EGU General Assembly 2020
© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Assimilating visible radiances to constrain aerosol properties in the ECMWF Integrated Forecast System.

Gareth Thomas1,2, Angela Benedetti3, Samuel Quesada Ruiz3, Julie Letertre-Danczak3, and Marco Matricardi3
Gareth Thomas et al.
  • 1RAL Space, Rutherford Appleton Laboratory, Didcot, United Kingdom (
  • 2National Centre for Earth Observation, Rutherford Appleton Laboratory, Didcot, United Kingdom
  • 3European Centre for Medium Range Weather Forecasting, Reading, United Kingdom

The Aerosol Radiance Assimilation Study (ARAS) has created a new approach for the assimilation of visible/near-IR radiances into the ECMWF’s Integrated Forecast System (IFS) for the constraining aerosol properties within the model. The capability is based on a new observation operator, based on the forward model used in the Optimal Retrieval of Aerosol and Cloud (ORAC) retrieval scheme, which predicts top-of-atmosphere radiances based on the model's aerosol field with sufficient accuracy while being computationally efficient enough to run in a operational analysis system such as that run at ECMWF. The system has been tested in the full IFS assimilation system, replacing the currently operational assimilation of MODIS AOD products, using MODIS radiances.
This presentation will give an overview of the new operator, show example results of its impact on the model output and discuss its merits and disadvantages compared to the AOD assimilation. 

How to cite: Thomas, G., Benedetti, A., Quesada Ruiz, S., Letertre-Danczak, J., and Matricardi, M.: Assimilating visible radiances to constrain aerosol properties in the ECMWF Integrated Forecast System., EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20358,, 2020

Comments on the presentation

AC: Author Comment | CC: Community Comment | Report abuse

Presentation version 1 – uploaded on 02 May 2020
  • CC1: Comment on EGU2020-20358, Franco Marenco, 04 May 2020

    Hello Gareth, this is very interesting indeed and at the Met Office we are considering doing something similar. I have seen your slide number 13 which compares a run with no DA (control) and a run with radiance assimilation. I was wondering whether you could also show a run with AOD assimilation, for comparison. Thanks!

    Franco Marenco

    • AC1: Reply to CC1, Samuel Quesada-Ruiz, 04 May 2020

      Dear Franco. Thanks for your comment. The plots shown in slide 13 correspond to 1 March 2017, which was a random day picked for the development of the observation operator for reflectances and is out of the evaluation period (JJA and SON 2017). We were lucky to have this dust event on that day. Unfortunately, we don't have a run for AOD. The goal of this slide is to show that the assimilation of visible reflectances provides useful information and improves the AOD model analysis.