Hi everyone,
I normally use Windy.com for day-to-day weather monitoring because part of my work involves keeping track of regional conditions in Ecuador and Argentina. We run a small streaming app website, and our traffic spikes or declines depending on storms, outages, and local weather events, so having accurate short-term forecasts really matters for planning.

Recently I’ve noticed something odd:

In Ecuador (coastal & Andean regions): The platform sometimes predicts heavier afternoon cloud buildup or rainfall, but on several days the actual conditions stay mostly clear.

In Argentina (mainly Buenos Aires and Córdoba): The opposite occurs — Windy occasionally shows minimal clouds, but local observations and ground reports show sudden overcast periods or brief showers.

To troubleshoot, I compared ECMWF, GFS and ICON layers, and the differences between them can be pretty significant in these areas.

So I’m wondering:

Has anyone else in South America noticed similar inconsistencies between model forecasts and real conditions on Windy?

Are certain models more reliable for Ecuador or Argentina specifically?

Are there recommended settings or layers (e.g., CAPE, humidity, satellite infrared) that help interpret these rapid-weather-shift regions more accurately?

I’m not reporting an error — just trying to understand whether these fluctuations are normal for the terrain, or if I should rely on different layers when planning our internal app-related operations.

Would appreciate any insights from users in the region.

Hi everyone,

I’ve been monitoring wind conditions on Windy for scheduling some outdoor recordings we use on our streaming website, but I’ve noticed that the forecasted gust levels (especially on ECMWF) often differ quite a bit from what we measure on-site. In a few cases, Windy predicted moderate gusts while our equipment sensors barely picked up anything. Other times, the opposite happened, and unexpected strong gusts forced us to delay a planned shoot.

I’m trying to figure out whether this variation is normal for short-range wind predictions or if there are particular models or layers that tend to be more reliable for open outdoor areas. Has anyone here noticed similar inconsistencies or found a model that performs better for real-world equipment setups?

Any advice would really help!