摘要
基于分段变维分形方法,本文提出加权分形的概念,即对某一区段内的数据点的r坐标乘以不同的加权系数,使这些数据点位于双对数坐标上的一条直线上.应用加权分形分析了西北太平洋30°N,125°E坐标点处的月平均海面水温资料,并根据该点某年1～8月的月平均海面水温(八点法)预测该点该年9～12月的月平均海面水温,或仅根据该点某年8月的月平均海面水温(一点法)预测该点该年9～12月及下一年1～7月的月平均海面水温.还根据分形相关现象用该点的水温分形参数和某些点8月的月平均气温预测这些点9～12目的月平均气温.
Based on sectioned variable dimension fractals, this paper presents the concept of weighted fractals .i.e. , for the data points in an interval, their r coordinates multiply by different weighted coefficients, μ making these data points locate at a straight-line on the double logarithmic coordinaters. By using weighted fractals, the monthly average sea surface temperature (MASST) data on the point 30°N, 125°E of Northwest Pacific are analysed. According to the MASST from January to August in a certain year, the MASST from September to December of that year has been forecasted by using eight-point-method. According to MASST of August merely in a certain year, the MASST from September of that year to July of the next year has been forecasted by using one-point-method. According to the phenomenon of fractal interrelation and the fractal coefficients of this point's MASST and the monthly average air temperature (MAAT) of August of some points, the MAAT of these points from September to December has been also forecasted.
出处
《海洋通报》
CAS
CSCD
北大核心
1996年第2期69-76,共8页
Marine Science Bulletin
关键词
海平面
水温
加权分形
weighted coefficient
sectioned variable dimension fractals
analysis
sea surface temperature
air temperature
forecast
