Validation of the EarthCARE ACM_RT product using surface solar irradiance measurements from BSRN stations and modeled values from CAMS.

The EarthCARE (ECA) satellite is currently in its second year in orbit, collecting new data every
day that could play a crucial role in advancing climate science. However, due to the advanced
technologies and retrieval approaches used in EarthCARE, the credibility of each instrument and
of their synergetic products must be verified. Significant effort has been devoted to this topic
both currently and in the past. Nevertheless, a substantial amount of publicly available data
that could improve validation has not yet been used. In this study, we use the ground-based
radiation measurements from the Baseline Surface Radiation Network (BSRN) to validate
1D surface solar radiation estimates from the EarthCARE ACM_RT product. Cloud effects are
analyzed separately using the cloud modification factor approach. Values from BSRN stations
are used if the station has less than 50 km distance to the satellite ground track. In addition,
an intercomparison with Copernicus Atmospheric Monitoring Service (CAMS) satellite based
surface solar radiation estimations has been performed. For the comparison with CAMS,
ECA values are averaged over time to obtain collocated grid cells. Due to limited gridded data
availability of the CAMS radiation service, this comparison is restricted to September-December
2024.

The ECA surface solar irradiance exhibits a Mean Bias Error (MBE) of −10.4 Wm^-2 and a
Root Mean Square Error (RMSE) of 191.7 Wm^-2 against ground based (BSRN) measurements.
Relative to CAMS, ECA surface solar irradiance exhibits a MBE of −23.3 Wm^-2 and a RMSE
of 103.3 Wm^-2. While some parts of South America, Northern Africa and Western Asia tend to
have higher EarthCARE irradiance, most of the available regions show higher CAMS irradiance.
This is especially the case in Oceania, middle part of Africa and Europe. Approximately 69%
of the difference between EarthCARE and CAMS can be contributed to differences in cloud
estimation, while 31% can be contributed to differences in clear-sky irradiance.

Future data releases from BSRN and CAMS are expected to enable a more robust assessment.
This analysis offers valuable insights relevant to the solar energy community.

Acknowledgements:

The authors acknowledge the project RACE-ECV (SBFI-633.4-2021-2024/PMOD - EarthCARE 202/2), supported by SBFI, the project Observe: Optimising 3D RT Earthcare product using geostationary observations and AI,  ESA Contract No. 4000147848/25/I/AG and the CERTAINTY (Cloud aERosol inTeractions & their impActs IN The earth sYstem) project funded from the Horizon Europe programme under Grant Agreement No 101137680