Identifying extreme hottest days from large scale upper air data: a pilot scheme to find California Central Valley summertime maximum surface temperatures

A pilot scheme uses upper air data from a few extreme hottest days to identify those and other extreme hottest days measured by 3 stations sampling the California Central Valley (CV). Prior work showed that CV extreme heat wave onsets have characteristic large scale patterns in many upper-air variables; those patterns also occur for the hottest days. A pilot scheme uses areas of two upper-air variables with high significance and consistency to forecast extreme surface temperatures. The scheme projects key parts of composite patterns for one or more variables onto daily weather maps of the corresponding variables resulting in a ??circulation index?? for each day. The circulation index measures how similar the pattern on that day is to the composite patterns in areas dynamically relevant to a CV extremely hot day, with a larger value for a stronger match and larger amplitude. The scheme is tested on the development period (1979?C1988) and on the subsequent 18?year ??independent?? period (1989?C2006). The pilot scheme captures about half of the rare events in the development period, with similar skill for the independent period. Based only on 16?days of extreme heat in the first 10?years, the scheme is not intended to represent the general distribution; however the circulation index has similar kurtosis, variance, and skewness as the observed maximum temperatures. Properties of the high end tail of the distribution are notably improved by adding the second predictor. The scheme outperforms simply using 850?hPa temperature above the CV.