Multi-Source Remote Sensing Sea Surface Temperature Fusion Product Generation And Frontal Detection

The sea surface temperature is an important indicator of the earth's ecosystem.It is widely used in weather change,atmospheric simulation and climate change research.It has great influence on marine dynamic environment and air-sea interaction.The acquisition of early SST data is mainly based on buoys,platforms and voluntary ships,but it is difficult to reveal the changes in the global ocean in the daily,monthly,seasonal,interannual and interdecadal scales of the ocean water body with great changes in the dynamic range of time and space.Satellite remote sensing technology is the only effective way to get global ocean SST observation data because of its fast,wide range synchronous and continuous observation technology advantages.This paper,based on infrared and microwave remote sensing data,exerts the advantages of multi source remote sensing,develops the global ocean SST fusion technology,and develops a global marine multi source remote sensing SST fusion product with high precision,high spatial coverage and high spatial and temporal resolution,and carries out the application of the Northwest Pacific frontal detection based on this product.The main work of this article is as follows:(1)Performed single satellite product accuracy evaluation of infrared radiometer SST data(AVHRR,MODIS(Aqua and Terra)and VIIRS)and microwave radiometer SST data(AMSR-E,AMSR2,TMI,GMI,WindSat,HY-2A RM).Except HY-2A RM data,the accuracy of other SST data is basically the same.The HY-2A RM data with larger error is linearly fitted and corrected,and the corrected data accuracy is improved to some extent.(2)A multi-source remote sensing SST fusion optimal interpolation algorithm was developed.Based on the development algorithm,a global ocean 0.1°/daily SST fusion product was produced from 2010 to 2016.The fusion product accuracy was verified based on Argo buoys,NDBC buoys,and tropical moored buoys.The results showed that the average deviations were-0.0455?,-0.0438?,and-0.0693?,and the standard deviations were 0.51?,0.8066?,and 0.46?,respectively;a consistent analysis with similar products of OISST and MISST was conducted,and the results showed that the SST was fused.The accuracy of the product is better than OISST,slightly lower than MISST,and the data consistency is better.The larger error area is distributed in the nearshore and the north and south poles.The comparison result with NDBC also reflects this phenomenon.(3)Using the multi-source remote sensing SST fusion product,the Sobel operator and the Canny operator are used to detect the temperature front in the Northwest Pacific Ocean.The detection results are basically consistent with the previous research.It shows that the developed SST fusion product is suitable for the frontal detection research,and the center line of the frontal surface is extracted,and the monthly and seasonal frequency of the frontal plane and the frequency of the year are calculated.Taking the typical frontal fronts such as the Kuroshio front,the Oyashio front,and the Zhejiang-Fujian coastal front as examples,the spatio-temporal change analysis of the front was carried out.