Particle layer not working correctly at high zoom level with low uv value

[Kanahe1800]

Issue: Particle movement only vertical at zoom level ~14.8

When the map zoom level is around 14.8, the particle moves only vertically, although it moves correctly at zoom level 14.6.

The first image shows the behavior at zoom 14.8, where the particle movement is incorrect.

The second image shows zoom ~14.6, where the movement works as expected.

The grid layer indicates that the particle should move from top-right to bottom-left, but at zoom 14.8 it moves strictly vertically. This issue mainly occurs in areas with low wind speed at high zoom levels.

[Kanahe1800]

I have noticed that when the wind direction (wd_deg) is uniformly from north to south (i.e., 0° or 360°), no particles or arrows are displayed in the visualization. This results in an empty or blank rendering even though the wind speed (ws) is non-zero.
Interestingly, when I set the wind direction to 180° (southward), the particles and arrows are visualized correctly.

[zakjan]

Hi, particle layer has a precision issue with larger zooms indeed. This is due to that particle positions are stored in world coordinates (to support smooth viewport interactions) but as Float32 instead of Float64. It's also described here weatherlayers/deck.gl-particle#3. There is no quick fix for this. Either Float64 support needs to be added, or a different coordinate system for particle positions needs to be selected. As a temporary measure, Particle layer has a default maxZoom = 15, which was selected for global data, and hides particles in higher zooms. Can you configure a lower maxZoom = 14.5 for your use case?

[Kanahe1800]

Thank you for the detailed explanation. I understand that the particle layer precision issue at higher zoom levels stems from storing particle positions in world coordinates as Float32 instead of Float64, which limits accuracy. I see that this is a known limitation described in weatherlayers/deck.gl-particle#3.

For now, I will configure the maxZoom parameter to 14.5 as a practical workaround for my use case.

Regarding the uniform north-to-south wind problem, this issue occurs consistently across zoom levels and not just at high zooms. It seems to be a separate concern that might require a different approach or fix.

Thanks again for your insights and support!