
Global cloud top height retrieval using SCIAMACHY limb spectra: model studies and first results
Author(s) -
KaiUwe Eichmann,
Luca Lelli,
Christian von Savigny,
H. Sembhi,
J. P. Burrows
Publication year - 2016
Publication title -
atmospheric measurement techniques
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.679
H-Index - 88
eISSN - 1867-8548
pISSN - 1867-1381
DOI - 10.5194/amt-9-793-2016
Subject(s) - sciamachy , nadir , cirrus , troposphere , cloud top , environmental science , remote sensing , radiative transfer , tropopause , lidar , cloud fraction , atmospheric radiative transfer codes , atmospheric sciences , optical depth , meteorology , aerosol , cloud computing , cloud cover , satellite , geology , physics , optics , astronomy , computer science , operating system
Cloud top heights (CTHs) are retrieved for the period 1 January 2003 to 7 April 2012 using height-resolved limb spectra measured with theSCanning Imaging Absorption SpectroMeter for Atmospheric CHartographY (SCIAMACHY) onboard ENVISAT (ENVIronmental SATellite). In this study, we present theretrieval code SCODA (SCIAMACHY cloud detection algorithm) based on a colourindex method and test the accuracy of the retrieved CTHs in comparison toother methods. Sensitivity studies using the radiative transfer model SCIATRAN show that themethod is capable of detecting cloud tops down to about 5 km and verythin cirrus clouds up to the tropopause. Volcanic particles can be detectedthat occasionally reach the lower stratosphere. Upper tropospheric ice cloudsare observable for a nadir cloud optical thickness (COT) ≥ 0.01, which is inthe subvisual range. This detection sensitivity decreases towards thelowermost troposphere. The COT detection limit for a water cloud top heightof 5 km is roughly 0.1. This value is much lower than thresholdsreported for passive cloud detection methods in nadir-viewing direction. Lowclouds at 2 to 3 km can only be retrieved under very clean atmosphericconditions, as light scattering of aerosol particles interferes with thecloud particle scattering. We compare co-located SCIAMACHY limb and nadir cloud parameters that areretrieved with the Semi-Analytical CloUd Retrieval Algorithm (SACURA). Onlyopaque clouds (τN,c > 5) are detected with the nadir passive retrievaltechnique in the UV–visible and infrared wavelength ranges. Thus, due to thefrequent occurrence of thin clouds and subvisual cirrus clouds in thetropics, larger CTH deviations are detected between both viewing geometries.Zonal mean CTH differences can be as high as 4 km in the tropics. Theagreement in global cloud fields is sufficiently good. However, the land–seacontrast, as seen in nadir cloud occurrence frequency distributions, is notobserved in limb geometry. Co-located cloud top height measurements of thelimb-viewing Michelson Interferometer for Passive Atmospheric Sounding(MIPAS) on ENVISAT are compared for the period from January 2008 to March 2012. The global CTH agreement of about 1 km is observed, which issmaller than the vertical field of view of both instruments. Lower stratospheric aerosols from volcanic eruptions occasionally interferewith the cloud retrieval and inhibit the detection of tropospheric clouds.The aerosol impact on cloud retrievals was studied for the volcanoesKasatochi (August 2008), Sarychev Peak (June 2009), and Nabro (June 2011).Long-lasting aerosol scattering is detected after these events in theNorthern Hemisphere for heights above 12.5 km in tropical and polarlatitudes. Aerosol top heights up to about 22 km are found in 2009 andthe enhanced lower stratospheric aerosol layer persisted for about 7 months.In August 2009 about 82 % of the lower stratosphere between 30 and70° N was filled with scattering particles and nearly 50 % in October2008