November 10, 2020 Meeting

The second meeting of the 2020-2021 year was held virtually via Google Meet. Our guest speaker was Dr. Jason Sippel of the Hurricane Research Division at NOAA’s Atlantic Oceanographic and Meteorological Laboratory.

Meeting Minutes

Call to Order
The second meeting of the 2020-2021 AMS/NWA chapter occurred on November 10, 2020 virtually via Google Meet. The meeting was called to order at 6:06pm by President Eric Carpenter.

Recording Secretary Joanne Culin took note of the number present. In total, there were 13 attendees.

Minutes Approval
Given the virtual format of the meeting, we did not cover minutes from the previous meeting.

New Business
David Cox gave a quick reminder of the treasury amount, which still remains at $452.29. He reminded everyone that dues would not be collected this year while we are doing virtual meetings.

Joanne then introduced the speaker, Dr. Jason Sippel. Dr. Sippel works at the Hurricane Research Division at NOAA’s Atlantic Oceanographic and Meteorological Laboratory in Miami, FL where he is the data assimilation team lead. He is a graduate of Texas A&M University and has previously worked at NASA’s Goddard Space Flight Center as well as the NOAA Environmental Modeling Center. Most of his current career is dedicated to contributing to operational HWRF upgrades as well as improving the use of reconnaissance data in operational NWP. He has also taken part in a number of hurricane reconnaissance flights, including hurricanes Ike, Karl, Harvey, Michael, and Dorian, and is one of the few people in the world who has seen the data from its point of origin on a flight through its operational assimilation. His talk was on the use of reconnaissance data in weather forecast models.

He discussed the different instruments used in recon data. These included flight-level instruments that measure winds, temperature and humidity as observed by the plane. Different entities fly the recon missions in different flight patterns, such as a butterfly pattern that the NOAA P3 flies. Then, dropsondes are used to measure the winds, temperature and humidity in the storm. A dropsonde can even go all around an eyewall. Some planes have tail doppler radar which estimates 3-D winds and also SMFR which estimates the surface wind speeds below the plane.

Overall, the hurricane hunters from the USAF 53rd WRS perform the bulk of the recon flights and are tasked by NHC to gather center fixes. They send real-time data including dropsondes, flight level observations and wind speeds and rain rates. NOAA also uses a G-IV and 2 P3 aircraft. The G-IV is used in environment and near storm while the P3 is used mostly in storm. Both aircraft transmit dropsonde, flight level winds, SFMR and doppler radar data.

In 2018, there were nearly 120 missions into 15 tropical systems. Over 1900 dropsondes were deployed and these sondes cost around $800 each. At one point during the season, there were simultaneous operations in Hawaii, Caribbean and the United States East Coast. During 2019, there were over 120 missions into 13 tropical systems with almost 2700 dropsondes deployed. There were over 50 flights into Hurricane Dorian, which accounted for just about half the flights for the whole season. Dropsondes improved the track and intensity forecasts for Dorian by 15%.

Dr. Sippel then provided a history of using reconnaissance data in models. The use of dropsondes in tropical cyclone model forecasts has occurred since 1997. Starting in 2008, it was known that using Doppler radar velocity data had potential for forecast improvement. Using this radar data significantly improved analysis and forecasts for Hurricane Humberto, which otherwise had not been forecast well from NHC. Subsequent work shows that forecast improvements occurred from assimilating Doppler velocity from reconnaissance and there began a dedicated effort to use this operationally. Hurricane Ike was an example of how this was done in real time. Doppler radar data began being assimilated in the HWRF in 2013. For weaker storms, there was substantial improvement in the HWRF but it is hard to make improvements with major hurricanes as compared to weaker storms. Nonetheless, recon has a clear positive impact on intensity with about 10% improvement through 72 hours.

Dr. Sippel proceeded to discuss current and ongoing developments. The G-IV has started two complete circumnavigations at 90 and 180nm which will likely have a stronger impact on track forecasts than that from distant dropsondes. Near vortex data also helps constrain the vortex structure. It will capture the outflow layer of the hurricane better which will help with intensity forecasts. The use of end point dropsondes also has started to occur rather than center fixes. Taking dropsonde measurements at the end points of the alpha pattern from C-130 missions began experimentally in 2017 and this has had a 10% impact on intensity skill and was implemented operationally in 2018. Unfortunately, this requires more dropsondes to be used but hopefully this will outweigh the cost with better forecasts and lives saved. NOAA has acquired another high altitude jet to replace or supplement the G-IV and there are smarter environmental targeting plans. There has been testing done with dropsondes, examining the impacts of inner core, in-vortex and environmental drops. Work has shown that drops improve intensity errors by 10% in the watch/warning time periods.

Finally, Dr. Sippel discussed the model performance, specifically with the HWRF. HWRF intensity errors have decreased significantly over the past decade. There is huge skill over water near the CONUS. Here, there is plenty of data and recon. He then went through storms from this year: Hanna, Laura, Zeta, Delta and Sally and spoke about how well the HWRF did. For Laura, once the track forecast got better, the intensity was spectacular. For Delta, it got the rapid intensification correct but the track was off over the Yucatan. Once this was corrected, the intensity forecast was good. For Zeta, there were big track problems early but the rapid intensification was good. It showed a couple of days out that the landfall could be of a category 2 storm, potentially a cat 3(which it made landfall just shy of a category 3). They are also working on real time monitoring of data that goes into the HWRF.

The meeting concluded around 7:32pm.

Minutes were submitted by Joanne Culin, Recording Secretary.