Logo GISdevelopment.net

GISdevelopment > Proceedings > GITA > 1997


GITA 2002 | GITA 2001 | GITA 2000 | GITA 1999 | GITA 1998 | GITA 1997
Sessions

Advanced Technical Topics

Building & Supporting Applications

Business Evolution & Platform Migration

Expanding the User Base -- Non-Traditional Applications

From the office to the Field

Fundamental & Economic Issues of AM/FM/GIS

Lessons Learned

Major Technology Trends and their Impacts

Project Planning, Implementation and Management

Re-Engineering and Integration Issues

Scada and Real-Time Systems

User Project Presentations

Best of the Rest

Invited Presentation


GITA 1997


Major Technology Trendus and Their Impacts


A Study of Areospace Reconnaissance of Earthquake Sites


Information Requirements

(1) The stage of extinguishing fire/emergency response (after 10 minutes, up to 3 hours) The disaster declaration was announced by Osaka Meteorological Observatory 10 minutes after the quake broke out. Prevention from fire and gas explosion is pressingly needed. Recent research shows that damage assessment information system is indispensable to determine appropriate mobilization activities in this stage.

(2) The stage of relief (after 3 hours, up to 3 days) Relief requests were issued by the municipalities in the disaster area within twelve hours of the quake. Pressing needs in this period were to locate the damaged spot and to recognize the damage scale. TV news broadcast images of the disaster area which were obtained by field survey and reconnaissance by helicopter. Without geographical information, it was too difficult to identify the disaster area and damage scale for people who do not have the feel of the place. Aerial reconnaissance began in this stage.

The government decided to enact the law of disaster relief to the damaged area and people started living in shelters. Information requirements about safety and relief became stronger. The aerial photographs of the whole Hanshin area were helpful for integration of fragmentary reports about the disaster and damage scale recognition.

(3) The stage of restoration (after 3 days, up to 1 month) Request of maps had increased for restoration use, and, as a result, bookstores ran out of their stock of atlases and maps. The Geographical Survey Institute and various organizations had made maps of damage distribution such as liquefaction, geographical displacement and collapsed houses by interpreting aerial photographs. The requirement for information accuracy had increased in the stage of restoration, and in some cases it exceeded the accuracy of aerial photography. For example, aerial photograph interpretations do not give enough information for classification of damage grade.

(4) The stage of redevelopment (after 1 month, up to 3 months and later) Notices for the town restoration plan began to be posted. The information of land height and control points ware required for the redevelopment of the darnaged area because the control points of the whole Kobe city had displaced in terms of deformation and destruction caused by liquefaction and/or sliding. Geographical Survey Institute had established twenty precise emergency primary control points surveyed with GPS and sixty-four emergency leveling bench marks (approximately 84km) in Kobe city, especially in the southern area. After the new control points took effect, the new survey results were compared with those from before the quake, and aerial photograph was required for measurement of displacement. Through these stages, fragmentary information had been integrated into ‘area information’ by using aerial photographs and had been refined up to ‘geocoded information’. Requirement for instant availability had been changed to that for accuracy with the passing of time.

Reconnaissance with Aerial Photography
Aerial photographs had also changed their roles to respond to the information requirements referenced above. However, the use of aerial photography resulted in a delayed response to the needs as shown in Table 1. For example, one day was spent for taking photos and developing them for interpretation, although much of the practical information was required within three hours of the quake; Maps of damage situations were released one week after the quake, although some recognition of the actual damage scale was required six hours after the quake. The delay was primarily caused by the required processing time for the analogue media. The process orthorectification (Nishimura, 1995).

It is concluded that aerial photography has physical limitation in instant availability and cannot be used for mobilization planning when a disaster occurs, although it may satisfy the functional requirement.

Applicability of high resolution satellite imagery to earthquake sites
High resolution satellite imagery, which will be one of the image resources from the air, is expected to complement aerial photography. We evaluated the applicability of this new information resource to earthquake sites with simulated images made from aerial photographs.

Outlines of Commercial Earth Observation Satellite
High resolution satellite imagery is one of the military technologies which have been commercialized since the end of the Cold War. EarthWatch communications, Inc. will provide 3m ground resolution satellite imagery beginning in March ’97, and 82cm ground resolution images in early 1998. Space Imaging, Inc. will also start to deliver data which have 82cm ground resolution in early 1998. Some other companies are also planning to launch commercial earth observation satellite after 1998 and large amounts of data will become available for practical use. Table 2 shows the specifications of representative high resolution satellite imageries.

Potential of Satellite Imagerv for Intermetation and Measurement
We evaluated the potential of high resolution satellite imagery in the analysis of disaster areas with simulated images created from aerial photographs. Fig.3 shows the sample of simulated images. Four people tried to interpret thirty-four simulated images with 3m and lm ground resolution. The following objects and conditions were recognizable with each referenced resolution:

Images with 3m ground resolution
  • Roads with over 20m width and more
  • Roads with 5m width (ambiguous)
  • Large constructions such as factories
  • Houses (small houses were ambiguous)
  • Damage situations;Landslides, Burned down area, Fall of elevated highway and railroads, Liquefaction of over 1,000m2, Collapse of shore protecting structures, Lineament, Coastal Pollution
  • Raih-oads (difference from reads was often ambiguous)
Images with lm ground resolution
Objects and disaster conditions which can be recognizable with 1/4,000 contact printed aerial photographs except collapse of small houses







Figure 3. Simulated images of damage situations created form aerial photographs




We have also evaluterd the accuracyof mapping with high resolution satellite imagery. A pair of 3m simulated images are created form 1/20,000 aerial photographs and processed by an automatic stereo matching program. Fig. 4. shows comparison between contours of criterion data and calculatd data. Residuals of reference points in orientation correspond vertical accuracy of 1/25,000 mapping using 1/40,000 aerial phot~graphs, and horizontal accuracy of 1/10,000 to 1/15,000 mapping using 1/30,000 aerial photographs. Ovals in Table 3 indicates the accuracy corresponding to 3-D measurement with 3m resolution satellite imageries. This result indicates that 3-D measurement accuracy with 1m resolution satellite imagery has an accuracy of 1/5,000 to 1/10,000 mapping.

Advantages and Disadvanta~es There are some advantages and disadvantages in application of high resolution satellite imagery to earthquake sites:




Page 2 of 3
| Previous | Next |

Applications | Technology | Policy | History | News | Tenders | Events | Interviews | Career | Companies | Country Pages | Books | Publications | Education | Glossary | Tutorials | Downloads | Site Map | Subscribe | GIS@development Magazine | Updates | Guest Book