Significant Changes to Weather Balloon Launches--Impacts
The text on the pictures isn't showing up well. The image on the left is from Mar 2025, and the image on the right is from May 2026. The bottom right images is an overlay of 18Z/UTC launches (orange) on top of 12Z/UTC launches (blue) from that May date, showing that offices aren't launching early morning balloons in many locations, opting for late morning launches because of more favorable staffing.
Taking a look at weather balloon launch coverage from 2025 from the SPC observed sounding analysis page and comparing it to recent launch summaries, it's easy to see that weather balloon launches are completely different from a year ago. Balloon launches are being skipped in the early morning hours (12 UTC/4am PT)) at many NWS offices, likely due to minimal staffing there at night. To keep the "two-balloon-launches-a-day" stat, those offices are launching them at 18 UTC/10am PT, later in the morning when more staff is available. 2 balloon launches a day, at varied times, represents degraded service compared to the global standard of 2 balloon launches a day at standard 12Z and 00Z times. That's why the World Meteorological Society guidance (14. WMO-1160-2023-Upd-2024_en.pdf) states in 5.3.8 "Upper‑air synoptic observations shall be made and reported at least at 0000 and 1200 UTC." When all sites launch at the same time across the world, the models ingest a clearer picture of the state of the atmosphere, resulting in more accurate model forecasts. If this weren't the case, the NWS wouldn't request additional balloon launches upstream from a hurricane or other severe weather.
Additionally, standardized balloon launches allow for a historical scientific record, so that people can access this data at sites like this (Storm Prediction Center Sounding Climatology Page) and understand how significant current conditions are, compared to standard times in the past. You'll note there are no options for 18Z/UTC soundings, because 12Z/UTC and 00Z/UTC are the standards.
To state that staffing has nothing to do with these significant balloon launch changes is very misleading. And I believe that these varied balloon launches are a significant degradation of service, both for model accuracy and for historical records. I'm curious what others think here.