Ensemble clustering has become a popular way for forecasters to visualize large ensemble forecast datasets. Ensemble clustering partitions the members of an ensemble into groups of members that have similar forecasts (i.e., discrete forecast scenarios), allowing forecasters to dig more into the data and get more complete information about forecast uncertainty in space and time without data overload. WPC created a popular ensemble clustering tool to assist forecasters in their preparation of 3-7 day forecasts. This tool groups together ensemble members with similar 500-hPa geopotential height forecasts. Verification has shown that these clusters adequately summarize the variability in temperature forecasts among the constituent ensemble members. However, this methodology can struggle to adequately depict the variability in the precipitation forecasts among constituent members. To address this, an ensemble clustering tool developed specifically for precipitation forecasting was developed. 
1. Evaluate and interpret the first two Empirical Orthogonal Functions (EOFs) of a relevant ensemble forecast variable over a region of interest.
2. Use k-means clustering on the Principal Components (PCs) for the EOFs to create clusters of similar members.
3. View the clusters.