S3Research Flight Predictor Update
There are 3 major changes.
- All times are now UTC (GMT). The localtime was just too confusing for me. It relied on the timezone set on the PC and that is not always correct. In addition if you were in a different time zone than the launch it got even more complicated. UTC is just more "universal".
- In order to improve accuracy I added an interpolation between the forecast just before and after the current flight time. This is in addition to the tri-linear interpolation of the 8 points around the current balloon location. This now more closely matches the habhub algorithm. There will be some differences in the results because of other differences in the tool. I use the 0.5x0.5 degree forecast vs. Habhub using 1.0x1.0 forecast. Also the update rate is different. I calculate a new point every 15 seconds vs their 50 second cycle time.
NOTE: This requires more weather data to be downloaded so it does slow down the first prediction.
- I added additional entry boxes for up to 3 each burst, ascent, & descent values. This makes it easier to run multiple prediction to create a landing zone window. You could always run multiple predictions without reloading weather data but you had to enter them one at a time.
NOTE: If you enter data in all 9 of the entry boxes it will result in 27 (3x3x3) predictions. It works but the map and table will not be very responsive with that many predictions.
A few other changes are listed below but these are the main ones.
If you try it let me know what you think or if you have any issues. I am still trying to decide if this is worth continuing to develop.
Hope to see you at the Hutchinson GPSL next month.
- All times are now UTC. Local times use time zones set on client and could result in incorrect times being used.
- Better forecast interpolation. Interpolation between forecast times
is also used in addition to tri-linear interpolation of the 8 points
around the balloon. .
Note: This requires more weather forecast data to be downloaded and may slow down the tool.
- Multiple burst, ascent, and descent rates can be entered for multiple forecast variations without re-downloading forecast data.
Note: The data in the first column will be used for overlay and saved predictions.
- Habhub.org predictions (csv) can be loaded for comparison.
- Flight data can be extracted from an aprs.fi or findu.com flight to make running a post flight correlation easier
Join GPSL@groups.io to automatically receive all group messages.