Hi HN, I started running an amateur weather forecast website 20 years ago as a hobby, and later turned it into a professional service during my studies. More recently, we built an advanced weather API after noticing that most existing APIs simply repackage NOAA data and miss key variables.
To improve accuracy, our forecasts pull from multiple sources and use proprietary numerical weather models, learning from historical data to reduce bias. This includes hyper-local forecasts, where each point is calculated based on surrounding data.
To improve accuracy, our forecasts pull from multiple sources and use proprietary numerical weather models, learning from historical data to reduce bias. This includes hyper-local forecasts, where each point is calculated based on surrounding data.