RGIS617: Urban and Environmental Applications of GIS/Remote Sensing PDF

RGIS617: Urban and Environmental Applications of GIS/Remote Sensing Remote Sensing and GIS in Urban Studies Dr. Elnazir Ramadan Department of Geography and Urban Sustainability UAE University [email protected] OUTLINE  Introduction  Identification and Delineation of Urban...

RGIS617: Urban and Environmental Applications of GIS/Remote Sensing Remote Sensing and GIS in Urban Studies Dr. Elnazir Ramadan Department of Geography and Urban Sustainability UAE University [email protected] OUTLINE  Introduction  Identification and Delineation of Urban Areas  Classification of Urban Areas  Measuring and Monitoring Physical Properties of Urban Areas  Analysis of Physical Characteristics and Demographic/Socioeconomic Patterns  Monitoring Urban Growth  Recent Applications and New Developments  References and Related Resources INTRODUCTION  Half of the world population lives in urban areas (United Nations 2001).  Urban growth has increasing impacts at local, regional and global scales (Berry, 1990).  Associated problems to urban expansion, e.g.,  loss of agricultural land and natural vegetation,  uncontrolled urban sprawl,  increased traffic congestion and,  degradation of air and water quality.  Have consequences for more distant regions. Microclimate of the human habitat, climate dynamics and environmental changes at local and regional scales. INTRODUCTION (CONT.)  Remote sensing is a consistent and efficient tool for characterization of the urban environment. Examples,  urban planning and decision making,  facilitates the study of local and regional environmental processes  the sustainability of cities and their hinterlands.  Satellite systems can provide timely and accurate information on existing land use and land cover. INTRODUCTION (CONT.)  Earlier studies were focused mainly on the use of photointerpreted data as  auxiliary data sources for the census, or  to predict socioeconomic variables such as poverty from housing density, structure type or vegetation cover.  With the advent of multiple spectral bands, including thermal infrared (Landsat MSS) in the 1970s and the subsequent Landsat TM and SPOT, virtually all research in urban areas focused on land use or land cover classification. INTRODUCTION (CONT.) In this lecture, the following topics are discussed, providing examples of uses of remote sensing in urban analysis:  Identification and delineation of the urban environment  Classification of urban areas  Measuring and monitoring physical properties of urban areas (vegetation, air quality, etc)  Analysis of physical characteristics and demographic/socioeconomic patterns of the urban environment  Monitoring changes and urban growth over time The first three topics intrinsically address relatively technical issues. The last two topics, on the other hand, report on studies that are clear social science applications. IDENTIFICATION AND DELINEATION OF URBAN AREAS  Remotely sensed data may provide a physically meaningful way to define urban areas that can then be utilized in urban and social science studies.  The main problem in delineation of urban areas in the social science context is the lack of a consistent definition of what is urban.  Definitions vary from country to country (United Nations, 2001) and are often based on different parameters. DEFINITION OF URBAN AREAS  Urban areas may be defined by administrative boundaries, or by population density, and this varies from country to country. The limitations in these approaches are:  the majority of urban areas have boundaries that don't coincide with administrative divisions, and defining cities based on a population density threshold that differs by country makes comparative studies more difficult.  Furthermore, such approaches do not include spatial extents of built-up areas. Satellite imagery may be used to define urban areas in a more consistent way and to produce spatially georeferenced urban extents. DELINEATION OF URBAN AREAS - LITERATURE  There is an extensive literature on urban delineation, although very often based on case studies of a single city, rather than on comparative studies. The book Remote Sensing and Urban Analysis (Donnay et al. 2001) shows efforts made to develop new methodologies and algorithms to improve delineation and characterization of urban features. It provides good examples of how to identify the different elements in the built-up environment based on their density and texture (e.g., Longley and Mesev 2001, Moller-Jensen 1990, Karathanassi et al. 2000). DELINEATION OF URBAN AREAS – LITERATURE (CONT.)  Another approach is one that looks at data fusion for urban analysis, which is based on the integration of data from different satellites, and with different spatial and spectral resolutions, to identify urban features, building types and building density (e.g., Proceedings of the IEEE/ISPRS Joint Workshop on Remote Sensing and Data Fusion over Urban Areas, 2001). CLASSIFICATION OF URBAN AREAS  If delineating urban areas is a difficult task, classifying different types of urban land use is even more so. The urban environment is characterized by a mixture of diverse material and land use classes, such as buildings, commercial infrastructures, transportation networks, and parks.  Because they are combinations of spectrally distinct land cover types, mixed pixels in urban areas are frequently misclassified as other land-cover classes.  Similarly, the definition of an "urban" spectral class will usually incorporate pixels of other non- urban classes. Such spectral heterogeneity severely limits the applicability of standard classification techniques, where it is assumed that the study area is comprised of a number of unique and internally homogeneous classes. CLASSIFICATION OF URBAN AREAS (CONT.)  For example, to be able to identify urban classes down to Level III of the Anderson classification system (that is, to differentiate between single- family and multi-family residential, for instance) a minimum ground resolution of 1-5 m is required.  Commercial satellites, like IKONOS, QuickBird and Worldview as well as aerial photography, are being used for Level IV classifications (identification of duplex, triplex or condominium units), while satellites like Landsat will allow a Level I classification (Residential vs. Commercial, for instance).  Higher spatial resolution normally comes at the price of lower temporal resolution and smaller areal coverages. For studies of urban growth or over large areas, such high ground resolutions might not be necessary. CLASSIFICATION OF URBAN AREAS (CONT.)  Urban classifications are often improved by integrating satellite-derived classifications with ancillary data in a GIS environment. Ancillary data might include a range of socioeconomic variables, such as population or housing density, derived from the census or similar data sources or variables like land use and digital elevation models (Stefanov et al. 2001). CLASSIFICATION OF URBAN AREAS (CONT.)  More recent techniques in urban classification rely on hyperspectral data.  However, there are some limitations in their applicability for the social sciences.  First, the majority of hyperspectral sensors have been airborne (e.g., AVIRIS, CASI, PROBE-1, AISA), with only two recent exceptions (NASA's Hyperion on EO-1 satellite and the US Air Force Research Lab's FTHSI sensor on the MightySat II satellite). This might limit temporal and spatial availability of data.  Second, the image classification process might not be trivial, in that it requires good spectral libraries, which in many cases need to be created beforehand and, in some cases, require complex sub-pixel analysis methods. MEASURING AND MONITORING PHYSICAL PROPERTIES OF URBAN AREAS  Urban areas exert an influence on local weather and climate, but they also affect wider regional and global atmospheric systems. Changes induced by urbanization include:  changes in solar radiation absorption,  surface temperature,  evapotranspiration,  water vapor and pollutants concentration, which in turn link to human health problems.  Remote sensing data is proving extremely useful for urban studies in terms of providing scientifically verifiable, routine measurements of physical properties that would be difficult or more expensive to obtain in situ, especially in developing countries. URBAN HEAT ISLAND  The urban heat island (UHI) effect, generally represented by the difference between urban and rural temperatures, has been studied since the 1930s. It is in the 1970s that the use of remote sensing to assess the urban heat island effect was initially explored.  Roth et al. (1989) analyze the reasons behind the differences between remotely-sensed heat island and air temperatures measured using standard or mobile stations. Such differences are related to:  the urban geometry (over representation of roofs and tree tops),  the lack of simple coupling between the surface and the air in the urban system, and  the failure to recognize and consider the different scales of climatic phenomena in the urban atmosphere. URBAN HEAT ISLAND (CONT.)  An increasing number of studies have been focusing on indirect measurements of the heat island effect. For instance,  Gallo et al. (1993) observed a correlation between a vegetation index (NDVI) and observed temperatures.  Gallo et al. (1995) suggest a combination of NDVI and night-time lights.  Karl (1988) suggests the use of urban population growth as a predictor of the urban heat island.  US EPA’s webpage provides useful information and resources about UHI: https://www.epa.gov/heatislands OTHER PHYSICAL PARAMETERS  Other measured physical parameters include: vegetation, ozone, dust and overall air quality in urban areas.  Spectral signature and interaction of electromagnetic radiation with matter can be used to derive an estimate of these parameters from remote sensing data. URBAN VEGETATION  Vegetation can substantially affect the wind, temperature, moisture, and precipitation regime of urban areas and is believed to have very important practical applications in urban planning, such as heating and cooling requirements of buildings, dispersion and concentration of pollutants, and urban weather (Avissar 1996).  Many published studies discussed the use of remote sensing and GIS to map and assess urban vegetation.  A nice review of methods used in mapping urban vegetation using high resolution satellite data is given in  Neyns, R.; Canters, F. Mapping of Urban Vegetation with High-Resolution Remote Sensing: A Review. Remote Sens. 2022, 14, 1031. https://doi.org/10.3390/rs14041031 AIR QUALITY  Remote sensing is used to provide Air Quality maps over extended areas allowing estimation of AQ in areas where ground measurements are not available. Multiple approaches and techniques varying from simple linear regression to deep learning methods were published in the literature.  MDPI’s special issue “Air Quality Research Using Remote Sensing” summarizes some approaches https://www.mdpi.com/journal/remotesensing/special_issues/AQ_RS  Some publications focused on the use of remote sensing and GIS to study the effect of urban morphological characteristics on air quality. E.g.:  Kokkonen, Tom V et al. “The effect of urban morphological characteristics on the spatial variation of PM2.5 air quality in downtown Nanjing.” Environmental science: atmospheres vol. 1,7 481-497. 26 Aug. 2021, doi:10.1039/d1ea00035g  Chenyu Huang, Tingting Hu, Yusen Duan, Qingyu Li, Nan Chen, Qi Wang, Mengge Zhou, Pinhua Rao,”Effect of urban morphology on air pollution distribution in high-density urban blocks based on mobile monitoring and machine learning”, Building and Environment,Volume 219, 2022, 109173, ISSN 0360-1323, https://doi.org/10.1016/j.buildenv.2022.109173. OZONE, DUST, SMOKE, AEROSOL  Ozone concentration has been measured by NASA's Total Ozone Mapping Spectrometer (TOMS) and subsequent sensors (OMI for example). These sensors provides long-term datasets of daily measurements over about three decades. The spatial resolution (about 100 km at the equator for TOMS) does not allow for a detailed characterization of air quality at the city level, but the data are extremely useful for global studies.  The launch of Aura in 2004 allows measurements of ozone, particulate, temperature etc, in the troposphere (from the ground to about 10 km), at a ground resolution of 12-24 km. ANALYSIS OF PHYSICAL CHARACTERISTICS AND DEMOGRAPHIC/SOCIOECONOMIC PATTERNS  Some of the past and on-going initiatives, especially in the remote sensing community, are focused on the integration of remote sensing with socioeconomic data to improve classification in urban areas (e.g., Harris and Ventura 1995, Mesev 1997, Vogelmann et al. 1998, and Chen 2001).  Studies of this type show that classification of satellite imagery alone sometimes does not produce adequate results for specific urban applications. Remote sensing provides repeat coverages of a given area, allowing great data availability, but often at moderate spatial resolutions, while some ancillary data may provide levels of detail that are not available through the satellite data. Combining the two proves to be an effective way to reduce misclassification errors and improve the specificity of the final classification DEMOGRAPHIC/SOCIOECONOMIC PARAMETERS  Several authors studied the correlation between population data from the census, or collected from social survey at the village level, and land cover characteristics derived from satellite imagery (e.g., Yuan et al, 1997, Radeloff et al. 2000).  Walsh et al. (1999) examine the correlation between biophysical and social variables. They also show the importance of scale dependence on the selected variables and that the relationships are not generalizeable across the sampled spatial scales.  Lo and Faber (1997) present another interesting case, where their study of the correlation between environmental variables extracted from Landsat data and socioeconomic data from the census shows that a combination of satellite data and census data can be used to determine Quality of Life assessment with an environmental perspective. DEMOGRAPHIC/SOCIOECONOMIC PARAMETERS (CONT.)  Research by Pozzi and Small (2002) looks at the correlation between population density (from the U.S. Census) and vegetation cover (extracted from Landsat) for a sample of cities in the United States. The authors show that for large cities there is a linear correlation between the two variables. But, given the difference in resolution of satellite and census data, and given the different urban structures and growth dynamics, it is difficult to consistently characterize urban areas at the 30 meter resolution of Landsat imagery. DEMOGRAPHIC/SOCIOECONOMIC PARAMETERS (CONT.)  A noteworthy effort is the Long Term Ecological Research (LTER) Network (LTER 2001). Two of the 24 sites included in the program are urban areas: Baltimore and Central Arizona-Phoenix. The objective is to analyze the interactions of ecological and socio-economic systems and the effect of infrastructure and development on fluxes of nutrients, energy, and water in urban environments. MONITORING URBAN GROWTH  The direct impacts of urban expansion on physical, ecological and social resources have made research on urban sprawl of increased interest.  Traditional census sources are extremely useful in that they capture changes in the socioeconomic and demographic structure of cities, but they lack spatial details and are not frequently updated.  Remote sensing, on the other hand, makes available a vast amount of data with continuous temporal and spatial coverage and can therefore provide a successful means for monitoring urban growth and changes.  Using remote sensing for change detection studies naturally requires that the different temporal images are atmospherically and zenith-angle corrected and carefully co-registered, in order to avoid errors in the estimation of land cover changes. MONITORING URBAN GROWTH (CONT.)  Despite the extensive literature of change studies available, most of these studies are based on more traditional land cover classifications (e.g., Wang and Zhang 1999, Esnard and Yang 2001, Stefanov et al. 2001), and only a few report examples of development of integrated datasets that can be used in planning and urban monitoring efforts.  Examples of how remote sensing data can be used in conjunction with socioeconomic data are those of Emmanuel (1997) and Wagner and Ryznar (1999). They find that changes in urban vegetation can be linked to urban social changes in the city of Detroit, and suggest the development of a vegetation-based urban environmental quality index to monitor physical and social changes in cities.  Many cities in developing countries are experiencing rapid increase in population and consequential urban expansion. Remote sensing may provide fundamental observations of urban growth that are not available from other sources (e.g., Balzerek 2001). MONITORING URBAN GROWTH (CONT.)  Many studies focused on the use of remote sensing and GIS in urban growth modeling. E.g.:  Luo, Jun & Yu, Danlin & Miao, Xin. (2008). Modeling Urban Growth Using GIS and Remote Sensing. Giscience & Remote Sensing - GISCI REMOTE SENS. 45. 426-442. 10.2747/1548-1603.45.4.426.  Xuecao Li, Peng Gong, “Urban growth models: progress and perspective”, Science Bulletin, Volume 61, Issue 21, 2016, Pages 1637-1650, ISSN 2095-9273, https://doi.org/10.1007/s11434-016-1111-1.  Hanoon, S.K.; Abdullah, A.F.; Shafri, H.Z.M.; Wayayok, A. Urban Growth Forecast Using Machine Learning Algorithms and GIS-Based Novel Techniques: A Case Study Focusing on Nasiriyah City, Southern Iraq. ISPRS Int. J. Geo-Inf. 2023, 12, 76. https://doi.org/10.3390/ijgi12020076  “A Deep Dive into Predicting Urban Growth using ArcGIS and R”: https://www.esri.com/arcgis- blog/products/arcgis-pro/analytics/a-deep-dive-into- predicting-urban-growth-using-arcgis-and-r/ RECENT APPLICATIONS AND NEW DEVELOPMENTS  The first area of new research is in the use of Spectral Mixture Analysis and linear mixture models to map urban extent and quantify physical properties (Small 2001, 2002a, 2002b).  The dominant spatial scale of individual features (roads, buildings, etc.) in urban mosaics is generally 10 to 20 meters.  Operational sensors like Landsat and SPOT do not have sufficient spatial resolution to discriminate individual features so most urban pixels image several different features with different reflectance. These mixed pixels are distinct from the more spectrally homogeneous pixels associated with most other types of land cover.  Spectral mixture analysis and linear mixture models quantify these mixed pixels on the basis of the fractional abundance of different spectral endmembers (e.g., vegetation, water, high albedo). RECENT APPLICATIONS AND NEW DEVELOPMENTS (CONT.)  The second area involves the use of Shuttle Radar Topography Mission (SRTM) data to identify urban infrastructure (Nghiem et al. 2001). The data include derived topography and backscatter intensity at a nominal resolution of 30 m.  Urban areas are generally characterized by very high backscatter intensity as a result of the abundance of corner reflectors (buildings).  Some of the potentially derivable parameters include  urban extent and boundaries,  urban/suburban vegetation height and distribution,  building height and volume, which could be used for various social science applications. In particular, if used in conjunction with data from other sensors (Landsat, AVIRIS), and from other sources (traditional census data), it may represent an excellent dataset to quantify economic development and transportation infrastructure, as well as to identify housing and other building stock. RECENT APPLICATIONS AND NEW DEVELOPMENTS (CONT.)  The third area involves the application of time series data from, e.g., the Defense Meteorological Satellite Program (DMSP) Operational Linescan System (OLS) to derive georeferenced inventories of human settlements (Elvidge et al. 1997a). The visible band of the OLS is intensified at night, permitting detection of nocturnal visible-near infrared emissions. The authors have developed a methodology to identify different light emissions sources and produced four separate datasets, at a nominal resolution of 1 km:  Stable City Lights,  Fires,  Gas Flares and,  Lights from Fishing Boats. RECENT APPLICATIONS AND NEW DEVELOPMENTS (CONT.)  The city lights dataset has been used to explore the relationship between the area lit by anthropogenic visible-near infrared emissions and socioeconomic variables such population, economic activity and electric power consumption (Elvidge et al. 1997b). Others have begun to use the city lights dataset to map urban areas in the U.S. (Imhoff et al. 1997) to estimate the global human population (Sutton et al. 2001), and to develop a spatially explicit map of GDP (Sutton and Costanza 2002).  Currently, the Center for International Earth Science Information Network (CIESIN) is utilizing OLS night- time lights data in combination with population data, high resolution spatial data and satellite imagery to derive a global dataset of populations and area extents for urban and rural areas. For more information on the Urban-Rural Database Project, see SEDAC's Urban Remote Sensing website.  https://sedac.ciesin.columbia.edu/ REFERENCES Avissar, R. 1996. Potential effect of vegetation on the urban thermal environment. Atmosphere Environment 30:437-448. Balzerek, H. 2001. Applicability of IKONOS Satellite scenes: monitoring, classification and evaluation of urbanization processes in Africa. Berry, B. L. 1990. Urbanization. In Turner B. L., Clark, W. C., Kates, R. W., Richards, J. F., Matthews, J. T. and Meyer, W. B. (eds.). The earth transformed by human action, Cambridge, UK: Cambridge University Press. Chen, K. 2001. An approach to linking remotely sensed data and areal census data. International Journal of Remote Sensing 23(1):37-48. Chenyu Huang, Tingting Hu, Yusen Duan, Qingyu Li, Nan Chen, Qi Wang, Mengge Zhou, Pinhua Rao,”Effect of urban morphology on air pollution distribution in high-density urban blocks based on mobile monitoring and machine learning”, Building and Environment,Volume 219, 2022, 109173, ISSN 0360-1323, https://doi.org/10.1016/j.buildenv.2022.109173. Donnay, J. P, Barnsley M. J. and Longley, P.A., (eds). 2001. Remote sensing and urban analysis. London, UK: Taylor & Francis. Elvidge C.D., Baugh, K.E., Hobson, V.R., Kihn, E.A., Kroehl, H.W., Davis, E.R. and Cocero, D. 1997. Satellite inventory of human settlements using nocturnal radiation emissions: A contribution for the global toolchest. Global Change Biology 3:387-395. Elvidge, C.D., Baugh, K.E., Kihn, E A., Kroehl, H.W., Davis E.R. and Davis, C.W. 1997. Relation between observed visible-near infrared emissions, population, economic activity and electric power consumption. International Journal of Remote Sensing 18(6):1373-1379. Emmanuel, R. 1997. Urban vegetational change as an indicator of demographic trends in cities: The case of Detroit. Environment and Planning 24:415-426. Esnard, A. and Yang, Y. 2001. Descriptive and comparative studies of 1990 urban extent data for the New York Metropolitan Region. Urisa Journal, accepted for future publication, version 10/16/01. Gallo K.P., Tarpley, J.D., McNab, A.L. and Karl, T.R. 1995. Assessment of urban heat island: A satellite perspective. Atmospheric Research 37:37-43. Gallo, K.P., McNab, A.L., Karl, T.R., Brown, J.F., Hood, J.J. and Tarpley, J.D. 1993. The use of a vegetation index for assessment of the urban heat island effect. International Journal of Remote Sensing 14(11):2223-2230. Hanoon, S.K.; Abdullah, A.F.; Shafri, H.Z.M.; Wayayok, A. Urban Growth Forecast Using Machine Learning Algorithms and GIS-Based Novel Techniques: A Case Study Focusing on Nasiriyah City, Southern Iraq. ISPRS Int. J. Geo-Inf. 2023, 12, 76. https://doi.org/10.3390/ijgi12020076 REFERENCES (CONT.) Harris, P.M. and Ventura, S.J. 1995. The integration of geographic data with remotely sensed imagery to improve classification in an urban area. Photogrammetric Engineering and Remote Sensing 61(8):993-998. Imhoff M.L., Lawrence, W.T., Stutzer, D.C. and Elvidge, C.D. 1997. A technique for using composite DMSP/OLS 'City Lights' satellite data to map urban area. Remote Sensing of the Environment 61:361-370. Karathanassi, V., Jossifidis C.H., and Rokos, D. 2000. A texture-based classification method for classifying built areas according to their density. International Journal of Remote Sensing 21(9):1807-1823. Karl, T.R., Diaz, H. and Kukla, G. 1988. Urbanization: Its detection and effect in the United States climate record. Journal of Climate 1:1099-1123. Kokkonen, Tom V et al. “The effect of urban morphological characteristics on the spatial variation of PM2.5 air quality in downtown Nanjing.” Environmental science: atmospheres vol. 1,7 481-497. 26 Aug. 2021, doi:10.1039/d1ea00035g Longley, P. A. and Mesev, V. 2001. Measuring urban morphology using remotely-sensed imagery. In Donnay, J. P., Barnsley, M. J. and Longley, P. A. (eds.). Remote sensing and urban analysis. London, UK: Taylor and Francis. Luo, Jun & Yu, Danlin & Miao, Xin. (2008). Modeling Urban Growth Using GIS and Remote Sensing. Giscience & Remote Sensing - GISCI REMOTE SENS. 45. 426-442. 10.2747/1548-1603.45.4.426. Mesev, V. 1997. Remote sensing of urban systems: Hierarchical integration with GIS. Computers, Environment and Urban Systems 21(3/4):175-187. Moller-Jensen L. 1990. Knowledge-based classification of an urban area using texture and context information in Landsat TM Imagery. Photogrammetric Engineering and Remote Sensing 65(6):899-904. Neyns, R.; Canters, F. Mapping of Urban Vegetation with High-Resolution Remote Sensing: A Review. Remote Sens. 2022, 14, 1031. https://doi.org/10.3390/rs14041031 Nghiem, S., Balk, D., Small, C., Deichmann, U., Wannebo, A., Blom, R., Sutton, P., Yetman, G., Chen, R., Rodriguez, E., Houshmand, B. and Neumann, G. 2001. Global infrastructure: The potential of SRTM data to break new ground. White paper produced by CIESIN and NASA's Jet Propulsion Laboratory. Pozzi, F. and Small, C. 2002. Vegetation and population density in urban and suburban areas in the U.S.A. Presented at the Third International Symposyium on Remote Sensing of Urban Areas, Istanbul, Turkey, 11-13 June. Radeloff, V.C., Hagen, A.E., Voss, P.R., Field D.R. and Mladenoff, D.J. 2000. Exploring the spatial relationship between census and land-cover data. Society and Natural Resources 13(6):599-609. Roth, M., Oke, T.R. and Emery, W.J. 1989. Satellite-derived urban heat islands from three coastal cities and the utilization of such data in urban climatology. International Journal of Remote Sensing 10(11):1699-1720. REFERENCES (CONT.) Small, C. 2002. Global analysis of urban reflectance. Presented at the Third International Symposium on Remote Sensing of Urban Areas, Istanbul, Turkey, June 2002. Small, C. 2002. High resolution spectral mixture analysis of urban reflectance. Ikonos Special Issue of Remote Sensing of Environment, Submitted 05/2002. Stefanov W., Ramsey, M. and Christensen, P. 2001. Monitoring urban land cover change: an expert system approach to land cover classification of semiarid to arid urban centers. Remote Sensing of the Environment 77:173-185. Sutton, P., Roberts, D., Elvidge C.D. and Baugh, K. 2001. Census from heaven: An estimate of the global human population using night-time satellite imagery. International Journal of Remote Sensing 22(16):3061-3076. Sutton, P. and Costanza, R. 2002. Global estimates of market and non-market values derived from nighttime satellite imagery, land cover, and ecosystem service valuation. Ecological Economics 41(3):509-527. United Nations. 2001. World urbanization prospects, 1999 Revision. United Nations. Vogelamnn, J.E, Sohl, T. and Howard, S.M. 1998. Regional characterization of land cover using multiple sources of data. Photogrammetric Engineering and Remote Sensing 64(1):45-57. Wagner, T.W. and Ryznar, R.M. 1999. Toward understanding urban processes with remotely sensed data: Examples from Detroit. Presented at GeoInformatics99, University of Michigan, Ann Arbor, Michigan, June 19-21 1999. Walsh S.J., Evans, T.P., Welsh, W.F., Entwisle B. and Rindfuss, R.R. 1999. Scale-dependent relationship between population and environment in Northeastern Thailand. Photogrammetric Engineering and Remote Sensing 65(1):97-105. Wang, Y. and Zhang, X. 1999. Land cover change of metropolitan Chicago area from 1972 to 1997 and the impact to natural communities in the region. Presented at GeoInformatics99, University of Michigan, Ann Arbor, Michigan, June 19-21 1999. Xuecao Li, Peng Gong, “Urban growth models: progress and perspective”, Science Bulletin, Volume 61, Issue 21, 2016, Pages 1637-1650, ISSN 2095-9273, https://doi.org/10.1007/s11434-016-1111-1 Yuan, Y., Smith, R.M. and Limp, W.F. 1997. Remodeling census population with spatial information from Landsat TM Imagery. Computers, Environment and Urban Systems 21(3/4):245-258. RELATED RESOURCES EPA’s webpage about Urban Heat Island: http://www.epa.gov/heatislands MDPI’s special issue “Air Quality Research Using Remote Sensing” summarizes some approaches https://www.mdpi.com/journal/remotesensing/special_issues/AQ _RS The Long Term Ecological Research (LTER) Network: http://lternet.edu/ Urban Remote Sensing website and Urban-Rural Database Project: http://sedac.ciesin.columbia.edu/urban_rs/ “A Deep Dive into Predicting Urban Growth using ArcGIS and R”: https://www.esri.com/arcgis-blog/products/arcgis- pro/analytics/a-deep-dive-into-predicting-urban-growth-using- arcgis-and-r/ SEDAC's Urban Remote Sensing website https://sedac.ciesin.columbia.edu/ RGIS617: Urban and Environmental Applications of GIS/Remote Sensing Urban Landscape and Socioeconomic Characteristics Dr. Elnazir Ramadan Department of Geography and Urban Sustainability UAE University [email protected] REMOTE SENSING OF URBAN LANDSCAPES  Urban landscapes are composed of diverse materials (concrete, asphalt, metal, plastic, grass, …) arranged in a complex way to build housing, transportation systems, utilities, commercial and industrial facilities, and recreational land-life.  Given the high value of urban area, many agencies, organizations and individuals need up-to-date information.  Remote sensing (aerial photos, satellite imagery) provides an overall view from above.  Urban/Suburban Temporal Resolution Considerations There are three types of temporal resolution to consider: 1) Urban/suburban development cycle whose stages can be identified by the presence or absence of the following factors: Partial or complete clearing Land subdivision Roads Buildings Landscaping  Stages of Development  Urban/Suburban Temporal Resolution Considerations 2) How often it is possible for the remote sensor to collect data of the urban landscape. Urban applications are often not time sensitive and can do with a yearly acquisition. 3) How often land managers require this information.  Urban Remote Sensing/ spatial resolution requirements Minimum spatial resolution depends on the information we seek. The higher the resolution, the more detailed information can be derived. Minimum of four pixels within an object to identify (one-half the width of the smallest dimension - i.e. 5 m mobile homes requires at least 2.5 m data)  Remote Sensing Resolution Requirements Clear ellipses represent the spatial and temporal characteristics of selected urban attributes Gray boxes depict the spatial and temporal characteristics of the remote sensing systems that may be used to extract the required urban information  Minimum Resolution Requirements Attributes Temporal Spatial Spectral Land Use/Land Cover L1 - USGS Level I 5 - 10 years 20 - 100 m V-NIR-MIR-Radar L2 - USGS Level II 5 - 10 years 5 - 20 m V-NIR-MIR-Radar Urban/Suburban Attributes and L3 - USGS Level III 3 - 5 years 1-5m Pan-V-NIR-MIR L4 - USGS Level IV 1 - 3 years 0.25 - 1 m Panchromatic Building and Property Infrastructure the Minimum Remote Sensing B1 - building perimeter, area, height and cadastral 1 - 5 years 0.25 - 0.5 m Pan-Visible information (property lines) Transportation Infrastructure T1 - general road centerline T2 - precise road width 1 - 5 years 1 - 2 years 1 - 30 m 0.25 - 0.5 m Pan-V-NIR Pan-V Resolutions Required to Provide Such Information T3 - traffic count studies (cars, airplanes, etc.) 5 - 10 min 0.25 - 0.5 m Pan-V T4 - parking studies 10 - 60 min 0.25 - 0.5 m Pan-V Utility Infrastructure U1 - general utility line mapping and routing 1 - 5 years 1 - 30 m Pan-V-NIR U2 - precise utility line width, right-of-way 1 - 2 years 0.25 - 0.6 m Pan-Visible U3 - location of poles, manholes, substations 1 - 2 years 0.25 - 0.6 m Panchromatic Digital Elevation Model (DEM) Creation D1 - large scale DEM 5 - 10 years 0.25 - 0.5 m Pan-Visible D2 - large scale slope map 5 - 10 years 0.25 - 0.5 m Pan-Visible Socioeconomic Characteristics S1 - local population estimation 5 - 7 years 0.25 - 5 m Pan-V-NIR S2 - regional/national population estimation 5 - 15 years 5 - 20 m Pan-V-NIR S3 - quality of life indicators 5 - 10 years 0.25 - 30 m Pan-V-NIR Energy Demand and Conservation E1 - energy demand and production potential 1 - 5 years 0.25 - 1 m Pan-V-NIR E2 - building insulation surveys 1 - 5 years 1-5m TIR Meteorological Data M1 - weather prediction 3 - 25 min 1 - 8 km V-NIR-TIR M2 - current temperature 3 - 25 min 1 - 8 km TIR M3 - clear air and precipitation mode 6 - 10 min 1 km WSR-88D Radar M4 - severe weather mode 5 min 1 km WSR-88D Radar M5 - monitoring urban heat island effect 12 - 24 hr 5 - 30 m TIR Critical Environmental Area Assessment C1 - stable sensitive environments 1 - 2 years 1 - 10 m V-NIR-MIR C2 - dynamic sensitive environments 1 - 6 months 0.25 - 2 m V-NIR-MIR-TIR Disaster Emergency Response DE1 - pre-emergency imagery 1 - 5 years 1-5m Pan-V-NIR DE2 - post-emergency imagery 12 hr - 2 days 0.25 - 2 m Pan-V-NIR-Radar DE3 - damaged housing stock 1 - 2 days 0.25 - 1 m Pan-V-NIR DE4 - damaged transportation 1 - 2 days 0.25 - 1 m Pan-V-NIR DE5 - damaged utilities, services 1 - 2 days 0.25 - 1 m Pan-V-NIR  Socioeconomic Characteristics Population Estimations Quality of Life Indicators Energy Demands and Conservation  Population Estimations Size and distribution of population are key factors in resource allocation by governments. Population estimates are essential in planning processes They are critical in decisions about building public facilities (schools, hospitals, utilities, transportation infrastructure) Also used by private sectors for customer-profile analysis, market analysis, site location. Provide input for many urban and regional studies such as land-use and transportation interaction models, urban sprawl analysis, environmental studies.  Population Estimations (cont.) Accurate and timely estimates of population are important. Accurate population data is available through census. Usually decadal Time consuming Frequency does not meet the needs in rapid growth environments Appropriate estimation methods are needed Many approaches exist in demography Most commonly used method relies on inventorying of occupied housing units. Remote sensing provides alternatives.  Remote Sensing Assisted Population Estimation Population estimation can be performed at the local, regional, and national level based on (Lo, 1995; Haack et al., 1997): Counts of individual dwelling units. Applicable at the local level (small area) and requires high resolution imagery. Aerial photography, IKONOS, Quickbird, WorldView Measurement of urbanized land areas Applicable at large regional scales using low resolution imagery. AVHRR, DMSP-OLS, MODIS, Landsat Land use classification Applicable to small to medium regional scales using moderate resolution imagery. Landsat, IRS, SPOT  Remote Sensing Assisted Population Estimation Counts of Individual Dwelling Unit The most accurate method uses Dwelling Unit Estimation Technique with the following assumptions (Lo, 1995; Haack et al., 1996): imagery must be of sufficient spatial resolution (0.3 - 5 m) to identify individual structures even through tree cover and whether they are residential, commercial, or industrial buildings; some estimate of the average number of persons per dwelling unit must be available, and it is assumed all dwelling units are occupied.  Counts of Individual Dwelling Units - Example Estimating Al Ain population using IKONOS imagery (Yagoub, 2006) Assuming 9 persons per individual dwelling unit. Al Khabisi Automated building counts  Population estimates from individual units count limitations  Results from this method correlate well with census estimates.  This technique is useful at the local level. However, it will be very costly to implement at the regional or national level because of the need for in-situ calibration data.  Remote Sensing Assisted Population Estimation Measurement of Urbanized Land Area  Applicable at large regional scales with low resolution imagery.  Based on ‘allometric’ modeling of a direct mathematical relationship between the population of an urban area and its size.  Example correlating population and the size of urbanized area (Tobler, 1969; Olorunfemi, 1984) r = a Pb (r=radius, P=population, a and b = coefficients)  Night time city lights (Sutton et al., 1997; Imhoff et al., 1997, Pozzi et al., 2002) and Synthetic Aperture Radar (Henderson and Xia, 1997) are popular data sets used in delineating urban areas. DMSP-OLS near-infrared nighttime data (1x1 km) was used to estimate the size of urban areas in the US and predict population. https://geostat.fcsa.gov.ae/gisportal/apps/MapJournal/index.html?appid=36b9de03560040a59814412373609d2f GLOBAL NIGHTTIME CITY LIGHTS, 1994-1995 (POZZI, 2002) GLOBAL GRIDDED POPULATION DENSITY, 1990 (POZZI,2002) CORRELATION BETWEEN POPULATION AND CITY LIGHTS FREQUENCY (POZZI, 2002)  Remote Sensing Assisted Population Estimation Land Use Classification  Applicable at small to medium regional scales with moderate resolution imagery.  A population rate per unit area is associated with each land use type. Population estimate is obtained by multiplying the rate with the area occupied by each type. (Lo, 2003; Langford, 2006)  Land use can also be used to disaggregate census data at a finer resolution. (Fischer and Langford, 1996) LAND USE CLASSIFICATION OF AL AIN FROM LANDSAT IMAGERY (YAGOUB) LIMITATIONS OF REMOTE SENSING ASSISTED POPULATION COUNT  Spatial resolution: the use of very high spatial resolution images bring with them some major problems such as availability in panchromatic mode and large size (storage, processing time)  Image classification: The heterogonous nature (mixed pixels) of urban environment and the possibility that identical spectral reflectance values can correspond to very different land uses and functions poses a classification problem  3D nature of urban areas: Imaging 3D buildings from satellites suffers from three problems. The first one is related to displacement of buildings from their true location (relief displacement), the second one is related to obscuring of lower buildings by higher ones, and the third is related to shadow  Remote Sensing Quality of Living Indicators Quality of living indicators such as house value, median family income, average number of rooms, average rent, education, and income can be estimated by extracting the following variables from high spatial resolution panchromatic and/or color imagery (Lindgren, 1985; Lo, 1986; 1995; Haack et al., 1997): building size (sq. ft.) lot size (acreage) existence of a pool (sq. ft.) vacant lots per city block frontage (sq. ft.) distance house is set-back from street existence of driveways existence of garages number of autos visible paved streets (%) street width (ft.) health of the landscaping (vegetation index signature) proximity to manufacturing and/or retail activity.  Remote Sensing Quality of Living Indicators  Inaddition to these site specific factors, situations factors can be used:  Proximity to community ameneties  Schools, shopping, hospitals, fire station, …  Proximity to nuisances or hazards  Heavy street traffic, airports, sewage-treatment plant, …  These factors have to be correlated with in situ census information to compute the quality of life indicators.  In some areas presence of green vegetation (grass) is correlated to the quality of life.  Socioeconomic Characteristics 0.3 1m 5 10 20 100 m 107 8 15 year 5 10 year S2 D1 D2 LII LI 3 5 year S3 S1 2 4 year E1 LIII 3 year 106 2 year U1 DE1, E2 T1 B1 LIV 8 T2 C1 U2 5 1 year U3 3 180 day SPOT HRV 1,2,3 and 4 (1998) 2 Pan 10 x10 JERS-1 MSS 20 x 20 105 SPOT HRG (2002) IRS-1 AB 100,000 min B2 C2 MSS 18 x 24 Radar 18 x 18 Pan 3 x 3; 5 x 5 (not shown) LISS-1 72.5 x 72.5 LISS-2 36.25 x 36.25 8 55 day MSS 10 x 10; MIR 20 x 20 IRS-1C 44 day Pan 5.8 x 5.8 5 30 day LISS-3 23 x 23; MIR 70 x 70 26 day EarthWatch 3 22 day Earlybird (1998) Pan 3 x 3 2 16 day MSS 15 x 15 NOAA AVHRR Temporal Resolution Quickbird (1998) LANDSAT 4,5 LAC 1.1 x 1.1 km GAC 4 x 4 km 9 day 0.82 x 0.82 MSS 79 x 79 104 3.28 x 3.28 TM 30 x 30 10,000 min LANDSAT 7 (1998) 8 5 day Pan 15 x 15 5 4 day 3 day 3 2 day DE3 2 DE4 MODIS* Land 0.25 x 0.25 km 1 day DE5 Land 0.50 x 0.50 km RADARSAT 103 DE2 Ocean 1 x 1 km 1,000 min M5 11 x 9 Atmo 1 x 1 km 8 12 hr 100 x 100 TIR 1 x 1 km Konso village in southern Ethiopia EOSAT/Space Imaging 5 IKONOS (1998) Pan 1 x 1 MSS 4 x 4 3 IRS-P5 (1999) Pan 2.5 x 2.5 2 GOES ORBIMAGE VIS 0.9 X 0.9 km M1 OrbView 3 (1999) TIR 8.0 X 8.0 km 102 100 min Pan 1 x 1 8 MSS 4 x 4 1 hr 5 M2 3 30 min Single and multiple family 2 T4 Aerial Photography 0.3 x 0.3 m (0.98 x 0.98 ft.) METEOSAT residences in Columbia, S. C. 1 x 1 m (3.281 x 3.281 ft.) VISIR 2.5 x 2.5 km TIR 5 x 5 km M3 10 10 min Ground 8 T3 Doppler M4 Radar 5 5 min 4 x 4 km Temporal Spatial 3 3 min 1 km Resolution Resolution 2 1m 2 3 5 10 15 20 30 100 m 1000 m 5 km 10 km S1 - local population estimation 5 - 7 years 0.3 - 5 m 0.2 0.3.5.8 1.0 2 3 5 10 2 3 5 102 2 3 5 103 2 34 5 8 104 S2 - regional/national population estimation 5 - 15 years 5 - 20 m S3 - quality of life indicators 5 - 10 years 0.3 - 0.5 m Spatial Resolution in meters Example of Quality of Living Indicators (Yagoub et al., 2022)  Energy demand and conservation  Localurban/suburban energy demand (heat- load density) can be predicted from remote sensing.  Building footage is determined from high resolution imagery.  Correlated to consumption for representative samples.  Regression relations are then used to predict the anticipated consumption in the region.  0.3 1m 5 10 20 100 m Energy Demand and Conservation 107 8 15 year 5 10 year S2 D1 D2 LII LI 3 5 year S3 S1 2 4 year E1 LIII 3 year 106 2 year U1 DE1, E2 T1 B1 LIV 8 T2 C1 U2 5 1 year U3 3 180 day SPOT HRV 1,2,3 and 4 (1998) 2 Pan 10 x10 JERS-1 MSS 20 x 20 105 SPOT HRG (2002) IRS-1 AB 100,000 min B2 C2 MSS 18 x 24 Radar 18 x 18 Pan 3 x 3; 5 x 5 (not shown) LISS-1 72.5 x 72.5 LISS-2 36.25 x 36.25 8 55 day MSS 10 x 10; MIR 20 x 20 IRS-1C 44 day Pan 5.8 x 5.8 5 30 day LISS-3 23 x 23; MIR 70 x 70 26 day EarthWatch 3 22 day Earlybird (1998) Pan 3 x 3 2 16 day MSS 15 x 15 NOAA AVHRR Temporal Resolution Quickbird (1998) LANDSAT 4,5 LAC 1.1 x 1.1 km GAC 4 x 4 km 9 day 0.82 x 0.82 MSS 79 x 79 104 3.28 x 3.28 TM 30 x 30 10,000 min LANDSAT 7 (1998) 8 5 day Pan 15 x 15 5 4 day 3 day 3 2 day DE3 2 DE4 MODIS* Land 0.25 x 0.25 km 1 day DE5 Land 0.50 x 0.50 km RADARSAT 103 DE2 Ocean 1 x 1 km 1,000 min M5 11 x 9 Atmo 1 x 1 km 8 12 hr 100 x 100 TIR 1 x 1 km EOSAT/Space Imaging 5 IKONOS (1998) Pan 1 x 1 MSS 4 x 4 3 IRS-P5 (1999) Pan 2.5 x 2.5 2 GOES ORBIMAGE VIS 0.9 X 0.9 km M1 OrbView 3 (1999) TIR 8.0 X 8.0 km 102 100 min Pan 1 x 1 8 MSS 4 x 4 1 hr 5 Daytime high resolution (0.3 Nighttime 0.3 x 0.3 m M2 3 30 min x 0.3 m) aerial photography thermal infrared 2 T4 Aerial Photography 0.3 x 0.3 m (0.98 x 0.98 ft.) METEOSAT of a gymnasium imagery (8 - 14 µm) 1 x 1 m (3.281 x 3.281 ft.) VISIR 2.5 x 2.5 km TIR 5 x 5 km M3 10 10 min Ground 8 T3 Doppler M4 Radar 5 5 min 4 x 4 km 3 3 min 1 km Temporal Spatial 2 1m 2 3 5 10 15 20 30 100 m 1000 m 5 km 10 km Resolution Resolution 0.2 0.3.5.8 1.0 2 3 5 10 2 3 5 102 2 3 5 103 2 34 5 8 104 E1 - energy demand and production potential 1 - 5 years 0.3 - 1 m E2 - building insulation surveys 1 - 5 years 1 - 5 m Spatial Resolution in meters Land suitability for solar energy in UAE Solar irradiance Relative suitability indices Land suitability Dust risk (Gherboudj & Ghedira, 2015) REFERENCES  Haack, B., Craven, D., and Jampoler, S. M., 1996, GIS tracks Kathmandu valleys urban explosion. GIS World, 9, 54–57.  Henderson, F. M., and Xia, Z. G., 1997, SAR applications in human settlement detection, population estimation and urban land use pattern analysis: a status report. IEEE Transactions on Geoscience and Remote Sensing, 35, 79–85.  Gherboudj, Imen & Ghedira, Hosni. (2015). Assessment of solar energy potential over the United Arab Emirates using remote sensing and weather forecast data. Renewable and Sustainable Energy Reviews. 55. 10.1016/j.rser.2015.03.099.  Imhoff, M. L., Lawrence,W. T., Stutzer, D. C., and Elvidge, C. D., 1997, A technique for using composite DMSP/OLS ‘city lights’ satellite data to map urban area. Remote Sensing of Environment, 61, 361–370.  Jensen, J.R., 2000. Remote Sensing of the Environment: An Earth Resource Perspective, Prentice-Hall, Upper Saddle River, NJ (Chapter 12: Remote Sensing The Urban Landscape).  Langford, M., Maguire, D. J., and Unwin, D. J., 1991, The areal interpolation problem: estimating population using remote sensing within a GIS framework. In Handling Geographical Information: Methodology and Potential Applications, edited by I. Masser and M. Blakemore (London: Longman) pp. 55–77.  Lo, C. P., 1995, Automated population and dwelling unit estimation from high-resolution satellite images: a GIS approach. International Journal of Remote Sensing, 16, 17–34.  Olerunfemi, J. F., 1986, Towards a philosophy of population census in Nigeria: remote sensing inputs. Remote Sensing Yearbook 1986, 117–125.  Pozzi, F., Small, C., Yetman, G., 2002, Modeling the distribution of human population with night-time satellite imagery and gridded population of the world, PECORA 15 Conference Processdings  Sutton, P., Roberts, D., Elvidge, C., and Meij, H. (1997) A comparison of nighttime satellite imagery and population density for the continental United States. Photogrammetric Engineering and Remote Sensing, 63, 1303–1313.  Yagoub, M.M., 2006, Application of remote sensing and Geographic Information Systems (GIS) to population studies in the Gulf: A case of Al Ain city (UAE), Journal of the Indian Society of Remote Sensing, 34, 1.  Yagoub, M.M.; Tesfaldet, Y.T.; Elmubarak, M.G.; Al Hosani, N. Extraction of Urban Quality of Life Indicators Using Remote Sensing and Machine Learning: The Case of Al Ain City, United Arab Emirates (UAE). ISPRS Int. J. Geo-Inf. 2022, 11, 458. https://doi.org/10.3390/ijgi11090458 RGIS617: Urban and Environmental Applications of GIS/Remote Sensing Application of Remote Sensing in Air Quality Monitoring Dr. Elnazir Ramadan Department of Geography and Urban Sustainability UAE University [email protected] Outline Importance of air quality Air Quality Index and its components Air quality monitoring stations Why use remote sensing Commonly used sensors and products PM2.5 and PM10 estimation Case study using MODIS AOD to estimate PM10 Conclusions UN Sustainable Development Goals (SDGs) A plan of action for people, planet, and prosperity All countries and all stakeholders, acting in collaborative partnership, will implement this plan 17 SDGs and 169 targets under this agenda Balance the three dimensions of sustainable development: economic, social, and environmental Source: “Transforming our world: the 2030 Agenda for Sustainable Development” https://sdgs.un.org/2030agenda Air Quality Index (AQI) - UAE https://airquality.ncm.ae/?lang=en  How is AQI calculated? The Air Quality Index (AQI) is calculated based on five major air pollutants: Particulate matter (PM2.5 and/or PM10) Nitrogen dioxide (NO2) Carbon monoxide (CO) Sulfur dioxide (SO2) Ground-level ozone (O3) UAE has established National Ambient Air Quality Standards to protect public health (see table). The AQI is determined by the component having the highest C/Climit ratio (%), where C is the concentration and Climit is the air quality standard for each component. AQI values below 100 are considered satisfactory. Whereas values above 100 are considered unhealthy for specific sensitive groups or for everyone based on their magnitude. Source: Environmental Agency – Abu Dhabi https://www.adairquality.ae/ AQI and Major Pollutant https://airquality.ncm.ae/?lang=en Traditional Air Quality Monitoring Source: www.aqicn.org  Why use Remote Sensing? Ground Sensor Network (openaq.org) Population Density Air Pollution Monitoring  Benefit of Remote Sensing Satellite estimated PM2.5 Ground Sensor Network Source: Van Donkelaar et al., 2010 What do satellites measure ? The Remote Sensing Process Commonly Used Sensors for Air Quality Satellite Sensor Launch date Derived Spatial resolution Temporal parameters resolution Terra Moderate 18-12-1999 Total column 10 km, 3 km, 1 km Daily Aqua Resolution 4-5-2002 Aerosol Optical Imaging Depth Spectroradiometer (MODIS) Aura Ozone Monitoring 15-7-2004 O3, SO2, NO2 13 x 24 km Daily Instrument (OMI) Suomi NPP Visible Infrared 28-10-2011 Total column 6 km Daily JPSS (NOAA-20) Imaging 18-11-2017 Aerosol Optical Radiometer Suite Depth (VIIRS) Aerosol type Sentinel 5P TROPOspheric 13-10-2017 Total and 3.5 x 5.5 km – Daily Monitoring Tropospheric O3, 7 x 7 km Instrument SO2, NO2, CO (TROPOMI) Example of a SO2 plume (Al Mishraq Sulfur Plant Fire on 20-10-2018) MODIS True Color Composite (22-10-2018) OMI SO concentration (24-10-2018) 2 Source: NASA Earth Observatory (https://earthobservatory.nasa.gov/images/88994/sulfur-dioxide-spreads-over-Iraq) SO2 from a volcanic eruption SO2 plume from the eruption of La Soufriere volcano on Saint Vincent island Source: TROPOMI website as captured by TROPOMI on 15-4-2021. The volcano erupted on 27-12-2020. http://www.tropomi.eu/data-products/sulphur-dioxide Global CO concentration by TROPOMI Source: TROPOMI website Average CO mixing ratio derived from data collected 13-19 Nov 2017 http://www.tropomi.eu/data-products/carbon-monoxide NO2 concentration from OMI Source: https://www.temis.nl/airpollution/no2col PM2.5 and PM10 Tiny solid and liquid particles suspended in the atmosphere are called aerosols or particulate matter (PM) Aerosol particles are very small. Particles with diameter 2.5 µm and less are considered fine particles and denoted PM2.5 while particles with diameter between 2.5 and 10 µm are considered coarse particles and denoted PM10. While the particles are too small to be seen, their effect on radiation can be perceived. Pictures in Pittsburg, PA taken from the same location. The one on the left with high concentration of aerosols reducing the visibility and making the sky look grey-ish. (source: caice.ucsd.edu/introduction-to-aerosols/) Aerosol Optical Depth (AOD) The optical depth expresses the quantity of light removed from a beam by scattering and absorption during its path through a medium (the atmosphere). The optical depth is higher for higher concentrations of PM The optical depth (τ) : τaer = AOD also called Aerosol Optical Thickness (AOT) AOD Values AOD is a unitless value AOD values are small for clear atmosphere and increase with the amount of pollution 0.02 – very clean isolated areas 0.08 – background over ocean 0.2 – fairly clean 0.6 – polluted 1.5 – heavy smoke or dust events > 3.0 sun’s disk obscured AOD is ‘spectral’: varies with wavelength The Angstrom exponent (α) describes how AOD depends on wavelength τλ / τλ0 = (λ / λ0)^α α < 1.0 => coarse particles α > 2.0 => fine particles Several algorithms exit to calculate AOD from radiance. They require radiance measurement at multiple wavelengths including the mid-infrared Aerosol Retrieval from MODIS MODIS implements 2 algorithms to retrieve AOD from radiance Dark Target that assumes that surface reflectance at different wavelengths are correlated over dark targets. Does not work over bright surfaces. Deep Blue that assumes that surface appears darker and aerosol signal stronger in the violet/blue (400-490 nm) than at longer wavelengths. (source: aerocenter.gsfc.nasa.gov) AERONET AERONET measurements of aerosol depth are considered http://aeronet.gsfc.nasa.gov/ ground truth and are used to validate satellite aerosol retrievals MODIS AOD 2/1/2018 Dark target algorithm Deep blue algorithm Converting AOD to PM Factors that can affect the conversion: - Aerosol type and atmospheric conditions (atmospheric pressure, humidity) - Vertical distribution of aerosol (boundary layer height, transport, production, loss) PM estimation methods Case study: Predicting PM10 from MODIS AOD for Al Ain region The study implements a simple two variable method to establish a correlation between PM10 measured at EAD monitoring stations and MODIS AOD derived using the Deep Blue algorithm. Used data collected in 2018 (Saleous et al., 2021)  Interpolated PM10 values from station data 15-1-2018 15-3-2018 15-6-2018 15-9-2018 Seasonal variations of PM10 and MODIS AOD Monthly average PM10 concentrations Monthly average of MODIS AOD Approach Regression: individual stations Regression: All stations Conclusions Remote Sensing can be used to predict air quality indicators and allow filling the gaps due to lack of air monitoring stations Limitations of remote sensing systems need to be recognized Acquisition during daytime only Low temporal resolution compared to air monitoring stations Total column retrieval More complex prediction models taking into account factors affecting the quality are to be considered Geostationary platforms with AOD retrieval capabilities can be used to increase temporal resolution (GEOS-16 + , Meteosat-8 +) RGIS617: Urban and Environmental Applications of GIS/Remote Sensing Climate Change: Causes & Impacts Dr. Elnazir Ramadan Department of Geography and Urban Sustainability UAE University [email protected]  The Greenhouse Effect  Infrared (IR) active gases, principally water vapor (H2O), carbon dioxide (CO2) and ozone (O3), naturally present in the Earth’s atmosphere, absorb thermal IR radiation emitted by the Earth’s surface and atmosphere.  The atmosphere is warmed by this mechanism and, in turn, emits IR radiation, with a significant portion of this energy acting to warm the surface and the lower atmosphere.  As a consequence the average surface air temperature of the Earth is about 30° C higher than it would be without atmospheric absorption and re-radiation of IR energy.  This phenomenon is popularly known as the greenhouse effect, and the IR active gases responsible for the effect are likewise referred to as greenhouse gases.  The Greenhouse Effect Incoming Solar 2 Radiation 343 W/m Long-wave Radiation 2 Reflected Solar 240 W/m 2 Radiation 103 W/m CO2 CH4, N2O, O3, “Blanket” of Greenhouse Gases Water vapour, aerosols, clouds Earth’s ground temperature, approx 13oC with greenhouse effect, approx ­ 20oC without it. Doubling CO2 increases temperature by between 1.5oC and 4oC. Sun Greenhouse Effect  The Greenhouse Effect The rapid increase in concentrations of greenhouse gases since the industrial period began has given rise to concern over potential resultant climate changes  Greenhouse Gases and Global Climate Change  The principal greenhouse gas concentrations that have increased over the industrial period are carbon dioxide (CO2), methane (CH4), nitrous oxide (N2O), and chlorofluorocarbons (CFCs).  The observed increase of CO2 in the atmosphere from about 280 ppm in the pre-industrial era to about 364 ppm in 1997 has come largely from fossil fuel combustion and cement production.  Of the several anthropogenic greenhouse gases, CO2 is the most important agent of potential future climate warming because of its large current greenhouse forcing, its substantial projected future forcing, and its long persistence in the atmosphere.  Recorded Worldwide Temperatures 0.8 0.6 ∆ Mean Temperature (°C) 0.4 0.2 0.0 -0.2 -0.4 -0.6 1880 1900 1920 1940 1960 1980 2000 Year  2005 Temperature Changes Compared to 1951-1980 -3 -2.5 -1.5 -1 -.5 -.1.1.5 1 1.5 2.5 3.4 Ozone Layer Depletion and Climate Change  The ozone layer absorbs harmful ultraviolet-B radiation from the sun. Over the past 30 years ozone levels over parts of Antarctica have dropped by almost 40% during some months and a 'hole' in ozone concentrations is clearly visible in satellite observations.  Ozone is been damaged mainly by: 1. Chlorofluorocarbons (CFCs) that are used in refrigerators, aerosols, and as cleaners in many industries. 2. Halons that are used in fire extinguishers. 3. Aircraft emissions of nitrogen oxides and water vapour.  As Ozone is considered to be a greenhouse gas, a depleted ozone layer may partially dampen the greenhouse effect. This may therefore lead to increased global warming.  Conversely, efforts to tackle ozone depletion may result in increased global warming!  Some Impacts of Climate Change  The hydropower-dependent energy sector in Tanzania has been seriously affected by drought. The country is turning to coal and natural gas as new sources of energy.  Some cities in Europe and USA experienced power shortages during summer of 2006 due to effects of increased temperatures (The infrastructures failed to cope with the record heat - unsuitable wires, pipes, etc not designed for higher temperatures).  About 82% of the icecap on mount Kilimanjaro in 1912 is now gone. If recession continues at the present rate, the majority of the mountain glaciers could vanish in the next 15 years.  The area covered by glaciers on the Rwenzori Mountains halved between 1987 and 2003, expected to disappear in the next 20 years. The Melting Snows of Mt Kilimanjaro  RELATIVE SEA LEVEL CHANGE  Sea level varies as a result of processes operating on a great range of time-scales, from seconds to millions of years, so that current sea level change is also related to past climate change.  The local change in sea level at any coastal location as measured by a tide gauge depends on the sum of global, regional and local factors and is termed relative sea-level change.  It is so called because it can come about either by movement of the land on which the tide gauge is situated or by the change in the height of the adjacent sea surface.  Relative sea levels are also measured by dating buried coastal vegetation (salt marshes, mangroves, etc.).  Most of the tide gauges are located in mid-latitude northern hemisphere, few in middle of oceans, and contaminated by earth movements.  The main source for the uncertainties in using tide gauge records still remain: poor historical distribution of tide gauges, lack of data from Africa and Antarctica, the GIA corrections used, and localized tectonic activity. CLIMATE CHANGE AND SEA LEVEL RISE  Sea-level rise due to global warming occurs primarily because water expands as it warms up. The melting ice caps and mountain glaciers also add water to the oceans, thus rising the sea level.  The contribution from large ice masses in Greenland and Antarctica is expected to be small over the coming decades. But it may become larger in future centuries.  Sea-level rise can be offset up by irrigation, the storage of water in reservoirs, and other land management practices that reduce run-off of water into the oceans. Changes in land-levels due to coastal subsidence or geological movements can also affect local sea-levels.  Average Rate of Sea Level Rise and the Estimated Contributions from Different Processes: 1910 - 1990 Factor Min Mid value Max Ocean thermal expansion 0.3 0.5 0.7 Glaciers and ice caps 0.2 0.3 0.4 Greenland – 20th Century effects 0.0 0.05 0.1 Antarctica – 20th Century effects -0.2 -0.1 0.0 Ice sheets – Adjustment since LGM 0.0 0.25 0.5 Permafrost 0.00 0.025 0.05 Sediment deposition 0.00 0.025 0.05 Terrestrial storage -1.1 -0.35 0.4 Total -0.8 0.7 2.2 Estimated from tide gauge records 1.0 1.5 2.0  CLIMATE CHANGE AND SEA LEVEL RISE  About 20,000 years ago during the LGM, large ice sheets melted causing a rise in sea level of about 100m, most of the melting occurred about 6,000 years ago.  Over the past 1,000 years and prior to the 20th century, the average global sea level rise was of the order of 0.2 mm/yr.  The rate of sea level rise climbed to about 1-2 mm/yr during the 20th century, with a central value of 1.5 mm/yr (IPCC TAR). The most recent estimate during the 20th century is 1.4 -2.0 mm/yr, with a central value of 1.7 ± 0.3 mm/yr (Church & White, 2006).  This significant rate of rise in sea level is attributed to global warming caused by industrialization during the second half of the 19th century.  CLIMATE CHANGE AND SEA LEVEL RISE  There is no evidence for any acceleration of sea level rise in data from the 20th century data alone. Mediterranean records show decelerations and even decreases in sea level in the latter part of the 20th century.  Most records show evidence of a gradual rise in global mean sea level over the last century. However, signals caused by land movements (e.g. uplift or submergence) can mask this signal due to actual changes in sea level.  The IPCC has estimated that, if the emission of greenhouse gases continues at the current rate, the level of the sea surface will rise by an additional 8-20 cm by 2030, 21-71 cm by 2070 and 31-110 cm by 2100. Global Sea Level Change Over the Last 140,000 Years (IPCC TAR)  THE PROSPECT OF SATELLITE ALTIMETRY IN SEA LEVEL STUDIES  Satellite altimetry provides near-global coverage of the world’s oceans and thus the promise of determining the global-averaged sea level rise, its regional variations, and changes in the rate of rise more accurately and quickly than is possible from the sparse array of in situ gauges.  TOPEX/Poseidon satellite altimeter mission with its (near) global coverage from 66°N to 66°S was launched in August 1992. Estimates of the rates of rise from the short T/P record are 2.5 ± 1.3 mm/ yr over the 6-yr period 1993–98 (Church et al, 2004).  Using a combination of tide gauge records and satellite altimetry, Jevrejeva et al. (2006) have estimated this rate to be 2.4 mm/yr over the same period.  THE PROSPECT OF SATELLITE ALTIMETRY IN MEAN SEA LEVEL STUDIES  Whether this larger estimate is a result of an increase in the rate of rise, systematic errors in the satellite and/or in situ records, the shortness of the satellite record, or a reflection of the large error bars is not clear.  Analysis of TOPEX/Poseidon satellite altimeter data has demonstrated that meaningful estimates of global averaged mean sea level change can be made over much shorter periods than possible with tide gauges because the global satellite data account for horizontal displacements of ocean mass.  However, achieving the required sub-millimeter accuracy is demanding and requires satellite orbit information, geophysical and environmental corrections and altimeter range measurements of the highest accuracy. It also requires continuous satellite operations over many years and careful control of biases.  PHYSICAL IMPACTS OF SEA LEVEL RISE PRIMARY IMPACTS  Inundation and displacement of wetlands and lowlands  Increased vulnerability to coastal storm damage and flooding  Shoreline erosion  Saltwater intrusion into estuaries and freshwater aquifers SECONDARY IMPACTS  Altered tidal ranges in rivers and bays  Changes in sedimentation patterns ?