EGU23-6365, updated on 08 Apr 2024
https://doi.org/10.5194/egusphere-egu23-6365
EGU General Assembly 2023
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Towards a Digital Twin of the Earth: ECMWF's effort to build a Kilometre-scale Earth System Model

Benoît Vannière, Irina Sandu, Estibaliz Gascon, Richard Forbes, Inna Polichtchouk, Annelize Van Niekerk, Birgit Suetzl, Michail Diamantakis, Peter Bechtold, and Gianpaolo Balsamo
Benoît Vannière et al.
  • ECMWF, Research, Germany (benoit.vanniere@ecmwf.int)

This presentation gives an overview of the work undertaken at ECMWF, in the  Destination Earth initiative of the European Comission, to build the global continuous component of the Weather-induced and Geophysical Extremes Digital Twin (Extremes DT) of the Earth. The Extremes DT aims to forecast and monitor extreme weather, globally, at a range of around 5 days, with unprecedented fidelity.

The Extremes DT utilizes the ECMWF Integrated Forecasting System cycle 48r1, with the Tco2559 grid, which has a horizontal resolution of 4.5 km. Our evaluation strategy is based on both 5-day forecasts of extreme weather cases and forecasts initialized daily over one summer and winter seasons. The performance of these simulations is compared to that of ECMWF’s operational deterministic 9km forecasts, using the following metrics: forecast skill as compared to the operational analysis or the observations, and ability to capture extreme weather events.

The results show a clear added-value of higher resolution for near-surface fields and the predicted precipitation amounts in regions of high orography. However, some aspects of the forecasts, which were initially degraded, have required additional developments and tunings. For instance, the atmospheric circulation over the Tibetan plateau has a clear dependence on time-step and resolution, which is linked to the treatment of the mean and sub-grid orography, as well as the representation of orographic gravity waves. Additionally, temperature biases in the lower troposphere in the Tropics are likely due to the sensitivity of moist physics to the model time-step. We will present the choices made to reduce those biases. 

This work can offer valuable insights into strategies for evaluating and improving kilometric-scale Earth System Models.

 

How to cite: Vannière, B., Sandu, I., Gascon, E., Forbes, R., Polichtchouk, I., Van Niekerk, A., Suetzl, B., Diamantakis, M., Bechtold, P., and Balsamo, G.: Towards a Digital Twin of the Earth: ECMWF's effort to build a Kilometre-scale Earth System Model, EGU General Assembly 2023, Vienna, Austria, 23–28 Apr 2023, EGU23-6365, https://doi.org/10.5194/egusphere-egu23-6365, 2023.