Topics

Flight/Prediction Correlations

Jerry
 

I am still trying to figure out how to do a better prediction.  To that end, I grabbed 5 of today's flights from aprs.fi & findu.com.  I imported them into my too and had it extract the details from the data.  Using that data I then ran a prediction using that data.  This created a prediction that should have all the ascent, descent, and burst errors removed.

Below is the basic info from this.  You can see the screen capture from this here. The csv files are there as  well if you want to play with them.

I don't see a pattern to the errors.  Most predictions were long but one came up came up 11 miles short.

If you see anything that might help reduce the error, please let me know.  About the only thing I can see would be to create a tool that corrects the prediction while the balloon is in flight.  I don't know if anyone would be interested in a new tool that does that or not.  Of course it would have to work offline if required.

Weather used: 2017061706  This forecast will be available for a few weeks if you want to rerun any of this.

Here are the basic info:
Callsign: N9XTN-11
    Burst: 98976 Feet
    Ascent: 873 F/M
    Descent: 998 F/M
    Error Distance: 3.55 Miles
    Prediction was Long
   
Callsign: N5XUL-11
    Burst: 96880 Feet
    Ascent: 640 F/M
    Descent: 1345 F/M
    Error Distance: 10.77 Miles
    Prediction was Short

Callsign: KE0BMV-11
    Burst: 98559 Feet
    Ascent: 733 F/M
    Descent: 1330 F/M
    Error Distance: 3.52 Miles
    Prediction was too far south

Callsign: KD4STH-13
    Burst: 103966 Feet
    Ascent: 1090 F/M
    Descent: 896 F/M
    Error Distance: 7.93 Miles
    Prediction was Long

Callsign: KD4STH-7
    Burst: 96890 Feet
    Ascent: 1136 F/M
    Descent: 699 F/M
    Error Distance: 6.15 Miles
    Prediction was Long




 
Jerry Gable
Balloon Flight Prediction tools
http://www.s3research.com

Bill Brown
 

Here's the prediction I did via the Habhub predictor the night before the flight for the KD4STH-13 and WB8ELK-11 balloon. It was 6 km south of the actual landing but right on the money in the east-west position.

- Bill WB8ELK

I'm wondering how accurate the upper level winds are (above 86,000 feet). That might make a difference due the westward track in the stratosphere.




-----Original Message-----
From: Jerry Gable jerrygable@... [GPSL]
To: GPSL List ; Mark Conner
Sent: Sat, Jun 17, 2017 7:26 pm
Subject: [GPSL] Flight/Prediction Correlations

 
I am still trying to figure out how to do a better prediction.  To that end, I grabbed 5 of today's flights from aprs.fi & findu.com.  I imported them into my too and had it extract the details from the data.  Using that data I then ran a prediction using that data.  This created a prediction that should have all the ascent, descent, and burst errors removed.

Below is the basic info from this.  You can see the screen capture from this here. The csv files are there as  well if you want to play with them.

I don't see a pattern to the errors.  Most predictions were long but one came up came up 11 miles short.

If you see anything that might help reduce the error, please let me know.  About the only thing I can see would be to create a tool that corrects the prediction while the balloon is in flight.  I don't know if anyone would be interested in a new tool that does that or not.  Of course it would have to work offline if r equired.

Weather used: 2017061706  This forecast will be available for a few weeks if you want to rerun any of this.

Here are the basic info:
Callsign: N9XTN-11
    Burst: 98976 Feet
    Ascent: 873 F/M
    Descent: 998 F/M
    Error Distance: 3.55 Miles
    Prediction was Long
   
Callsign: N5XUL-11
    Burst: 96880 Feet
    Ascent: 640 F/M
    Descent: 1345 F/M
    Error Distance: 10.77 Miles
    Prediction was Short

Callsign: KE0BMV-11
    Burst: 98559 Feet
    Ascent: 733 F/M
    Descent: 1330 F/M
    Error Distance: 3.52 Miles    Prediction was too far south

Callsign: KD4STH-13
    Burst: 103966 Feet
    Ascent: 1090 F/M
    Descent: 896 F/M
    Error Distance: 7.93 Miles
    Prediction was Long

Callsign: KD4STH-7
    Burst: 96890 Feet
    Ascent: 1136 F/M
    Descent: 699 F/M
    Error Distance: 6.15 Miles
    Prediction was Long




 
Jerry Gable
Balloon Flight Prediction tools
http://www.s3research.com