Research to Operations: Global Numerical Weather Prediction

The deterministic Global Model and the Global component of the Met Office Global and Regional Ensemble Prediction System (MOGREPS-G) produce highly performant forecasts that underpin many other components of the operational forecast suite. To maintain their performance relies on a good understanding of the main components of the system and a sound process to pull through scientific advances into operations in an expedient and coordinated way.

Key aims

  • To enhance the 4D-hybrid-Var data assimilation scheme and to exploit new and existing observations to improve both global analyses and forecasts
  • To develop our numerical modelling capability with a focus on physically realistic forecasts of near-surface quantities
  • Maintaining state-of-the-art ensemble prediction techniques, including the treatment of model uncertainty through stochastic parametrization schemes such that ensemble spread aims to match forecast error at all forecast lead-times
  • Using evaluation and verification tools to properly diagnose and measure forecast performance such that changes to the NWP system can be suitably tested before implementation
  • Diagnosing and tracking causes of forecast error with the aim to improve future NWP forecasts by prioritising specific activities during the model development cycle

Current activities