The Quantitative Study Of Sea Surface Temperature-Based On The Time Series Analysis

The sea surface temperature is an important parameter for monitoring ocean phenomenon,it has a great influence and effect on the marine ecosystem,has important research value,which is widely used in the fields of ocean dynamics,air sea interaction,fishery economy research and pollution detection.Many previous researches on sea surface temperature in China and abroad were based on remote sensing inversion and remote sensing data reconstruction,and did not make further research.The forecasting methods of sea surface temperature are mainly based on empirical method,numerical forecast method and statistical forecast method.The accuracy of these methods is limited.Because of the influence of various factors on the sea surface temperature,the time series of sea surface temperature showed obvious seasonal variation characteristics.In this paper,the method of time series analysis is used to study the sea surface temperature and its prediction in the East China Sea,the Gulf of Taiwan,the Hangzhou Strait and the South China Sea.The research content mainly includes the following three aspects:First,the sea surface temperature time series pretreatment.The first is the clustering analysis of the sea surface temperature time series,and the samples with high similarity are clustered into one class.576 samples in each study area are divided into two classes,taking the sample points as an example to study the Sea surface temperature.In this paper,we put forward the idea of climate month,The time series data are averaged by monthly climate,which improves the accuracy of prediction.Second,from the point of view of time domain analysis,the monthly mean sea surface temperature time series data after model identification,model estimation and model diagnosis test,established the SARIMA model and forecast.Third,from the perspective of frequency domain analysis,spectral analysis of monthly mean sea surface temperature time series data,observing its periodogram,establishing the corresponding latent period model or mixed latent period model and making prediction.The monthly mean sea surface temperature(SST)from March 2010 to February2011 is predicted by the fitted SARIMA model and the latent period model respectively.By comparing the actual value with the predicted value,it is found that the two forecasting methods have higher prediction precision,Which can provide reference for the study of sea surface temperature in these areas.