Spatial And Temporal Distribution Characteristics Of Sea Surface Temperature Based On Ensemble Kalman Filtering Fusion In The Northwest Pacific Ocean

The sea surface temperature is one of the basic ocean physical parameters.The change of sea surface temperature will affect the marine climate and the distribution of fishery resources.Therefore,the research on the variation of long-term sequence sea surface temperature is particularly important for us to predict the changes of climate and marine fishery resources.Satellite remote sensing is one of the effective means to monitor sea surface temperature.There are many remote sensing data,choose suitable methods for multi-source remote sensing data fusion is very important for improving the accuracy and coverage of sea surface temperature.Sea surface temperature(SST)provides by hybrid coordinate model(HYCOM)was the background field for data fusion in this dissertation,AMSR2 SST and AVHRR SST in November 26,2012 in Northwest Pacific Ocean region(120°~165°E,0°~45°N)was used for ensemble Kalman filter fusion,and the accuracy of fusion results were validated.The image quality of fusion results were analyzed by the mean,variance,entropy and gradient of fusion results,and Word Ocean Database(WOD)was applied to verify the accuracy of the fusion results.The result even contained more information.Its variance and entropy are37.7960℃~2 and5.0774 respectively.The gradient is 0.2916℃/0.25°,which is higher than the original data.It means that the merged SST has better accuracy and more details.The result’s maximal absolute error and the average absolute error are 1.1919℃and 0.2848℃respectively.Its root mean square error and the average relative error are 0.3762℃and1.47%respectively.And the result of Ensemble Kalman Filter has a higher quality both in details and accuracy compares with the result of Optimal Interpolation.After validating the fusion results of ensemble Kalman filter temperature,we used the ensemble Kalman filtering method to fuse monthly AMSR2 SST and AVHRR SST from January 2013 to December 2017 in Northwest Pacific Ocean,and the fusion results were transformed into the form of anomaly.Then we used the empirical orthogonal function for further analysis of temporal and spatial characteristics.The results showed that the sea surface temperature in the northern part of the Northwest Pacific Ocean is more intense than that in the southern.In the time distribution,from 2013 to 2015,sea surface temperature in the Northwest Pacific Ocean was cooling down,from 2015 to2017,the time coefficient mainly for positive anomaly,overall sea surface temperature was rising.The second mode in spatial distribution showed,35°~45°N,130°~140°E waters were the high positive anomalies,consistent with the area of Western Pacific Ocean subtropical,and there was a strong correlation between the second mode of time coefficient and SOI(Southern Oscillation Index).The third mode spatial coefficient had high positive anomalies in the 30°~45°N,145°~160°E,other areas were negative anomalies,North Pacific Ocean westerly drift region may had a relationship with the third mode spatial coefficient.In addition,the quarterly sea surface temperature gradient in the Northwest Pacific Ocean showed that the area in the north of 20°N was higher than that of others,and the gradient was the most small in the third quarter in the past 5years,while the largest in the first quarter.