Effect of Vertical Profile of Aerosols on the Local Shortwave Radiative Forcing Estimation

In this work, the effect of the aerosol vertical distribution on the local shortwave aerosol radiative forcing is studied. We computed the radiative forcing at the top and bottom of the atmosphere between 0.2 and 4 microns using the libRadTran package and compared the results with those provided by...

Full description

Bibliographic Details
Published in:Atmosphere
Main Authors: Francisco Molero, Alfonso Javier Fernández, María Aránzazu Revuelta, Isabel Martínez-Marco, Manuel Pujadas, Begoña Artíñano
Format: Article in Journal/Newspaper
Language:English
Published: MDPI AG 2021
Subjects:
Boa
Online Access:https://doi.org/10.3390/atmos12020187
https://doaj.org/article/ec9d2083530f417dacdafd50654160dc
Description
Summary:In this work, the effect of the aerosol vertical distribution on the local shortwave aerosol radiative forcing is studied. We computed the radiative forcing at the top and bottom of the atmosphere between 0.2 and 4 microns using the libRadTran package and compared the results with those provided by AERONET (AErosol RObotic NETwork). Lidar measurements were employed to characterize the aerosol vertical profile, and collocated AERONET measurements provided aerosol optical parameters required to calculate its radiative forcing. A good correlation between the calculated radiative forcings and those provide by AERONET, with differences smaller than 1 W m −2 (15% of estimated radiative forcing), is obtained when a gaussian vertical aerosol profile is assumed. Notwithstanding, when a measured aerosol profile is inserted into the model, differences between radiative forcings can vary up to 6.54 W m −2 (15%), with a mean of differences = −0.74 ± 3.06 W m −2 at BOA and −3.69 W m −2 (13%), with a mean of differences = −0.27 ± 1.32 W m −2 at TOA due to multiple aerosol layers and aerosol types. These results indicate that accurate information about aerosol vertical distribution must be incorporated in the radiative forcing calculation in order to reduce its uncertainties.