SPIE Startup Challenge 2015 Founding Partner - JENOPTIK Get updates from SPIE Newsroom
  • Newsroom Home
  • Astronomy
  • Biomedical Optics & Medical Imaging
  • Defense & Security
  • Electronic Imaging & Signal Processing
  • Illumination & Displays
  • Lasers & Sources
  • Micro/Nano Lithography
  • Nanotechnology
  • Optical Design & Engineering
  • Optoelectronics & Communications
  • Remote Sensing
  • Sensing & Measurement
  • Solar & Alternative Energy
  • Sign up for Newsroom E-Alerts
  • Information for:
    Advertisers
SPIE Startup Challenge 2015 Lead Sponsor - HAMAMATSU

2014 Asia-Pacific Remote Sensing | Register Today

Gold Open Access is coming to SPIE Journals

SPIE PRESS




Print PageEmail PageView PDF

Remote Sensing

Mapping atmospheric pollution using remote sensing

Satellite imagery offers a lower-cost and less time consuming alternative to traditional methods for mapping atmospheric concentrations of aerosols.
19 October 2007, SPIE Newsroom. DOI: 10.1117/2.1200710.0853

Air pollution affects human health and reduces the quality of our land and water. We cannot escape from it, even in our own homes. In Malaysia, in particular, environmental pollution is a persistent problem. In recent years, air quality has been degraded almost annually during haze episodes generated by forest fires in Sumatra and Kalimantan, Indonesia. The worst of these events caused Malaysia to declare an emergency in 1997 for Kuching, Sarawak, and in 2005 for Port Klang, a district in Kuala Selangor. The declarations were made after Air Pollution Index (API) values reached dangerous levels.


Figure 1. Map of PM10 (particular matter up to 10(μ)m in diameter) over the land surface around Penang Island, Malaysia. (a) 30 July 2000, (b) 15 February 2001, (c) 17 January 2002, (d) 6 March 2002, (e) 5 February 2003, (f) 19 March 2004, and (g) 2 February 2005. Blue <40μg/m3, green = 40–80μg/m3, yellow = 80–120μg/m3, orange = 120–160μg/m3, red = >160μg/m3, and black = water and cloud area.

Smoke haze belongs to a class of pollutants called aerosols, liquid or solid particles suspended in the air from natural or man-made sources. Aerosol particles also affect climate. Since the early 1900s, the Earth's surface temperature has increased by 0.6°C, reaching its highest level of the last thousand years. This rapid temperature change is attributed to a shift of less than 1% in the energy balance between the absorption of incoming solar radiation and the emission of thermal radiation from the Earth. Greenhouse gases in the atmosphere, such as carbon dioxide and ozone, affect this energy balance, but aerosols also play a role. They do so directly, by interacting with solar and terrestrial radiation, and indirectly, by their effect on cloud microphysics, albedo (reflectivity), and precipitation.2 The effects of aerosols on climate differ from those of greenhouse gases. Since most are highly reflective, they raise the planet's albedo, thereby cooling the surface and effectively offsetting greenhouse gas warming by anywhere from 25 to 50%.

The traditional sampling method for environmental monitoring of aerosols is time consuming and expensive. Furthermore, field measurements cannot provide the fine spatial resolution needed to show detailed distribution patterns over a large area, or allow for continuous monitoring. Remote-sensing techniques can overcome these problems. Radiant energy reflected and emitted by the Earth carries with it a signature of the Earth's atmospheric and surface characteristics. By measuring the wavelength and the angular and polarization properties of this energy, satellite sensors can quantify several atmospheric and surface features.1 Satellite data have traditionally been used to detect air pollution and can be used for qualitative measurements over a large coverage area. Mapping air quality by using remote-sensing data provides better results at a relatively cheaper cost. This study used high spatial resolution digital imagery to estimate air quality.

The objective of this study was to test the use of digital satellite imagery to detect particulate matter up to 10μm in diameter (PM10). PM10 is one of the most harmful components in the atmosphere because of its ability to penetrate into the human respiratory system and embed in the lungs, where it is retained for a long time. Smaller particles also offer more surface area to absorb other pollutants compared with an equivalent mass of larger particles.

Landsat satellite images from seven dates were analyzed to map the PM10 concentration over land. PM10 measurements were collected simultaneously with the images. Only one date coincided with the measurements of PM10 data over water. Digital numbers for each wavelength band corresponding to the ground-truth data were determined and then converted into radiance and reflectance values.

The value at the top of the atmosphere is the sum of the surface and atmospheric reflectance. The signals measured in each of these visible bands represent the combination of these two effects, usually in different proportions depending on the conditions. Hence, we must determine the surface contribution from the total received at the sensor. The reflectance detected at the top of the atmosphere, ρ (TOA), was decreased by the surface amount to obtain the atmospheric reflectance. The surface values in the visible bands (red and blue) were retrieved using ATCOR2 software. In this study, a new algorithm was developed to detect and map air pollution levels. The algorithm was derived using the reflectance model, which is a function of the inherent optical properties of the atmosphere and which can be related in turn to the concentration of its constituents. The details of the algorithm have been described elsewhere.3–5 In short, our algorithm for particle concentration is given by A = e0 + e1 Ratm1 + e2 Ratm3, where A is the particle concentration of PM10, Ratmi is the atmospheric reflectance (i = 1 and 3 are the band numbers), and ej is the algorithm coefficient (j = 0, 1, and 2 are empirically determined).


Figure 2. Map of PM10 over the water surface around Penang Island, Malaysia, 9 March 2006. Blue > 40μg/m3, green = 40–80μg/m3, yellow = 80–120μg/m3, orange = 120–160μg/m3, red = >160μg/m3, and black = water and cloud area.

We used Landsat Thematic Mapper (TM) signals as independent variables in our analysis. The atmospheric reflectance values were combined and related to the corresponding PM10 amounts, and the data was used to calibrate the algorithm. Coefficients were determined using a regression technique, and PM10 maps were generated using the calibrated algorithm.

Then color-coded maps were produced for visual interpretation and geometrically corrected for displaying distribution patterns of PM10 over land and water as shown in Figures 1 and 2, respectively. Generally, PM10 concentrations in the industrial and urban areas were higher compared to other areas based on the generated PM10 maps shown here.

High air pollution levels also originated from stationary sources such as chemical and manufacturing plants. Furthermore, many mobile sources such as motor vehicles raised the PM10 concentrations in urban areas. The particles and dust in Penang, Malaysia, mostly come from motor vehicles, smoke from forest fires in the neighboring countryside, and particles and dust generated by industrial activities.

Conclusion

This study indicates the potential applications of digital satellite data for air-quality studies. The time required for traditional sampling methods for environmental monitoring can be reduced by using satellite images. Applying the algorithm we developed for mapping PM10 using the present data set produced reliable and accurate results over a large study area.

This project was made possible by Universiti Sains Malaysia (USM) short-term grants and Science Fund grant 04-01-05-SF0035. We would like to thank the technical staff who participated in this project, and USM for support and encouragement. Thanks are due also to Dr. Rudolf Richter, DLR-German Aerospace Center, Remote Sensing Data Center, Germany, who provided the papers related to the ATCOR2 software. Finally, we would like to express our special appreciation to the Global Land Cover Facility of the Institute for Advanced Computer Studies, University of Maryland, College Park, MD, for providing free online data for the satellite image used in this study.


Lim Hwee San, Mohd. Zubir Mat Jafri, Khiruddin Abdullah, Nasirun Mohd. Saleh  
School of Physics
Universiti Sains Malaysia (USM)
Penang, Malaysia

Lim Hwee San is a lecturer at USM, where he also obtained his BSc in geophysics (2001), his MSc in remote sensing (2003), and his PhD in environmental remote sensing (2006). His research interests include remote-sensing applications for water- and air-quality monitoring, land surface properties, and digital image classification. He is a member of SPIE.

Mohd. Zubir Mat Jafri is a lecturer who obtained his BSc in physics from Universiti Kebangsaan Malaysia in 1984, his MSc in microprocessor technology and application from Brighton Polytechnic, UK, in 1991, and his PhD in 1996 from University College of Swansea, Wales, in the research area of algorithm development for detecting curves from digital images. He has more than 10 years of teaching experience in the areas of physics, optical communication, digital and analog electronics, and microprocessors. He is also active in research work on current-based systems, automated visual inspection systems, digital image processing, and remote sensing. He has published more than 100 articles in these areas. He is a life member of the Malaysian Institute of Physics and a member of SPIE.

Khiruddin Abdullah received his BSc in physics from Bedford College, University of London, in 1982, his MSc in geophysics from Imperial College, London, in 1984, and his PhD in remote sensing from the University of Dundee, Scotland, in 1994. He is currently a faculty member at USM, where he teaches courses in geophysics and remote sensing. His research interests include remote-sensing applications in marine and coastal environments. Presently, he is working on a remote-sensing algorithm for retrieving water-quality parameters and sea surface temperature.

Nasirun Mohd. Saleh received his BSc in geophysics from USM in 1982, his MSc in meteorology from Reading, UK, in 1985, and his MPhil in boundary layer meteorology from East Anglia University, UK, in 1997. He is currently director of USM's Islamic Centre and coordinator of the university's Astronomy and Atmospheric Science Research Unit. He is also a faculty member in the School of Physics, where he teaches courses in geophysics, meteorology, and remote sensing.