After the model runs, weather data often undergo additional processing steps to improve their quality. The main reasons for post-processing are:
- Quality control: detection and removal of errors, filling of gaps
- Accuracy: Improving accuracy by quality control, bias correction, downscaling or other methods
- Transformation: turning a signal (e.g. reflection) into a meteorological value (radiation)
Post-processing methods may be applied to different data sources:
- Measurements: quality control, others
- Observations: transformation, interpolation,
- Weather model simulations: downscaling, ensemble computation, multi-model, etc.
- or a combination thereof (e.g. MOS; mLM)
Some post-processing methods are described in the sub-pages.