Fig. 7. PCA correlation circle of Table 3. Variation of numeric variables is maximized along two components. Meanwater depth (‘‘depth’’), riverine ASi and DSi (‘‘asisch’’ and ‘‘ dsisch’’) and mean water temperature (‘‘t’’) are shown with ASi and DSi percent delivery (‘‘asi’’, ‘‘dsi’’). Parallel vectors are highly correlated along the two components, while squared vectors are not correlated. Opposite direction indicates a negative correlation. Explicit R-squared and p-values depend on the amount of variation explained by the components.
...read more