Abstract/Summary

The spectrum of reflected solar radiation emerging at the top of the atmosphere is rich with Earth system information. To identify spectral signatures in the reflected solar radiation and directly relate them to the underlying physical properties controlling their structure, over 90,000 solar reflectance spectra are computed over West Africa in 2010 using a fast radiation code employing the spectral characteristics of the SCanning Imaging Absorption spectroMeter for Atmospheric CHartographY (SCIAMACHY). Cluster analysis applied to the computed spectra reveals spectral signatures related to distinct surface properties, and cloud regimes distinguished by their spectral short-wave cloud radiative effect (SWCRE). The cloud regimes exhibit a diverse variety of mean broadband SWCREs, and offer an alternative approach to define cloud type for SWCRE applications that does not require any prior assumptions. The direct link between spectral signatures and distinct physical properties extracted from clustering remains robust between spatial scales of 1, 20 and 240 km, and presents an excellent opportunity to understand the underlying properties controlling real spectral reflectance observations. Observed SCIAMACHY spectra are assigned to the calculated spectral clusters, showing that cloud regimes are most frequent during the active West African monsoon season of June–October in 2010, and all cloud regimes have a higher frequency of occurrence during the active monsoon season of 2003 compared with the inactive monsoon season of 2004. Overall, the distinct underlying physical properties controlling spectral signatures show great promise for monitoring evolution of the Earth system directly from solar spectral reflectance observations.