Skip to content

Hybrid filter for post-processing forecasts provided by NWP models.

License

Notifications You must be signed in to change notification settings

aliakatas/Kalman_Bayesian_filter

Repository files navigation

Hybrid Kalman-Bayesian filter

Hybrid filter for post-processing forecasts provided by Numerical Weather/Wave Prediction (NWP) models. This work is based on this publication: A hybrid Bayesian Kalman filter and applications to numerical wind speed modeling

It implements each module of the combined system, as well as the integrated version of the full Hybrid filter. The aim of this tool is to remove systematic and random errors from model predictions as much as possible, leading to more accurate forecasts.

Currently, it is able to handle data following Normal, Log-normal and Weibull (min) distributions. The error of the model predictions is always assumed to follow Normal distribution.

About

Hybrid filter for post-processing forecasts provided by NWP models.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages