RICERCA ITALIANA     

Characterization of the optical and microphysical aerosol properties by several experimental techniques and calculation of the aerosol induced radiative effects : a key tool to asses aerosol climatology

Università degli Studi di Lecce

Abstract

We propose a project between three Units, to realize complementary measurements and studies to characterize the optical and microphysical properties of atmospheric aerosols and infer the aerosol radiative forcing at the top (TOA) and bottom (BOA) of the atmosphere, considering different surface albedo conditions.

To this end, Lecce's Unity will make use :

a) of a Raman lidar operating at 351 nm, to characterize the aerosol vertical distribution by means of the vertical profiles of the extinction and backscattering coefficients, of the lidar ratio, of the aerosol depolarization ratio, and of the relative humidity;

b) of an AERONET sunphotometer working at 8 wavelengths to measure aerosol optical depths (AOD) and some aerosol parameters such as: dimensional distributions, refraction indices, single scattering albedos, and Angstrom coefficients;

c) of a particulate sampler (FH 95 KF, Thermo ESM Andersen) equipped with 3 measurement heads, to monitor ground particulate matter concentrations and to characterize its microphysical properties (dimensional distribution, elemental composition, complex refractive index) by scanning electron microscopy (SEM) and spectroscopic techniques.

The University of Basilicata Unity:

a) will make use of high resolution and high spectral range radiometers (Mechelle-9000 and Avantes USB-2100) to measure aerosol optical thicknesses versus wavelengths and get by locally developed inversion techniques, Angstrom coefficients and aerosol dimensional distributions.

b) The Unity will also acquire and use satellite data (MODIS and TOMS) to get aerosol indices and/or aerosol optical thicknesses.

The Ferrara's Unity:

a) will make use of two inertial separators (INSPEC) to monitor the ground particulate matter and separate on filters the particles versus their aerodynamic diameter within the 0.5-6 micrometer range. The filters will then be used to better characterize the properties of particles of different aerodynamic diameter.

b) The Unity will also realize within the Project a nephelometer to characterize the particle radiative properties and to find out the particle refractive indices.

c) An important activity of the Ferrara's Unity will also be the calculation of instantaneous direct aerosol-induced radiative forcing at the top (TOA) and bottom (BOA) of the atmosphere by means of the 6S computer code (Vermote et al.' 1997), for atmospheric models assumed to be with and without aerosol particles.

The aerosol induced radiative forcing will be determined on days characterized by different advection patterns and within the same day as a function of the solar zenith angle.

Measurements and studies performed within the Project will refer only to Lecce's site, since one of the main goals is establishing working procedures to be used in future to define a large-scale aerosol climatology. To this end, it is worth mentioning that Lecce is in a climatological significant Mediterranean location, since it is on a flat peninsula and therefore directly affected from advection patterns coming from North Africa, from Mediterranean and Atlantic Seas, and from North and North-East Europe.

Analytical backtrajectories and cluster analysis will be used to correlate aerosol optical and microphysical properties and their spatial and temporal evolutions to origins and characteristics of the monitored air masses and therefore to the main aerosol types (marine, continental, desert). These last activity will be performed by all Units. <<<

Principal Investigator

Maria Rita PERRONE Università degli Studi di LECCE

Research Objectives

Aerosol particles are an important atmosphere constituent that interacts directly and indirectly with solar and terrestrial radiation to influence the energy balance of the earth-atmosphere. As a consequence, is at present, well established among aerosol scientists and climate modelers, the importance of the atmospheric aerosol particles on regional or global climate. They also show that there is an urgent need to quantitatively estimate the real effects of atmospheric aerosol on climate, and to accomplish this with the same vigor so far reserved for the study of changes due to minor and trace atmospheric gases.

Beyond the need to provide climate modelers with an appropriate data base, there are several further motivating issues that support the establishment of a global climatology of aerosols, as the following two:

1) A reference climatology is a ‘sine qua non" condition for determining possible trends due to natural and man-made processes, as well as feedback related to global warming.

2) Satellite-borne radiometers, sun photometers and other remote sensing devices, such as elastic lidars, require current aerosol data to remove atmospheric effects or to be used in algorithms for determining non-aerosol properties, such as surface albedos.

Complementary measurements and studies, to characterize optical and microphysical aerosol properties, their spatial and temporal evolution, and to infer the aerosol radiative forcing, will be performed within this project by 3 Research Units, to support the establishment of an aerosol climatology. To this end, numerical models and remote sensing techniques will be used, besides the characterization of ground collected particulate samplers.

The direct measurement of all required optical quantities as function of space and time, and the modeling of the microphysical aerosol properties, using data of different origins and types, and subsequently computing the average radiative characteristics of the aerosols, represent the two main approaches that will be followed within this project, to contribute to the design of a data set suitable for climate modeling purposes. In accordance to these last issues, the main objectives of the project are:

A) Provide a statistical significant data base of the optical and microphysical aerosol properties, and of their spatial and temporal evolution, retrieved:

1) by a lidar, a CIMEL sun photometer, and two high resolution and spectral range radiometers,
2) by MODIS and TOMS satellite measurements, and
3) by measurements on particles collected by 2 inertial separators and a ground particulate sampler.

B) Investigate the correlation between aerosol properties retrieved by different methods and experimental techniques, understand main characteristics and limits of each measurement technique, and show to what extent different techniques and data set are required to properly characterize aerosol properties and get statistical significant data to design an aerosol climatology.

The results on these last correlation studies will allow a better understanding of each measurement technique and will provide ground truth for present and future satellite missions.

C) Relate optical and microphysical aerosol properties to different aerosol types (mainly, marine, desert, and continental) and to air mass source regions by making use of analytical backtrajectories and cluster analysis, and satellite images as those provided by SeaWiFS.

D) Implement a numerical model based on the Mie theory, literature data and/or on data provided by the Project, to compute average radiative properties of marine, desert, and continental aerosols and then, investigate the correlation of numerical results and experimentally data determined by lidar, photometer and radiometric measurements for the same aerosol type, accordingly to aerosol source regions. These investigations should allow to better relate experimental determined aerosol properties (optical and microphysical) to aerosol type and then to source regions.

E) Calculate the instantaneous direct aerosol-induced radiative forcing at the top (TOA) and bottom (BOA) of the atmosphere by using the 6S numerical code (Vermote et al., 1997) for atmospheric models assumed to be with and without aerosol particles, including also the contribution furnished by multiple reflection effects between the surface and the atmospheric aerosols. Then, calculate the radiative effects induced by different aerosol types for various surface albedo conditions and aerosol optical thicknesses.

It is worth mentioning that aerosols will be monitored on a fixed schedule, to reduce effects due to the choice of preferential experimental conditions and therefore, to get statistically significant data sets (Point A). In particular, 4 measurements per week in two different days are planned. We believe that this choice will be sufficient to get statistically significant data sets by considering the available man power. Then, lidar measurements should provide vertical profiles of aerosol extinction and backscatter coefficients, of lidar ratios, of aerosol depolarization ratios, and of relative humidity, sufficient to define their spatial and temporal evolution, as well as to get the seasonal evolution of the aerosol mean radiative properties and of the aerosol load.

The Cimel sun photometer measurements performed on the same days of lidar measurements, will provide complementary data to the lidar ones, such as dimensional aerosol distributions, real and imaginary refractive indices, Angstrom coefficients, and single scattering albedos. These last data will allow evaluating the seasonal evolution of mean microphysical aerosol properties. On the contrary, the aerosol optical thicknesses retrieved by sun photometer, and MODIS and TOMS satellite measurements will allow a better evaluation of the seasonal evolution of the aerosol load and possibly, of how the aerosol load of different aerosol types vary with seasons.

The ground measurements performed by a particulate sampler and two inertial separators, performed on the same days of lidar measurements, will allow at first evaluating the correlation between ground particle concentration and aerosol optical thicknesses. To this end, it is worth mentioning that the aerosol monitoring will be performed on a rural area away from industrial sites and urban traffic. Moreover, the analysis on the ground collected particles by SEM, nephelometer and spectroscopic techniques will provide data on their microphysical properties such as: dimensional distributions, refraction indices, and elemental composition. These last data will be used to investigate the correlation with the corresponding parameters referring to atmospheric aerosols and provided by remote sensing techniques.

Finally, it is worth mentioning that the measurement campaigns, during which the lidar, the Cimel sun photometer and the high spectral resolution radiometers will simultaneously used, will allow evaluating benefits and limits of all used experimental techniques. <<<

First Results

The data referring to the aerosol radiative and microphysical properties, retrieved by various remote sensing technique (lidar photometric, radiometric, and satellite), by the sampling and characterization of the ground collected particles, and by the numerical model represent the main results of the first 18 months of the Project's activities.

These data are made of :

A) The vertical profiles of the aerosol extinction, and backscatter coefficients, of the lidar ratio, of the aerosol depolarization ratio, of the water vapor mixing ratio, and of the relative humidity, retrieved from lidar measurements performed at least on 100 different measurement days (~ 2 per week).

B) The data retrieved by sun photometer measurements, mainly the ones referring to measurements performed on the same days of the lidar measurements and on the days of the radiometer campaigns. These data will refer to: aerosol size distributions, water vapor columnar contents, and to aerosol optical thicknesses, single scattering albedos, and refraction indices at various wavelengths.

C) The data referring to aerosol size distributions and Angstrom coefficients, retrieved by various experimental techniques and numerical procedures, along the campaigns performed with the spectroradiometers of the Basilicata's Unity.

D) The data referring to the ground collected particles concentrations (PTS, PM10, PM2.5), to the particles size distributions, to their elemental composition and chemical-physical properties, and finally to the retrieval of the particles refraction index by nephelometer measurements and by the Kubelka-Munch method.

E) The data provided by the aerosol radiative forcing computations and in particular, the ones referring:

a) to the direct radiative forcing at the TOA and BOT calculated for different surface albedo conditions;
b) to the radiative forcing dependence on the concentration and type of the atmospheric particles.

F) The Monte Carlo model functional relationships providing extinction coefficients, lidar ratios, aerosol areas and volumes versus the backscattering coefficients for marine, desert, and continental aerosols.Some of the most important results of the Project's PHASE 2 are reported below:

A) Results on the seasonal and spatial evolution of the main lidar parameters: aerosol extinction and backscattering coefficients, lidar ratios, and aerosol depolarization ratios.

B) Results on the dependence of the above mentioned lidar parameters on relative humidity.

C) Results on the seasonal and spatial evolution of the water vapor mixing ratio and relative humitidy.

D) Results on the seasonal dependence of the main parameters retrieved by sun photometers measurements: aerosol optical thickness, Angstrom coefficient, single scattering albedo, refraction index, and size distribution.

E) Results on the seasonal dependence of PTS, PM10, and PM2.5 concentrations.

F) Results on the correlation between the AOT values retrieved by lidar, sun photometer, spectraradiometer, and MODIS and TOMS satellite data.

G) Results on the dependence of the main determined aerosol parameters on the air mass origins: AOT, Angstrom coefficient, single scattering albedo, refraction index, size distribution, lidar ratio, depolarization ratio, water vapor columnar content, relative humidity, ground particles concentration.

These data will allow evaluating the correlation between optical and microphysical aerosol properties and main aerosol types. It is worth mentioning that lidar ratios depend on aerosol shape, size distribution and refraction index, and as a consequence are generally used to characterize different aerosol types.

H) Results on the dependence of the elemental composition of the ground collected particles on the air mass origins.

I) Results on the aerosol radiative forcing at the TOA and BOA calculated for various aerosol types, and different aerosol optical thicknesses, and surface albedo conditions.

These last results will be rather useful to evaluate the conditions determining positive or negative aerosol raditive forcing.

L) Results on the correlation between the radiative parameters retrieved by the implemented numerical model and the corresponding ones determined by lidar measurements. <<<

Timescale

24 months

National and international background

Aerosol is an atmospheric component whose importance is increased in the last years because of their role in the energy budget of the earth-atmosphere system. Aerosols have a direct effect on this budget because they can absorb and scatter the solar radiation, and an indirect effect because they act as a nucleation center for water vapour, thus contributing to clouds formation and modifying their optical properties. Some experimental evidencies of the direct effect are well known, particularly the cooling induced by the increasing in the atmospheric albedo[Robock1995]; the warming due to light absorption from black carbon compounds is instead more difficult to demonstrate [Jacobson2000]. There are at present some experimental evidencies of the first indirect effect [Penner2004], which is the increase in the number of water droplets and a decreasing of their radius, and of the second indirect effect, which is a decreasing of precipitations [Ramanathan2001]. There is a broad agreement in the scientific community about the importance of the aerosol in climate changes and the fact that future efforts shoud be concentrated on the reduction of the uncertainty of estimation on aerosol radiative forcing [Anderson2003]. The chapter dedicated to aerosols of the 2001 report of "International Panel on Climate Change" [IPCC2001] gives a detailed account of assessed results.

The intrinsec difficulty of aerosol studies is due to the variability of the sources (both natural and anthropogenic) and of the chemical-physical and morphological properties, together with their short average lifetime, compared to that of the typical greenhouse gases. All these characteristics make the aerosol properties extremely variable both in space and in time on a world scale. As a consequence, the aerosol radiative forcing will be variable and the climate evolution model should properly account for this.

The present trend in global warming has stimulated developping of sophisticated models, but the role of aerosols is still uncertain, in particular it is not yet known if aerosols contribute with a positive or negative forcing to the climate evolution. This fact has stimulated setting up several measurement campaigns and permanent measurement networks at a national, continental or worldwide level, that use different technologies, to create a reliable database to constrain results of climate models.

Also the theoretical analysis of particle radiation-interaction has experienced a great development, starting from the Mie theory for spherical particles. Models for the creation and transport of aerosols are also available.

Thus, there are at present experimental and theoretical techniques to contribute to the assessment of an aerosol climatology, i.e. a collection of aerosol data linked to the season and the meteorological situation to get the correct contribution to the climate evolution.

In the following we will describe the most common measurement techniques and the principal current projects. Talking about measurements technologies, first of all we separate local measurements from remote measurements. Remote measurements mostly rely on measurements of the radiative properties of the atmosphere. Since the molecular composition and the density of the atmosphere is known to a good approximation, it is possible to isolate from the performed measurements the molecular contribution to get informations about the aerosols. Since the linear dimensions of the atmospheric aerosols are comparable to the optical wavelenghts there is a measurable dependence of the optical properties from the wavelenght.
Optical remote measurements by passive sensors are made by radiometers at ground or on a satellite (in some cases on an aircraft) measuring the sun direct or diffuse irradiance. These instruments have a poor spatial resolution, since they measure quantities that are integrated on the solar path, but they have the advantage to be easily transported and relatively easy to use.
Their operation can be automatised, giving measurements with a good temporal resolution on the long period. Furthermore, they can operate over a broad spectral interval. In particular, at the Lecce unit an automatic Cimel photometer operating over 8 wavelength is operating, while at the unit of Potenza an high resolution (0.5 nm) radiometer is operating. From irradiance measurements at the different wavelengths we can obtain directly the atmosphere optical thickness; using suited inversion algorithms it is possible to deduce the microphysical properties of the aerosols, i.e. the complex refraction index and the dimensional distribution.


This kind of measurements can be obtained also by satellite mounted radiometers. We mention here the TOMS and MODIS radiospectrometers that give a global monitoring of the Earth, with a limited spacetime resolution.


The active sensors for aerosol optical properties measurements are LIDARs. It has been shown in the last years that lidar systems using visible or ultraviolet laser, like excimers laser or the harmonics of the Nd:Yag laser, can give measurements of the exinction and backscattering coefficients, with high spatial and temporal resolution [Whiteman92]. Signals at the same wavelength than the laser (elastic) depends on molecular and aerosol backscattering and extinction processes. Signals originating from vibrational Raman effect from the molecules of the atmosphere (nitrogen,oxygen,water vapor) depends only on the molecular density and the total extinction; thus, they can be used to obtain the aerosol extinction and the water vapor mixing ratio [Ansmann1990]. From this last quantity it is possible to obtain the relative humidity profile, which is an important parameter for the characterization of the aerosols, because they can absorb water and increase their dimension.
The use of several wavelengths (tipically the different harmonics of a Nd-Yag laser) allows to get informations about microphysical parameters of the particles [Bockmann2001b]. The detection of the depolarisation of the backscattered light gives informations about the simmetry of the particles.

Local measurements give very detailed measurements about atmospheric aerosols. Analysis on the ground collected particulate give informations about their concentration, size distribution, morphology,chemical composition and optical properties (refraction index) [Levin1979], using SEM,TEM, spectrophotometry techniques. Other optical properties like total extinction, phase function, asimmetry parameter can be obtained can be obtained by scattering measurements using suited instruments like "nephelometers". Local measurements can be, in particular conditions, correlated with remote measurements [Smirnow2000].

All these different techniques are used, if possible, in measurement campaigns. In large scale coordinated measurement network similar instruments are instead preferred.

The international relevance of the study of the spatial and temporal evolution of the aerosols is demonstrated by several past or current projects on this topic. Several intensive, but necessarily limited in space and time measurements campaigns have been performed. The intensive campaign called "Tropospheric Aerosol Radiative Forcing Observational Experiment" (TARFOX), performed off the east shore of United States between 10 and 31 july, 1996, has been programmed to reduce the uncertainties on the estimation of the anthropic aerosol climate effects [Ferrare2000].

Three intensive campaigns called "Aerosol Characterization Experiment" (ACE) have been conducted in these last years to investigate aerosol properties by all the different possible means
[ACE1999, ACEAsia2003].

Global measurements are instead of two kinds: satellite measurements and coordinate measurements organized on worldwide networks. The demonstration of the feasibility of a lidar monitoring of the global distribution of aerosols has been obtained by the "Lidar In-space Technology Experiment" (LITE) organized by NASA [Winker96] in 1994, using a three-wavelength lidar on a Space Shuttle. Similar experiments with a larger temporal duration should be mounted on a satellite (or on an orbiting station, like it has been the case for the russian MIR station that hosted a lidar for clouds monitoring )and operate automatically.

We mention, as a current project, the space lidar CALIPSO, whose launch is planned in 2005, that will sound atmosphere at two wavelenghts (1021 and 510 nm)and detect depolarization of the signal at 510 nm. This system will fly in formation with the Aqua satellite as part of the Aqua constellation, a system of 5 satellites to get correlated measurements on atmosphere and Earth.

The first coordinated network of lidar measurements, "Aerosol Lidar Network" has been instituted in 1998 in Germany by six research institutions, all of them equipped by a lidar system [Boesenberg2001]. This project have originated the european project EARLINET (Contract N. EVRI-CT1999-40003), which has operated from February 2000 to February 2003 [Boesenberg2003]. The goal of the EARLINET project has been to obtain a data base of the horizontal and vertical distribution of aerosols, using 22 lidar stations spreaded on Europe. The Unit of Lecce has participated to this project. Even if the project has terminated, the network is still operating in view of a possible renewal of funding.

The largest effort for global monitoring of aerosol using ground based passive systems is at present the NASA network "Aerosol Robotic Network", AERONET [Dubovik2000], which is a worldwide network of similar sun photometers operating in continuous and automated mode. The results are available to the scientific community via the Internet. The unit of Lecce is part of this network. On a specific site it is possible to obtain the optical thickness at different wavelengths and the diffuse irradiance. If spherical particles and homogeneous distribution of aerosols are assumed it is possible to retrieve the concentration, size distribution and complex refraction index of the particles. Inversion algorithms are frequently updated in order to correct the initial hypothesis, like the assumption of spherical particles.

Thus, it is clear that the development of numerical models to treat the light-aerosol interaction should proceed on the same way than the development of measurements method. From a theoretical point of view the treatment of atmospheric aerosol - light interaction proceeds in two steps:

a) study of the interaction of light with a single particle, extended to the interaction with a population of different particles ; from this it is possible to obtain models for the radiative transfer in atmosphere.

b) the modelisation of the chemical-physical and morphological properties of aerosols, their production, their transport and their transformations. A good modeling should thus give a theoretical prediction of the aerosol kind in a given place as a function of the meteorological situation and the air masses history.

In the following we will focus on a) because point b) is less relevant for the present project.


Point a) is based on the Mie theory for spherical homogeneous particles [Bohren83]. Even if this is an exact theory, it requires numerical methods for the complete calculation of the optical properties of an arbitrary spherical particles populations. Recently there has been great efforts to describe the interaction of non spherical or not homogeneous particles. It has been shown in [Bockmann2001] that the scattering properties of inohomogeneous and weakly absorbing particles can be very different from those of the same particles hogeneously mixed and or not absorbing. A treatment on nonspherical particles can be found in [Mischenko97]

A numerical study on the "lidar ratio", the ratio of the extinction coefficient to backscattering coefficient, has been made in [Ackermann2000]. A correct estimation of the lidar ratio is important to solve the lidar equation and to investigate the aerosol impact on climate. A "data-base" of the optical properties of the aerosols (extinction, scattering, absorption coefficient, phase function, asimmetry factor, single scattering albedo) has been presented in [Levoni97]. In this paper also the effect of relative humidity (RH)has been considered, so the calculation has been performed in dry (RH=0%) and wet (RH=99%)conditions. In the cited paper it is shown
that the optical properties of the aerosols, calculated from the data base, allow the simulation of the solar spectrum from the atmosphere.
The properties of a Saharan dust transport event over Mediterranean Sea have been investigated in [Gobbi2000] using a numerical model to estimate the extinction and the surface of aerosols. Desert dust absorbs at UV, visible, infrared wavelengths. Thus, their presence in the atmosphere can lead either to a cooling or a warming effect following some properties like the single scattering albedo, the altitude of the layer and the surface albedo [Liao98].

A numerical model that uses the Monte Carlo technique has been developed in [Barnaba2001] to determine analitical relationship linking the backscattering coefficient to the extinction coefficient, the area and volume of tropospheric marine and desert dust aerosols. This model can be extended to non spherical particles using the results of [Mischenko97].


The calculation of radiative forcing is obtained from scattered solar irradiance and radiative transport models like the 6S [Vermote97], that can account for the presence of aerosols and different surface albedo conditions. In various radiative models [Hanel1999, Vitale2000] it has been shown that a given columnar quantity of particles, with a low content of absorbing particles, can induce the cooling of the underlying surface if the reflectivity is low, like an ocean surface, but it can have an opposite effect if the reflectivity is larger than 0.15. Consequently, besides aerosol properties, a careful modelisation of the surface albedo is also needed. <<<