Bulletin Board

Hi Omar

VRTool worked beautifully for me for Leg 5 of VORG. Now that it's finished here are a couple of minor suggestions you might like to consider:

1. Sometimes the wind interpolation is quite different to the game. This can happen when there is a big direction change across the cell. It may be something to do with the order in which the x and y interpolations are done. In these situations Zezo is always agrees with the game, which suggests a small difference in the algorithms.

2. If we do an early weather update before 6 or 18 UTC the current wind is lost. Then the only way to keep the routing going is to project a position forward to the update hour and start the route from there. This is OK but it's a bit messy and it leaves a 1-2 hr gap in the route. Is there any way to retain the old grid for the current period, while using the new grid for the future periods, so that there is no gap?

mclean wrote:a couple of minor suggestions

mclean, thanks for the suggestions.

1. I will look into this. I suppose the problem is more in the wind direction
than wind speed.

2. When I wrote the NOAA module I set it to use the latest available forecast,
according to NOAA publication schedule. But Games delay wind publication
up to a couple hours. This is why there is a difference between games and vrtool
if you download before the wind shift time. Of course this can be corrected.

omar wrote:.......I suppose the problem is more in the wind direction
than wind speed.

Hi Omar. Correct, the wind speed is fine. It seems to me the problem happens when the direction difference across the cell passes through 180 deg, so the rotation along the line between the end-points has to change direction. When this happens the game produces a big discontinuity in the wind field (like a big wind shift across the line, of up to 180 deg). It's in this situation that the game seems to handle the interpolation differently to VRTool. Lucklily it's rare - for my boat only once in Leg 5, and twice in Leg 3 in the Malacca Straits when the wind was light and all over the place. But is sure messes up the routing if you don't spot it. I hope that makes some sense?!

I posted version 2.02 of vrtool in:

http://www.tecepe.com.br/nav/vrtool/vrtool.zip
md5 checksum 93A500D19B8225ECD8692F57FB45325E

This contains the program executable only (vrtool.exe).
Unzip over the previous version (2.0).
It is supposed to correct two problems with winds:

1-Interpolation was changed to fit VORG game more closely.
I used the method similar to uGrib, described in page:

http://www.grib.us/Forum/tabid/55/forum ... fault.aspx

I also created an object called WindField that shows wind
barbs at multiple points (configurable grid resolution)
This helps to see the interpolation at work.
To create a wind field, specify resolution (p.e. 0.1 degrees),
set the vrtool zoom to the desired viewport and click the button
[Wind field] in the Interpolation panel. The field will be the size of the
viewport.

Tip: To remove a wind field, use Marks > Delete > 'WindField'

2-Game winds downloaded between 4Z and 6Z and
between 16Z and 18Z now use the same forecast as VR and VORG games.
Previous vrtool version downloaded a newer forecast at these times
(different from the game). I hope this is correct now.

Thanks mclean for the help and test data.

There was still a problem in version 2.01 which
I hope is correct in 2.02 (use the download link in previous post)

Game winds are tricky and it took a while to get them right.
The idea is to use the same forecast 00 (current wind) as the game,
but the latest forecast available for 12,24,36 etc (all "future" winds)
Note that vrtool future winds (12,24,etc) are different from game's
most of the time, because they are newer predictions (and probably better).

The table below shows which files are downloaded from NOAA for each forecast:

Code: Select all

time          NOAA dir   file      
0Z-4Z  - latest=18z of previous day (Dt-1)
   00        (Dt-1)12z  file=06
   12        (Dt-1)18z  file=12   
   24        (Dt-1)18z  file=24   
4Z-6Z  - forecast 00z becomes available
   00        (Dt-1)12z  file=06
   12        (Dt  )00z  file=06   
   24        (Dt  )00z  file=18   
6Z-10Z - Game publication 1
   00        (Dt  )00z  file=06
   12        (Dt  )00z  file=18   
   24        (Dt  )00z  file=30   
10Z-16Z - forecast 06z available
   00        (Dt  )00z  file=06
   12        (Dt  )06z  file=12   
   24        (Dt  )06z  file=24   
16Z-18Z - forecast 12z available
   00        (Dt  )00z  file=06
   12        (Dt  )12z  file=06   
   24        (Dt  )12z  file=18   
18Z-22Z - Game publication 2
   00        (Dt  )12z  file=06
   12        (Dt  )12z  file=18   
   24        (Dt  )12z  file=30   
22Z-24Z- forecast 18z available
   00        (Dt  )12z  file=06
   12        (Dt  )18z  file=12   
   24        (Dt  )18z  file=24   


So, if you are using weather routing, you should update
NOAA winds right after the following times: 4Z, 6Z, 10Z, 16Z, 18Z, 22Z (Z=GMT)
and recalculate your route. Winds for 4Z and 16Z are particularly important, because
the 12h forecast is an accurate preview of winds that will be used in the game
two hours latter...

Great work Omar, that is a very nice improvement. Thanks for your efforts.

Yet another version is available for download (2.03)
This time, I corrected a couple problems in the boat data for VORG:

• corrected the value of wind angle for your boat (note that this is calculated locally, so it requires winds loaded)
• Added wind angle and wind speed for friends, followed and top 10 boats
• Added color to the Wind Field object (based on wind speed)

http://www.tecepe.com.br/nav/vrtool/vrtool.zip
md5 checksum 9A7B0A67A448371C8BB3C5CC23312D33