Wind interpolation modified?

[climbator]

Hi all,

recently I see some quite strange wind interpolation values in the VR game with no equivalent setting in VRTool, especially in situations with extremely circulating winds.

As an example, the current (02/23 20h CET) 1° wind grid shows the raw sampling points
18°S 163°E: 7.0kn/234°
18°S 164°E: 6.5kn/1°
in both the VR game and in the VRTool 1° NOAA grid.

However, the exact intermediate point (interpolated) shows
18°S 163.5°E: 6.1kn/293° (VR game)
18°S 163.5°E: 6.7kn/293° (VRTool, bilinear, U/V interpolation)
18°S 163.5°E: 6.7kn/298° (VRTool, bilinear, no U/V interpolation).

I've seen the same effect (less wind in the VR game) in many situations now, which messes up the entire route planning. Is this a VR game tool bug or a misconfiguration of VRTool?

[climbator]

Hi all,

a similar question was raised in the zezo forum: http://sail.zezo.org/forum.pl?tid=5716. Zezo interpolates the winds exactly as in the VR game. According to the zezo site programmer, zezo uses
- direction: bi-linear u/v (same as VRTool)
- speed: "weird formula on u/v, average and magnitude interpolation" (different from VRTool)

I would highly appreciate if this "weird formula" would be available in VRTool too. Thanks in advance!

[omar]

speed: "weird formula on u/v, average and magnitude interpolation"

I found the formula changes little when compared to normal magnitude interpolation.
Changes are larger in areas with high vorticity.

In the image below, I inverted the direction of the 25.4 wind arrow to create an extreme situation.
The left side uses regular magnitude interpolation and the right side uses the weird formula.
While magnitude interpolation keeps the magnitude (speed) of the wind vector when
we move between two arrows, the formula goes to 0 and back.

Which one looks better ?

[climbator]

Hi Omar,

thank you for your fast response. I guess the rationale behind the "weird formula" is to obtain a more natural wind field of real-world cyclones (zero wind in the eye of the cyclone).

Below is a snapshot of the current wind grid. The magnitude of the wind field with bilinear interpolation (VRTool, left image) is always above 4.4 knots, even in the eye of the cyclone. As a consequence, VRTool tends to route right trough the eye of the cyclone. On the contrary, the VR game (right image) applies the "weird formula" with near-zero winds in the vicinity of the eye (actuallay, there is a lower limit of 0.5 knots).

Consequence: Applying the routing suggested by VRTool to the actual VR game may result in beeing trapped for hours in a near-zero wind hole. It happend to me a few times over the last days.

[omar]

..trapped for hours in a near-zero wind hole..

Not so happy...

The figure below illustrates what u say, sailing on a cyclone.
Pink lines using magnitude interpolation.
Blue lines using the formula.

[climbator]

Excellent illustration, this is exactly what I experienced several times: Too optimistic route prediction in VRTool (pink) leading to disastrous performance in the VR game (blue). Unfortunately, I was extremely successful in stumbling into those rare pitfalls over the last few days.

BTW: Is it possible to upload images to the forum? Since I couldn't figure out how, I just added links to an external server. The temp folder hosting these images might be deleted in the future.

[omar]

see viewtopic.php?f=2&t=3103&p=4957#p4957