Abnormally high mixing heights may lead to the potential for large wildfire growth due to the downward transport of air with relatively high potential temperature and momentum and upward transport of higher humidity air. However, mixing height observations are not readily available outside of 12Z and 00Z at radiosonde observation locations. While work has been done to draw a connection between high mixing heights and the spread of fire, numerical modeling with high-resolution, convective-allowing models could give a clearer sense of the relationship. This study uses the High-Resolution Rapid Refresh (HRRR) model in conjunction with radiosonde observations of the mixing height to assess the HRRR’s ability to correctly forecast the mixing height. If the HRRR can accurately represent the mixing height or if HPBL biases become known, they could be used to give a clearer sense of the relationship between high mixing heights and fire spread. Additionally, it could be used to quickly get a sense of the mixing height in the field, and whether that mixing height was abnormally high when compared to climatological averages.
Data were collected for all radiosonde locations in the continental United States (CONUS) with focus given to the 00Z launches from 2 December 2020 through 26 September 2022. Calculations of mixing height were made using the Stull and bulk Richardson methods. Corresponding HRRR data (HRRRv4) were collected at each station for study. The 00Z Planetary Boundary Layer height (HPBL) from the HRRR was compared to calculated 00Z sounding mixing heights at each radiosonde location. Output from the previous day at 23Z was also compared to these calculated mixing heights, as the radiosondes that produce the soundings are generally launched at this time.
Results demonstrate that the HPBL was similar to, though typically slightly higher than, the mixing height calculated from radiosonde data using the Stull method over central and western CONUS. These similarities were expected as the way the HRRR determines HPBL is similar to that of the Stull method. Larger differences between the methods were found over the east coast. Agreement was less between HPBL and mixing height calculated from radiosonde data using the bulk Richardson method. With the similarities between the HRRR output and the observations, we have increased confidence that the HRRR can be used to represent the mixing height at any time of day.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner