Next Article in Journal
Deep Unfolding Sparse Bayesian Learning Network for Off-Grid DOA Estimation with Nested Array
Next Article in Special Issue
Increased Warming Efficiencies of Lake Heatwaves Enhance Dryland Lake Warming over China
Previous Article in Journal
Prediction of Sea Surface Reflectivity under Different Sea Conditions Based on the Clustering of Marine Environmental Parameters
Previous Article in Special Issue
Analysis of Land Surface Temperature Sensitivity to Vegetation in China
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Remote Sensing
Volume 15
Issue 22
10.3390/rs15225319
remotesensing-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles thumb_up
...
Endorse textsms
...
Comment
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Integrated Remote Sensing Observations of Radiative Properties and Sources of the Aerosols in Southeast Asia: The Case of Thailand

by
Arika Bridhikitti
1,2,*,
Pakorn Petchpayoon
3 and
Thayukorn Prabamroong
4
1
Environmental Engineering and Disaster Management Program, School of Interdisciplinary Studies, Mahidol University Kanchanaburi Campus, Kanchanaburi 71150, Thailand
2
Earth Science Research Center, Mahidol University Kanchanaburi Campus, Kanchanaburi 71150, Thailand
3
Geo-Informatics and Space Technology Department Agency (GISTDA), Bangkok 10210, Thailand
4
Faculty of Environment and Resource Studies, Mahasarakham University, Mahasarakham 44150, Thailand
*
Author to whom correspondence should be addressed.
Remote Sens. 2023, 15(22), 5319; https://doi.org/10.3390/rs15225319
Submission received: 2 September 2023 / Revised: 13 October 2023 / Accepted: 17 October 2023 / Published: 10 November 2023
(This article belongs to the Special Issue Climate and Environmental Changes Monitored by Satellite Remote Sensing III)
Browse Figures
Review Reports Versions Notes

Abstract

:
Aerosols in Southeast Asia (SEA) are entangled with complex land–sea–atmosphere–human interactions, and it is difficult for scientists to understand their dynamic behaviors. This study aims to provide an insightful understanding of aerosols across SEA with respect to their radiative properties using several lines of evidence obtained from remote sensing instruments, including those from onboard Earth observation satellites (MODIS/Terra and MODIS/Aqua, CALIOP/CALIPSO) and from ground-based observation (AERONET). The findings, obtained from cluster analysis of aerosol optical properties, showed seven aerosol types which were dominant across the country, exhibiting diverse radiative forcing potentials. The light-absorbing (prone to warm the atmosphere) aerosols were likely found in mainland SEA, both for background and high-aerosol events. The light-scattering aerosols were associated with aging processes and hygroscopic growth. The neutral potential, which comprised a mixture of oceanic and local anthropogenic aerosols, was predominant in background aerosols in insular SEA. Further studies should focus on carbonaceous aerosols (organic carbons, black carbon, and brown carbon), the aging processes, and the hygroscopic growth of these aerosols, since they play significant roles in the regional aerosol optical properties.

Graphical Abstract

1. Introduction

Understanding atmospheric aerosols in Southeast Asia (SEA) presents several challenges due to the region’s unique geographical, meteorological, and anthropogenic characteristics [1,2]. The region often experiences high levels of aerosol concentration from various sources, such as biomass burning, urbanization, and industrial activities [3]. The high aerosol levels have adversely affected human health, visibility, and the climate. There is a diversity of aerosol types in this region, including organic carbon, black carbon, mineral dust, and sea salt, and interactions between these aerosols further complicates the understanding of their properties and their radiative forcing potentials [4,5,6]. The climate in this region is influenced by monsoon systems, resulting in distinct wet and dry seasons, which significantly influence aerosol production and transport. Aerosols in SEA can undergo long-range transport, affecting neighboring countries and reaching the Pacific region [1,5]. Furthermore, the location of a tropical heat source affects aerosol dispersion, vertical mixing, and interactions between aerosols and clouds [4,7,8]. Understanding aerosol sources, transportation, interaction, and effects is essential for developing effective mitigation strategies for climate change and air pollution in SEA.
The net radiative forcing of aerosols depends on the balance between their scattering and absorption properties, as well as their spatiotemporal distribution in the atmosphere. The radiative forcing of aerosols can significantly affect the Earth’s climate, surface temperatures, atmospheric stability, cloud formation, and precipitation patterns [9]. Aerosols can scatter incoming solar radiation back to space, reducing the amount of sunlight reaching the Earth’s surface, leading to a cooling effect. The backscattering efficiency depends on the aerosol’s size, shape, and refractive index [10], with particles of about 300 to 700 nm (similar to the wavelength of light) in size, and higher real-part refractive indices generally scatter more effectively. Single-scattering albedo (SSA) quantifies the fraction of radiation scattered relative to the total radiation interaction with the aerosols. Higher SSA values indicate aerosols that scatter more efficiently, whereas lower SSA values indicate aerosols that exhibit stronger light absorption. Some aerosols absorb solar radiation, converting it into heat, resulting in a warming effect. The absorption efficiency depends on the aerosol’s composition, with materials like black carbon (soot) being particularly effective absorbers [10]. Understanding these optical properties is essential for assessing the climate impacts of aerosols and developing effective climate mitigation and adaptation strategies.
Due to geographical constraints, there are gaps in the spatiotemporal coverage of aerosol monitoring networks, especially in remote and rural areas in Southeast Asia. Remote sensing, therefore, plays a crucial role in monitoring aerosols on a regional scale. However, uncertainties in satellite retrievals, especially in the presence of clouds, can introduce errors in the estimation of aerosol properties [3]. The Aerosol Robotic Network (AERONET), equipped with ground-based sun–sky scanning radiometers, has become a prominent tool for investigating atmospheric aerosols. In a study by Dubovik et al. (2000) [11], the AERONET inversion algorithm for retrieving aerosol optical properties was tested, and the analysis shows successful retrieval of all aerosol characteristics, although limited accuracy was noted under conditions of low aerosol loading [11]. Moreover, it is noteworthy that the ground-based aerosol data are exclusively available from AERONET stations, providing localized information rather than comprehensive spatial coverage.
Therefore, integrated multi-satellite retrievals and ground observations could overcome these weaknesses. Remote sensing technology provides aerosol and gaseous compositions on a total atmospheric column basis. The spectrophotometer is designed to observe solar radiation and measure the extinction (both scattering and absorption) of the solar beam attributed to aerosol, called aerosol optical depth (AOD). Satellite observations provide strong connections between hot spots and pollution plumes, or between urban land cover and heat island zones. Some satellite sensors, such as the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP), provide aerosol vertical profiles beneficial for the identification of aerosol types and sources [6]. Furthermore, ground-based remote sensing, such as AERONET, is typically used to validate satellite products. It can provide the scattering and extinction properties of the atmospheric aerosols, with a high temporal resolution, by tracking the direct sun and sky radiances [12].
Remote sensing technologies have been universally used to study air pollution, particularly biomass-burning smoke aerosols in SEA. Nguyen et al. (2019) [1] conducted a spatiotemporal analysis of ground and satellite-based AOD for air quality assessment in SEA. They found high AOD concentrations in major urban areas. The AOD levels could be related to biomass burning and the transportation of Indonesian forest fire smoke, in some countries [1]. Feng and Christopher (2013) [3] employed satellite retrievals, including datasets of the Moderate Resolution Imaging Spectroradiometer (MODIS), the Multi-Angle Imaging Spectroradiometer (MISR), and the Clouds and The Earth’s Radiant Energy System (CERES) from the Terra satellite to assess the spatiotemporal distributions of aerosol properties across the SEA region. They discovered dominant fine aerosol particles and their net negative forcing potentials. Lin et al. (2014) [7] demonstrated the advantages of integrated spaceborne and ground-based remote sensing instruments for smoke aerosol–cloud interaction studies in SEA.
This study took advantage of integrated spaceborne and ground observations to provide the spatiotemporal characteristics of the radiative properties and the potential sources of the atmospheric aerosols in SEA, in both mainland and insular regions, through a case study of Thailand. The findings could be beneficial for policymakers and environmental agencies to prioritize effective mitigation strategies to reduce the impacts of climate change and minimize aerosol pollution.

2. Methodology

2.1. Study Area

Thailand is a country located in SEA. It consists of the continental region, representing “mainland Thailand”, and the maritime region, representing “insular Thailand”. The mainland is dominated by mixed deciduous and dry deciduous forests and cropland, primarily rice paddies [13]. The insular region is dominated by tropical lowland forests, broadleaved evergreens, and mixed cropland and plantations, primarily oil palm and rubber farms [13].
As shown in Figure 1, northern and western Thailand display north-south lying mountain ranges up to 2577 m above mean sea level, separated by alluvial valleys. The lower central region possess alluvial plains, with many rivers flowing from north to south, making this region highly productive for cultivation. The northeast region is a plateau, characterized by vast, flat plains. Southern Thailand is characterized by a long, narrow coastal strip featuring various mountain ranges.
According to the Köppen climate classification, the mainland exhibits a tropical savanna climate, whereas the insular region is classified as possessing a tropical rainforest climate. The southwest monsoon brings warm moist air from the Indian Ocean to Thailand from May to October, developing a wet season over most of the country, with the exception of the eastern part of southern Thailand. Starting in October, the northeast monsoon brings cold, dry air from the Chinese mainland, developing a dry season in mainland Thailand and a wet season in the eastern part of southern Thailand. The mean seasonal precipitation from the years 1991 to 2020 was 37.85 mm for December through February, 325.79 mm for March through May, 702.90 mm for June through August, and 489.68 mm for September through November, and the mean temperature was 24.91 °C, 28.69 °C, 27.68 °C, and 26.47 °C, respectively [14].

2.2. Data Description

Both remote sensing and ground-based datasets for the years 2007 to 2019 were acquired for use in this study. The datasets are detailed below.
  • The Version 2 AERONET (Aerosol Robotic Network) inversion aerosol L2 (cloud-screened and quality-assured) product from 12 ground stations across Thailand was acquired interactively via the AERONET Aerosol Robotic Network, NASA website, https://aeronet.gsfc.nasa.gov/, accessed on 1 August 2020. The stations are listed in Table 1. The AERONET aerosol products are measured using sun and sky photometers to calculate direct and diffuse radiation. The AERONET product used in this study includes the following parameters [15]:
    ○
    The Angstrom exponent (AE) is the log–slope exponent of the spectral aerosol optical depth between 440 nm and 870 nm wavelengths. It indicates the sensitivity of the aerosol scattering properties to the wavelength of the incoming electromagnetic radiation. If the AE approaches zero, it suggests a prominent coarse-sized aerosol. In contrast, an AE close to four suggests that the aerosol size is comparable to that of air molecules and can better scatter light at shorter wavelengths.
    ○
    The single-scattering albedo (SSA) is the ratio of the scattering coefficient to the extinction coefficient (combined scattering and adsorption effects). An aerosol with an SSA close to one tends to scatter light well, whereas the SSA closer to zero exhibits light absorbing potential [16]. In this study, the AERONET SSA was referred to at 440, 675, 870, and 1020 nm wavelengths.
    ○
    The complex refractive index indicates the aerosol’s light scattering (real part) and light absorbing (imaginary part) properties. This study measured the refractive index for 440, 675, 870, and 1020 nm wavelengths.
    ○
    The asymmetry parameter indicates the asymmetry of light scattering for all directions around the aerosol. When the factor is close to zero, it suggests the perfect symmetry of light scattering, whereas forward light scattering is associated with a positive asymmetry parameter of about 0.85. In this study, the asymmetry parameter was measured for 440, 675, 870, and 1020 nm wavelengths.
    ○
    The effective radius is estimated, based on the size distribution of the particle number.
    ○
    The aerosol optical depth (AOD) is the extinction (both scattering and absorption) of the solar beam attributed to the aerosol. In this study, the AERONET AOD was measured at 500 nm using the direct sun algorithm. The AOD is widely used to imply atmospheric aerosol loading [15].
2.
This study utilized several products from the Moderate Resolution Imaging Spectroradiometer (MODIS), an instrument developed and operated by the United States’ National Aeronautics and Space Administration (NASA). The MODIS is onboard both the Terra and Aqua satellites, passing the equator twice a day at about 10:30 a.m. and about 1:30 p.m., respectively. Thus, the combined products from both satellites could help determine the daily dynamics of the land and atmosphere across Thailand. Details on the MODIS products used in this study are as follows.
1.
The land cover classification for 2018–2019 from the MODIS sensor onboard the Terra and Aqua satellites (MCD12Q1 Version 6) were used in this study. The products are provided at annual time steps, with a 500 m spatial resolution. In this study, we used the University of Maryland (UMD) legend, and the land cover for Thailand is shown in Figure 1. The land cover was reclassified into four types before analysis: forest (combined evergreen forests, deciduous broadleaf forests, and mixed forests), savannas (woody savannas, savannas, and grasslands), croplands (croplands and cropland mosaics), and urban (urban and built-up lands). The analysis relied on the % grid of each land cover type within a 5 km radius centered at the studied sites.
2.
The daily Level 3 MODIS-based PM2.5 (particulate matter with a size of 2.5 μm or smaller) air quality index (AQI) was acquired from the Geo-Informatics and Space Technology Development Agency, Thailand (GISTDA). The product was generated using the Level 2 MODIS aerosol optical depth, AOD, from the Terra satellite (MOD04). The product was downscaled to 66 meteorological stations.
3.
The near real-time MODIS Thermal Anomalies/Fire Locations, Collection 6 (MCD14DL) has a spatial resolution of 1 km, and was acquired from the Fire Information of Resource Management System (FIRMS) via https://firms.modaps.eosdis.nasa.gov/active_fire/, accessed on 15 August 2020. This study used the location of active fires to assess the fraction of fire pixels within the 5 km radius centered at studied sites.
3.
The Shuttle Radar Topography Mission (SRTM) digital elevation model (DEM) was acquired online from the USGS Earth Explorer via https://earthexplorer.usgs.gov/, accessed on 15 August 2020. The product has a spatial resolution of 1 arc second, or about 30 m. The DEM for Thailand is illustrated in Figure 1.
4.
The ground-based meteorological dataset was acquired from 66 stations of the Thai Meteorology Department (see the locations in Figure 1) located across the country. The dataset includes daily rainfall, relative humidity, air temperature, air pressure, and wind speed about 10 m above ground level.
5.
The Lidar vertical aerosol type dataset was retrieved by the Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) instrument onboard the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) through CALIPSO’s web-based search and sub-setting web application (https://subset.larc.nasa.gov/calipso/login.php, accessed on 23 October 2020). The CALIOP retrieval algorithm reports seven tropospheric aerosol subtypes by considering the aerosol Lidar ratio at 532 nm and 1064 nm. The aerosol subtypes are clean marine, dust, polluted continental/smoke, clean continental, polluted dust, elevated smoke, and dusty marine [17]. In this study, the CALIOP aerosol vertical profiles were retrieved for the areas close to (±1°) of the five locations (see Table 1), which were the Chiang Mai Meteorological Station (in the North) for the years 2010–2016, Silpakorn University (Central) for the years 2010–2019, Nong Khai (Northeast) for the years 2015–2019, Ubon Ratchathani (Northeast) for the years 2010–2019, and the Songkhla Meteorological Station (South) for the years 2010–2019. These stations were selected for their spatial correspondence to the AERONET stations and the air quality monitoring stations operated by the Pollution Control Department (PCD), Thailand. In this study, the CALIOP aerosol subtype product illustrates seasonal vertical aerosol profiles at the stations. There profiles provide evidence for inferring or supporting conclusions about possible sources for each aerosol cluster, based on their seasonal preferences.

2.3. Source Identification Using the HYSPLIT Backward Trajectory Model

The 5-day backward air trajectories were acquired from the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model, developed using the global reanalysis archived meteorological dataset. The trajectory illustrated an air parcel’s path at an arrival height of 10 m above the ground at the PCD air pollution monitoring stations during the high aerosol events (PM2.5 > 50 μg m−3) from 2012 to 2019. The simulation was assessed interactively through the NOAA Air Resource Laboratory READY system website, https://www.ready.noaa.gov/HYSPLIT_traj.php, accessed on 13 September 2021.

2.4. Data Analysis

  • Cluster analysis was used for classifying aerosol types using the AERONET aerosol optical parameters. In this study, Ward’s minimum variance method was used for clustering. In the Ward’s method, each observation is treated as a separate cluster, and subsequently, the clusters are iteratively combined in a way that minimizes the total within-cluster variance. Seven aerosol types were chosen in this study, according to dendrogram analysis. The dendrogram is a tree-like diagram that visually represents the process of merging clusters in the hierarchical clustering analysis (refer to the dendrogram for this study in Appendix A). The analysis was performed using MATLAB R2020b from MathWorks. The inherent patterns of aerosol optical parameters are reported and discussed for each aerosol cluster.
  • Kriging interpolation was used to interpolate the point observations of the meteorological parameters at 66 stations to the spatial information at 12 AERONET stations. A variogram model in the kriging interpolation is a mathematical representation of the spatial relationship among data points, and it helps in the estimation of values at unsampled locations, based on spatial autocorrelation. In this study, the variogram model was developed using either Gaussian or exponential functions, depending on which provided the better fit, based on visual inspection. The spatial information was later used for discriminant analysis. The interpolation was conducted using the Spatial Analyst Tools of ArcMap 10.5.1 in Esri’s ArcGIS 10.5.1 software.
  • Logit-transformed linear discriminant analysis was used to assess the possibility of occurrence for individual aerosol types. The models were developed using the spatial information from 12 AERONET stations, which were the MODIS land cover fractions (% forest, % savannas, % croplands, and % urban), the MODIS active fire fraction, the MODIS AOD-derived PM2.5 AQI, the DEM, and the meteorological parameters (surface temperature, relative humidity, and wind speed at approximately 10 m). Later, the models were used to assess the aerosol types at 66 meteorological stations across the country. The analysis was performed using MATLAB R2020b from MathWorks.
  • The validation confusion matrix was employed to assess the performance of the logit-transformed linear discriminant model. The model is trained using data from the years 2015 and earlier, and then tested using the remaining dataset. The confusion matrix offers a comprehensive breakdown of the correct and incorrect classifications made by the model. The user’s accuracy quantifies how well each of the simulated aerosol clusters was correctly classified, which implies reliability in the model application. The producer’s accuracy measures how well each of the AERONET aerosol clusters was correctly classified, which reflects the model performance.
  • Pearson’s linear correlation analysis was conducted using AERONET precipitable water and the corresponding effective radius of atmospheric aerosol particles in both fine and coarse modes. Precipitable water indicates atmospheric moisture or relative humidity levels. A strong positive correlation coefficient, indicating that aerosol particles grow in size with increasing humidity, implies hygroscopic growth.

3. Result and Discussion

3.1. Classification of Aerosol Optical Properties

Aerosol classification was analyzed using cluster analysis, and seven aerosol clusters were chosen to represent aerosol types across Thailand. The mean characteristics for each aerosol cluster are detailed in Table 2.
In this study, we classified aerosols into three groups based on their radiative properties: light-absorbing, light-scattering, and neutral. This classification relied on aerosol single-scattering albedo (SSA) and the complex refractive index. While specific universally agreed upon cutoff values for SSA and the complex refractive index do not exist, we classified aerosols based on their overall optical characteristics. The “neutral” classification was assigned to aerosols with SSA values close to 0.9.

3.1.1. Light-Absorbing Aerosols

Cluster 1 (L1), Cluster 2 (L2), Cluster 4 (L4), and Cluster 5 (L5) exhibited comparatively low SSA (~0.88 and ~0.89 at 675 nm) and high imaginary-part refractive indices (~0.02 at 675 nm). These characteristics suggest that those aerosols exhibit light absorption. Nonetheless, there were diverse properties of these aerosols, as discussed below.
  • Light scattering properties—the real-part refractive index was the highest for L1 (~1.53 at 440 nm), followed by L5 (~1.48 at 675 nm), L2 (~1.47 at 675 nm), and L4 (~1.45 at 440 nm).
  • Fine size—the L1 aerosol exhibited the finest size, as suggested by a low fine-mode effective radius (~0.15 μm), a high Angstrom exponent (~1.72), and a low fine-mode asymmetry factor (~0.56 at 675 nm).
  • Coarse size—the L5 aerosol exhibited the coarsest size, as suggested by a high coarse-mode effective radius (~3.13 μm) and a high coarse-mode asymmetry parameter (~0.96 at 675 nm).
  • Aerosol concentration—the L5 aerosol was often found under the highest aerosol optical depth (~1.72), followed by L1 (~0.74), L4 (~0.61), and L2 (~0.57).
In addition to the optical properties, the fine-mode aerosol particles exhibited hygroscopic growth, as suggested by positive linear correlation coefficients between the aerosol effective size and precipitable water, ranging from 0.15 (L4) to 0.47 (L5), as presented in Table 3. For the coarse-size aerosols, significant hygroscopic growth was only found for L2 (r = 0.28, p < 0.05) and L4 (r = 0.42, p < 0.05). Negative correlations were found for coarse-mode L1 (r = −0.14, p < 0.05), suggesting that it was a less-hygroscopic particle, possibly implying the presence of non-oxidized organic compounds [18] or black carbon aerosol [19].

3.1.2. Light-Scattering Aerosol

Cluster 6 (L6) and Cluster 7 (L7) exhibited a high SSA of ~0.93 at 675 nm and a comparatively low imaginary-part refractive index (~0.01 at 675 nm). Their sizes were comparatively coarser, as evidenced by a larger fine-mode effective radius (~0.19 μm and ~0.17 μm for L6 and L7, respectively), a lower Angstrom exponent (~1.49 and ~1.52, respectively), and a high fine-mode asymmetry parameter (~0.64 and ~0.61, respectively at 675 nm). These were likely found under high precipitable water (~3.80 and ~4.29 cm, respectively) and exhibited slightly hygroscopic properties for fine-mode particles (r = 0.188 for L6 and 0.131 for L7). Based on these properties, the L6 and L7 aerosol clusters should induce more negative radiative forcing, and their hygroscopic growth enhanced their light scattering properties. Nonetheless, the L6 aerosol was coarser than the L7, and the L6 was likely found under higher aerosol concentrations (AOD ~1.04).

3.1.3. Neutral Aerosol

Cluster 3 (L3) exhibited moderate light scattering properties (SSA ~0.91 at 675 nm) and relatively low light absorbing properties (imaginary-part of the refractive index of ~0.01 at 675 nm). Its size was comparatively fine, as suggested by an effective radius of ~0.16 μm (fine-mode) and ~2.43 μm (coarse mode), and a low asymmetry parameter (~0.86 for coarse mode at 675 nm). The L3 was often found under high precipitable water conditions (~3.45 cm).

3.2. Assessing Potential Sources for the Aerosol Clusters

Geographical characteristics (latitude, land cover, elevation, active fires), meteorological conditions (surface wind speed, temperature, relative humidity), and aerosol magnitude (represented by MODIS AOD-derived PM2.5) were assessed for individual aerosol clusters, downscaled to the 12 AERONET stations, and their mean values are reported in Table 4. Spatiotemporal distributions of the aerosol clusters were created based on the environmental characteristics using the linear discriminant model with logit transform. The model was used to assess the probability of monthly aerosol clusters; the results are shown in Figure 2 and Figure 3. It is noted that the model was used to depict spatiotemporal distributions of the aerosol clusters for the subsequent qualitative justification of the sources and the aerosol radiative forcing potential on a regional basis. The results from the models may not be accurately employed for local implementation for source control, since the model development was based on a limited number of AERONET stations, and a low overall accuracy of 48% was obtained from the model (see validation confusion matrix in Table 5). As shown in Table 5, both user and producer accuracies were the highest (79% and 65%, respectively) for classifying L1 aerosols, whereas the lowest accuracies (32% and 28%, respectively) resulted for classifying L3 aerosols.
Analysis of the CALIOP aerosol profile dataset provides information about seasonal aerosol types and their vertical information from five corresponding AERONET and PCD air quality monitoring stations across Thailand. There were seven CALIOP aerosol subtypes found in this study: (1) dust, (2) polluted continental/smoke, (3) clean continental, (4) polluted dust, (5) smoke, (6) dusty marine, and (7) clean marine. The results are illustrated in Figure 4.

3.2.1. Aerosol Cluster 1 (L1): Dry-Season Forest Fire Smoke Aerosols in the Northern Forests

The L1 was often found in the uppermost region of Thailand, adjacent to Myanmar and China, from December through March (see Figure 2 and Figure 3), in the local dry season, corresponding to low precipitable water levels of 1.15 cm (Table 2). From Table 4, the L1 aerosol likely appeared in the dominant forest (43.9%) and savanna (44%), with low wind speed (17.2 km h−1), and surrounded by active fires (active-fire areal ratio = 0.70). The evidence suggests that the local forest fires should be a source of the L1 aerosol. The fires should result in very fine-size L1 aerosols (effective radius of 0.15 μm and a high Angstrom exponent, 440–870 nm of 1.72).
This fine aerosol exhibited hygroscopic growth, as suggested by a significant positive correlation coefficient (r = 0.354, p < 0.01) between the fine-mode effective radius and precipitable water (see Table 3). Thus, the fine-mode L1 aerosol was expected to serve as a cloud condensation nuclei. However, an increase in aerosol concentrations may result in a greater number of smaller raindrops, potentially delaying the onset of heavy precipitation. Consistent with this conclusion, the coarse-model dust did not exhibit hygroscopic growth, as suggested by the negative correlation coefficient (r = −0.143) observed between the coarse-mode effective radius and precipitable water (see Table 3). Our finding aligns with the African smoke particles observed over the Caribbean. The smoke increased the cloud condensation nuclei number concentration, but slightly decreased the overall aerosol hygroscopicity [20]. Similar finding was also found in the study conducted by Liu et al. (2020) [21] on Amazonian biomass-burning aerosols, which demonstrated that the aerosols enhance the formation and prolong the lifetime of low-level clouds by providing more cloud condensation nuclei. However, they also suppress the formation of high-level clouds by reducing updrafts.
The L1 aerosol exhibited a low SSA of 0.88 (at 675 nm) and complex refractive indices of 1.53 + 0.02i (at 675 nm). These characteristics were similar to those of the aerosols found in a boreal forest of Northern Europe, which showed SSA values of 0.87 for PM1 (particulate matter with a size of 1 μm or smaller) and 0.89 for PM10 (particulate matter with a size of 10 μm or smaller), along with complex refractive indices of 1.487 + 0.021i at 550 nm for PM1 and 1.525 + 0.014i for PM10 [22]. The fine-mode L1 aerosol could primarily consist of aged biomass-burning and secondary organic carbonaceous particles, as commonly observed during many wildfires [23,24]. The coarse-mode L1 aerosol could originate primarily from soil dust, which easily disperses from dry soil, as also found in the burning particles of the coarse-mode Amazonian biomass [22].
Furthermore, the L1 forest fire smoke aerosol was obtained concurrently with the CALIOP smoke aerosol subtype in northern Thailand (Chiang Mai Meteorological Station) from December through February (DJF) at low elevation (from 0 to 4 km), as shown in Figure 4.

3.2.2. Aerosol Cluster 2 (L2): Dry-Season Background Aerosol in Mainland SEA

The L2 aerosol was generally found across northern to northeastern Thailand, far from the shoreline, throughout the dry season. The L2 aerosol often appeared in savanna (34.8%) and urban areas (34.5%), which exhibited active fires nearby (the fire areal ratio = 0.60). Furthermore, low winds (17.2 km h−1) and moderate temperatures (24.9 °C) were expected for the L2 aerosol. The aerosol was typically found under low aerosol levels (AERONET AOD~0.57). Thus, the L2 aerosol should be the dry-season background aerosol, contributed to by savanna open burning, mixed with emissions from urban activities.
The L2 aerosol exhibited an SSA of 0.88 and complex refractive indices of 1.47 + 0.02i (at 675 nm). The aerosol was coarser (fine-mode effective radius = 0.16 μm and the coarse-mode = 2.62 μm) than the L1 aerosol and exhibited hygroscopic growth in both fine- and coarse-mode (r = 0.28 and 0.20, respectively, p < 0.01). The L2 aerosol scattering property was similar to the PM1 boreal forest aerosol during wood burning, as reported by Luoma et al. (2019) [22] (SSA~0.85 and complex refractive indices ~1.484 + 0.025i at 550 nm) and comparable to the SEA biomass burning aerosol found in March through May, as reported by Park et al. (2020) [25] (SSA~0.86 at 550 nm). Compared with the L1 forest fire aerosol, the L2 aerosol possessed a lower real-part refractive index and overall water-absorbing properties, indicating that the L2 aerosol showed more water-soluble organic content [26].
The L2 aerosol could be represented by the CALIOP polluted continental/smoke aerosol subtype, found in the northern and the northeastern stations at a low elevation from 0 to 3 km (see Figure 4).

3.2.3. Aerosol Cluster 3 (L3): Background Aerosols in Urban Areas in Insular SEA

The L3 aerosol was dominant in urban lands (% urban = 0.415 in Table 4) in southern, central, and eastern Thailand, areas which experience oceanic influences in the SEA peninsula. The L3 aerosol was found in both dry and wet seasons under strong winds (19.4 km hr−1), as well as hot (28.2 °C) and humid (69% relative humidity and precipitable water = 3.45 cm) conditions. Thus, the L3 aerosol should exhibit the typical characteristics of the background aerosols in urban areas in insular SEA.
As shown in Table 2, the L3 aerosol optical properties exhibited an SSA of 0.91 (at 675 nm) and complex refractive indices of 1.44 + 0.01i (at 675 nm), similar to the results for the organic carbon aerosol (1.40 + 0i) reported by Schkolnik et al. (2007) [26]. The aerosol exhibited a comparatively smaller coarse-mode size (effective radius of 2.30 μm) relative to the other clusters and demonstrated moderately hygroscopic growth in both coarse mode (r = 0.207) and fine mode (r = 0.187). The L3 aerosol was found under moderate aerosol levels (AERONET AOD = 0.62).
Based on the optical properties, we hypothesized that the L3 aerosol should be an oceanic aerosol mixed with urban aerosols. The SEA urban aerosols, such as those found in Bangkok, central Thailand [27], Nakhon Ratchasima, Northeast Thailand [28], Songkhla, South Thailand [29], Chiang Mai, North Thailand [30,31], were primarily contributed by traffic and biomass burning. The studies in the major Southeast Asian cities of Chiang Mai (Thailand), Penang (Malaysia), Kuala Lumpur (Malaysia), and Singapore, also reported abundant UV-neutral smoke aerosols across the atmospheric column, followed by polluted dust and dust [6]. Feng and Christopher (2013) [3] confirmed that megacities in the southern region of Southeast Asia (e.g., Singapore and Jakarta) are typically polluted by aerosols from anthropogenic and urban activities—mainly emissions from automobiles, combustion of biofuels at low temperatures, and biomass burning—during the winter and spring seasons. In the case of Thailand, the urban aerosols are predominantly inorganic species (ammonium, nitrate, sulfate, potassium, sodium, chloride), carboxylates, elemental carbons, and organic carbons [27,28,29,30,31]. The oceanic aerosol typically consists of sea salt, exhibiting very high light scattering properties (SSA of 0.99 [32]). When mixed with black carbon, the sea salt’s light scattering properties decreased significantly, as found in coastal industrial areas of France (SSA of 0.75 for internally mixed and 0.85 for externally mixed particles [32]). Park et al. (2020) [25] investigated the SSA of the aerosols in the Pacific Ocean sampled at the remote island of Mauna Lao in Hawaii and found that the dominant fine aerosol showed a low SSA (0.86 at 550 nm). Similar SSA results were obtained for the long-range transport biomass burning smoke (0.86 at 550 nm) from SEA to Hawaii, whereas a slightly high SSA (0.87 at 550 nm) was obtained for the long-range transport of polluted dust from Northeast Asia [25].
The overall hygroscopic properties of these mixed oceanic-anthropogenic aerosol particles could also enhance their light scattering properties. Freshly emitted black carbon from anthropogenic combustion sources exhibited poor hygroscopicity [33] and light-absorbing capacities [34]. The particles become hydrophilic after mixing with soluble matter in the atmosphere [34]. This hygroscopic growth has been observed in various studies, including those regarding anthropogenic aerosol particles in the Pearl River Delta [35], anthropogenic aerosol particles in southern Spain [36], and haze aerosol particle in Beijing, China [37]. The growth also enhances the aerosol light scattering properties [34], resulting in a higher SSA.
It is noted that the CALIOP aerosol subtypes corresponding to the L3 aerosol were likely not classified as marine aerosols, but as dust, polluted dust, and polluted continental/smoke aerosol subtypes. These aerosol subtypes appeared in different altitudes. As shown in Figure 4, during the rainy season from June through August (JJA), the southern station (Songkhla Meteorological Station) observed the polluted dust at a lower altitude (0–3 km), possibly due to cumulonimbus cloud development at high altitudes. The polluted continental/smoke subtype was often observed in low levels from 0 to 3 km elevation, but was absent in JJA. The dust alone was usually detected at elevated levels of 10 km or higher, while the polluted dust was found at 2 to 4 km from the ground from December to May. The intrusion of the dust subtype was also found at a >6 km elevation over Penang and Kuala Lumpur, Malaysia, which was hypothesized to be attributable to wind-blown dust of transboundary origin [6]. There have been reports regarding the downward transportation of Asian dust (originating from the Gobi Desert and Inner Mongolia) to Taiwan [38,39], and the aging processes could enhance aerosol hygroscopicity [40]. The dust, however, was less likely to be dominant in L3 aerosol, since the origin was very far from insular Thailand. Thus, the source of the CALIOP elevated dust layers across insular Thailand remains unknown. Further studies should focus on assessing the accuracy of using CALIOP aerosol subtypes and evaluating long-range transported aerosol particles, including dust, in this region.

3.2.4. Aerosol Cluster 4 (L4): Wet-Season Background Aerosols in Mainland SEA

The L4 aerosol was found over the Thai mainland and was even more pronounced in the wet season from June through October (see Figure 2 and Figure 3). The aerosol was less likely to be found close to the forest area (3.1% forest), and there was a comparatively low number of active fires (fire areal ratio = 0.45), factors which corresponded to moderate aerosol levels (AERONET AOD ~0.61). Thus, we hypothesized that the L4 aerosol is a background aerosol across the SEA mainland during the wet season.
The wet season background aerosol (L4) displayed optical properties similar to those of the dry season aerosol (L2), but the L4 showed a slightly higher SSA (0.89 at 675 nm) and a somewhat lower real-part refractive index (1.45 at 675 nm), likely attributable to different air mass trajectories. During the wet season, the trajectory paths of air masses toward the mainland SEA originated from the Indian and Pacific oceans. In the dry season, the near-surface air masses were primarily from E Asia. The coarse-mode L4 aerosol exhibited strong hygroscopic growth (r = 0.424, see Table 3), and sea salt could be responsible for this behavior. Park et al. (2020) [25] reported an SSA of 0.86 (at 550 nm) for oceanic aerosols on a remote island in the Pacific Ocean. The maritime aerosols could later modify their optical properties after either mixing with local aerosols or undergoing the aging processes [25,32,41]. The aging processes involve gas-to-particle conversions, internal or external mixing with organic compounds, and hygroscopic growth [42].
The CALIOP dust aerosol subtype was periodically found at elevated levels across mainland Thailand (Central, North, Northeast Thailand; see Figure 4) from September through November, transitioning from a southwest monsoon to a northeast monsoon. Given their observation in the wet season, the CALIOP dust subtype layers could be involved with cloud formation. However, the origins of the lofted dust cannot be clearly explained. Further studies should prioritize investigating the effects of cloud interference on the CALIOP aerosol subtype products. The HYSPLIT backward air trajectory model reveals that the elevated dust layers could also be found along with prevailing northeasterly to easterly upper winds, as shown in Appendix B. The winds exhibited long trajectories over the Pacific Ocean. One possible source is the Asian dust in Mongolia and North China. The downward trajectories of the Asian dust to Taiwan were reported in previous studies [38,39]. Furthermore, there was evidence of the export of atmospheric Hg from the East Asian continent to the Northern South China Sea, including the areas of Da Nang, Vietnam, and Chiang Mai, Thailand [43]. Evidence of Asian dust transport to deep mainland SEA has not yet been found.

3.2.5. Aerosol Cluster 5 (L5): Dry-Season Fresh Biomass Burning Smoke Aerosols in Mainland SEA

The L5 aerosol was exclusively found in the dry season from February to the beginning of the cropping season in May (see Figure 2 and Figure 3), with strong winds (19.4 km hr−1), many active fires (areal ratio = 0.782), and high aerosol levels (AERONET AOD = 1.72). The L5 aerosol was often dominant in urban areas (% urban = 0.471). Therefore, the L5 aerosol should be fresh biomass-burning smoke from nearby savanna grasslands and agricultural lands, mixed with urban aerosols.
The L5 aerosol optical properties exhibited an SSA of ~0.89 (at 675 nm) and complex refractive indices of ~1.48 + 0.02i. This L5 aerosol showed lower light scattering properties than did the biomass-burning smoke aerosol reported in the Amazon basin (SSA = 0.92 [44]). In addition, the fine-mode L5 aerosol exhibited strong hygroscopic growth (r = 0.469). We hypothesize that the L5 aerosol was water-soluble brown carbon composed of organic matter, black carbon, sulfate, and nitrate. This brown carbon exhibits strong light absorption and is typically found during biomass-burning episodes in SEA [45] and South Asia [46]. Similar to the situation with the L5 aerosol, Luoma et al. (2009) [22] reported refractive indices of 1.484 + 0.025i and an SSA of 0.85 ± 0.08 (at 550 nm) for the brown carbon from wood burning mixed with sea salt and soil dust found in boreal forests.
The CALIOP vertical aerosol profiles for northeastern stations, shown in Figure 4, also showed a dominant smoke aerosol subtype in March–April–May at a lower level (from 0 to 5 km). The CALIOP smoke subtype could represent both the fresh L5 and aged L6 aerosol clusters, which will be discussed in the next section.

3.2.6. Aerosol Cluster 6 (L6): Dry-Season Aged Biomass Burning Smoke Aerosols in Mainland SEA

The L6 aerosol was found during the dry season from February through April (see Figure 2 and Figure 3), when the prevalent winds were from the northeast and east (see examples of the HYSPLIT backward air trajectory in Appendix C), and the aerosol was highly pronounced in the Chao Phraya basin of central Thailand, where rice cultivation fields are prominent (see Figure 1). Consistently, the L6 aerosol was prevalent in croplands (% crop ~20.5) and found under high aerosol levels (AERONET AOD ~1.04). Nonetheless, comparatively few active fires were found nearby (active fire areal ratio ~0.462). Thus, the L6 aerosol should represent aged biomass burning smoke aerosol, which is typically produced in agricultural fields (primarily rice and sugarcane) in the SEA mainland.
The L6 aged biomass burning smoke aerosol exhibited higher light scattering properties than did fresh smoke (L5), as suggested by an SSA of ~0.93. The magnitude of SSA was similar to that found in urban central China in summer [47] and in Delhi, India, in winter [48], where secondary aerosols were predominant in the atmosphere [47,48]. The L6 aerosol also exhibited coarser sizes, with a fine-mode effective radius of ~0.19 μm and a coarse-mode radius of approximately ~2.96 μm. As implied from Table 3, the water-absorbing property for the coarse-mode L6 dust was insignificant (negative r), whereas the fine-mode aerosol exhibited moderately hygroscopic growth (r = 0.188, p < 0.01). Comparatively low complex refractive indices (1.43 + 0.01i, at 675 nm) could imply the prominence of organic compounds (1.4 + 0i [26,49]) over the inorganic components, including sulfate, nitrate, and ammonium (1.53 + 0i [49]). We hypothesized that the high light scattering properties of the L6 aerosol should be more or less attributed to the aging processes.

3.2.7. Aerosol Cluster 7 (L7): Polluted Oceanic Aerosols in Insular SEA

The L7 aerosol could be observed throughout the year in the SEA peninsula (highlighted in South, Central, and East Thailand, as shown in Figure 2 and Figure 3), but was strongly influenced by the prevailing southwesterly winds from the Indian Ocean and southeasterly winds from the Pacific Ocean (see examples of the HYSPLIT backward air trajectory in Appendix D) from June through October. Concurrently, the SEA haze episodes usually intensified from September through October, often attributed to forest fires and peat land burning over southern Malaysia and Indonesia [6]. The aerosol was likely found under strong winds (19.8 km h−1) and high atmospheric water conditions (RH ~70.8% and precipitable water ~4.29 cm). The L7 aerosol appeared to have a moderate aerosol level (AERONET AOD ~0.68) and was less likely to be found with active fires nearby (active fire areal ratio ~0.429). Based on the evidence, we hypothesize that the L7 aerosol could be transboundary biomass-burning smoke aerosol transported to insular Thailand.
The L7 aerosol exhibited similar optical properties to those of the aged biomass-burning smoke aerosol in mainland Thailand (L6). According to geographical preference, the L7 aerosol should also be involved with oceanic aerosols. The lowest complex refractive indices for the L7 (1.42 + 0.01i, at 675 nm) among all aerosol clusters agreed with the maritime aerosol characteristics reported by Levoni et al. (1997) [50] at 1.381+ <0.001i. In addition, the higher SSA (~0.93) implies its prominence in light scattering, which is a characteristic commonly observed in various types of aerosols such as sulfate oceanic aerosol (SSA = 0.96 ± 0.03 at 550 nm) [51], and sea salt (SSA = 0.99) [32]. This L7 transboundary biomass burning smoke aerosol in insular Thailand demonstrated much higher light scattering properties than that from the remote islands in the Pacific Ocean (SSA ~0.86 at 550 nm) [25]. Nonetheless, the light scattering of the L7 transboundary smoke aerosol was similar to that of the smoke plumes from the 1997 Indonesia fire, promoting atmospheric stability, weakening atmospheric circulation, and inducing drought [52].

4. Conclusions

Understanding the Southeast Asian aerosol radiative properties has challenged many scientists due to the complexity of land–sea interconnections, a unique monsoon system, and multiple pollution sources, including agricultural biomass burning, forest fires, and fossil fuel combustion from traffic vehicles and industries. Furthermore, strong solar radiation could enhance secondary aerosol formation along the transportation pathways, as well as the mixing states of the aerosols, which could alter their optical properties. This study aimed to provide an insightful understanding of Southeast Asian aerosols in respect of their optical properties through the Thailand case study. The evidence used in this study included land and atmosphere products from the Earth’s observation satellites (MODIS/Terra and MODIS/Aqua, CALIOP/CALIPSO), aerosol optical properties from the ground-based remote sensing network (AERONET), and ground-based meteorological parameters acquired from the Thai Meteorology Department. Cluster analysis was used to group aerosol types by considering their similarity in regards to optical properties. Linear discriminant analysis with logit-transformed data provided the model for assessing the aerosol types using satellite retrievals and ground observations. The findings illustrated the spatiotemporal distributions of aerosol types. Finally, aerosol sources were justified, based on several lines of evidence: aerosol optical properties, land information, spatiotemporal characteristics, and vertical profiles of aerosol subtypes.
The findings in the case of Thailand suggested that seven aerosol clusters were dominant across SEA, of which four clusters were categorized as possessing light-absorbing potential (L1, L2, L4, and L5), two clusters were associated with light-scattering potential (L6 and L7), and one was neutral (L3).
The four groups of light-absorbing aerosols were as follows:
  • L1—dry-season forest fire smoke aerosols in the northern forests. These were less hygroscopic and were carbon-rich.
  • L2—dry-season background aerosol in mainland SEA, especially in the north and the northeast. These were predominantly water-soluble organic aerosols from biomass open burning and emissions from urban activities.
  • L4—wet-season background aerosol in mainland SEA. This exhibited hygroscopic growth in the coarse mode and could be a mixture of oceanic and local aerosols. Furthermore, the CALIOP dust subtypes were often found in elevated layers, but the source of this elevated dust remains unknown.
  • L5—dry-season fresh biomass burning smoke aerosols in mainland SEA. These were water soluble in the fine mode, and the main composition was brown carbon.
The groups of light-scattering aerosols were as follows:
  • L6—dry-season aged biomass burning smoke aerosols in mainland SEA, especially in central Thailand. This exhibited slight water solubility in the fine mode and were likely formed during transportation.
  • L7—polluted oceanic aerosols in insular SEA. These aerosols could be attributed to transboundary biomass burning smoke, since they were likely found within the CALIOP smoke subtype, and they appeared in the elevated layers.
The neutral aerosol was L3, representing background aerosols in insular SEA. The aerosol could be found in all seasons from oceanic and urban emissions. Based on the CALIOP profiles of aerosol subtypes, the aerosols likely originated from polluted centers (anthropogenic emissions) near the ground, polluted dust near the ground or up to 4 km elevation, and elevated dust from 10 km or higher.
Integrated satellite retrievals can provide an understanding of aerosol radiative forcing potentials in complex and dynamic environmental systems. In Southeast Asia, represented by Thailand, the aerosols exhibited various radiative properties. Further studies should focus on carbonaceous aerosols (organic carbons, black carbon, and brown carbon) and the aging processes of these aerosols, since they play significant roles in regional aerosol optical properties.

Author Contributions

A.B.—conceptualization, methodology, data analysis, and writing; P.P.—data provision and funding acquisition; T.P.—project administration. All authors have read and agreed to the published version of the manuscript.

Funding

This research was financially supported by the National Research Council of Thailand, NRCT.

Data Availability Statement

The datasets generated and/or analyzed during the current study are available from the authors upon request.

Acknowledgments

The National Research Council of Thailand (NRCT) financially supported this research during the fiscal years 2020–2021 under the project titled “Assessing Sources of Ground-Level PM2.5 using Near Real-Time Satellite Observation and Ground Measurement”. The authors offer sincere thanks to their colleagues from GISTDA, Thailand, and all researchers from the Earth Science Research Cluster (MUKA) for their technical comments. The authors also appreciate the help from Adrian R. Plant, Mahasarakham University, for his comments and language editing.

Conflicts of Interest

All authors declare that they have no conflict of interest/competing interest.

Appendix A

Remotesensing 15 05319 g0a1
Figure A1. Dendrogram from cluster analysis referred to in this study.
Figure A1. Dendrogram from cluster analysis referred to in this study.
Remotesensing 15 05319 g0a1

Appendix B

Remotesensing 15 05319 g0a2
Figure A2. HYSPLIT 5-day backward air trajectories during a haze episode in Bangkok (13.774380°N, 100.614580°E); the initial trajectory started on 24 September 2019 (A) at 10 m above ground level and (B) at 12 km above ground level. These model outputs were acquired from https://www.ready.noaa.gov/HYSPLIT.php, accessed on 7 October 2023.
Figure A2. HYSPLIT 5-day backward air trajectories during a haze episode in Bangkok (13.774380°N, 100.614580°E); the initial trajectory started on 24 September 2019 (A) at 10 m above ground level and (B) at 12 km above ground level. These model outputs were acquired from https://www.ready.noaa.gov/HYSPLIT.php, accessed on 7 October 2023.
Remotesensing 15 05319 g0a2

Appendix C

Remotesensing 15 05319 g0a3
Figure A3. HYSPLIT 5-day backward air trajectories at 10 m above ground level during a haze episodes (initial trajectory started on 13 February 2019) (A) at Sakaew (13.636540°N, 102.468620°E) and (B) at Bangkok (13.774380°N, 100.614580°E). These model outputs were acquired from https://www.ready.noaa.gov/HYSPLIT.php, accessed on 9 November 2021.
Figure A3. HYSPLIT 5-day backward air trajectories at 10 m above ground level during a haze episodes (initial trajectory started on 13 February 2019) (A) at Sakaew (13.636540°N, 102.468620°E) and (B) at Bangkok (13.774380°N, 100.614580°E). These model outputs were acquired from https://www.ready.noaa.gov/HYSPLIT.php, accessed on 9 November 2021.
Remotesensing 15 05319 g0a3

Appendix D

Remotesensing 15 05319 g0a4
Figure A4. HYSPLIT 5-day backward air trajectories at 10 m above ground level during a haze episode (initial trajectory started on 23 September 2019) (A) at Satun (6.619180°N, 100.073560°E) and (B) at Songkhla (7.01216°N, 100.483190°E). These model outputs were acquired from https://www.ready.noaa.gov/HYSPLIT.php, accessed on 10 November 2021.
Figure A4. HYSPLIT 5-day backward air trajectories at 10 m above ground level during a haze episode (initial trajectory started on 23 September 2019) (A) at Satun (6.619180°N, 100.073560°E) and (B) at Songkhla (7.01216°N, 100.483190°E). These model outputs were acquired from https://www.ready.noaa.gov/HYSPLIT.php, accessed on 10 November 2021.
Remotesensing 15 05319 g0a4

References

  1. Nguyen, T.T.; Pham, H.V.; Lasko, K.; Bui, M.T.; Laffly, D.; Jourdan, A.; Bui, H.Q. Spatiotemporal analysis of ground and satellite-based aerosol for air quality assessment in the Southeast Asia region. Environ. Pollut. 2019, 255, 113106. [Google Scholar] [CrossRef] [PubMed]
  2. Reid, J.S.; Hyer, E.J.; Johnson, R.S.; Holben, B.N.; Yokelson, R.J.; Zhang, J.; Campbell, J.R.; Christopher, S.A.; Di Girolamo, L.; Giglio, L. Observing and understanding the Southeast Asian aerosol system by remote sensing: An initial review and analysis for the Seven Southeast Asian Studies (7SEAS) program. Atmos. Res. 2013, 122, 403–468. [Google Scholar]
  3. Feng, N.; Christopher, S.A. Satellite and surface-based remote sensing of Southeast Asian aerosols and their radiative effects. Atmos. Res. 2013, 122, 544–554. [Google Scholar] [CrossRef]
  4. Khan, R.; Kumar, K.R.; Zhao, T. The climatology of aerosol optical thickness and radiative effects in Southeast Asia from 18-years of ground-based observations. Environ. Pollut. 2019, 254, 113025. [Google Scholar] [CrossRef]
  5. Adam, M.G.; Chiang, A.W.J.; Balasubramanian, R. Insights into characteristics of light absorbing carbonaceous aerosols over an urban location in Southeast Asia. Environ. Pollut. 2020, 257, 113425. [Google Scholar] [CrossRef]
  6. Banerjee, T.; Shitole, A.; Mhawish, A.; Anand, A.; Ranjan, R.; Khan, M.F.; Srithawirat, T.; Latif, M.T.; Mall, R. Aerosol climatology over South and Southeast Asia: Aerosol types, vertical profile, and source fields. J. Geophys. Res. Atmos. 2021, 126, e2020JD033554. [Google Scholar] [CrossRef]
  7. Lin, N.-H.; Sayer, A.M.; Wang, S.-H.; Loftus, A.M.; Hsiao, T.-C.; Sheu, G.-R.; Hsu, N.C.; Tsay, S.-C.; Chantara, S. Interactions between biomass-burning aerosols and clouds over Southeast Asia: Current status, challenges, and perspectives. Environ. Pollut. 2014, 195, 292–307. [Google Scholar] [CrossRef]
  8. Ross, A.D.; Holz, R.E.; Quinn, G.; Reid, J.S.; Xian, P.; Turk, F.J.; Posselt, D.J. Exploring the first aerosol indirect effect over Southeast Asia using a 10-year collocated MODIS, CALIOP, and model dataset. Atmos. Chem. Phys. 2018, 18, 12747–12764. [Google Scholar] [CrossRef]
  9. Bellouin, N.; Quaas, J.; Gryspeerdt, E.; Kinne, S.; Stier, P.; Watson-Parris, D.; Boucher, O.; Carslaw, K.S.; Christensen, M.; Daniau, A.L. Bounding global aerosol radiative forcing of climate change. Rev. Geophys. 2020, 58, e2019RG000660. [Google Scholar]
  10. Liu, Y.; Jia, R.; Dai, T.; Xie, Y.; Shi, G. A review of aerosol optical properties and radiative effects. J. Meteorol. Res. 2014, 28, 1003–1028. [Google Scholar]
  11. Dubovik, O.; Smirnov, A.; Holben, B.; King, M.; Kaufman, Y.; Eck, T.; Slutsker, I. Accuracy assessments of aerosol optical properties retrieved from Aerosol Robotic Network (AERONET) Sun and sky radiance measurements. J. Geophys. Res. Atmos. 2000, 105, 9791–9806. [Google Scholar] [CrossRef]
  12. Bridhikitti, A.; Overcamp, T.J. Optical characteristics of southeast Asia’s regional aerosols and their sources. J. Air Waste Manag. Assoc. 2011, 61, 747–754. [Google Scholar] [CrossRef] [PubMed]
  13. Stibig, H.J.; Belward, A.; Roy, P.; Rosalina-Wasrin, U.; Agrawal, S.; Joshi, P.; Hildanus; Beuchle, R.; Fritz, S.; Mubareka, S. A land-cover map for South and Southeast Asia derived from SPOT-VEGETATION data. J. Biogeogr. 2007, 34, 625–637. [Google Scholar] [CrossRef]
  14. Work Bank Group. Climate Change Knowledge Portal for Development Practitioners and Policy Makers: Thailand. Available online: https://climateknowledgeportal.worldbank.org/country/thailand/climate-data-historical#:~:text=Thailand%20enjoys%20a%20tropical%20climate,country%2C%20especially%20the%20mountainous%20regions (accessed on 2 August 2023).
  15. Bridhikitti, A. Applications of Remote Sensing for Air Pollution Monitoring in Thailand: An Early Warning for Public Health. In Earth Data Analytics for Planetary Health; Springer Nature Singapore Pte Ltd.: Singapore, 2023; pp. 3–31. [Google Scholar]
  16. Di Biagio, C.; di Sarra, A.; Meloni, D.; Monteleone, F.; Piacentino, S.; Sferlazzo, D. Measurements of Mediterranean aerosol radiative forcing and influence of the single scattering albedo. J. Geophys. Res. Atmos. 2009, 114. [Google Scholar] [CrossRef]
  17. Falah, S.; Mhawish, A.; Omar, A.H.; Sorek-Hamer, M.; Lyapustin, A.I.; Banerjee, T.; Kizel, F.; Broday, D.M. Intercomparison of aerosol types reported as part of aerosol product retrieval over diverse geographic regions. Remote Sens. 2022, 14, 3667. [Google Scholar] [CrossRef]
  18. Ming, Y.; Russell, L.M. Predicted hygroscopic growth of sea salt aerosol. J. Geophys. Res. Atmos. 2001, 106, 28259–28274. [Google Scholar] [CrossRef]
  19. Liu, X.; Gu, J.; Li, Y.; Cheng, Y.; Qu, Y.; Han, T.; Wang, J.; Tian, H.; Chen, J.; Zhang, Y. Increase of aerosol scattering by hygroscopic growth: Observation, modeling, and implications on visibility. Atmos. Res. 2013, 132, 91–101. [Google Scholar] [CrossRef]
  20. Royer, H.M.; Pöhlker, M.L.; Krüger, O.; Blades, E.; Sealy, P.; Lata, N.N.; Cheng, Z.; China, S.; Ault, A.P.; Quinn, P.K. African smoke particles act as cloud condensation nuclei in the wintertime tropical North Atlantic boundary layer over Barbados. Atmos. Chem. Phys. 2023, 23, 981–998. [Google Scholar] [CrossRef]
  21. Liu, L.; Cheng, Y.; Wang, S.; Wei, C.; Pöhlker, M.L.; Pöhlker, C.; Artaxo, P.; Shrivastava, M.; Andreae, M.O.; Pöschl, U. Impact of biomass burning aerosols on radiation, clouds, and precipitation over the Amazon: Relative importance of aerosol–cloud and aerosol–radiation interactions. Atmos. Chem. Phys. 2020, 20, 13283–13301. [Google Scholar] [CrossRef]
  22. Luoma, K.; Virkkula, A.; Aalto, P.; Petäjä, T.; Kulmala, M. Over a 10-year record of aerosol optical properties at SMEAR II. Atmos. Chem. Phys. 2019, 19, 11363–11382. [Google Scholar] [CrossRef]
  23. Gunsch, M.J.; May, N.W.; Wen, M.; Bottenus, C.L.; Gardner, D.J.; VanReken, T.M.; Bertman, S.B.; Hopke, P.K.; Ault, A.P.; Pratt, K.A. Ubiquitous influence of wildfire emissions and secondary organic aerosol on summertime atmospheric aerosol in the forested Great Lakes region. Atmos. Chem. Phys. 2018, 18, 3701–3715. [Google Scholar] [CrossRef]
  24. Pio, C.; Legrand, M.; Alves, C.; Oliveira, T.; Afonso, J.; Caseiro, A.; Puxbaum, H.; Sánchez-Ochoa, A.; Gelencsér, A. Chemical composition of atmospheric aerosols during the 2003 summer intense forest fire period. Atmos. Environ. 2008, 42, 7530–7543. [Google Scholar] [CrossRef]
  25. Park, J.-U.; Kim, S.-W.; Sheridan, P.J.; Williams, A.; Chambers, S.D. Long-term variability of aerosol optical properties at Mauna Loa. Aerosol Air Qual. Res. 2020, 20, 1700–1711. [Google Scholar] [CrossRef]
  26. Schkolnik, G.; Chand, D.; Hoffer, A.; Andreae, M.; Erlick, C.; Swietlicki, E.; Rudich, Y. Constraining the density and complex refractive index of elemental and organic carbon in biomass burning aerosol using optical and chemical measurements. Atmos. Environ. 2007, 41, 1107–1118. [Google Scholar] [CrossRef]
  27. Chuersuwan, N.; Nimrat, S.; Lekphet, S.; Kerdkumrai, T. Levels and major sources of PM2. 5 and PM10 in Bangkok Metropolitan Region. Environ. Int. 2008, 34, 671–677. [Google Scholar] [CrossRef]
  28. Li, C.; Tsay, S.-C.; Hsu, N.C.; Kim, J.Y.; Howell, S.G.; Huebert, B.J.; Ji, Q.; Jeong, M.-J.; Wang, S.-H.; Hansell, R.A. Characteristics and composition of atmospheric aerosols in Phimai, central Thailand during BASE-ASIA. Atmos. Environ. 2013, 78, 60–71. [Google Scholar] [CrossRef]
  29. Phairuang, W.; Inerb, M.; Furuuchi, M.; Hata, M.; Tekasakul, S.; Tekasakul, P. Size-fractionated carbonaceous aerosols down to PM0. 1 in southern Thailand: Local and long-range transport effects. Environ. Pollut. 2020, 260, 114031. [Google Scholar] [CrossRef]
  30. Tsai, Y.I.; Sopajaree, K.; Chotruksa, A.; Wu, H.-C.; Kuo, S.-C. Source indicators of biomass burning associated with inorganic salts and carboxylates in dry season ambient aerosol in Chiang Mai Basin, Thailand. Atmos. Environ. 2013, 78, 93–104. [Google Scholar] [CrossRef]
  31. Pengchai, P.; Chantara, S.; Sopajaree, K.; Wangkarn, S.; Tengcharoenkul, U.; Rayanakorn, M. Seasonal variation, risk assessment and source estimation of PM 10 and PM10-bound PAHs in the ambient air of Chiang Mai and Lamphun, Thailand. Environ. Monit. Assess. 2009, 154, 197–218. [Google Scholar] [CrossRef]
  32. Wang, H.; Wang, X.; Yang, X.; Li, W.; Xue, L.; Wang, T.; Chen, J.; Wang, W. Mixed chloride aerosols and their atmospheric implications: A review. Aerosol Air Qual. Res. 2017, 17, 878–887. [Google Scholar] [CrossRef]
  33. Wang, Y.; Henning, S.; Poulain, L.; Lu, C.; Stratmann, F.; Wang, Y.; Niu, S.; Pöhlker, M.L.; Herrmann, H.; Wiedensohler, A. Aerosol activation characteristics and prediction at the central European ACTRIS research station of Melpitz, Germany. Atmos. Chem. Phys. 2022, 22, 15943–15962. [Google Scholar] [CrossRef]
  34. Zeng, C.; Liu, C.; Li, J.; Zhu, B.; Yin, Y.; Wang, Y. Optical properties and radiative forcing of aged BC due to hygroscopic growth: Effects of the aggregate structure. J. Geophys. Res. Atmos. 2019, 124, 4620–4633. [Google Scholar] [CrossRef]
  35. Liu, X.; Cheng, Y.; Zhang, Y.; Jung, J.; Sugimoto, N.; Chang, S.-Y.; Kim, Y.J.; Fan, S.; Zeng, L. Influences of relative humidity and particle chemical composition on aerosol scattering properties during the 2006 PRD campaign. Atmos. Environ. 2008, 42, 1525–1536. [Google Scholar] [CrossRef]
  36. Titos, G.; Lyamani, H.; Cazorla, A.; Sorribas, M.; Foyo-Moreno, I.; Wiedensohler, A.; Alados-Arboledas, L. Study of the relative humidity dependence of aerosol light-scattering in southern Spain. Tellus B Chem. Phys. Meteorol. 2014, 66, 24536. [Google Scholar] [CrossRef]
  37. Bai, Z.; Ji, Y.; Pi, Y.; Yang, K.; Wang, L.; Zhang, Y.; Zhai, Y.; Yan, Z.; Han, X. Hygroscopic analysis of individual Beijing haze aerosol particles by environmental scanning electron microscopy. Atmos. Environ. 2018, 172, 149–156. [Google Scholar] [CrossRef]
  38. Schepanski, K. Transport of mineral dust and its impact on climate. Geosciences 2018, 8, 151. [Google Scholar] [CrossRef]
  39. Lin, C.-Y.; Sheng, Y.-F.; Chen, W.-N.; Wang, Z.; Kuo, C.-H.; Chen, W.-C.; Yang, T. The impact of channel effect on Asian dust transport dynamics: A case in southeastern Asia. Atmos. Chem. Phys. 2012, 12, 271–285. [Google Scholar] [CrossRef]
  40. Kim, J.; Yoon, S.-C.; Jefferson, A.; Kim, S.-W. Aerosol hygroscopic properties during Asian dust, pollution, and biomass burning episodes at Gosan, Korea in April 2001. Atmos. Environ. 2006, 40, 1550–1560. [Google Scholar] [CrossRef]
  41. Meskhidze, N.; Petters, M.D.; Tsigaridis, K.; Bates, T.; O’Dowd, C.; Reid, J.; Lewis, E.R.; Gantt, B.; Anguelova, M.D.; Bhave, P.V. Production Mechanisms, Number Concentration, Size Distribution, Chemical Composition, and Optical Properties of Sea Spray Aerosols; Wiley Online Library: Hoboken, NJ, USA, 2013. [Google Scholar]
  42. Lesins, G.; Chylek, P.; Lohmann, U. A study of internal and external mixing scenarios and its effect on aerosol optical properties and direct radiative forcing. J. Geophys. Res. Atmos. 2002, 107, AAC 5-1–AAC 5-12. [Google Scholar] [CrossRef]
  43. Sheu, G.-R.; Lin, N.-H.; Lee, C.-T.; Wang, J.-L.; Chuang, M.-T.; Wang, S.-H.; Chi, K.H.; Ou-Yang, C.-F. Distribution of atmospheric mercury in northern Southeast Asia and South China Sea during Dongsha experiment. Atmos. Environ. 2013, 78, 174–183. [Google Scholar] [CrossRef]
  44. Chand, D.; Guyon, P.; Artaxo, P.; Schmid, O.; Frank, G.P.; Rizzo, L.V.; Mayol-Bracero, O.L.; Gatti, L.V.; Andreae, M.O. Optical and physical properties of aerosols in the boundary layer and free troposphere over the Amazon Basin during the biomass burning season. Atmos. Chem. Phys. 2006, 6, 2911–2925. [Google Scholar] [CrossRef]
  45. Zhang, Y.; Peng, Y.; Song, W.; Zhang, Y.-L.; Ponsawansong, P.; Prapamontol, T.; Wang, Y. Contribution of brown carbon to the light absorption and radiative effect of carbonaceous aerosols from biomass burning emissions in Chiang Mai, Thailand. Atmos. Environ. 2021, 260, 118544. [Google Scholar] [CrossRef]
  46. Wu, G.; Ram, K.; Fu, P.; Wang, W.; Zhang, Y.; Liu, X.; Stone, E.A.; Pradhan, B.B.; Dangol, P.M.; Panday, A.K. Water-soluble brown carbon in atmospheric aerosols from Godavari (Nepal), a regional representative of South Asia. Environ. Sci. Technol. 2019, 53, 3471–3479. [Google Scholar] [CrossRef] [PubMed]
  47. Zhang, M.; Ma, Y.; Gong, W.; Liu, B.; Shi, Y.; Chen, Z. Aerosol optical properties and radiative effects: Assessment of urban aerosols in central China using 10-year observations. Atmos. Environ. 2018, 182, 275–285. [Google Scholar] [CrossRef]
  48. Tiwari, S.; Pandithurai, G.; Attri, S.; Srivastava, A.; Soni, V.; Bisht, D.; Kumar, V.A.; Srivastava, M.K. Aerosol optical properties and their relationship with meteorological parameters during wintertime in Delhi, India. Atmos. Res. 2015, 153, 465–479. [Google Scholar] [CrossRef]
  49. Ebert, M.; Weinbruch, S.; Rausch, A.; Gorzawski, G.; Helas, G.; Hoffmann, P.; Wex, H. Complex refractive index of aerosols during LACE 98# x2010; as derived from the analysis of individual particles. J. Geophys. Res. Atmos. 2002, 107, LAC 3-1–LAC 3-15. [Google Scholar]
  50. Levoni, C.; Cervino, M.; Guzzi, R.; Torricella, F. Atmospheric aerosol optical properties: A database of radiative characteristics for different components and classes. Appl. Opt. 1997, 36, 8031–8041. [Google Scholar] [CrossRef]
  51. Reidmiller, D.R.; Hobbs, P.V.; Kahn, R. Aerosol optical properties and particle size distributions on the east coast of the United States derived from airborne in situ and remote sensing measurements. J. Atmos. Sci. 2006, 63, 785–814. [Google Scholar] [CrossRef]
  52. Chou, M.-D.; Chan, P.-K.; Wang, M. Aerosol radiative forcing derived from SeaWiFS-retrieved aerosol optical properties. J. Atmos. Sci. 2002, 59, 748–757. [Google Scholar] [CrossRef]
Remotesensing 15 05319 g001
Figure 1. The digital elevation model, the MODIS land use/cover map, the 66 meteorological stations, and the 12 AERONET stations referred to in this study.
Figure 1. The digital elevation model, the MODIS land use/cover map, the 66 meteorological stations, and the 12 AERONET stations referred to in this study.
Remotesensing 15 05319 g001
Remotesensing 15 05319 g002
Figure 2. Monthly probability maps for the individual aerosol clusters (L1 to L7) for January through June, along with prevalent aerosol clusters.
Figure 2. Monthly probability maps for the individual aerosol clusters (L1 to L7) for January through June, along with prevalent aerosol clusters.
Remotesensing 15 05319 g002
Remotesensing 15 05319 g003
Figure 3. Monthly probability maps for the individual aerosol clusters (L1 to L7) for July through December, along with the prevalent aerosol clusters.
Figure 3. Monthly probability maps for the individual aerosol clusters (L1 to L7) for July through December, along with the prevalent aerosol clusters.
Remotesensing 15 05319 g003
Remotesensing 15 05319 g004
Figure 4. Vertical profiles of aerosol subtypes acquired from CALIOP/CALIPSO from five locations.
Figure 4. Vertical profiles of aerosol subtypes acquired from CALIOP/CALIPSO from five locations.
Remotesensing 15 05319 g004
Table 1. Details regarding the twelve AERONET stations (see map location in Figure 1) referred to in this study.
Table 1. Details regarding the twelve AERONET stations (see map location in Figure 1) referred to in this study.
AERONET StationLatitudeLongitudeSampling PeriodNumber of ObservationsCALIOP Observing
  • Chiang Mai Meteorological Station
18.77112598.9724676 January 2010 to 26 May 2016580Y
2.
Silpakorn University
13.819308100.0411831 January 2010 to 30 December 2019667Y
3.
Doi Ang Khang
19.9324599.045416 February 2013 to 1 May 2019164N
4.
Doi Inthanon
18.5998.4855564 April 2017 to 28 April 20176N
5.
Mukdahan
16.606667104.676119 January 2007 to 5 December 2009159N
6.
Nong Khai
17.8772102.7167121 January 2015 to 1 October 2019257Y
7.
Omkoi
17.79833398.4316673 March 2014 to 14 March 2018174N
8.
Pimai
15.181944102.56416710 January 2007 to 28 March 200869N
9.
Songkhla Meteorological Station
7.184387100.6045839 January 2010 to 24 March 201984Y
10.
Ubon Ratchathani
15.245518104.87101119 March 2010 to 11 March 2019387Y
11.
Lopburi
15.266667101.18736128 September 2018 to 9 January 20197N
12.
Bangkok
13.7491100.517730 January 2007 to 25 January 201953N
Table 2. AERONET aerosol optical properties for the seven aerosol clusters.
Table 2. AERONET aerosol optical properties for the seven aerosol clusters.
Aerosol ClusterNumber ObservedAngstrom Exponent (440–870 nm)SSA (440/675/870/1020 nm)Real-Part Refractive Indices (440/675/870/1020 nm)Imaginary-Part Refractive Indices (440/675/870/1020 nm)Asymmetry Factor (Fine) (440/675/870/1020 nm)Asymmetry Factor (Coarse) (440/675/870/1020 nm)Effective Radius-Fine, µmEffective Radius-Coarse, µmAerosol Optical Depth, AODPrecipitable Water, cm
L11941.720.89/0.88/0.85/0.821.53/1.53/1.53/1.520.02/0.02/0.02/0.020.65/0.56/0.49/0.441.03/0.88/0.83/0.810.152.430.741.15
L24931.570.89/0.88/0.86/0.851.47/1.47/1.47/1.480.02/0.02/0.02/0.020.68/0.59/0.53/0.491.06/0.90/0.83/0.820.162.620.571.93
L34091.570.91/0.91/0.89/0.881.44/1.44/1.44/1.460.01/0.01/0.01/0.010.69/0.59/0.52/0.470.99/0.86/0.83/0.800.162.300.623.45
L44881.590.90/0.89/0.88/0.861.45/1.45/1.45/1.470.02/0.02/0.02/0.020.69/0.59/0.52/0.481.03/0.88/0.83/0.810.162.510.612.76
L5861.730.89/0.89/0.87/0.851.48/1.48/1.48/1.500.02/0.02/0.02/0.020.68/0.57/0.50/0.461.16/0.96/0.83/0.860.163.131.722.59
L63321.490.93/0.93/0.92/0.911.43/1.43/1.43/1.450.01/0.01/0.01/0.010.72/0.64/0.58/0.541.08/0.91/0.83/0.830.192.961.043.80
L76051.520.93/0.93/0.92/0.911.42/1.42/1.42/1.450.01/0.01/0.01/0.010.71/0.61/0.54/0.490.98/0.85/0.83/0.800.172.400.684.29
Table 3. Linear correlation coefficient (r) between the AERONET precipitable water vapor and the effective radius, for both coarse mode (Reff-C) and fine mode (Reff-F).
Table 3. Linear correlation coefficient (r) between the AERONET precipitable water vapor and the effective radius, for both coarse mode (Reff-C) and fine mode (Reff-F).
Aerosol
Cluster		Number of
Observations		Reff-CReff-F
rp-Valuerp-Value
1194−0.1430.0460.354<0.001
24930.280<0.0010.201<0.001
34090.207<0.0010.187<0.001
44880.424<0.0010.1500.001
586−0.1300.2320.469<0.001
6332−0.0620.2620.1880.001
76050.174<0.0010.1310.001
Table 4. The mean values of the parameters used for aerosol cluster (L1 to L7) identification using the logit linear discriminant model.
Table 4. The mean values of the parameters used for aerosol cluster (L1 to L7) identification using the logit linear discriminant model.
N = 1184Latitude% Forest% Savannas% Croplands% UrbanElevation, mWind Speed,
km h−1		RH, %Temperature, °CActive Fire FractionPM2.5 AQI
L119.2270.4390.4400.0220.085100517.261.725.40.700.76
L217.7630.1120.3480.1120.34542517.265.925.40.600.45
L315.2270.0130.2930.1100.41511919.469.028.20.470.64
L416.6620.0310.2830.1790.37621418.167.826.70.450.54
L518.4200.0520.2770.1180.47130619.460.927.40.781.34
L615.8190.0090.2470.2050.38212618.368.427.50.460.93
L714.9910.0090.3180.0810.4559419.870.828.70.430.71
Table 5. Validation confusion matrix for the simulated aerosol clusters compared to the AERONET-based aerosol clusters.
Table 5. Validation confusion matrix for the simulated aerosol clusters compared to the AERONET-based aerosol clusters.
Simulated Aerosol TypeProducer Accuracy, %
1234567Total
AERONET
Aerosol Type		18936173 13665
224179155914428663
3 350646114517928
4 614110011111824241
5 12 432615558
6 1119328421813032
7 28292168115652
Total11330215429563901671184
User Accuracy, %79593234514749 48
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Bridhikitti, A.; Petchpayoon, P.; Prabamroong, T. Integrated Remote Sensing Observations of Radiative Properties and Sources of the Aerosols in Southeast Asia: The Case of Thailand. Remote Sens. 2023, 15, 5319. https://doi.org/10.3390/rs15225319

AMA Style

Bridhikitti A, Petchpayoon P, Prabamroong T. Integrated Remote Sensing Observations of Radiative Properties and Sources of the Aerosols in Southeast Asia: The Case of Thailand. Remote Sensing. 2023; 15(22):5319. https://doi.org/10.3390/rs15225319

Chicago/Turabian Style

Bridhikitti, Arika, Pakorn Petchpayoon, and Thayukorn Prabamroong. 2023. "Integrated Remote Sensing Observations of Radiative Properties and Sources of the Aerosols in Southeast Asia: The Case of Thailand" Remote Sensing 15, no. 22: 5319. https://doi.org/10.3390/rs15225319

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop