|
|
|
Hazard Mitigation
|
Validity Study of EDES Application to Taiwan Chi-Chi Earthquake Disaster
Figure 9 Electric load data from September 21 to 23, 1999
Figure 10 Aggregation areas on July 23, 1999
Figure 11 Aggregation areas on September 21 to 23, 1999
5. Gain and Radiance Estimation Method
If the observation year of VNIR image is near RCI creation years, that is, there are few changes of light emission source between observed image and RCI, the gain G can be estimated based on DR in edges of Dv-saturated areas (that means Dv equals to 63) using the following equation:
Equation (4)
Generally, Dv saturates in highly urbanized cities and Dv starts to saturate in peripherals of cities. The DR contour lines become longer along with the distance from centers of cities, which should have larger DR. Thus, DR in edges of Dv-aturated areas is the largest in pixel number. This DR can be found as the mode of histogram of DR in areas where Dv equals to 63. When G is given, R can be expressed by Dv from (1) as follows:
Equation (5)
6. Gain Estimation before and after the Earthquake
Figures 12 and 13 are histograms of DR in Dv-saturated areas of the PreI and Post I shown in figures 1
and 2, respectively. The area ranging from 30° N to 40° N latitude and from 129° E to 143° E longitude,
which covers Taiwan, is excluded from sample datasets to eliminate the earthquake influence. Conditions:
DV, DS and DR > 0 are used to select urban areas where DS is DN of SLI. To eliminate cloud influence, a
condition: DT > 219 (corresponding to 20° C) where DT is DN of TIR band, is also used. Estimated gains
are shown in Table 1.
7. Verification of Estimated Damaged Areas
Figure 14 shows the relation between cumulative radiances (CRs) and ELs in Taiwan. CRs directly
calculated from RCI are also plotted with pre-event ELs. Evaluated CR of an observed PreI can be smaller
than that of RCI because large radiances cannot be expressed due to DV saturation. Though the north
aggregation area become larger in area after the event, the EL become smaller. There was power cut and
restriction of power supply due to the damage of the extra-high voltage transmission facilities in
Chungliao and Tienlun. The fact that CR becomes smaller proves the lights decrease in this area.
Similarly, the south area become larger in area and the EL is smaller after the event. This indicates
there was the reduction or loss of lights in Nantou or Hualien County, which are the increment parts of
south aggregation areas. These results support the validity of the estimation results of EDES.
Figure 14 Cumulative radiances and electric loads in Taiwan
8. Conclusions
Radiance calibrated nighttime image can be used to evaluate an artificial pre-event image better than
stable lights image. Estimation method of gain and radiance of VNIR image using radiance calibrated image
is proposed. The relation between the cumulative radiance and electric load in Taiwan before and after the
Chi-Chi Earthquake reflects the power cut due to the damage of transmission facilities, and this
supports the validity of the estimation results of EDES.
Acknowledgements
- We wish to thank National Oceanic and Atmospheric Administration's National Geophysical Data Center
for DMSP/OLS images, SLI and RCI, and also Taiwan Power Company for electric load data before and after
the Chi-Chi Earthquake.
- References
- Elvidge, C. D. et al. 1997, Mapping City Lights With Nighttime Data from the DMSP Operational
Linescan System, Photogrammetric Engineering & Remote Sensing, Vol. 63, No. 6, pp. 727-734.
- Elvidge, C. D., et al., 1999, Radiance Calibration of DMSP-OLS Low-Light Imaging Data of Human Settlements, Remote
Sensing of Environment, Vol. 68, pp. 77-88.
- Hayashi, H. et al., 2000, International Collaboration for the Early Damaged
Area Estimation System Using DMSP/OLS Nighttime Images, Proceedings of IEEE 2000 International
Geoscience and Remote Sensing Symposium.
- Kohiyama, M. et al., 2000, Development of Early Damaged Area Estimation System (EDES)
Using DMSP/OLS Nighttime Imagery, Journal of the Institute of Social Safety Science, No. 2
(in press, in Japanese).
|
|
|
|
|
|
|
