Publications

Santese, M, De Tomasi, F, Perrone, MR (2007). Moderate resolution Imaging spectroradiometer (MODIS) and aerosol robotic network (AERONET) retrievals during dust outbreaks over the Mediterranean. JOURNAL OF GEOPHYSICAL RESEARCH-ATMOSPHERES, 112(D18), D18201.

Abstract
Moderate Resolution Imaging Spectroradiometer ( MODIS) - ocean aerosol products and corresponding Aerosol Robotic Network ( AERONET) data retrieved during Sahara dust outbreaks that occurred over the central Mediterranean basin have been analyzed, to contribute to the validation of MODIS retrievals during dust-loading conditions. It is shown that MODIS aerosol optical depths (AOD(M)) retrieved at 550 nm for window sizes of 50 x 50 km(2) centered at different Mediterranean AERONET sites are in satisfactory accordance to corresponding AERONET optical depths ( AOD(A)) colocated in space and time: Correlation factor and slope of the linear regression are 0.86 and 1.04, respectively, and 42% of data points are within the prespecified accuracy range. However, the plot of the differences ( AOD(M) - AOD(A)) versus AOD(A) has revealed that the differences ( AOD(M) - AOD(A)) are almost positive and increase with AOD(A), and has led assuming that MODIS overestimates AODs mainly during high dust load conditions. The comparison of MODIS-ocean and corresponding AERONET fine-mode fractions ( eta(M) and eta(A), respectively) colocated in space has revealed that the MODIS-ocean inversion algorithm overestimates the contribution of fine-mode particles for dust-dominated aerosols: (eta(M) - eta(A)) difference values vary within the 0.11-0.24 range. In addition, the comparison of MODIS-ocean and corresponding AERONET volume size distributions has revealed that the latter results are mainly due to the poor sensitivity of the MODIS-ocean inversion algorithm to the concentration of coarse-mode particles during dust outbreaks. The frequency distributions of MODIS fine and coarse volume modal radii are found to be rather close to corresponding AERONET frequency distributions.

DOI:
10.1029/2007JD008482

ISSN:
0148-0227