I am trying to understand singular value decomposition analysis. I compared two gridded atmospheric data.
The Mode 1 has 79.5% squared covariance fraction. Modes 2 and 3 have 3% and 2%, respectively. From Mode 1 to 3, the Varf are 39, 12, 5, respectively. Then, there is heterogeneous correlation at 0.5, 0.4, 0.6 for Modes 1 to 3.
My question is how to interpret these values or what are these values/numbers? What is scf and how is it different from varf? What is a heterogeneous correlation and how is it different from the homogeneous one? I get frustrated by the lack of online reference on this.