EOFs of FSU winds and BMRC Z20

In all these EOFs, the spatial fields have units of the data, while the time functions are dimensionless. That is, the sum over all EOFs of the spatial functions squared equals the total variance at each location, while the sum over time of each time function squared is the number of time realizations.

EOFs of FSU Winds for 1961-Aug 1999
Input winds were anomalies from the average annual cycle smoothed by a 13-monnth running mean.

Figures showing space and time fields

Tau-X: EOF 1 2 3 4 5
Tau-Y: EOF 1 2 3 4 5
Curl: EOF 1 2 3 4 5
Percent variance of first 5 EOFs of Tau-X, Tau-Y and Curl
Overlay time functions for Tau-X, Tau-y and Curl EOF 1 2 3 4 5
Since time functions for EOF 1 of Tau-X and -Y are so well correlated, show the EOFs as vectors
As long as we're playing this game, form the "Curl" from these vector fields. Compare EOF 1 of real curl

EOFs of BMRC Z20 for 1980-Aug 1999
Input Z20 was anomalies from the average annual cycle smoothed by a 13-monnth running mean.

Figures showing space and time fields

EOF 1 2 3 4 5
Percent variance of first 5 EOFs of Z20