Improving the McClear model estimating the downwelling solar radiation at ground level in cloud-free conditions – McClear‑v3
International audience The fast McClear clear-sky model estimates the downwelling shortwave direct and diffuse irradiances received at ground level under cloud-free conditions. Several improvements are presented. They focus on the modeling of changes in irradiances with the solar zenithal angle and...
Published in: | Meteorologische Zeitschrift |
---|---|
Main Authors: | , , , , , |
Other Authors: | , , , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
HAL CCSD
2019
|
Subjects: | |
Online Access: | https://minesparis-psl.hal.science/hal-02164312 https://minesparis-psl.hal.science/hal-02164312/document https://minesparis-psl.hal.science/hal-02164312/file/2019_metzet_mcclearv3_gschwind.pdf https://doi.org/10.1127/metz/2019/0946 |
Summary: | International audience The fast McClear clear-sky model estimates the downwelling shortwave direct and diffuse irradiances received at ground level under cloud-free conditions. Several improvements are presented. They focus on the modeling of changes in irradiances with the solar zenithal angle and on a better exploitation of the aerosol properties offered by the Copernicus Atmosphere Monitoring Service (CAMS). Irradiances from this new version McClear-v3 were compared to 1 min measurements made in cloud-free conditions at 11 stations belonging to the Baseline Surface Radiation Network and being located in various climates. The correlation coefficient ranges between 0.982 and 0.999 for the global irradiance. The bias is positive (overestimation) and ranges between 1 W m −2 (0.1 % of the mean observed irradiance) and 20 W m −2 (3.2 %), with the exception of Barrow in Alaska (18 W m −2). The standard deviation ranges between 16 W m −2 (2.3 %) and 30 W m −2 (3.8 %). The correlation coefficient for the direct irradiance ranges between 0.902 and 0.995. As expected, since the direct in McClear does not comprise any circumsolar contribution, the bias is negative (underestimation) and ranges between 49 W m −2 (7.7 %) and 5 W m −2 (0.7 %), with two exceptions: Sede Boqer (79 W m −2) and Brasilia (13 W m −2). The standard deviation is comprised between 34 W m −2 (5.3 %) and 69 W m −2 (10.7 %). These results are similar to those obtained with McClear version 2. Compared to the latter, McClear-v3 removes several artifacts and its estimates are continuous in space and time. |
---|