Show HN: Meteosource – Hyper-local weather API based on improved ML models
I am excited to share the new version of my weather API with you. It offers even more accurate forecasts than before, thanks to updates to our proprietary ML model.
My weather forecast website initially started as a personal project. Later, it became more professional, and recently, we have evolved into an advanced weather API. Our goal from the beginning has been to offer affordable yet superior services.
To achieve high accuracy, we incorporated machine learning into our calculations. Our forecasts pull data from multiple sources and use proprietary numerical weather models. These models also learn from historical data to reduce bias. This applies to hyper-local forecasts as well, where each point is further refined based on surrounding data.
In addition, you will find advanced features like a minute-by-minute forecast for the next hour, updated every ten minutes.
I would be grateful for all your thoughts and any ideas!
If you can predict visibility (in statue miles) and ceiling height (in feet AGL), you can predict the flight category (VFR, MVFR, IFR, LIFR) which would be quite useful for aviation.
Good luck!