Linus Magnusson

Senior Scientist
Forecast, Evaluation, Diagnostics


Linus Magnusson obtained his PhD from Stockholm University in 2009 and joined ECMWF shortly after. He started his work on ocean initialisation and ENSO forecasting in the Seasonal forecasting section where he also worked on sea-ice modelling. In 2011 he moved to the Ensemble forecasting section working on diagnostics and has now a similar job the Evaluation section. He is leading the model intercomparison collaboration DIMOSIC ( He is since 2023 a member of the scientific steering committee for the WMO World Weather Research Programme (WWRP). His research interests include severe weather evaluation, tropical cyclone forecasting, medium-range forecast error propagation and diagnostics for processes in the Arctic.

Professional interests:
  • Predictability
    • The research focuses on understanding periods of low predictability in the extra-tropics. For such investigations several diagnostic tools are used, such as relaxation experiments that constrain the model inside a specified box. The methods are described in Magnusson (2017).
    • Leading the project DIMOSIC (DIfferent Models, Same Initial Conditions), where global medium-range forecast models are initialised from the same initial conditions to evaluate the role on model differences on forecast skill, biases and forecast uncertainties.

  • Severe weather evaluation
  • Tropical cyclones
    • The research focuses on model sensitivity for tropical cyclone intensity forecasts. Recent publications includes ECMWF activities for improved hurricane forecasts (Magnusson et. al. ,2019 and ECMWF Technical Memorandum 888) and the effect on coupled modelling  (Mogensen et al. 2017).
  • Arctic processes
    • Linus was involved in H2020 project APPLICATE, and contribute to the work on diagnostic tools for the Arctic and predictability studies.


Career background:

2005 Master degree Uppsala University

2009 PhD degree Stockholm University

2009- ECMWF