# Direct and diffuse shortwave radiation in ERA5 vs ERA5-Land

**URL:** <https://forum.ecmwf.int/t/direct-and-diffuse-shortwave-radiation-in-era5-vs-era5-land/15131>\
**Category:** C3S - Datasets and Usage\
**Tags:** era5, era5-land\
**Created:** [6 July 2026 13:05 UTC](https://forum.ecmwf.int/t/direct-and-diffuse-shortwave-radiation-in-era5-vs-era5-land/15131 "2026-07-06T13:05:16Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Emilie\_Ferreira](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/e/a587f6/32.png) [@Emilie\_Ferreira](https://forum.ecmwf.int/u/Emilie_Ferreira)\
**Post date:** [6 July 2026 13:05 UTC](https://forum.ecmwf.int/t/direct-and-diffuse-shortwave-radiation-in-era5-vs-era5-land/15131/1 "2026-07-06T13:05:17Z")

</div>

Hello,

I am working with ERA5 and ERA5-Land datasets for a study on surface solar radiation and I would like to clarify the physical consistency of direct and diffuse components.

In ERA5, I use the following variables:

- Mean surface downward short-wave radiation flux (W m⁻²)
- Mean surface direct short-wave radiation flux (W m⁻²)

From this, I understand that the diffuse component could potentially be estimated as:  
diffuse = downward shortwave radiation − direct shortwave radiation ([https://www.ecmwf.int/sites/default/files/elibrary/2015/18490-radiation-quantities-ecmwf-model-and-mars.pdf](https://www.ecmwf.int/sites/default/files/elibrary/2015/18490-radiation-quantities-ecmwf-model-and-mars.pdf)).

In ERA5-Land, I only have access to:

- Surface solar radiation downwards (J m⁻²)

which represents total shortwave radiation (direct + diffuse), but without a direct component.

My questions are:

1. Is there any recommended or official method within ECMWF products to obtain direct and diffuse surface shortwave radiation in ERA5-Land?
2. Given the differences between ERA5 and ERA5-Land (resolution, land surface scheme, radiation diagnostics), is it scientifically valid to combine both datasets to reconstruct direct/diffuse components, or would this introduce inconsistencies?

I want to ensure a physically consistent methodology before using these variables in my analysis.

Thank you very much for your help.

---

<div class="post-metadata">

**Author:** ![Robin\_Hogan](https://forum.ecmwf.int/letter_avatar_proxy/v4/letter/r/a88e4f/32.png) [@Robin\_Hogan](https://forum.ecmwf.int/u/Robin_Hogan)\
**Post date:** [10 July 2026 18:50 UTC](https://forum.ecmwf.int/t/direct-and-diffuse-shortwave-radiation-in-era5-vs-era5-land/15131/2 "2026-07-10T18:50:53Z")

</div>

Dear Emilie,

You are right in your interpretation of downward shortwave flux (SSRD) and its direct component (FDIR) and that you can obtain the diffuse component as SSRD - FDIR. However, ERA5-Land does not calculate surface downwelling fluxes independently of ERA5 because to do that it would need an explicit high-resolution representation of the atmosphere. Rather, it is a land-only reanalysis that is forced by downwelling fluxes from ERA5. Although I haven’t checked, I am fairly sure that SSRD from ERA5-Land is simply SSRD from ERA5 but interpolated to the higher resolution of ERA5-Land. So if it is downwelling fluxes you are interested in, you need only look at ERA5. However, if you are interested in upward fluxes then ERA5-Land would have information at higher resolution than ERA5 because it can resolve variations in surface temperature and surface albedo.

Kind regards,

Robin Hogan.