| Since 2007,the green tide gathered by floating Ulva prolifera has broken out in the Yellow Sea in successive years,causing economic losses to coastal cities of Shandong Peninsula,and real-time monitoring of the green tide has become an urgent matter.The technology of remote sensing plays an important role in the monitoring of green tide in the Yellow Sea with its advantages of high resolution,large monitoring range and high observation frequency.Since most of the green tide pixels monitored by satellite images are mixed pixels,images with different resolutions are affected differently by the mixed pixels,and the monitoring area of green tide is significantly different,which brings uncertainly to the estimation of green tide coverage area.Therefore,the study compares the coverage area and affected area monitored by different resolution images,and reveals the influencing factors of the information extraction of green tide,so as to realize multi-scale observation and inter-annual seasonal change monitoring of green tide.Based on multi-source remote sensing satellite images(including MODIS,GF-WFV,HJ-CCD,Sentinel-2 MSI,Landsat,HY-1C CZI and Tiangong-2 MWI),synchronous satellite images are firstly used to analyze the variation trend of coverage area,affected area and aggregation degree of green tide with resolution of satellite images.The missing phenomenon of green tide in low resolution satellite images is evaluated and the effects of spatial resolution on the extraction results of green tide are explored.By constructing the scale conversion model of green tide coverage area with different resolutions,the scale conversion of green tide coverage area with different resolutions can be realized.Meanwhile,the application of multi-source remote sensing data in green tide monitoring can improve the observation frequency of green tides.Based on these results,the temporal and spatial distribution and inter-annual seasonal variation characteristics of green tide in the Yellow Sea from2007 to 2021 are obtained,and the relationship between the scale of green tide and the raft harvesting process of Porphyra culture in recent years is analyzed.The main conclusions are as follows:(1)The variation trend of coverage area,affected area and aggregation degree of green tide with spatial resolution of satellite image is obtained by synchronous satellite image.The results show that with the increase of spatial resolution,the coverage area and aggregation density of green tide decrease,while the area affected satellite remote sensing images by green tide increases.The omission rate of green tide increases with the decrease of resolution,and the omission rate of the area with lower aggregation density of green tide is higher than that of the area with higher aggregation density.The difference of spatial resolution of satellite images with different resolutions and the mixed pixel effect caused by them are the main reasons for the difference of green tide monitoring results from multi-source remote sensing images.(2)This study analyzes the influence of dynamic threshold setting on the monitoring results of green tide,and finds that in addition to the spatial resolution of satellite images,dynamic threshold setting is also an important factor in estimating the coverage area of green tide.In this study,two recognition methods,Linear and Equalization,are proposed,and the coverage area of green tide with different recognition methods of images with different resolutions is explored.The results show that the overall variation trend of green tide coverage obtained by images with different resolutions is consistent,and it is found that the coverage area of green tide monitored by Equalization is about 20%-50% more identified than Linear method.(3)In terms of interannual variation of green tide,MODIS images with a resolution of 250 m are used to obtain the interannual variation of green tide in the Yellow Sea from 2007 to 2021.The results show that the variation trend of annual and daily maximum coverage is roughly the same for different identification methods,and the difference of green tide coverage obtained by different indices(DVI,NDVI and RVI)is within 10%.From 2007 to 2021,the interannual scale of green tide in the Yellow Sea changed significantly.There are three peaks in the interannual scale of green tide,separately,2009,2016 and 2019,and the annual maximum daily coverage area of green tide is more than 2500 km~2.(4)In terms of seasonal variation of green tide,WFV,CCD and CZI images are used to supplement MODIS data to obtain the spatial and temporal distribution and seasonal variation of green tide coverage area in the Yellow Sea from 2007 to 2021.It is found that the occurrence time,extinction time and annual scale of green tide differ significantly in each year,but the overall trend is similar: It originated in the waters near the Northern Shoal of Jiangsu province around May and drifted along the coast of Shandong Peninsula with the prevailing wind direction,and the coverage area of green tide increased rapidly.After reaching the annual daily maximum coverage area,the coverage area of green tide decreases rapidly and generally disappears in the coastal waters of Shandong Peninsula at the end of July and early August.(5)WFV optical and Sentinel-1 microwave data are used in this study to extract Porphyra cultivation area from 2019 to 2021,and analyze the Porphyra cultivation equipment recycling process in May.The results show that in 2019 and 2021,the annual maximum coverage area of green tide based on CZI reach 2290 km~2 and 1949km~2,while in 2020,the annual maximum coverage area is only 336 km~2.This study further confirms the conclusion that delaying the recovery process of laver culture raft can control the annual maximum coverage area of green tide.In this study,multi-source remote sensing satellite data are used to obtain the interannual and seasonal variations of green tide in the Yellow Sea from 2007 to 2021,which provides a basis for scale transformation and quantitative observation of green tide disasters. |
PDF Full Text Request