Synoptic and Mesoscale Characteristics of Extreme Heat Events in Southern California

Southern California has periodically received extreme heat events during summer where temperatures reached or exceeded 46 degrees C in the western valleys (west of the peninsular mountain ranges, including the Inland Empire and San Diego County valleys) and/or 50 degrees C in the lower deserts (including the Coachella Valley). Extreme heat had a higher frequency between 2016 and 2018. This included 6 July 2018 when five locations, including official climate stations and ASOS locations, set or tied all-time record maximum temperatures for any date of the year. These events were associated with strong upper tropospheric ridges and anomalously high temperatures in the lower to mid troposphere. However, temperatures aloft and pressure heights were insufficient at predicting the surface temperatures as surface and lower-tropospheric wind flow was a critical factor, especially for events west of the mountains. The Weather Research and Forecasting (WRF) model, run by the National Weather Service in San Diego, California with 3.7-km grid spacing and used to examine these events from runs that used real-time and reanalysis data, showed that during the most extreme heat events of the western valleys of Southern California, a surface convergence zone 20-40 kilometers inland from the coast separated temperature-moderating westerly coastal flow from downsloping easterly flow with compressional heating and a deep mixed layer (typically to 4000-4500 meters above sea level). Along and east of the surface convergence zone was where the extreme heat occurred. Surface and lower-tropospheric relative humidity was also a factor, but less so than the winds, and may be more of a result of the heat due to deep tropospheric mixing and large temperature-dew point spreads than a cause. Model output statistics (MOS) for the Global Forecast Systems (GFS) and North American (NAM) models had widely variable performance during these heat waves, with some maximum temperatures successfully predicted within 1 degree C, but others with errors of as much as 8 degrees C.