Dear @bczernecki,
Thank you for your work on developing and maintaining the Thunder library. The combined Skew-T, hodograph, and parameter plot functionality is excellent, and I sincerely appreciate the effort that went into creating it.
While testing the library, I encountered a potentially unexpected behavior. This may be a bug, or it could result from a misunderstanding on my part regarding the documentation. I would greatly appreciate your clarification. Additionally, I’ve noticed that in some of my plots, the height axis occasionally has missing labels - I'll report that separately in another issue.
Description of the Issue
The radiosonde we use provides data in geopotential height above sea level. According to the Thunder documentation, users can provide either height above ground level or altitude above sea level as input for generating Skew-T diagrams, hodographs, and parameter tables.
I tested four different approaches for providing altitude data:
- PLOT 1: Geopotential altitude above sea level (raw altitude data from the radiosonde)
- PLOT 2: Geopotential altitude above ground level (raw altitude data minus the lowest level)
- PLOT 3: Geometric altitude above sea level (approximated from geopotential altitude)
- PLOT 4: Geometric altitude above ground level (geometric altitude minus the lowest level of geometric altitude)
Observations
- PLOT 1 vs. PLOT 2: No differences in plots or calculated parameters, which is expected and aligns with the documentation.
- PLOT 1 vs. PLOT 3: Minor differences in both plots and parameter values due to the small, expected difference between geopotential and geometric altitude. The change in the "mean wind Sfc–2 km" parameter is particularly noticeable.
- PLOT 3 vs. PLOT 4: The plots are visually identical, but many parameter values change—most of them becoming identical to those in PLOT 1.
This last result seems inconsistent. Since the only difference between PLOT 3 and PLOT 4 is the subtraction of the initial level (just as in PLOT 1 vs. PLOT 2, where no such change occurred), I would have expected no difference in the calculated parameters.
Request
Could you please clarify whether this behavior is expected? Is there a difference in how the library internally handles geometric versus geopotential heights that could explain this inconsistency?
I’ve attached a zip file with radiosounding .cor file along with the code used to reproduce the plots. I've also attached plots in svg format as i have some troubles exporting them as png (graphics package issue allready discussed in closed issues - "A rectangle appears around the wind barbs profile")
If I can help further by clarifying the issue or providing additional information, please don’t hesitate to let me know.
Kind regards,
Andrej Iskra
thunder_radiosounding_bug.zip
Dear @bczernecki,
Thank you for your work on developing and maintaining the Thunder library. The combined Skew-T, hodograph, and parameter plot functionality is excellent, and I sincerely appreciate the effort that went into creating it.
While testing the library, I encountered a potentially unexpected behavior. This may be a bug, or it could result from a misunderstanding on my part regarding the documentation. I would greatly appreciate your clarification. Additionally, I’ve noticed that in some of my plots, the height axis occasionally has missing labels - I'll report that separately in another issue.
Description of the Issue
The radiosonde we use provides data in geopotential height above sea level. According to the Thunder documentation, users can provide either height above ground level or altitude above sea level as input for generating Skew-T diagrams, hodographs, and parameter tables.
I tested four different approaches for providing altitude data:
Observations
This last result seems inconsistent. Since the only difference between PLOT 3 and PLOT 4 is the subtraction of the initial level (just as in PLOT 1 vs. PLOT 2, where no such change occurred), I would have expected no difference in the calculated parameters.
Request
Could you please clarify whether this behavior is expected? Is there a difference in how the library internally handles geometric versus geopotential heights that could explain this inconsistency?
I’ve attached a zip file with radiosounding .cor file along with the code used to reproduce the plots. I've also attached plots in svg format as i have some troubles exporting them as png (graphics package issue allready discussed in closed issues - "A rectangle appears around the wind barbs profile")
If I can help further by clarifying the issue or providing additional information, please don’t hesitate to let me know.
Kind regards,
Andrej Iskra
thunder_radiosounding_bug.zip