Section III.

The DMSP-OLS (http://dmsp.ngdc.noaa.gov/html/sensors/doc_ols.html) data have historically been used to monitor the global distribution of clouds and cloud top temperatures. However, the data acquired by the visible sensor (0.4 - 1.1 micrometers) at night under cloud-free conditions can provide a view of the light emitted from the Earth's surface. Gallo et al. (1995) demonstrated that this data might be useful for identification of urban and rural locations as it identifies the light associated with urban locales. Caution must be used when these data are used for identification of urban areas. In some areas of the world, on any single night, the burning of forests (e.g., burning biomass) and fossil fuels can be misinterpreted to repesent the location of an urban locale.

The DMSP-OLS data utilized in this module was mapped to a 1 km projection. Calibrated values of the OLS data are not yet available. The National Geophysical Data Center has provided a prototype dataset based on data aquired during 1994 and 1995. The values of the OLS data were computed as:



OLS = (number of scenes in which light was observed at a given pixel ) /(number of cloud free scenes for a given pixel)

(Equation 3.1)

Generally, the greater the OLS value the more likely the pixel is located in an urban area. The lights associated with urban features contribute to the greater likelihood that a pixel in an urban area would have a large OLS value. The OLS data are available at dfwols.dat (160 lines, 200 samples). Note that the number of lines and samples differ from that of the AVHRR data. Although the OLS data have been mapped to 1km resolution, the map projection utilized (Interrupted Goode Homolosine) is different from that of the AVHRR (Lambert Azimuthal Equal Area).

3.P1. Use your software display system to view the OLS image for the DFW region: dfwols.dat or examine Figure 3.1 (See implementation note 3.P.1).

3.Q1. Visually examine the OLS data. The data may need to be stretched, similar to that of Figure 3.1, to enhance the contrast of the values. Do there appear to be some areas outside of the DFW region that might be considered urban?

The station ID and locational information for the OLS data are provided in Table 3.1.

Table 3.1. Station location and OLS data extraction information.

3.P2. Use your software to extract the OLS values for the pixels associated with the stations identified in Table 3.1. The OLS values for the stations should be averaged over the pixels of each station for the OLS image. Fill out Table 3.2 (See implementation note 3.P.2) .

Table 3.2. OLS values computed for the pixels associated with the weather observation stations.

3.Q2. Based on the OLS data, which meteorological observation stations are likely surrounded by urban, compared to rural, land surface features? How do these stations compare to those selected for question 1.Q9?

Summary

This module demonstrates the value of the use of multiple sensors for the identification of urban heat islands. The repartitioning of surface energy fluxes due to urban land use change can be implicitly linked to lower values of NDVI and higher values of radiant surface temperature with the AVHRR data. Higher resolution land cover data from classified Landsat MSS data can be used to relate land cover classes to observed changes in NDVI and radiant surface temperature. The DMSP-OLS exploits anthropogenic nighttime light sources as a further source in the distinction between urban and rural locales. The information provided by a single sensor, while valuable, can clearly be enhanced through the use of multiple sensors.

Credits:

This module was written by Dr. Kevin Gallo, Office of Research and Applications, NOAA, and Tim Owen of the National Climatic Data Center, NOAA. The data were provided by NOAA's National Climatic Data Center and National Geophysical Data Center, and the USGS/EROS Data Center.