Diablo Winds in the California Bay Area: Their Climatology, Extremes, and Predictability

Diablo winds (DWs), which are characterized as hot, dry, and intense north-northeasterly winds in Northern California, have been linked to occurrences of several intense firestorms, such as the devastating Wine Country fire in 2017 and the Paradise fire in 2018. Despite DWs' strong linkage to wildfires in Northern California, there have not been many studies dedicated to the DWs, especially compared to its close cousin, the Santa Ana winds in Southern California. Furthermore, a long-term climatology of DWs is still lacking and relationships with large-scale climate variabilities have not been clearly elucidated. This study aims to document the DW climatology with a particular attention to extremes, and to explore the underlying mechanisms of DWs and their predictability from a climatological perspective.

In this study, 3-hourly NCEP North American Regional Reanalysis (NARR) data of September to February 1979-2018 are used for all analyses. Results show that DW events (DWEs) have significant upward long-term trends in frequency and intensity for October as well as in duration for November. In addition, the frequency of DWEs, especially the extreme cases, for September and October has nearly doubled in the last 20-year period (1999-2018) compared to the previous 20-year period (1979-1998). Meanwhile, the opposite is found for January and February. Considering the fact that late fall in California is when abundant dry grass becomes available fuel for DW-driven wildfires, more frequent DWEs at that time imply a higher likelihood of wildfires in the last 20-year period compared to the previous 20-year period. This situation is projected to continue in the future based on climate modeling results under a global warming scenario.

To explore the potential impact of low-frequency and intra-seasonal climate variabilities on the development of DWEs, five daily climate indices (NAO, PNA, WPO, EPO and AO) and the Madden-Julian oscillation (MJO) are examined. For the purpose of predictability, time lags from zero to 12 days before the occurrence of DWEs are considered. Results show that the daily Pacific/North American teleconnection pattern (PNA) and MJO demonstrate a promising potential for DWEs predictability. The underlying mechanisms on how the PNA and MJO affect DWEs are further studied in detail in this study.