| Chlorophyll a（Chl-a）concentration has a strong correlation with marine primary productivity,global carbon cycle and upwelling research.The complex marine environment in the South China Sea leads to the extremely complex spatiotemporal distribution of the concentration of Chl-a.However,there are many cloud-covered areas in the South China Sea,resulting in a high loss rate of water color remote sensing Chl-a data.The missing data may cause spatiotemporal discontinuity and loss of important data information,which has an important impact on the study of the spatiotemporal distribution,change trend and related impact factors of Chl-a concentration in the South China Sea and its upwelling area.Therefore,in this paper,data interpolation empirical orthogonal function（DINEOF）is used to reconstruct the whole and missing points of the monthly average medium resolution imaging spectrometer（MODIS）Chl-a and fused Chl-a remote sensing data in the South China Sea from 2010 to 2021,and the quality of the reconstructed data is analyzed,Then use the reconstructed data to carry out the relevant application research of the upwelling in the South China Sea.First of all,through the comparative analysis of the overall reconstructed data quality,the reliability of the reconstructed data with missing points is discussed.The results show that the root-mean-square error between the MODIS Chl-a and the fused Chl-a reconstruction data and the original data are 0.1238 mg·m-3and 0.1202 mg·m-3,respectively,and the correlation coefficients are 0.93 and 0.94,respectively.The quality of the fused Chl-a reconstruction data is better than that of the MODIS Chl-a reconstruction data.On the data missing points,the reconstruction data obtained from the overall reconstruction and the reconstruction of the missing points are identical,which means that the data reconstruction can only be carried out for the missing points,and the remote sensing data set of chlorophyll a without missing can be established.It can be seen from the spatiotemporal distribution of data reconstructed by fused Chl-a that the higher the data loss rate,the more obvious the smoothness,but the more obvious the overall reconstructed data will be.The proportion of data points with reconstruction relative error within 20%has a certain negative correlation with data missing rate data.The higher the data missing rate is,the lower the proportion is,indicating that the reliability of reconstruction data of missing points is worse.Secondly,combined with the fusion of Chl-a reconstruction data,sea surface temperature（SST）and sea surface wind（SSW）data and Ni（?）o3.4 index,the change trend and spatiotemporal change characteristics of Chl-a,SST and SSW are discussed,and the significant EOF modes of the three are analyzed.The results show that:（1）the areas with high Chl-a concentration are mainly distributed on the offshore continental shelf,which is characterized by high near-shore Chl-a value and low offshore,The northwest is higher than the southeast,and the central area of the basin has little change and basically remains the same.（2）Except in the late summer and early autumn,SST is higher in the southeast direction of the South China Sea than in the northwest direction,and lower in the north of the South China Sea than in the south,with obvious gradient changes as the dimension increases.（3）The wind speed and the spatial distribution of wind direction of SSW are affected by the monsoon.The northeast monsoon runs through the whole South China Sea,forming a northeast-southwest wind axis.The wind speed at both ends of the wind axis is greater than the middle of the wind axis and the north end is greater than the south end.The southwest monsoon wind direction is basically parallel to the coastline of Vietnam below 12°N,forming a southwest-northeast wind axis,but the wind axis does not run through the whole South China Sea.In terms of time,Chl-a,SST and SSW all have obvious seasonal changes.（1）Chl-a generally reaches the annual maximum around January in winter,and the annual minimum around April in spring.（2）SST is the lowest in winter and the highest in summer.（3）SSW has two high peaks throughout the year,and the wind speed of the northeast monsoon is higher than that of the southwest monsoon.（4）There is a lag between the sea surface temperature anomaly in the South China Sea and the Ni（?）o3.4index.The lag time is generally 6-7 months,and its correlation coefficient is the largest in the lag of 6 months.（5）On the whole,the annual average concentration of Chl-a shows an obvious downward trend,with a rate of about 0.008 mg·m-3·year-1,reflecting the continuous decline of primary productivity in the South China Sea.（6）On the whole,SST shows an obvious upward trend,rising at the rate of about 0.036°C·year-1.It should be noted that 2011-2012 is the year of La Nina,and the annual average temperature does not decrease but increases,which needs to be further explored.（7）On the whole,the annual average SSW value shows a slight upward trend,rising at a rate of about 0.013m·s-1·year-1.From the EOF mode analysis,the first two EOF modes,Chl-a,SST and SSW,are independent and not closely related to each other,while the third EOF mode is closely related,which is reflected in the waters of the Mekong estuary in Vietnam.From the perspective of correlation,Chl-a has a weak negative correlation with SST,a significant positive correlation with SSW,and a significant negative correlation between SST and SSW.Finally,based on the monthly average SSW data from 2010 to 2021,the temporal and spatial distribution of the average Ekman pumping velocity in the same month for many years in the South China Sea is calculated,and the main location areas of the upwelling in the South China Sea are analyzed.（1）Under the influence of the northeast monsoon,the upwelling area is mainly distributed in the coastal waters of the east boundary of the South China Sea on the right side of the wind direction,and the downwelling area is mainly distributed in the coastal waters of the west boundary of the South China Sea on the left side of the wind direction.（2）Under the influence of the southwest monsoon,it is opposite to the northeast monsoon.（3）During the monsoon transition period,the up-and-down flow formed by Ekman pumping has the characteristics of both winter monsoon and summer monsoon.In order to analyze the temporal and spatial changes of Chl-a in the upwelling area,sea level anomaly（SLA）,sea surface current（SSC）,SST and WOA23 water temperature data at different depths are used as the relevant environmental fields.Select the Vietnam Mekong Estuary area under the influence of the southwest monsoon and the northwest Luzon Island under the influence of the northeast monsoon as the typical upwelling area,and look for the threshold value of the Chl-a boundary to define the boundary of the upwelling area.The results show that in the northwest Luzon Island area,0.5mg·m-3can be used as the threshold value to define the significant area of the high value of the upwelling Chl-a concentration.In the Vietnam Mekong Estuary area,The estuarine and coastal waters greater than 2.0mg·m-3are regarded as the areas with significant impact of terrestrial nutrients.After the data in the area are eliminated,0.15mg·m-3can be used as the threshold to define the areas with significant high value of upwelling Chl-a concentration. |
PDF Full Text Request