10, 2018) were released (https:// docserver.gesdisc.eosdis.nasa.gov/public/project/OCO/OCO2_DUG.V9.pdf).
Only high quality data (i.e. unflagged data) were employed in our case studies.8-day averaging of XCO 2 data was implemented in this study.

E. huxleyi bloom masking.
The 1998-2018 time series of E. huxley blooms in the Subarctic Atlantic and Arctic oceans was retrieved from Ocean Color Climate Change Initiative (OC CCI) data through the analysis of spectra of remote sensing reflectance, R rs (λ).The methodology is described in detail in [Kondrik et al., 2017[Kondrik et al., , 2019]].Concisely, a typical R rs spectrum from a E. huxleyi bloom exhibits a maximum at λ = ~490 nm at the late stage of its development, when the surface water is predominantly populated by coccoliths whereas E. huxleyi cells have already mostly died off.Accurate delineation of E. huxleyi blooms was based on fulfillment of the requirement that the spectral values of Rrs (sr -1 ) in the OC CCI standard spectral channels exceed the following statistically established thresholds: 0.001 at 412 nm, 0.008 at 443 nm, 0.01 at 490 nm, 0.008 at 510 nm, and ~ 0 at 670 nm.On this basis, masks of E. huxleyi blooms were plotted for the target Subarctic and Arctic seas to restore the chronicle of spatiotemporal variations of the bloom areas between 1998 and 2018.

results and discussion
Here we present the results of eight satellitebased case studies from the Barents, Iceland, Greenland and North seas (Table, Fig., a-h).Note that red lines show the limits of the beyond-bloom areas used in this study for assessing ΔXCO 2 ; black arrows indicate the force and direction of wind over the bloom area; black areas are E. huxleyi blooms.
As the and 3.0 ppm.These numbers are fully consistent with the results we have obtained in the study of E. huxleyi -induced XCO 2 in the Black Sea as registered in 2016-2017.However, in three cases (the South Iceland, and North seas), no XCO 2 enhancement was found.A combined OCO-2 and wind data analysis has shown that the explanation of the apparent absence of E. huxleyi blooming impact upon XCO 2 might reside in the effect of above water air mass advection.Indeed, for cases 4, 6, and 8 the meteorological and E. huxleyi blooming conditions were specific.In case 4 the blooming area was essentially inhomogeneous/fractionized, and the wind direction was southern, i. e. bringing air masses from the parts of the sea free of any E. huxleyi bloom influence.
In case 8 there were very similar conditions in terms of wind-driven advection of above-water air from marine tracts void of E. huxleyi blooming.It is also worth mentioning that the blooming area was also significantly fractionized.
A special consideration should be given to case 6.At first glance, it appears that the E. huxleyi-driven ΔXCO 2 signal should not be zero: the footprint trajectory traverses the bloom area, and the advected air comes from a large portion of bloom.However, the number of OCO-2 pixels within the bloom area is rather small, rendering the ΔXCO 2 retrievals insufficiently reliable.

concluding remarks
Produced in the reaction of calcification inside the cell of E. huxleyi, CO 2 becomes available for the reaction of photosynthesis with a result of a reduced uptake of dissolved CO 2 from ambient water.Thus, surface marine waters within the bloom of E. huxleyi turn out to be less CO 2 depleted.Moreover, the thus enhanced partial pressure of dissolved CO 2 can either nullify the flux of atmospheric CO 2 or even reverse it.This has been proven in shipborne surveys, and through spaceborne observations over the Black Sea: the enhancement of CO 2 content in the atmospheric column proved to be within 1-2 ppm.
The eight case studies conducted with the employment of OCO-2 satellite data and presented in this concise report have shown that the impact of E. huxleyi blooming phenomenon on the atmospheric CO 2 partial pressure over the North, Iceland, Greenland, and Barents seas proved to be of the same order of magnitude as over the Black Sea (0.6-3 ppm).It is also shown that the magnitude of CO 2 enhancement in the atmospheric column is significantly controlled by the air advection in the boundary layer.
Arguably, this might be an indication of some inherent property of E. huxleyi, and the obtained results on the increment of CO 2 in the atmospheric column over the blooms of this alga can be considered as representative of this phenomenon across the oceanic tracts, at least, in the Northern Hemisphere.