Our presentation focuses on the large scale meteorological patterns (LSMPs) that persist before and during the onset of northern California (USA) extreme precipitation (PEx) events. An LSMP is more than an average of a variable across extreme events because it also shows those parts of the pattern that are both significant and consistent. Hence, LSMPs are useful for forecasting, predictability, model assessment, and simulating future changes of PEx. PEx events can occur for different weather patterns, so we apply k-means clustering on the two leading principal components of daily 500 hPa geopotential height anomalies (Za500). We shall compare and contrast use the longer-period NOAA-CIRES-DOE 20th Century Reanalysis V3 (based on 2-D input data) with shorter-period, widely-used reanalyses (e.g. ERA5, based on 3-D input data). Using different reanalyses improves our confidence in the LSMPs by quantifying uncertainty. Significant areas in the cluster mean fields are identified using bootstrapping while consistency is assessed using sign counts (as in prior works). The optimal number of clusters is three, as determined by a Monte Carlo experiment of statistical significance. Plots of the LSMPs at PEx onset are quite similar near California, but differ greatly elsewhere and at times before the onset. Accordingly, we found that applying the cluster analysis to data two days before PEx onset efficiently separated the three clusters. We shall show animations of how each cluster mean evolves. The Cluster 1 Za500 LSMP has a ridge-trough-ridge pattern that evolves from a southwest-northeast orientation rotating counter-clockwise to northwest-southeast over a week. This cluster has the strongest link to the tropics. The Cluster 2 Za500 LSMP develops rapidly over a few days into a trough over the Gulf of Alaska bounded by a horseshoe-shaped ridge to the south that extends from the western North Pacific to eastern North America. The Cluster 3 Za500 LSMP is a middle to high latitude wave train undergoing downstream development from a Siberian low and Alaskan high more than a week before PEx onset. We shall discuss the contrasting synoptics of these three clusters using Za500 and other relevant variables, like: outgoing longwave radiation, winds, moisture, and moisture transport anomalies.