**Aim:**
In many circumstances, climate studies involve more than just two time series. Sometimes, spatial maps of a parameter, such as the SLP, are used instead of just one station value. The geographically distributed climate quantities are often stored in *gridded* data files, consisting of ** grid boxes (or points), each which represent the mean value for that particular region.
**

**An example of multivariate analysis is the map of correlation scores between the SLP and Bergen temperature. Such maps can be used to identify teleconnections, and lagged-correlation maps can be used in the study of propagating signals.
**

- The geographical distribution of correlation
- Problem of multiplicity
^{*} - The data matrix and SVD
- Empirical Orthogonal Function analysis
- CCA, SVD and Multivariate regression
^{*} - Common EOFs
^{*} - Cluster analysis
^{*} - Further reading
- Exercises