Call number:
AWI A11-22-94735
Description / Table of Contents:
Over the last decades, the rate of near-surface warming in the Arctic is at least double than elsewhere on our planet (Arctic amplification). However, the relative contribution of different feedback processes to Arctic amplification is a topic of ongoing research, including the role of aerosol and clouds. Lidar systems are well-suited for the investigation of aerosol and optically-thin clouds as they provide vertically-resolved information on fine temporal scales. Global aerosol models fail to converge on the sign of the Arctic aerosol radiative effect (ARE). In the first part of this work, the optical and microphysical properties of Arctic aerosol were characterized at case study level in order to assess the short-wave (SW) ARE. A long-range transport episode was first investigated. Geometrically similar aerosol layers were captured over three locations. Although the aerosol size distribution was different between Fram Strait(bi-modal) and Ny-Ålesund (fine mono-modal), the atmospheric column ARE was similar. The latter was related to the domination of accumulation mode aerosol. Over both locations top of the atmosphere (TOA) warming was accompanied by surface cooling.
Subsequently, the sensitivity of ARE was investigated with respect to different aerosol and spring-time ambient conditions. A 10% change in the single-scattering albedo (SSA) induced higher ARE perturbations compared to a 30% change in the aerosol extinction coefficient. With respect to ambient conditions, the ARETOA was more sensitive to solar elevation changes compared to AREsur f ace. Over dark surfaces the ARE profile was exclusively negative, while over bright surfaces a negative to positive shift occurred above the aerosol layers. Consequently, the sign of ARE can be highly sensitive in spring since this season is characterized by transitional surface albedo conditions.
As the inversion of the aerosol microphysics is an ill-posed problem, the inferred aerosol size distribution of a low-tropospheric event was compared to the in-situ measured distribution. Both techniques revealed a bi-modal distribution, with good agreement in the total volume concentration. However, in terms of SSA a disagreement was found, with the lidar inversion indicating highly scattering particles and the in-situ measurements pointing to absorbing particles. The discrepancies could stem from assumptions in the inversion (e.g. wavelength-independent refractive index) and errors in the conversion of the in-situ measured light attenuation into absorption. Another source of discrepancy might be related to an incomplete capture of fine particles in the in-situ sensors. The disagreement in the most critical parameter for the Arctic ARE necessitates further exploration in the frame of aerosol closure experiments. Care must be taken in ARE modelling studies, which may use either the in-situ or lidar-derived SSA as input.
Reliable characterization of cirrus geometrical and optical properties is necessary for improving their radiative estimates. In this respect, the detection of sub-visible cirrus is of special importance. The total cloud radiative effect (CRE) can be negatively biased, should only the optically-thin and opaque cirrus contributions are considered. To this end, a cirrus retrieval scheme was developed aiming at increased sensitivity to thin clouds. The cirrus detection was based on the wavelet covariance transform (WCT) method, extended by dynamic thresholds. The dynamic WCT exhibited high sensitivity to faint and thin cirrus layers (less than 200 m) that were partly or completely undetected by the existing static method. The optical characterization scheme extended the Klett–Fernald retrieval by an iterative lidar ratio (LR) determination (constrained Klett). The iterative process was constrained by a reference value, which indicated the aerosol concentration beneath the cirrus cloud. Contrary to existing approaches, the aerosol-free assumption was not adopted, but the aerosol conditions were approximated by an initial guess. The inherent uncertainties of the constrained Klett were higher for optically-thinner cirrus, but an overall good agreement was found with two established retrievals. Additionally, existing approaches, which rely on aerosol-free assumptions, presented increased accuracy when the proposed reference value was adopted. The constrained Klett retrieved reliably the optical properties in all cirrus regimes, including upper sub-visible cirrus with COD down to 0.02.
Cirrus is the only cloud type capable of inducing TOA cooling or heating at daytime. Over the Arctic, however, the properties and CRE of cirrus are under-explored. In the final part of this work, long-term cirrus geometrical and optical properties were investigated for the first time over an Arctic site (Ny-Ålesund). To this end, the newly developed retrieval scheme was employed. Cirrus layers over Ny-Ålesund seemed to be more absorbing in the visible spectral region compared to lower latitudes and comprise relatively more spherical ice particles. Such meridional differences could be related to discrepancies in absolute humidity and ice nucleation mechanisms. The COD tended to decline for less spherical and smaller ice particles probably due to reduced water vapor deposition on the particle surface. The cirrus optical properties presented weak dependence on ambient temperature and wind conditions.
Over the 10 years of the analysis, no clear temporal trend was found and the seasonal cycle was not pronounced. However, winter cirrus appeared under colder conditions and stronger winds. Moreover, they were optically-thicker, less absorbing and consisted of relatively more spherical ice particles. A positive CREnet was primarily revealed for a broad range of representative cloud properties and ambient conditions. Only for high COD (above 10) and over tundra a negative CREnet was estimated, which did not hold true over snow/ice surfaces. Consequently, the COD in combination with the surface albedo seem to play the most critical role in determining the CRE sign over the high European Arctic.
Type of Medium:
Dissertations
Pages:
x, 136 Seiten
,
Illustrationen, Diagramme, Karten
URL:
https://doi.org/10.25932/publishup-53036
URL:
https://nbn-resolving.org/urn:nbn:de:kobv:517-opus4-530366
URL:
https://d-nb.info/1248561724/34
URN:
urn:nbn:de:kobv:517-opus4-530366
DOI:
10.25932/publishup-53036
Language:
English
Note:
Dissertation, Universität Potsdam, 2021
,
CONTENTS
1 INTRODUCTION
1.1 Motivation: Aerosol and cloud relevance to Arctic amplification
1.2 Theoretical background
1.2.1 Atmospheric aerosol
1.2.2 Aerosol in the Arctic
1.2.3 Cirrus clouds
1.3 Research questions
2 METHODS
2.1 lidar remote sensing techniqu
2.1.1 Elastic and Raman lidar equations
2.1.2 lidar signal corrections
2.1.3 Derivation of particle optical properties and related uncertainties
2.2 Lidar systems
2.2.1 Ground-based system KARL
2.2.2 Air-borne system AMALi
2.2.3 Space-borne system CALIOP
2.3 Ancillary instrumentation
2.3.1 Radiosondes
2.3.2 Sun-photometers
2.3.3 Radiation sensors
2.4 Modeling tools
2.4.1 Air mass backward trajectories
2.4.2 Aerosol microphysics retrieval algorithm
2.4.3 Radiative transfer model SCIATRAN
2.4.4 Multiple-scattering correction model
2.4.5 Simplified cloud radiative effect model
3 ARCTIC AEROSOL PROPERTIES AND RADIATIVE EFFECT (CASE STUDIES)
3.1 Aerosol in the upper troposphere (Spring)
3.1.1 Overview of aerosol observations and air mass origin
3.1.2 Modification of aerosol optical and microphysical properties
3.1.3 Aerosol radiative effect (ARE)
3.2 Sensitivities of the spring-time Arctic ARE
3.2.1 Sensitivity on aerosol related parameters
3.2.2 Sensitivity on ambient conditions
3.3 Aerosol in the lower troposphere (Winter)
3.3.1 Overview of remote sensing and in-situ measurements
3.3.2 Aerosol properties from the remote sensing perspective: KARL and CALIOP
3.3.3 Aerosol microphysical properties from in-situ and remote sensing perspectives
3.4 Discussion and Conclusions
4 DEVELOPMENT OF A CIRRUS CLOUD RETRIEVAL SCHEME
4.1 Fine-scale cirrus cloud detection
4.1.1 Selection of cirrus clouds
4.1.2 Wavelet Covariance Transform method
4.1.3 Revised detection method: Dynamic Wavelet Covariance Transform
4.2 Comparison of dynamic and static cirrus detection
4.3 Cirrus cloud optical retrievals
4.3.1 Existing cirrus optical retrievals: double-ended Klett and Raman
4.3.2 Temporal averaging within stationary periods
4.3.3 Revised optical retrieval: constrained Klett method
4.4 Comparison to established optical retrievals
4.5 How uncertainties in cirrus detection affect the optical retrievals?
4.6 Discussion
4.6.1 Limitations of cirrus retrieval schemes
4.6.2 Strengths of the revised retrieval scheme
4.7 Conclusions
5 LONG-TERM ANALYSIS OF ARCTIC CIRRUS CLOUD PROPERTIES
5.1 Overview of cirrus occurrence and meteorological conditions over Ny-Ålesund
5.2 Quality assurance of optical properties
5.2.1 Specular reflection effect
5.2.2 Investigation of extreme cirrus lidar ratio values
5.2.3 Multiple-scattering correction
5.3 Overview of cirrus optical properties over Ny-Ålesund
5.4 Inter-relations of cirrus properties
5.5 Dependence on meteorological conditions
5.5.1 Cirrus clouds in the tropopause
5.6 CRE estimation at TOA: sensitivity analysis
5.7 Conclusions
6 CONCLUSIONS AND OUTLOOK
A CIRRUS DETECTION SENSITIVITIES
a.1 Wavelet Covariance Transform - dilation sensitivity
a.2 Wavelet Covariance Transform - wavelength dependency
B CIRRUS OPTICAL CHARACTERIZATION SENSITIVITIES
b.1 Reference value accuracy and limitations
b.2 Inherent uncertainties of constrained Klett
C MULTIPLE-SCATTERING CORRECTION FOR CIRRUS CLOUDS
D SEASONAL CIRRUS PROPERTIES: DESCRIPTIVE STATISTICS
BIBLIOGRAPHY
Location:
AWI Reading room
Branch Library:
AWI Library
Permalink
