Summary

Particulate matter suspended in the atmospheric environment is one of the air pollutants that has been a problem for society since the middle of the 1960s. Particulate matter in the open air has become a concern for its adverse effects on the human body. On a global basis, moreover, particulate matter suspended in the troposphere and stratosphere has recently been found to directly and indirectly affect the earth's climatic fluctuations. Fine particulate matter consists mainly of soil granules, organic substances and minute particulates from combustion in the troposphere, and minute particulates, such as sulfates and nitrates in the stratosphere. Direct effects of particulate matter are considered to be the shutting off of the sun's rays and depletion of the ozone layer. Degradation of visibility caused by microscopic particulates, moreover, reduces atmospheric temperature because it blocks sunlight. Fine particulate characteristics of size distribution and weight concentration serve as information for us in controlling pollution and simulating climate fluctuation. Since fine particulate types are highly diverse and their composition and configuration is unstable, solving their weight concentration and size distribution will be a major challenge.

The Japan Environment Agency has determined environmental standards for the weight concentration of fine particulate pollutants in the atmosphere and has approved a standard or automatic measuring method equivalent. Air is drawn into the device and measurements taken in it. It is very difficult, however, to draw fine particulates suspended 20 kilometers up, such as stratospheric PSCs, into such an instrument. Attempts have been made to put rider techniques, which employ lasers, into practical use. Rider techniques used so far measure relative concentration, and do not allow the inferences to be made about particulate size distribution that are necessary for predicting climatic fluctuations.

Rider technology allows analyses of attenuated and scattered light obtained by irradiating light against remotely suspended fine particulates, and is considered a promising means of determining particulate size distribution.

To establish a remote measuring method, the authors numerically analyzed light that had been scattered by irradiating a laser against an aerosol, and which included fine particulates suspended in hydroso1. They studied means of estimating size distribution of fine particulates. Their challenge was to find a way to analyze scattered light. If that light comes from an aerosol containing fine particulates of varying size, it is measured in such a manner that each particulate's scattered light information is integrated. Deriving particulate size distribution from such measured values is done through the use of integral equations. Similar attempts have been made by many researchers. The nucleic functions caused by the light-scattering phenomenon, however, have little independence. As a result, some unreasonable cases were obtained with vibrating solutions or negative values.

For the study reported here, therefore, a specific solution was examined by a singular value decomposition of light scattered from fine particulates. The present report mainly comprises:

First, remote sensing to estimate the distribution of particulate size in an aerosol. Singular value decomposition was adopted as the technique for inverting measurement data of scattered light into a particulate size distribution. It is this singular value decomposition that is described here. Singular value decomposition has been used as a method for determining the independence of a matrix. It is also used to obtain the least square minimum norm sulutions.. The singular value decomposition was, therefore, also considered applicable to methods of solving simultaneous equations. With this method, a minimum norm solution can be calculated as long as there is no error in measure-
ment. If the error is significant, however, the solution will be negative solution. The solution, therefore, was calculated by manipulation that ignored the minimum singular value. Nevertheless, that solution had a longer norm than that of the preceding solution.

Chapter 2 describes the results of studying the behavior of a solution obtained by reversal from an obtainable attenuation spectrum, with a light irradiated against the aerosol, and of examining, in the laboratory, a hydrosol containing a highly stable distribution of fine particulates. For practicality, the sun was used as the light source for obtaining an attenuation spectrum of fine particulates in a vertical air column (the space between the ground and the sun that is taken as the light source, and is reckoned as a column having unit area) from the ground on the surface to the stratosphere. From such a spectrum, the reversal was calculated and it was confirmed that the fine particulates had a typical Junge distribution.

In Chapter 3, to acquire a size distribution of fine particulates from an angular distribution of the laser light irradiated against the aerosol, numerical experiments were made assuming a certain number of distributions and the reversal was studied by singular value decomposition, with the hydrosol of polystyrene latex (PSL) prepared in laboratory size,.

Chapter 4 describes an assumption of the size distribution of particulates in multi-wave form laser-applied scattering methods by wavelength to use the bistatic rider method to determine the size distribution of particulates in real aerosol. To study the probability of reversal to the particulate size distribution, two model distributions were assumed as numerical experiments; one is the Junge distribution, which approximates a particulate size distribution of fine particulates suspended in open air; the other is the log-normal distribution as a monodispersible distribution. For this distribution, Mie's scattering theory was used to calculate nucleonic function and scattered light intensity by wavelength. Singular value decomposition was used for reversal to investigate the reliability of an application on actual data measurement, with the effects of measurement error studied numerically on particulate distribution and weight concentration. For the experiments, hydrosol with PSL particulates having two distributions was prepared for measuring intensities by wavelength and for assuming distribution of particulate sizes. Moreover, distribution of fine particulate sizes in the atmosphere for assumed particulate radii of 0.15µm-5.0µm, was also studied. This confirmed that the haze had a log-normal distribution and that fine particulates in the atmosphere had a Junge distribution of particulate sizes.

Returns to the Report list.