Coupled Atmosphere–Land Surface Data Assimilation in ECCC's Medium-Range NWP Forecast Systems

The Canadian Land Data Assimilation System (CaLDAS) is run operationally in offline mode at ECCC since 2014. Screen-level observations are assimilated using an Ensemble Kalman Filter approach on a global yin-yang grid at a resolution of 25 km. In an effort to improve the quality of the surface analyses produced by CaLDAS, the Interaction Soil Biosphere Atmosphere (ISBA) surface prediction model is being replaced by the Soil, Vegetation and Snow (SVS) model. The impact of that change on numerical weather prediction forecasts will be shown using standard surface and atmospheric objective scores.

CaLDAS was also tested in a two-way interactive coupled mode with the Global Deterministic Prediction System as well as the Global Ensemble Prediction System. The impact of the coupling on both systems will be presented.