Processing Of Surface Temperature Data In East China Sea Based On Kriging Interpolation And Effect Analysis

Water temperature, one of the major elements in hydrological survey, is characterized with two fundamental features of space and time. Due to limiting survey conditions, systematic or random error in data and low sampling density, data is usually not uniformly distributed in terms of space. As to survey time, it is difficult to obtain data from all locations simultaneously. These bring forward the necessity of interpolating the original data and analyzing among specific period of time. Therefore, this paper will focus on temperature data interpolation and the methods of ordinary Kriging and Co-Kriging are introduced into the processing and analysis of sea surface temperature data and put forward a new Kriging based sea surface temperature data processing and analysis method, which is applied in processing surface temperature data in East China Sea and the results are discussed.The average monthly temperature data in experimental area of East China Sea for the months of January, February and March is used in this paper. The experimental data is rarefied and decomposed into physical trend component and residual component which are calculated. The sum of two estimated components is finally used to realize gridding interpolation on the sea surface temperature data.For univariate interpolation on single time domain data, three methods of trend analysis, moving average and ordinary Kriging are adopted and it is discovered that the best effect of interpolation is with Kriging method. For multivariate co-interpolation on single time domain data, co-Kriging method is adopted for interpolation, the result obtained is compared with that of single variable ordinary Kriging interpolation on single time domain data, and the former is better in effect. For adjoining time domain, co-Kriging interpolation in staggered gridding is better in effect than single time domain ordinary Kriging interpolation in fixed gridding. Keywords: ordinary Kriging, Co-Kriging, SST, staggered survey, local variogram fitting...