Satellite-based Estimation Of Particulate Organic Carbon Export In The Northern South China Sea

A major term in the global carbon cycle is the ocean’s biological carbon pump（BCP）.Particulate organic carbon（POC）vertical flux refers to the amount of POC export from euphotic layer to the interior within one day or one year and it can be used to characterize the efficiency of marine BCP.Accurately quantifying the magnitude of the POC export flux is essential to understand the global carbon cycle and,by extension,the impact of anthropogenic perturbation on atmospheric CO₂ levels,especially in the context of sustained changes in the marine ecosystem.However,there is still considerable uncertainty（which is equivalent in magnitude to annual anthropogenic CO₂ emissions）in the estimates of global POC export flux from euphotic layer.And the uncertainty of the POC export flux in marginal sea is considered to be one of the main reasons for the high variation of global ocean POC export flux estimate.The South China Sea（SCS）is the largest marginal sea in the Western Pacific.Since the 1980s,POC export flux have been estimated by means of sediment traps or isotope tracer(²³⁴Th).However,limited by the small spatial coverage of the measured data,the measured results still have large uncertainties.In addition,the measured data with short-time sampling period also cannot meet the demand for the study of longtime changes of POC export flux in the SCS.With the large-scale and long-term advantages of satellite remote sensing,it is helpful to quantitatively estimate the effect of climate change on POC export flux of euphotic layer in the SCS.The main work of this paper are as follows:（1）Spatiotemporal variation and regulation mechanism of POC export efficiency in the northern South China Sea（NSCS）The annual average POC export efficiency were about 0.37 and 0.24 in the shelf（50-200m depth）and basin（depth greater than 200m）of the NSCS,respectively,based on measured POC export flux data using ²³⁴Th method and 16-days composited satellite-derived NPP data prior to ²³⁴Th sampling date.Spatially,the high POC export efficiency mainly distributed in the coastal areas,near the Pearl River estuary area and shelf regions.POC export efficiency generally decreased with the increase of water depth.In addition,POC export efficiency was the highest in autumn,followed by spring and summer,and the lowest in winter.On the other hand,NPP had different regulating mechanisms on POC export efficiency in the shelf and basin.With a shallow euphotic depths（50m）and relative high proportion of large-size algae（diatom）in euphotic layer,the increase of NPP can significantly improve the POC export efficiency in the shelf areas.Under the regulation of DOC export,zooplankton grazing and the decoupling of NPP and POC export,the increase of NPP would cause an decrease of POC export efficiency in the basin.（2）Estimation of POC export flux in the shelf area by satellite remote sensingObservation results have shown that there was a significant correlation between POC export flux and POC reserve（IPOC）in the euphotic layer of the shelf regions in the NSCS.Accurate sea surface POC concentration and POC profile structure were prerequisites for IPOC estimation.Based on measured data,we established an empirical relationship between satellite-derived particle backscattering coefficienct(b_bp)and sea surface POC concentration in the shelf regions.In addition,POC concentration in the shelf area is significantly correlated with the measured particle attenuation coefficient at660nm（c_p660）.According to the profile distribution of c_p660,the POC profile showed a linear increasing structure with depth in spring and summer,and a mixed uniform distribution structure in autumn and winter.The shelf area IPOC can also be characterized as a power function of the sea surface POC concentration（r²～0.8）.And the validation results show that the IPOC estimation results based on sea surface POC concentration are in good agreement with the observation results（the root mean square error is small）.On the basis of satellite-derived IPOC,the spatial and temporal distribution of POC export flux in the shelf area were analyzed:for spatial distribution,the POC export flux could be partly affected by the terrestrial POC input,and the POC export flux were higher in summer and winter;with the increase of depth,the POC export flux in the outer shelf area decreased significantly.Generally,the POC export flux was small than 100 mmol·m^-2·d^-1.On the seasonal average,the satellite-derived POC export flux in spring,summer,autumn and winter were 11.24±9.46,14.59±13.30,40.92±27.52 and 21.50±17.06 mmol·m^-2·d^-1,with a relative deviation of less than 36%.On the annual average scale,the satellite-derived and in situ results POC export flux were 21.70 and 18.32 mmol·m^-2·d^-1,respectively,with a relative deviation of 18.45%.（3）Estimation of POC export flux in the NSCS based on a semi-analytical food web model with all input parameter could get from satellite remote sensingThe inverse relationship between NPP and POC efficiency in the basin of NSCS determined the inapplicability of traditional empirical POC export flux estimation methods in this area.While the food web model can quantitatively estimate the POC export flux,it can better characterize the regulation mechanism of POC export.Comparing the satellite-simulated results with observed data,it was shown that the satellite-simulated POC export flux was significantly correlated with the observed data both on the shelf and basin region,although there were some differences in the magnitude.After using the empirical threshold and linear fitting correction,on the annual average scale,the satellite-simulated POC export flux were 21.64 and 6.46mmol·m^-2·d^-1 for the shelf and basin regions,respectively,with the relative deviation was less than 8.04%.The validation based on independent dataset（2004-2005 cruises）in the basin region showed that the relative difference of annual averaged satellite-simulated POC was about 24%.Generally,the POC export flux in the NSCS was highest in winter,followed by spring and autumn,and the smallest in summer based on food-model method.Comparing to the direct sinking flux of large phytoplankton and associated aggregates by gravity,the zooplankton grazing（especially the grazing of micro-zooplankton）may be the main contributor of POC export in the NSCS.The POC export flux remote sensing inversion method based on food web model can better characterize the seasonal,inter-annual and spatial variations of the POC export flux,and can partially explain the regulatory process of POC export in the NSCS.（4）Long-term variation of phytoplankton in the SCS under climate changeTaking the upwelling area in the western of the Luzon Strait and the central basin area as the representative regions,the 1998-2007 period as the research period,the modified food web model was used to simulated the POC export flux products in the SCS,combined with the satellite-derived sea surface chlorophyll concentration,phytoplankton size structure and primary production products,to comprehensively analysis of changes in carbon sequestration parameters of phytoplankton in the SCS under climate change.In terms of seasonal variation,influenced by the winter monsoon and the grazing of zooplankton,POC export flux in the upwelling region reached a high value in March and November while the lowest happened in July.The POC export flux in the central basin region was affected more by the southwest monsoon,and reached a high value in July-August.For inter-annual variation,except in 1998,the POC export flux in the SCS remained relatively stable,and there was no significant inter-annual variation trend.The results of long-term variation analysis based on monthly mean data also showed that there was no significant change trend in Chla,NPP and POC export in the SCS.ENSO events will also affect the POC export from euphotic layer.Analyzed results showed that E1 Ni?o would reduce the POC export flux while La Ni?a could promote the export of POC in the SCS.In addition,the time scales of the impact of ENSO events on the regulation of POC export in the SCS were also different:in the upwelling region,the ENSO mainly affected the POC export a year after the ENSO happened;while in the central basin,the time scale was about six months.Compared with traditional carbon cycle research,remote sensing research of marine carbon cycle can combine microscopic biogeochemical processes with remote sensing large-scale macroscopic observations.The marine POC export flux involves a series of continuous processes from primary production of phytoplankton to POC export,which urgently need the support of long-term observation data.Based on observational data and remote sensing products,this study found different regulation mechanism of primary productivity on the POC export efficiency on the shelf and basin region in the NSCS.In addition,based on the analytical food-web model,a POC export flux remote sensing estimation method was established,which laid foundation for the study of long-term variation of phytoplankton carbon sequestration of the SCS in the context of climate change.At the same time,the results of remote sensing large-scale observation data can further promote micro-scale biogeochemical processes,such as:1)the formation mechanism of negative correlation between primary production and POC export efficiency in the basin area;2)regional adaption of important parameters in the food-web model.Since the profile information of POC can not be obtained directly from ocean color remote sensing,how to combine numerical model,ocean color remote sensing and active lidar remote sensing to realize the analysis of 3-D dynamic process of POC export flux is a research direction that needs further development.