论文部分内容阅读
Principal component analysis (PCA) as used by meteorologists and oceanographers is a powerful tool for analysis of the spatial and temporal variability of physical fields.This study was aimed at applying “quasi-local PCA for singular factor” to make the cumulative percentage for the first principal component as great as possible, so that a multi-dimensional problem can be reduced to a single-dimensional one, and then stepwise regression analysis can be used to parameterize the relationship between El Nino events and the hydrographic factor anomalies along 137 癊 in winter.The results show that the salinity anomalies on 30 m level, the sigmat on 250 m level, and the temperature on 300 m level at 8 癗 are most closely related with El Nino events because of thermocline movement caused by enhanced upwelling in this area during El Nino years.
Principal component analysis (PCA) as used by meteorologists and oceanographers is a for tool for analysis of the spatial and temporal variability of physical fields. This study was aimed at applying “quasi-local PCA for singular factor ” to make the cumulative percentage for the first principal component as great as possible, so that a multi-dimensional problem can be reduced to a single-one, and then stepwise regression analysis can be used to parameterize the relationship between El Nino events and the hydrographic factor anomalies along 137癊 in winter.The results show that the salinity anomalies on 30 m level, the sigmat on 250 m level, and the temperature on 300 m level at 8 癗 are most closely related with El Nino events because of thermocline movement caused by enhanced upwelling in this area during El Nino years.