Back to search

National Technical University of Athens (NTUA)

Aerosol typing based on active/passive remote sensing technique

Abstract

dc:description

Aerosol particles directly affect the Earth’s radiation budget by interacting mainly with ultraviolet solar radiation through absorption and scattering (aerosol-radiation interaction). For example, purely scattering aerosols include sulfates, nitrates, ammonium, and sea-salt particles. On the other hand, absorbing aerosols are primarily black carbon, while organic carbon and dust aerosols partly absorb in the ultraviolet spectrum. Furthermore, aerosols affect cloud formation and lifetime serving as seeds (cloud condensation nuclei, ice nuclei) upon which cloud droplets and ice crystals form, changing the cloud albedo and the Earth’s radiation equilibrium once more. The processes of aerosol-radiation interaction depend on the aerosol size, shape, morphology, absorption, solubility, etc.; thus, knowledge of the aerosols’ chemical composition and mixing state in the atmosphere is essential for modeling their impact on the Earth’s radiation budget. It is well established that the Mediterranean basin is, frequently, affected by significant aerosol emissions from Africa, Europe and Turkey and it is characterized by high evaporation, low precipitation, and remarkably high solar activity. The specific role of aerosols, which are modified by atmospheric aging and mixing processes, makes this area an interesting case for climate change studies.Atmospheric aerosols are emitted directly as particles (primary aerosol) or formed in the atmosphere by gas-to-particle conversion processes (secondary aerosol). Primary aerosols arise from natural sources, such as windborne dust, sea spray, and volcanoes, and anthropogenic activities, such as the fuel combustion. Aerosol classification refers to the processes by which aerosol observations are associated with an aerosol type. In practice, several aerosol types are characterized by their origin, e.g., urban, dust, marine, and biomass burning. Thus, one way to characterize an aerosol layer is based on tools that identify the source region of the observed aerosols (i.e., air-mass backward trajectory analysis ending over the observation site, atmospheric models, concurrent satellite products, and ground-based photometric data, etc.). An emerging technique for aerosol detection and characterization which received much attention in the last decades is the light detection and ranging (lidar). Advanced systems, such as multiwavelength Raman lidars, can retrieve near-real-time information on the vertically resolved extensive (mass-dependent) and intensive (type-dependent) aerosol optical properties (namely the backscatter and extinction coefficient, lidar ratio, Ångström exponent and linear particle depolarization ratio) with high spatial (few meters) and temporal (from seconds to minutes) resolution.Towards this direction, two automated methods have been developed: (1) the Mahalanobis distance aerosol classification algorithm (MD) and (2) a neural network aerosol classification algorithm (NATALI), both based on aerosol intensive properties in order to classify aerosol layers observed by lidars into different categories. However, in reality, the intensive aerosol optical properties can vary greatly even for a single aerosol type due to the atmospheric aging and the mixing processes particles may undergo during their lifetime. For example, fresh biomass burning aerosols present higher Ångström exponent values than aged ones. Additionally, the lidar ratio values of dust, can vary significantly depending on the source mineralogy and the aging processes. Considering that both NATALI and MD are based on the aerosol intensive optical properties, they experience several limitations, thus, a more generic aerosol classification code free from these defects was needed.To fill this scientific lack, an improved automated aerosol layer classification algorithm known as Source Classification Analysis (SCAN) has been introduced, based on air-mass backward-trajectory analysis and satellite data.. SCAN is based on the amount of time that the air parcel spends over an already characterized aerosol source region and several additional criteria. The advantage of this algorithm is that its classification process is not affected by the overlapping values of the optical properties representing more than one aerosol type (e.g., clean continental, continental polluted, smoke). Moreover, it has no limitations concerning its ability to classify aerosol mixtures and it can be useful for all lidar systems (regardless of the number of their channels) as well as for other network-based systems (radar profilers, sun photometers).Ιn Part A of this Ph.D. Thesis, we compare the three independent aerosol classification methods: NATALI, MD, and SCAN revealing their strengths and limitations. Additionally, the aerosol optical properties of 5 free-tropospheric aerosol mixtures and the aerosol microphysical properties of 2 free-tropospheric aerosol mixtures are provided, giving valuable insights into the properties of aerosol mixtures observed over Europe, emphasizing those observed over the eastern Mediterranean region (Greece). The latter are, later, used for estimating the direct radiative effects of the most representative aerosol types observed over Athens, Greece. Additionally, an innovative method of discriminating between the smoke and continental polluted aerosol types is discussed, which may be a prohibitive procedure when only the aerosol optical properties are used because smoke and continental polluted aerosols frequently exhibit common optical property values. Finally, tropospheric aerosol remote sensing observations over Volos, Greece, and the corresponding aerosol particle linear depolarization ratio values are presented. Apart from the aerosols’ perturbation on the Earth’s radiation budget, the inhalable particulate matter (PM) is considered responsible for several health problems, from short-term respiratory diseases to premature mortality. However, the severity of the hazard caused by inhaled particles depends not only on the inhalable PM10 or PM2.5 (aerosols with aerodynamic diameter <10 μm or <2.5 μm, respectively) mass concentrations but also on the area at which they deposit within the respiratory system. This area is variable due to differences in the individual's airway geometry and breathing pattern, as well as, the particle’s size and shape. Additionally, recent studies suggest that PMs with different chemical composition behave differently inside the human body after inhalation, depending on their solubility and toxicity. The latter interactions are complex due to aerosols’ variable physical properties and chemical composition, depending on meteorological conditions, emission sources and atmospheric aging. More precisely, inhalation of PMs containing transition metals (such as Fe, Cr, Cu, Co, Zn, Ni) or organic compounds (including quinones, polycyclic aromatic hydrocarbons (PAHs)) can induce the formation of Reactive Oxygen Species (ROS) in vivo, which may lead to oxidative stress, inflammation or even cell death, leading later to various diseases. In other words, PM damage to human health is not an isolated process, as their chemical composition plays a crucial role.In this regard, there has been insufficient information to make clear conclusions about the health effects among different aerosol components or sources (sulfates, nitrates, organics, crustal materials, black carbon) and, especially, smoke from biomass combustion, including their capacity to induce oxidative stress to body cells. Moreover, all computational aerosol deposition models require additional verification with real input data. Thus, Part B of this Thesis aims on: a) providing a quantitative evidence of the risks that PMs (of different aerosol sources) can have on the human respiratory tract (RT), b) giving an insight into the toxicology of different aerosol types and finally, c) comparing the results of the three different aerosol dosimetry models.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2023

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Mylonaki, Maria
  • Μυλωνάκη, Μαρία

Subjects

dc:subject × 18

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/53393

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Mylonaki, Maria; Μυλωνάκη, Μαρία. Aerosol typing based on active/passive remote sensing technique. National Technical University of Athens (NTUA), 2023. http://hdl.handle.net/10442/hedi/53393