An investigation of seasonal variation in aerosol optical properties from ground based and satellite measurements over the region of tainan taiwan

Since the atmospheric aerosols has diverse temporal variation, this study presents an integration of both ground-based AERONET and satellite measurements MODIS AOD with regards to the se

THAI NGUYEN UNIVERSITY

UNIVERSITY OF AGRICULTURE AND FORESTRY

ERIKA ROMERO PADLAN

AN INVESTIGATION OF SEASONAL VARIATION IN AEROSOL OPTICAL PROPERTIES FROM GROUND-BASED AND SATELLITE MEASUREMENTS OVER THE REGION OF TAINAN, TAIWAN

BACHELOR THESIS

Study Mode: Full time

Major: Environmental Science and Management

Faculty: International Programs Office

Batch: 2014- 2017

Thai Nguyen, 20/11/2017

DOCUMENTATION PAGE WITH ABSTRACT

Thai Nguyen University of Agriculture and Forestry

Degree Program Bachelor of Environmental Science and Management Student Name Erika Romero Padlan

Thesis Title An Investigation of Seasonal Variation in Aerosol

Optical Properties from Ground-Based and Satellite Measurements Over the Region of Tainan, Taiwan Supervisor (s) Assoc Prof Tang Huang Lin, Ph.D

Dr Do Thi Ngoc Oanh Abstract:

Atmospheric aerosols suspended in the air have a range of hardly a nanometer (less than the width of the smallest virus) to a several micrometers that makes them inhalable easily Therefore, different health effects could be caused with respect to the particle sizes As a result, the air pollution (atmospheric aerosols) assessment drew great attention to the people At present, the ground- based measurements and remote sensing observations are the general approaches for air quality monitoring Since the atmospheric aerosols has diverse temporal variation, this study presents an integration of both ground-based (AERONET) and satellite measurements (MODIS AOD) with regards to the seasonal variation from a 5-year worth of data throughout Tainan, Taiwan (22.9997° N, 120.2270° E) The collected data suggested various aerosol accumulations and dominance in each season with corresponds to precipitable water With attention to the yearly AOD, 2014 was shown with the highest record of AOD (675 nm) at 1.142 On the other hand, 2011 was shown with the lowest record of AOD (675 nm) at 0.225 In addition, months of March and April (spring) were both found with the highest peak of AOD (675 nm) at 0.943 and 1.142, respectively In contrast, June and July (summer) were the months recorded with accordingly lowest peak of AOD (675 nm) at 0.224 and 0.225 High density of larger Ångström exponent (fine mode particles more than 90% from its total mode derivation from Standard Deconvolution Algorithm (SDA) were detected all year round which indicates the fine particulate matter (PM) dominance in all seasons The magnitude of AOD (aerosol loading) are found low in correspond with aerosol removal due to high moisture content in the months of May to July Meanwhile, the air mass flows of back trajectory monthly map from HYSPLIT model suggested that aerosols observed within study area are primarily influenced by the transportations from East China Sea as well as South East Asia regions

Keywords AERONET, MODIS AOD, Ångström Exponent, Precipitable

Water, SDA, HYSPLIT Trajectory

ACKNOWLEDGEMENT

First and foremost, I am entirely grateful to The Almighty God as well as my family

(Nanay Rosie, Tatay Jing, Kuya Aba and Kuya Edward) and friends (PTAN, Anne, Kat, Mishel, Ken and Martina) for giving me the strength and provision that helped brought me

in completion of this research

I wish to express my sincerest thanks to my research professor, Assoc Professor Tang Huang Lin, Ph.D who is in spite of busy schedule still manages to supplement me the additional knowledge I ought for, along with the necessary facilities needed for my research at Center for Space and Remote Sensing (CSRSR) of National Central University (NCU) I am also entirely grateful to Dr Do Thi Ngoc Oanh who guided me thoroughly with passion in order to present my paper ideally to the public

I place on record, my deepest thanks to Advance Education Program (AEP) for I consider myself a lucky individual, given the chance to meet and be part of the Environmental Sensing Laboratory who gave me such wonderful lab mates including, Mr Wei Hung Lien and Ms Chang Yi-Ling who supported me with great patience throughout my research My salute goes to all the coding you’ve done with different software just to retrieve a 5 value point (thought they were 12, sorry 100x) that I didn’t event got to use in the end I am really sorry guys I’ll make sure to make your teas next time we meet, I promise

Great appreciation also goes to my family in Christ at NCU International Fellowship for the endless prayer and support, especially to Yu Tang Chien for helping me out a lot with

my data despite her own busy schedules I would also like to include a special note of ―謝謝我醜陋的朋友們‖ to Mayor, Mayora, Ate Shawie, Batang Hamog, Mr Right, Walao

Eh Mommyta, and 美国人 (solely educational purposes) for making me fat and keeping

me sane as I probably could have just turned into a complete skinny psycho because of how stubborn my data are

Thai Nguyen, 25/09/2017 ERIKA ROMERO PADLAN

TABLE OF CONTENTS

ACKNOWLEDGEMENT iii

LIST OF FIGURES vi

LIST OF ABBREVIATIONS ix

PART I INTRODUCTION 1

1.1 Research Background and Rationale 1

1.2 Objectives of the Study 4

1.3 Scope of the Study 4

1.4 Statement of the Problem 5

1.5 Limitations 6

1.6 Requirement 6

PART II LITERATURE REVIEW 7

2.1 Air Pollution: Atmospheric Aerosols 7

2.2 AERONET– Ground-Based Measurement 10

2 3 MODIS – Satellite Imagery of Aerosol Optical Depth 14

2.3.1 NASA EOSDIS: Worldview Application 15

2.3.2 NAAPS- Aerosol Modelling 16

19

2.4 Air Mass Trajectory 20

2.4.1 Concept of Trajectory 20

2.4.2 Application of Trajectory 21

PART III DATA AND METHODOLGY 24

3.1 Data Collection 24

3.1.1 AERONET – Ground-based measurement of Aerosol Optical Depth (AOD) 24

3.1.2 MODIS – Satellite Imagery of Aerosol Optical Depth and Ångström

Exponent 26

3.1.3 Navy Aerosol Analysis and Prediction System (NAAPS) – Aerosol Modelling 27 3.1.4 Hybrid Single Particle Lagrangian Integrated Trajectory Model (HYSPLIT) READY – Back Trajectory Modelling 27

3.1.5 Wind field 28

3.2 Methodology 28

PART IV RESULT AND DISCUSSION 32

4.1 Annual and Monthly Aerosol Optical Depth 32

4.2 Spectral Deconvolution Algorithm (SDA) 41

4.3 NAAPS Aerosol Modelling 44

4.4 HYPSLIT READY Backward Trajectory 48

PART V CONCLUSION 51

CITED REFERENCES 54

LIST OF FIGURES

Figure 1 Number of deaths by thousands attributed to combined household (HAP) and

ambient air pollutant (AAP)………8

Figure 2 Distributed AERONET sites throughout the globe 11

Figure 3 Chen-Kung_Univ AERONET site Version 2 Level 2 Aerosol Optical Depth

at 340nm, 380nm, 440nm, 500nm, 675nm, 870nm and 1020nm from year 2009 12

Figure 4 NAAPS 4-panel surface aerosol concentration model with total optical depth

(Sulfate: Orange/Red; Dust: Green/Yellow; Smoke: Blue) at the upper left, sulfate at the upper right, dust at the lower left and smoke at the lower right corner at 00:00Z 15

August, 2017 in South East Asia 19

Figure 5 Applications of air mass trajectories in different fields (Umesh K and Bablu

K., 2014) 22

Figure 6 (a) Chen-Kung_Univ AERONET site’s availability of Level 2 daily average

of AOD by year (b) Chen-Kung_Univ AERONET site’s relative contribution of day to

total record (%) 25

Figure 7 Flow chart of the methodology used in the study 31

Figure 8 (a) Monthly AOD mean at 675nm obtained from year 2009, 2010, 2011,

2013, and 2014 at Chen-Kung AEROENET site 32

Figure 8 (b) Yearly AOD mean at 675nm obtained from year 2009, 2010, 2011, 2013,

and 2014 at Chen-Kung AEROENET site 33

Figure 9 Ångström exponents at 440-870 nm vs AODs at 675 nm in year 2009,

2010, 2011, 2013 and 2014 respectively 34

Figure 10.a Taiwan MODIS Merged Dark Target/Deep Blue AOD (Land and

Figure 11.c Taiwan MODIS Deep Blue Ångström Exponent from July, 2014 40

Figure 11.d Taiwan MODIS Deep Blue Ångström Exponent from October, 2014 40

Figure 12 Monthly SDA retrievals from years 2009-2011 and 2013-2014 from

Chen-Kung_Univ AERONET Site 41

Figure 13 Mean bar graph of total mode aerosol and PW in each month of 2009-2011

and 2013-2014 43

Figure 14 NAAPS 4-panel surface aerosol concentration model with total optical

depth (Sulfate: Orange/Red; Dust: Green/Yellow; Smoke: Blue) at the upper left, sulfate at the upper right, dust at the lower left and smoke at the lower right corner at

00:00Z 13 January 2014 in South East Asia 44

Figure 15 NAAPS 4-panel surface aerosol concentration model with total optical

depth (Sulfate: Orange/Red; Dust: Green/Yellow; Smoke: Blue) at the upper left,

sulfate at the upper right, dust at the lower left and smoke at the lower right corner at

00:00Z 13 April 2014 in South East Asia 45

Figure 16 NAAPS 4-panel surface aerosol concentration model with total optical

depth (Sulfate: Orange/Red; Dust: Green/Yellow; Smoke: Blue) at the upper left, sulfate at the upper right, dust at the lower left and smoke at the lower right corner at

00:00Z 13 July 2014 in South East Asia 46

Figure 17 NAAPS 4-panel surface aerosol concentration model with total optical

depth (Sulfate: Orange/Red; Dust: Green/Yellow; Smoke: Blue) at the upper left, sulfate at the upper right, dust at the lower left and smoke at the lower right corner at

00:00Z 13 October 2014 in South East Asia 47

Figure 18 Air parcel trajectory map of Tainan, Taiwan from months of January,

April, July and October in year 2014 49

Figure 19 Seasonal Wind Map at 850 hPa 50

LIST OF TABLES

Table 1 Back Trajectory Model Parameters Selected 30

Table 2 Yearly AOD (675 and 440nm), and AE statistics obtained from

Chen-Kung_Univ AERONET site 36

Table 3 Monthly SDA statistics from 2009, 2010,2011,2013 and 2014 at

Chen-Kung_Univ AERONET site 43

LIST OF ABBREVIATIONS

AOD Aerosol Optical Depth

AOT Aerosol Optical Thickness

AE Ångström Exponent

PW Precipitable Water

SDA Standard Deconvolution Algorithm

NRL Naval Research Laboratory

NAAPS NRL Aerosol Analysis and Prediction System

ARL Air Resources Laboratory

AERONET Aerosol Robotic Network

MODIS Moderate Resolution Imaging Spectroradiometer

READY Real-time Environmental Applications and Display

HYSPLIT Hybrid Single-Particle Lagrangian Integrated Trajectory

WHO World Health Organization

USGS United States Geological Survey

EOSDIS Earth Observing System Data and Information System

DT Dark Target

DB Deep Blue

GIBS Global Imagery Browse Services

AGL Above Ground Level

NOAA National Oceanic and Atmospheric Administration

NASA National Aeronautics and Space Administration

GDAS Global Data Assimilation System

hPa hectopascals

UTC Coordinated Universal Time

τa aerosol optical depth

PART I INTRODUCTION

1.1 Research Background and Rationale

Over the past decades, research has provided ample support for an ever-present ten-fold of inhalable solid particles which are drifting through the air just above the deserts, oceans, forests, mountains, ice and in every ecosystem entwined Known as aerosols, these particles float across from the Earth’s stratosphere down to the atmosphere and have a range of hardly a nanometer— smaller than the width of the smallest virus that makes it easily inhalable — to a several micrometers that is about the size of a single human hair (Adam Voiland, 2010) Aerosol properties are emitted from the surface of the Earth by means of both naturally (e.g., sea-salt, dust, biogenic, emissions), as well as an event of human activities (e.g., combustion of fossil fuels, land cover change and, etc.) Thus, directly related with their sizes, particles have different health effects (Wilson R & Spengler J., 1996)

Assigned as one of the major contributors to a wide variety of health problems than any other pollutant worldwide, World Health Organization (WHO) estimated that particle pollution contributes to roughly 7 million premature deaths per year; establishing it as the leading cause of mortality worldwide As particulate matter (PM) develops high concentration, it alters risky health issues in particular with decline in lung functions that can result to asthma and other respiratory issues (Nwafor et al., 2007)

Alongside the thriving economic growth of USA, Western, Europe, Southeast and East Asia, regional emissions of air pollutants are rapidly increasing over the past decades (Kato and Akimoto, 1992; Streets et al., 2003) As a result, great interest has been invested on observing air quality Monitoring approaches such as ground-based and satellite measurements are taken as a methodical solution to the worsening rapid air ambient quality

Ground-based measurements such as AERONET (AErosol RObotic NETwork)

that has been established by both NASA and PHOTONS (PHOtométrie pour le

Traitement Opérationnel de Normalisation Satellitaire; Univ of Lille 1, CNES, and CNRS-INSU) provides better time resolution and accuracy, however, its effectivity in terms of observation range is restricted by the given settled location of the instrument On the contrary, satellite remote sensing is now a rapidly developing field for its favorable properties in particular with global coverage, satisfactory spatial resolution and near-simultaneous view over a wide region (Bicen Li and Lizhou Hou, 2015; P.Veefkind, et al., 2007)

Thus, to carry out an assessment throughout a region, physical modelling is usually employed which heavily depends on meteorological modelling and emission inventories Correspondingly, remote sensing facilitated and developed it through the use of satellites (P Veefkind, et al., 2007)

MODIS (Moderate Resolution Imaging Spectroradiometer) is the vital payload

aboard the Aqua (primarily known as EOS PM-1) and Terra (primarily known as EOS

AM-1) satellites As both Aqua and Terra MODIS views the entire Earth’s surface every 1 to 2 days which receive 36 spectral bands of data, these data enhances the comprehension of global dynamics as well as various episodes occurring on the oceans, lands, and within the atmosphere Through MODIS advancement, simulation/prediction of global change can be now foreseen accurately for the policy makers in making decisions regarding the environmental protection

Though both data from ground-based and satellite provides an excitement on global air quality observations, satellite has a disadvantage of providing spatial and temporal resolution In other words, to greatly benefit from the space borne measurements, combination of both ground-based and satellite measurements is suggested (P Veefkind, et al., 2007)

In this paper, an inspection of air pollution based on both ground-based and satellite measurements are supported to investigate the current extent of present atmospheric aerosols along with its transporting origins The approach is to examine and compare the states of the atmospheric aerosols in each season over the past 5 years’ worth of data in Tainan, Taiwan using AERONET and MODIS for both ground-based and satellite measurements With this application, NAAPS model plays

a major role in estimating the surface concentrations of sulfate, dust and smoke along the region of interest

1.2 Objectives of the Study

The objectives of the study are given as follow:

1 To inspect the aerosol concentration in terms of aerosol optical depth in Tainan, Taiwan during the study period

2 To figure out which season and year presents significant variance in aerosol optical properties

3 To find out the majority of aerosol mode within the study region

4 To demonstrate which aerosol concentration (level of aerosol loading) is majorly displayed in the region

5 To determine the initial/potential pathways of the aerosols before it reached the region of interest (long range transportation)

6 To examine the efficiency of both ground-based and satellite observation in evaluating the air pollution in Tainan, Taiwan

1.3 Scope of the Study

This study is designed to address the current air pollution that is present within

2009, 2010, 2011, 2013 and 2014 at Tainan, Taiwan by integrating both ground-based and satellite measurements The study focused on both AOD and precipitable water in order to measure the present atmospheric aerosols within the vicinity of Tainan, Taiwan For detecting the source regions and pathways of the observed transported air pollutants, the wind fieldin each season of year 2009 to 2011 and year 2013 to 2014

a 7 day back trajectory from Air Resources Laboratory’s (ARL) new web based HYSPLIT system called READY (real-time environmental applications and display) was also taken into account of detecting the source regions of the air pollutants Accordingly, NAAPS’ aerosol modelling was carried out to present the estimated surface concentrations of sulfate, dust and smoke observed within the study area

1.4 Statement of the Problem

For the study to be considered a success, it should be able to answer each following questions:

1 How much aerosol concentration is present over Tainan, Taiwan within the study period?

2 Which year and month can be seen bringing a noticeable effect on its aerosol optical properties?

3 What mode of aerosol is majorly presented in the study site?

4 Among the estimated surface concentrations, which one is abundant?

5 Where are the source regions of all the air pollutants coming from?

6 How does the integration of ground-based and satellite measurements in favor

of evaluating air pollution in Taiwan region?

1.5 Limitations

 As the integration of both ground-based and satellite measurements is still viable, the paper limited itself in categorizing the present aerosol types using its ground-based measurements within the study area

 The paper only presented data from a single station

 Aerosol transports across the international boundaries were only put into focus and thus, setting aside local boundaries

1.6 Requirement

i Acquisition of a 5 year (2009,2010,2011,2013 and 2014) worth of ground-base

measurement data from AERONET which includes Aerosol Optical Depth

(AOD), Precipitable Water (PW), and Ångström Exponent(AE)

ii Acquisition of MODIS AOD products with 3km Land and Ocean through EOSDIS Worldview for the seasonal months of 2014

iii Generation of systematic AOD visualization map and its parameters through

ArcMap and Matlab

iv Recovery of seasonal back trajectories from NOAA ARL’s READY HYSPLIT

v Creation of seasonal wind field maps throughout the whole study period using

GRIB software

PART II LITERATURE REVIEW

2.1 Air Pollution: Atmospheric Aerosols

Alongside the thriving economic growth of USA, Western, Europe, Southeast and East Asia, regional emissions of air pollutants are rapidly increasing over the past decades (Kato and Akimoto, 1992; Streets et al., 2003) Alteration of the aerosol environment through combustion of fossil fuels, land cover change, and the contact to particulate and gas species to the atmosphere is introduced by man (Holben, B N., et

al 2001) As an outcome, an ever-present ten-fold of inhalable solid particles are recorded to be drifting through the air just above the deserts, oceans, forests, mountains, ice and in every ecosystem linked These inhalable solid particles, also

known as atmospheric aerosols float across from the Earth’s stratosphere down to the

atmosphere and have a range of hardly a nanometer (smaller than the width of the smallest virus) that makes it easily inhalable, to a several micrometers that is about the size of a single human hair (Adam Voiland, 2010)

Seinfield and Pandis in year 2006 claimed that aerosol is not only directly diffused through natural resources and anthropogenic agents but also by means of various physical and chemical processes Given as sea salt, windblown dust, volcanic ash, pollution from factories and smoke from wildfires, aerosols can either cool or warm the surface defending on their size, type and location; they are proven to impact the climate system by absorbing and reflecting solar radiation as well as altering the

cloud properties Stated as particulate matter (PM), aerosols were also known to expose human to serious health damages (Pope et al., 2009)

In support to Pope et al., (2009), World Health Organization (WHO) in year

2012 named particle pollution as one of the major contributor to a wide variety of health problems than any other pollutant worldwide Contributing to roughly 7 million premature deaths per year, it was established as the leading cause of mortality worldwide

Figure 1 Number of deaths by thousands attributed to combined household (HAP) and

ambient air pollutant (AAP)

HAP: Household air pollution; AAP: Ambient air pollution; Amr: America, Afr: Africa; Emr:

Eastern Mediterranean, Sear: South-East Asia, Wpr: Western Pacific; LMI: Low- and

middle-income; HI: High-income (Public Health, Social and Environmental Determinants of Health

Department, World Health Organization, 2014)

In relation to the health problems, atmospheric aerosols also play a vital role in the regional and global climate change (IPCC, 1996, 2001; Haywood and Boucher, 2000) Kaufman’s findings (2002) lend support that although aerosol particles displays

a significant importance physically (human health) and climatically, understandings of the impact of the atmospheric aerosols underlying on the Earth’s climate structure is poorly recognized; and thus, brings great interest in the field of science as well as policy making

Aerosol optical depth (AOD) is the only variable that can be easily understood to remotely assess the aerosol loading from ground-based instruments (B N Holben et al., 2001) It is the measured extinction of the solar beam by haze and dust That is to say, particles that are in the atmosphere can either scatter or absorb sunlight Aerosol optical depth tells the extent of the prevented direct sunlight from reaching the ground

by these aerosol particulates In particular, AOD is a dimensionless measurement which is linked to the aerosol concentration within the vertical column in the atmosphere along the observed region (SURFRAD Aerosol Optical Depth, n.d.)

AOD is a variable used in local analysis to assess atmospheric pollutions, characterize aerosols, and make atmospheric alterations to satellite remotely sensed data

Accordingly, the growing continuous aerosol measurements are one of the most effective ways of monitoring regional and global aerosol properties, calibrating satellite remote sensing instruments and examining model results (Huebert et al., 2003; Holben

et al., 1998) It contributes in creating a global imagery of aerosol distributions for operations such as, assistance from surface measurements (e.g., Dubovik et al., 2002), satellite aerosol products (e.g., Remer et al., 2005; Torres et al., 1997), and aerosol transport models (e.g., Tegen et al., 1997; Chin et al., 2002)

2.2 AERONET– Ground-Based Measurement

Noted as the simplest and fundamentally the most accurate and easiest to manage monitoring systems, ground-based measurements draws to provide aerosol optical depth which is the single most understandable parameter in examining the aerosol burden present in the atmosphere This parameter is widely integrated to characterize aerosols as well as to assess atmospheric pollution in order to make atmospheric adjustments to satellite remotely sensed data (Holben, B N., et al 2001)

Supported by NASA’s Earth Observing System and widened by many

non-NASA institutions, the monitoring system AErosol RObotic NETwork (AERONET)

is granted to monitor and archive ground-based aerosol optical measurements through network hardware that consists of alike automatic sun-sky scanning spectral radiometer governed by universities, institutes, individual scientists, partners and national agencies The data presents globally distributed near real time measurements of aerosol size distributions, aerosol optical depths, and precipitable water in various aerosol system

Each data is set to see preliminary processing (real time data) and reprocessing that includes a final calibration within 6 to 12 months after the data are collected In addition, the data are also set to have undergone a quality assurance, archiving and dispersion from NASA’s Goddard Space Flight Centre master archive and other similar data based maintained globally As a result, AERONET’s affiliation supports the globally dispersed observational measurements that are calculated for three data quality levels: Level 1.0 (unscreened), Level 1.5 (cloud-screened), as well as Level 2.0 (cloud screened and quality assured) From those levels, precipitable water, inversion products, and other AOD-dependent products are adopted and may grant supplementary quality inspection

Figure 2 Distributed AERONET sites throughout the globe (https://aeronet.gsfc.nasa.gov)

Designed to all AERONET stations, spectral observations of sun and radiance are given at various nominal wavelengths, with AOD derived from sun measurements The certainty of every attained AOD measurement is predicted to be on the line of

±0.03 in regards of level 1.5 Moreover, in view of validating forecast, the AOD at 550

nm is retrieved through linear interpolation of the two nearest wavelengths

On the other hand, while AOD is considered as the optical thickness of an aerosol, the Ångström Exponent (AE) mainly presents the spectral dependence of AOD Thus, for AE at 440-870 nm, the wavelengths include 440,500,670,870 nm AOD measurements In addition, for a polarized instrumentation of AE at 440-870 nm, the wavelengths only include 440,670 and 870 nm AOD data as polarized instruments

do not provide 500 nm channels

Figure 3 Chen-Kung_Univ AERONET site Version 2 Level 2 Aerosol Optical Depthat 340nm, 380nm, 440nm, 500nm, 675nm, 870nm and 1020nm from year 2009.

At a given wavelength, AOD relies on the aerosol column concentration while

AE mainly relies on aerosol size distribution The correlation between fine and coarse aerosols right in the atmosphere can be pursued through AE recognition One feature is that, since Ångström exponent is inversely proportional to the average size of a particle

in an aerosol, the larger the exponent, the smaller the particles is Thus, an increase of fine particles signifies low dust plume amid an event

A Smirnov et al., (2002) analysed the diurnal variability of aerosol optical depth observed at AERONET sites Through the study, it was found that an existing increase pattern of aerosol optical depth by 10-40% over the region of major urban/industrial areas are displayed during the day Particularly, a range of 10% dust was recorded as the major contributor to aerosol optical depth

Investigated by T F Eck et al., (2005), a column-integrated aerosol optical properties were put into focused within Asia and mid tropical Pacific using sun/sky radiometer measurements done at AERONET The result showed that at sites in Mongolia, China, South Korea, and Japan, aerosol optical depth was at its maximum in summer/spring and at its minimum in winter Accordingly, the annual ångström wavelength exponent showed with a minimal activity during spring time Regardless of the result, the ångström exponent 440–870 was found to be greater than 0.8 which suggest fine mode pollution at Eastern Asian sites that originated from Western source regions

Moreover, a study done by C Toledano et al., (2007) conducted a routine aerosol measurement, of which, had been done at INTA-EL Arenosillo (Huelva, Spain) from February 2000 through a Cimel sun photometer used within the AERONET network With a 5-year aerosol database, it allowed aerosol properties to be characterized and classified, which then, helped determined the local aerosol climatology The AOD results showed two peaks during the year, one during the summer and one at the end of the winter, of which, said to be the desert dust aerosols that has been transmitted from North Africa to South-Western Iberian Peninsula In addition, the ångström exponent showed coastal marine aerosols and desert dusts as the two main aerosol types that were presented over El Arenosillo Influenced by local sources of continental and polluted aerosols, 66% of the main aerosol scenarios were found to be coastal marine aerosols On the other hand, while 11% was determined as continental aerosols, desert dust showed a very significant amount of 20% of the data

2 3 MODIS – Satellite Imagery of Aerosol Optical Depth

Moderate-resolution Imaging Spectroradiometer (MODIS) offers 36 spectral bands (0.4-14.4 micrometer) in a high radiometric sensitivity Provided that, it gathers measurements all day, every day, to which it is accompanied by a wide field of view, this continuous and thorough coverage grants MODIS to finish an electromagnetic image of the Earth in every two days,

Jun Wang along with Sundar A Christopher (2005) explored the relationship between column aerosol optical thickness (AOT) that has been derived from MODIS

Terra/Aqua satellites, and the hourly fine particulate mass ( ) gathered from surfaces of the seven selected locations within Jefferson county, Alabama for 2002 The results introduced that between MODIS AOT and , a good correlation does not exist (linear correlation coefficient, R= 0.7) which signified that most of the observed aerosols are well-mixed in the lower boundary layer during the satellite overpass

2.3.1 NASA EOSDIS: Worldview Application

A rapidly growing literature on Aerosol Optical Thickness or Aerosol Optical Depth (AOD) displays the level in which the particles present in the air forbids light from roaming through the atmosphere (Frank G and Chester O., 1949) As aerosols scatter and absorb incoming sunlight, it also decreases visibility Viewed from an observer on the ground, an AOD with less than 0.1 is considered to be ―clean‖ – which has a characteristic of a bright sun, clear blue sky, and maximum visibility Mentioned by multiple researches, as AOD increases

by 0.5, 1.0, and <3.0, aerosols grow so dense that it can block the sun Moreover,

as aerosols are proven hard to be retrieved and identified when they appear in

various types of ocean and land surface, worldview application supports several

various kinds of imagery layer to support AOD classification

The application implements the ability to interactively scan global, resolution satellite imagery and later download the underlying inputs Additionally, the application also offers an approximate 400 plus available data products that are

full-updated within the next three hours of observation In fact, worldview helps assist

time-critical utilization areas such as air quality measurements, wildfire management and flood monitoring tool that can be easily access through tablet and

smartphones In particular, it also employs Global Imagery Browse Services

(GIBS) for a rapid interactive imagery retrieval experience With GIBS imagery, worldview’s layers can be easily accessed through NASA World Wind, Google Earth, and several other clients

The merged MODIS Dark Target (DT)/ Deep Blue (DB) AOD layer supports a more global, synoptic view of MODIS AOD over ocean and land The layer is made from three algorithms: two DT assigned for retrieval of (1) over dark-soiled land/ vegetated land (dark in visible) and (2) over ocean (dark in visible and longer wavelengths) and the DB algorithm which is initially made for retrieving (3) over arid land/desert (bright in the visible wavelengths)

2.3.2 NAAPS- Aerosol Modelling

Developed by Naval Research Laboratory (NRL) in Monterey, California, the NRL Aerosol Analysis and Prediction System (NAAPS) were designed as a near-operational system for predicting tropospheric aerosol distributions Modified by Christensen (1997), NRL utilizes global meteorological fields from Navy

Operational Global Atmospheric Prediction System (NOGAPS) (Hogan and

Rosmond, 1991; Hogan and Brody 1993) investigations and thus, forecasts with a

1 x1 degree grid at a given 6-hour intervals and 24 vertical levels stretching up to a

hundred megabytes (12 levels for the initial runs while 18 levels starting from

1998062412 as well as 24 levels starting from 2002091700.) In particular, NRL

used northern hemispheric with 12-hourly ECMWF fields on its 2.5 X 2.5 degree grid from its original model

As an illustration, NAAPS displays regional plots in a 4-panel format as single figures or as a 2-day loop having 9 figures (6-hourly plots) in it The optical depths at a wavelength of 0.5 micron were taken for the three components: sulfate, dust and smoke Since 1999, NAAPS had used a specific extinction formula based

on Nemesure et al 1995 (JGR, vol 100, p26105-26116)

For sulfate, the formula extends to present values of 4.5, 5.1, 7.2, 15.0, and 31.g with respective relative humidity of 30, 50, 70, 90, and 98% and a maximum granted value of 40.3 at a relative humidity of 100% Considering the application used, the approach yielded optical depths that are up to

10 times larger than those retrieved using the previous application (assuming particles with 0.1 microns in radius, no swelling as a result of humidity aftermath and refractive index of 1.55 + 01i, the Mie theory derived the specific extinction

of less than 1.0 ) and thus present a good qualitative concurrence with NOAA/NESDIS AOD data off China and the east coast of United States The sulfate contouring starts at 0.01 and doubles in magnitude for individual succeeding contour

As for dust, a 1 micron radius and a refractive index of 1.55 + 001i particles are assumed with specific extinction of 0.56 The contouring for dust starts at 0.05 and doubles in magnitude for individual succeeding contour as well Subsequently, smoke particles at 0.01 microns in radius and a refractive index of 1.55 + 0.1i were estimated with exact extinction of 7.1 For its contouring, it begun at 0.01 and also doubles with its succeeding contour

Figure 4 NAAPS 4-panel surface aerosol concentration model with total optical depth (Sulfate:

Orange/Red; Dust: Green/Yellow; Smoke: Blue) at the upper left, sulfate at the upper right, dust

at the lower left and smoke at the lower right corner at 00:00Z 15 August, 2017 in South East Asia

2.4 Air Mass Trajectory

2.4.1 Concept of Trajectory

J.A Dutton (1986), H R Byers (1974), D.J Thomson along with J D Wilson (2012) put forward the view that air flow can be described into two different manners: (i) Eulerian (coined after the Swiss mathematician Leonhard Euler) and (ii) Lagrangian (coined after the French mathematician J.L Lagrange) Thomson’s and Wilson’s (2012) findings lend support to differentiate that the air flows through fixed points in the space, and while in Lagrangian approach, individual air parcel is selected and gets

to be followed while it moves through space and time

Whereas, the Eulerian model is stated to be the most followed one when it comes

to chemistry modelling as a number of papers claim that it to be a convenient tool in explaining various physical and chemical processes In terms of the Eulerian model, calculations of the chemical reactions are solely based on the pollutant’s concentration which has been diluted over the whole grid scale Most dispersion and transport models use Lagrangian approach because of some minor limitations accompanied by the Eulerian model; for instance are the convective transport and boundary layer on top entrainment (Seibert P., 2004; Davis K J et, al., 1997)

Along similar lines, K.J Davis and his coworkers (1997) supported that an advantage of a Lagrangian model is that it has minimal numerical diffusions Through calculation of air mass trajectory, an infinitesimal air parcel over a centerline of an air mass advection with horizontal and vertical dispersion path can be displayed

Following an air parcel pathway upwind the chosen coordinates is described as

―backward air trajectory,‖ whereas, the calculation of the finest potential pathway to be pursued downwind from the chosen coordinates through the course time is termed

―forward trajectory‖

Integrating the calculation of backward air trajectory through a Lagrangian approach is simpler as well as computationally cheap for it eliminate the upwind influence to the receptor site (A Eliassen et al., 1982; D Simpson, 1983)

2.4.2 Application of Trajectory

Trajectories were formerly used to find the source region and the transport movement of air pollution However, along with knowledge advancement is the discovery of useful various trajectory aspects that are helpful in relation to atmosphere

and environment, as displayed in (Figure 5) Trajectories are integrated in a number of

fields including climatology (R.R Draxler and A.D Taylor, 1982), meteorology (R.R Draxler and A.D Taylor, 1982), transport of pollutants (E F Danielsen and R Bleck, 1967; A Stohl, 1998; D Wen et al., 2012; J.L Moddy et al., 1989), residence time inspection, source apportionment (B de Foy et.al., 2009; R.M Harrison, 2009), aerosol measurements (T.W Chan and M Mozurkewich, 2007; L A K Reddy et al., 20008;

P Salvador et al., 2008), precipitation chemistry (L Granat et al., 2002; J Satyanarayana et al., 2010), and policies (A Dvorska et at., 2009)

Figure 5 Applications of air mass trajectories in different fields (Umesh K and Bablu K.,

2014)

Utilizing backward and forward trajectories for achieving ozone maps in Northern America, Tarasick and coworkers (2010) developed an effective approach through linking ozonesonde data to the grind points of the trajectory pathways The approach resulted to a successful contribution of trajectory mapped ozone valued with justifiable accordance with the actual soundings Trajectories have been utilized in setting apart periods of stratosphere-troposphere (S-T) transfer in several regions of the United States (A S Lefohn et al., 2011)

Attempted by Ashrafi and his coworkers (2014), trajectory modelling have been used to investigate circulation of pollutants over Iran through dust model of HYSPLIT modelling which is primarily from dust emissions over desert areas Through their findings, the dust motion simulations gathered from the model were similar to one found from MODIS imageries

Applications

of Trajectories

Climate Science

Meteorology

Transport

Policy Quality Air

Aerosols and Precipitation Measurements Residence Time Analysis

Source Apportionm ent

Granted for a quick and accessible tool that allows them to investigate and predict the transport and dispersion of pollutants within the atmosphere, are air quality forecasters, aviation interests, government, emergency responders, international agencies and atmospheric researcher communities Thus, in order to support these services in a convenient format and timely manner, a newly web-based system was originally created in 1997 by National Oceanic and Atmospheric Administration's (NOAA) Air Resources Laboratory (ARL) and was termed Real-time Environmental Applications and Display system (READY, http://www.ready.noaa.gov)

READY has been deliberately expanded and maintained since then to support access to an advance new way of illustrating meteorological forecast and archive data

as well as creating air parcel trajectory and dispersion model by means of a series of user-interactive webpage

Thereupon, through READY, a HYSPLIT model can be presented with a cross calculation between Lagrangian approach and Eulerian methodology

PART III DATA AND METHODOLGY

3.1 Data Collection

3.1.1 AERONET – Ground-based measurement of Aerosol Optical Depth (AOD)

Majorly used for dust forecast validation on a daily basis, the AERONET

(AErosol RObotic NETwork) serves as an overlook monitor for optical based aerosol

as well as a data collection authorized by NASA’s Earth Observing system

In this paper, retrieved aerosol optical properties from ground-based measurements

of AOD from AERONET were selected from Chen-Kung_Univ site with coordinates (Latitude and Longitude) and elevation of 23.00000° north, 120.21667° East and 50.0 meters, respectively The data were collected from Version 2 AOD with Level 2 data type (quality assured) which is already cloud cleared, pre- and post-field calibration applied and has been manually inspected

Thus, in order to avoid biases in the study, years that were selected were year’s equivalent with at least half a year worth of data (or available data with more than 170 observations at least) As a result, from years of 2009, 2010, 2011, 2013 and 2014, measurements of aerosol optical properties such as AOD (τ) with ångström parameter

at 440- 870nm (α), precipitable water (PW), and SDA were collected