Use of the CALIOP vertical feature mask for evaluating global aerosol models

E. P. Nowottnick,P. R. Colarco,A. da Silva,E. J. Welton

doi:10.5194/amt-8-3647-2015

Abstract

Abstract. We use observations from the space-based Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) to evaluate global aerosol distributions simulated in the NASA Modern Era Retrospective Analysis for Research and Applications aerosol reanalysis (MERRAero). We focus particularly on an evaluation of aerosol types, using the CALIOP vertical feature mask (VFM) algorithm, and look especially at Saharan dust distributions during July 2009. MERRAero consists of an aerosol simulation produced in the Goddard Earth Observing System version 5 (GEOS-5) Earth system model and incorporates assimilation of MODIS-derived aerosol optical thickness (AOT) to constrain column aerosol loadings. For comparison to the CALIOP VFM we construct two synthetic VFMs using the MERRAero aerosol distributions: a CALIOP-like VFM in which we simulate the total attenuated backscatter and particle depolarization ratio from the MERRAero output and pass those into the CALIOP VFM typing algorithm (MERRAero-CALIOP), and an extinction-based VFM in which we use the MERRAero-simulated species-resolved extinction to map the MERRAero species to the CALIOP VFM types (MERRAero-Extinction). By comparing the MERRAero-CALIOP VFM to CALIOP VFM, we can diagnose the aerosol transport and speciation in MERRAero. By comparing the MERRAero-CALIOP and MERRAero-Extinction-simulated VFM, we perform a simple observing system experiment (OSE), which is useful for identifying limitations of the CALIOP VFM algorithm itself. We find that, despite having our column AOT constrained by MODIS, comparison to the CALIOP VFM reveals a greater occurrence of dusty aerosol layers in our MERRAero-CALIOP VFM due to errors in MERRAero aerosol speciation. Additionally, we find that the CALIOP VFM algorithm is challenged when classifying aerosol features when multiple aerosol types are present, as our application of the CALIOP VFM algorithm to MERRAero aerosol distributions classified marine-dominated aerosol layers with low aerosol loadings as polluted dust when the contribution of dust to the total extinction was low.

Highlights

Mineral dust aerosols directly affect Earth’s atmospheric radiative balance by absorbing and scattering light (Colarco et al, 2014a; Ryder et al, 2013; Balkanski et al, 2007; Zhu et al, 2007; Haywood et al, 2003; Tegen and Miller, 1998; Sokolik and Toon, 1996)
We focus our analysis on July 2009 over and downwind of the dust-laden Sahara, which is dominated by the CloudAerosol Lidar with Orthogonal Polarization (CALIOP) vertical feature mask (VFM) aerosol type that had best agreement with AERONET (Mielonen et al, 2009) and High Spectral Resolution Lidar (HSRL) (Burton et al, 2013)
We provide an evaluation of the MERRAerosimulated Saharan dust plume during July 2009 using spacebased satellite imagery and lidar, and ground-based sun photometer observations

Summary

Introduction

Mineral dust aerosols directly affect Earth’s atmospheric radiative balance by absorbing and scattering light (Colarco et al, 2014a; Ryder et al, 2013; Balkanski et al, 2007; Zhu et al, 2007; Haywood et al, 2003; Tegen and Miller, 1998; Sokolik and Toon, 1996). Because of its radiative impacts, dust may affect tropical storm dynamics by enhancing atmospheric stability and inducing low-level wind shear that is not favorable for storm development (Reale et al, 2009, 2014; Dunion and Velden, 2004), with observational evidence suggesting that tropical storm activity and Saharan dust events are anti-correlated (Lau and Kim, 2007; Evan et al, 2006). Understanding of the roles of dust in all of these processes remains incomplete owing to the heterogeneous spatial, temporal, and compositional nature of dust and other related aerosols, and overall aerosol interactions within the Earth system remain a driving source of uncertainty in assessing the current climate and projecting future climate (IPCC, 2014)

Objectives

Methods

Results

Conclusion