High Spatial Resolution Airborne Multispectral Thermal Infrared Data to Support Analysis and Modeling Tasks in EOS IDS Project ATLANTA

Dale A. Quattrochi (dale.quattrochi@msfc.nasa.gov), NASA, Global Hydrology and Climate Center, Huntsville, AL

Jeffrey C. Luvall (jeff.luvall@msfc.nasa.gov) , NASA, Global Hydrology and Climate Center, Huntsville, AL


Project ATLANTA (ATlanta Land-use ANalysis: Temperature and Air-quality) as a newly-funded NASA EOS Interdisciplinary Science (IDS) investigation in 1996, seeks to observe, measure, model, and analyze how the rapid growth of the Atlanta, Georgia metropolitan area since the early 1970's has impacted the region's climate and air quality. The primary objectives for this research effort are: 1) To investigate and model the relationship between Atlanta urban growth, land cover change, and the development of the urban heat island phenomenon through time at nested spatial scales from local to regional; 2) To investigate and model the relationship between Atlanta urban growth and land cover change on air quality through time at nested spatial scales from local to regional; and 3) To model the overall effects of urban development on surface energy budget characteristics across the Atlanta urban landscape through time at nested spatial scales from local to regional. Our key goal is to derive a better scientific understanding of how land cover changes associated with urbanization in the Atlanta area, principally in transforming forest lands to urban land covers through time, has, and will, effect local and regional climate, surface energy flux, and air quality characteristics. Allied with this goal is the prospect that the results from this research can be applied by urban planners, environmental managers and other decision-makers, for determining how urbanization has impacted the climate and overall environment of the Atlanta area. It is our intent to make the results available from this investigation to help facilitate measures that can be applied to mitigate climatological or air quality degradation, or to design alternate measures to sustain or improve the overall urban environment in the future. Project ATLANTA is a multidisciplinary research endeavor and enlists the expertise of 8 investigators: Dale Quattrochi (PI) (NASA/Global Hydrology Center); Jeffrey Luvall (NASA/Global Hydrology and Climate Center); C.P. Lo (University of Georgia); Stanley Kidder (Colorado State University); Haider Taha (Lawrence Berkeley National Laboratory); Robert Bornstein (San Jose State University); Kevin Gallo (NOAA/NESDIS); and Robert Gillies (Utah State University).

Atlanta Urban Growth and Effects on Climate and Air Quality

In the last half of the 20th century, Atlanta, Georgia has risen as the premier commercial, industrial, and transportation urban area of the southeastern United States. The rapid growth of the Atlanta area, particularly within the last 25 years, has made Atlanta one of the fastest growing metropolitan areas in the United States. The population of the Atlanta metropolitan area increased 27% between 1970 and 1980, and 33% between 1980-1990 (Research Atlanta, Inc., 1993). Concomitant with this high rate of population growth, has been an explosive growth in retail, industrial, commercial, and transportation services within the Atlanta region. This has resulted in tremendous land cover change dynamics within the metropolitan region, wherein urbanization has consumed vast acreages of land adjacent to the city proper and has pushed the rural/urban fringe farther and farther away from the original Atlanta urban core. An enormous transition of land from forest and agriculture to urban land uses has occurred in the Atlanta area in the last 25 years, along with subsequent changes in the land-atmosphere energy balance relationships.

Air quality has degenerated over the Atlanta area, particularly in regard to elevations in ozone and emissions of volatile organic compounds (VOCs), as indicated by results from the Southern Oxidants Study (SOS) which has focused a major effort on measuring and quantifying the air quality over the Atlanta metropolitan region. SOS modeling simulations for Atlanta using U.S. Environmental Protection Agency (EPA) State Implementation Plan guidelines suggest that a 90% decrease in nitrogen oxide emissions, one of the key elements in ozone production, will be required to bring Atlanta into attainment with the present ozone standard (SOS, 1995).

Project ATLANTA Science Approach

The scientific approach we are using in relating land cover changes with modifications in the local and regional climate and in air quality, is predicated on the analysis of remote sensing data in conjunction with in situ data (e.g., meteorological measurements) that are employed to initialize local and regional-level numerical models of land-atmosphere interactions. Remote sensing data form the basis for quantifying how land covers have changed within the Atlanta metropolitan area through time from the mid-1970's, when Atlanta's dramatic growth began in earnest, to the present. These remotely sensed data will be used to provide input to numerical models that relate land cover change through time with surface energy flux and meteorological parameters to derive temporal models of how land cover changes have impacted both the climatology and the air quality over the Atlanta region. Current remote sensing data (i.e., data obtained during 1997) will be used to calibrate the models and as baseline data for extending the models to predict how prospective future land cover changes will effect the local and regional climate and air quality over the Atlanta-north Georgia region. Additionally, remote sensing data will be used as an indirect modeling method to describe urbanization and deforestation parameters that can be used to assess, as well as predict, the effects of land use changes on the local microclimate.

In concert with the remote sensing-based analysis and modeling of land cover changes is an extensive numerically-based modeling effort to better understand the cause-and-effect relationships between urbanization and trends in climatology and air quality. Sophisticated numerical meteorological models can complement extensive field monitoring projects and help improve our understanding of these relationships and the evolution of the urban climate on a location-specific basis. Measured data alone cannot resolve the relationships between the many causes of urban heat islands/urban climates and observations. For example, measured data cannot directly attribute a certain fraction of temperature rise to a certain modification in land use patterns, change in energy consumption, or release of anthropogenic heat into the atmosphere. These are aspects that numerical modeling can help resolve. Similarly, monitored air quality data cannot be used to establish a direct cause-and-effect relationship between emission sources, activities, or urbanization and observed air quality (e.g., smog). In this sense, photochemical models can be used in testing the sensitivity of ozone concentrations to changes in various land-use components, emission modifications and control, or other strategies. Thus, we are incorporating an assessment of land cover/land use change as measured from remote sensing data, with temporal numerical modeling simulations to better understand the effects that the growth of Atlanta has had on local and regional climate characteristics and air quality.

ATLAS Data: Role and Characteristics

To augment the quantitative measurements of land cover change and land surface thermal characteristics derived from satellite data (i.e, Landsat MSS and TM data for assessment of land cover change; Landsat TM thermal, and AVHRR and GOES data for land surface thermal characteristics), we are employing high spatial resolution airborne multispectral thermal data to provide detailed measurements of thermal energy fluxes that occur for specific surfaces (e.g., pavements, buildings) across the Atlanta urban landscape, and the changes in thermal energy response for these surfaces between day and night. This information is critical to resolving the underlying surface responses that lead to development of local and regional-scale urban climate processes, such as the urban heat island phenomenon and related characteristics. (Quattrochi and Ridd, 1994, 1997). These aircraft data will also be used to develop a functional classification of the thermal attributes of the Atlanta metropolitan area to better understand the energy budget linkages between the urban surface and the boundary layer atmosphere. This will be performed using the Thermal Response Number (TRN) (Luvall and Holbo, 1989; Luvall, 1997) which is expressed as


Where Rn is total net radiation and T change in surface temperature for time period t1 to t2.

Because urban landscapes are very complex in composition, the partitioning of energy budget terms depends on surface type. In natural landscapes, the partitioning is dependent on canopy biomass, leaf area index, aerodynamic roughness, and moisture status, all of which are influenced by the development stage of the ecosystem. In urban landscapes, however, the distribution of artificial or altered surfaces substantially modifies the surface energy budget. Thus, one key component of Project ATLANTA is to measure and model surface energy responses in both space and time, to better understand the processes-responses of urban climate and air quality interactions across the Atlanta metropolitan area.

The airborne sensor used to acquire high spatial resolution multispectral thermal infrared data over Atlanta is the Advanced Thermal and Land Applications Sensor (ATLAS), which is flown onboard a Lear 23 jet aircraft operated by the NASA Stennis Space Center. The ATLAS is a 15-channel multispectral scanner that basically incorporates the bandwidths of the Landsat TM (along with several additional channels) and 6 thermal IR channels similar to that available on the airborne Thermal Infrared Multispectral Scanner (TIMS) sensor (Table 1). Of particular importance to the Atlanta study is the multispectral thermal IR capability of the ATLAS instrument. ATLAS thermal IR data, collected at a very high spatial resolution, have been used to study urban surface energy responses in a previous study over the Huntsville, Alabama metropolitan area with excellent results (Lo et al., 1997).

Table 1. ATLAS sensor system specifications


BandWidth Limits (µm)

mW/cm-2 µm


2 mrad
10.45-0.52<0.008 N/A0.5Ambient
20.52-0.60<0.004 N/A0.5Ambient
30.60-0.63<0.006 N/A0.5Ambient
40.63-0.69<0.004 N/A0.5Ambient
50.69-0.76<0.004 N/A0.5Ambient
60.76-0.90<0.005 N/A0.5Ambient
71.55-1.75<0.05 N/A0.577 oK
82.08-2.35<0.05 N/A0.577 oK
93.35-4.20N/A <0.30.577 oK
108.20-8.60N/A <0.20.577 oK
118.60-9.00N/A <0.20.577 oK
129.00-9.40N/A <0.20.577 oK
139.60-10.2N/A <0.20.577 oK
1410.2-11.2N/A <0.20.577 oK
1511.2-12.2N/A <0.30.577 oK

ATLAS Data Collection

ATLAS data were collected over a 48 x 48 km2 area, centered on the Atlanta Central Business District (CBD) on May 11 and 12, 1997. An early May data acquisition window was selected to facilitate the collection of ATLAS data during the spring when vegetation canopy was filled out, surface temperatures were high enough to permit substantial heating of the urban landscape, and there was a high probability that cool fronts would still be moving through the Atlanta area to permit clear skies, as opposed to later in the spring or summer when increased cloud cover or convective storms become limiting factors in obtaining aircraft data. ATLAS data were collected at a 10m pixel spatial resolution during the daytime, between approximately 11:00 a.m. and 3:00 p.m. local time (Eastern Daylight Time) to capture the highest incidence of solar radiation across the city landscape around solar noon. ATLAS 10m data were also obtained the following morning (May 12) between 2:00-4:00 a.m. local time (Eastern Daylight Time) to measure the Atlanta urban surface during the coolest time of the diurnal energy cycle. Eleven flight lines were required to cover the 48 x 48 km2 area at a 10m spatial resolution. To permit the derivation of TRN values, all 11 daytime flight lines were flown and then repeated later at about a 2 hour interval. Nighttime overflights were not repeated because of the relative invariance in thermal energy fluxes at night which obviated the need to calculate TRNs.

Sky conditions at the time of the daytime overflights were mostly clear with some cirrus clouds present. The Lear jet aircraft flew at an altitude of 5,063m above mean terrain to achieve a 10m pixel resolution which was well below the cirrus clouds. Cirrus clouds covered the entire Atlanta metropolitan area during the night flights. The presence of cirrus cloud cover at night did, to some extent, dampen the cooling effect of thermal energy release to a clear sky, but air temperatures were still sufficiently cool to provide ample difference with daytime heating. Maximum air temperatures during the daytime overflights were approximately 25oC, while air temperature during the nighttime flights was around 10oC. Sample surface temperatures for tree-shaded grass, tree canopy, and asphalt in full sunlight recorded with a hand-held infrared thermometer (8-14 lm) during the afternoon were 28oC, 21oC, and 50oC, respectively. Daytime temperatures for a commercial building roof comprised of rock/membrane coating ranged from 49oC to 52oC. This illustrates that although air temperatures were cooler than optimal for development of the urban heat island effect, there was still significant heating by artificial urban surfaces to permit good contrast with nighttime cooling.

Atmospheric radiance must be accounted for in order to obtain calibrated surface temperatures. Although the ATLAS thermal channels fall within the atmospheric window for atmospheric longwave transmittance (8.0-13.0 m), the maximum transmittance is only about 80%. The amount of atmospheric radiance in the atmospheric window is mostly dependent on the atmospheric water vapor content, although there is an ozone absorption band around 9.5 m. To assist in obtaining accurate thermal surface energy response measurements from the ATLAS data, radiosonde launches were made concurrently with both the daytime and nighttime overflights. The atmospheric profiles obtained from these radiosonde data are then incorporated into the MODTRAN3 model for calculation of atmospheric radiance (Berk et al., 1989). The output from MODTRAN3 is combined with calibrated ATLAS spectral response curves and blackbody information recorded during the flight, using the Earth Resources Laboratory Applications Software (ELAS) module TRADE (Thermal Radiant Temperature) (Graham et al., 1986), to produce a look-up table for pixel temperatures as a function of ATLAS values (Anderson, 1992).

One pyranometer and one pyrgeometer were also stationed on a rooftop within one of aircraft flight lines for use in measuring incoming shortwave and longwave radiation within the study area. Additionally, two shadowband radiometers were placed in strategic locations within the flight path for use in measuring shortwave visible radiation for determining visibility parameters for input into MODTRAN3. The output from MODTRAN3 is combined with calibrated ATLAS spectral response curves and onboard calibration lamp information recorded during the flight in TRADE to produce calibrated at-sensor radiance for the visible wavelengths.

ATLAS Data: Some Examples

Approximately 5 Gb of raw (unprocessed) ATLAS data were collected during the May 11-12 aircraft overflights. In addition to the digital ATLAS data, color infrared aerial photography at 1:32,000 scale was obtained during daytime mission. Figure 1 illustrates daytime thermal (channel 13 - 9.60-10.2 m) ATLAS data collected over the Atlanta CBD area. Figure 2 provides an example of ATLAS data (channel 13) acquired during the night over the Atlanta CBD. Both images are oriented with north at the top. Excluding the effects of the highly variable emissivites of urban building materials, an empirical observation of the images presented in Figures 1 and 2 illustrates the wide range of thermal energy responses present across the Atlanta city landscape, as well as the detail that can be discerned from the 10m data. The Georgia Dome, an enclosed football stadium, appears as the large square-shaped structure due west of the Atlanta city center. Interstate highways 75/85 which traverse in a north-south direction around the city center, are seen as a dark "ribbon" on the day data (Figure 1) just to the east of downtown Atlanta. Just south of the city center, is the junction of Interstate Highways 75/85 and 20. Shadows from tall buildings located in the Atlanta city center can also be observed on the daytime data. In Figure 1, the intense thermal energy responses from buildings, pavements and other surfaces typical of the urban landscape, as well as the heterogeneous distribution of these responses, stand in significant contrast to the relative "flatness" of Atlanta thermal landscape at night (Figure 2). Also, the damping effect that the urban forest has on upwelling thermal energy responses is evident, particularly in the upper right side of the daytime image where residential tree canopy is extensive. In Figure 2, there is still evidence, even in the very early morning, of elevated thermal energy responses from buildings and other surfaces in the Atlanta CBD and from streets and highways. It appears that thermal energy responses for vegetation across the image are relatively uniform at night, regardless of vegetative type (e.g., grass, trees).

ATLAS Data Analysis: The Next Steps

From the images in Figures 1 and 2, it is apparent that high resolution ATLAS data offer a unique opportunity to measure, analyze and model the state and dynamics of thermal energy responses across the Atlanta metropolitan landscape. In addition to deriving energy balance measurements for day and night, and TRN values for specific urban surfaces to better understand the thermal characteristics that drive the development of the urban heat island phenomena and the overall Atlanta urban climate, these multispectral ATLAS data also exist as database record of current land cover/land use conditions for the Atlanta metropolitan area. Along with the extensive meteorological data available via a network of mesonet stations that are currently operating across the Atlanta area, the ATLAS data will be used to initialize and calibrate the meteorological and air quality models that will be run for the time period when the airborne data were collected. Moreover, one of the key facets from Project ATLANTA is to work with local planning agencies, such as the Atlanta Regional Commission (ARC), to model how the continued growth of Atlanta will impact the climate and air quality of the north Georgia region. The ARC is currently developing a 20-year growth plan for a 10 county area around Atlanta. Using the ATLAS data obtained in May, 1997 as a baseline for land cover/land use, our objective is to perform some "prospective" modeling on how meteorological conditions and air quality will change, predicated on the ARC's 20-year plan. By doing so, we hope to provide the ARC and other planning or decision-making bodies, with model output that can be used to modify or revise growth plans for the Atlanta metropolitan area, and to help mitigate or ameliorate the expansion of the urban heat island effect or the further deterioration in air quality.


Anderson, J. E., 1992. Determination of water surface temperature based on the use of thermal infrared multispectral scanner data. Geocarto International 3:3-8.

Berk, A., L. S. Bernstein, and D. C. Robertson., 1989: Modtran: A Moderate Resolution Model for Lowtran 7. U.S. Air Force Geophysics Laboratory, Environmental Research Papers GL-TR-89-0122, Hanscom Air Force Base, MA, 37 pp.

Graham, M.H., B.G. Junkin, M.T. Kalcic, R.W. Pearson and B.R. Seyfarth, 1986. ELAS - Earth resources laboratory applications software. Revised Jan.1986. NASA/NSTL/ERL Report No. 183.

Lo, C.P., D.A. Quattrochi, and J.C. Luvall, 1997. Application of high-resolution thermal infrared remote sensing and GIS to assess the urban heat island effect. International Journal of Remote Sensing 18:287-304.

Luvall, J.C., and H. R. Holbo, 1989: Measurements of short-term thermal responses of coniferous forest canopies using thermal scanner data. Remote Sensing of Environment, 27, 1-10.

Luvall, J.C., 1997. The use of remotely sensed surface temperatures from an aircraft-based thermal infrared multispectral scanner (TIMS) to estimate the spatial and temporal variability of latent heat fluxes and thermal response numbers from a white pine (Pinus strobus L.) plantation. In Scale in Remote Sensing and GIS, D.A. Quattrochi and M.F. Goodchild, eds. CRC/Lewis Publishers, Boca Raton, FL, pp.169-185.

Quattrochi, D.A. and M.K. Ridd, 1994. Measurement and analysis of thermal energy responses from discrete urban surfaces using remote sensing data. International Journal of Remote Sensing 15:1991-2022.

Quattrochi, D.A. and M.K. Ridd, 1997. Analysis of vegetation within a semi-arid urban environment using high spatial resolution airborne thermal infrared remote sensing data. Atmospheric Environment (In press).

Research Atlanta, Inc., 1993: The Dynamics of Change: An Analysis of Growth in Metropolitan Atlanta over the Past Two Decades. Policy Research Center, Georgia State University, Atlanta.

SOS, 1995. The State of the Southern Oxidants Study: Policy-Relevant Findings in Ozone Pollution Research 1988-1994. Southern Oxidants Study: Raleigh, NC, 94 pp.

Figure Captions

Figure 1. ATLAS daytime thermal image (channel 13 -- 9.60-10.2 m) of the Atlanta central business district area. These data have not been geometrically or atmospherically corrected.

Figure 2. ATLAS nighttime thermal image (channel 13 -- 9.60-10.2 m) of the Atlanta central business district area. These data have not been geometrically or atmospherically corrected.

Global Hydrology and Climate Center

Responsible Official: Dr. Steven J. Goodman (steven.goodman@nasa.gov)
Page Curator: Paul J. Meyer (paul.meyer@msfc.nasa.gov)

Last Updated: July 8, 1997