Trajectory forecast vs track flown: a case study

During the recent Sagrantino Cup event I created daily trajectory forecasts with the windy.com traj plugin for the morning flights and asked pilots to share actual GPS tracks so I could compare actual versus forecast. Here is one example:

ECMWF forecast versus actual track – click to enlarge

The KML file for this screenshot is available here, and it is instructive to play with it – have a look at it yourself in Google Earth.

  • Most of the flight was between 800m and 1000m altitude, so I selected the 2000ft (amber), 2500ft (red) and 3000ft (blue) trajectories.
  • The ECMWF model had the best match, so only ECMWF forecasts are shown.
  • The flight started on the valley floor, within an inversion and a slight northerly flow.
  • The match is excellent – especially with the 2500ft trajectory, and the descent phases around 11km and 15km into the flight show the northeasterly flow forecast by the amber line.
  • As to be expected, the track is shorter than the forecast – slightly more than half the distance.
  • my gut feeling after comparing several forecasts to tracks: ECMWF wind speed forecasts are a tad on the high side.

Here is a second example:

  • trajectory colors: red: 2500ft, blue: 3000ft, purple: 5000ft
  • most of the flight was 1000-1200m altitude
  • the track’s match with the 3000ft trajectories (all models) again is very good!

Author: mhaberler

nethead emeritus