MSS mitomi@edm.bosai.go.jp matsuoka@edm.bosai.go.jp yamazaki@edm.bosai.go.jp taniguchi@manage.nitech.ac.jp 1 MSS MSS 2 2 1 m MSS CCT CCT Fig.1 CCT b02-b0 b0-b0b-b b-b1 CCT Landsat/TM MSS S/N 21x21 21x21 %30% Fig.2 MSS ( CCT) - ( CCT for "burned") 0 40 20 0-20 -40-0 Burned Severely damaged Slightly damaged Liquefaction Bare ground Railway Vegetation b02 b04 b0 b0 b0 b0 b0 b b b b1 b1 Fig. 1. Relative spectral characteristics for training data Fig. 2. Estimated result of the areas with heavy building damage. Gray and black represent the area with severely damaged and burned buildings, respectively. Dotted polygon indicates the area of seismic intensity VII in the JMA scale.
MSS DETECTION OF THE AREAS WITH BUILDING DAMAGE DUE TO THE KOBE EARTHQUAKE USING AIRBORNE MSS IMAGES 1 2 3 4 1 Hajime Mitomi, Earthquake Disaster Mitigation Research Center, NIED, mitomi@edm.bosai.go.jp 2 Matsuoka Masashi, Earthquake Disaster Mitigation Research Center, NIED, Dr. Eng., matsuoka@edm.bosai.go.jp 3 Fumio Yamazaki, Earthquake Disaster Mitigation Research Center, NIED, Dr. Eng., yamazaki@edm.bosai.go.jp 4 Hitoshi Taniguchi, Professor, Nagoya Institute of Technology, Dr. Eng., taniguchi@manage.nitech.ac.jp Yujiro Ogawa, College of Environment and Disaster Research, Fuji Tokoha University, Dr. Eng. SUMMARY In multi-stage remote sensing performed by NASDA just after the Kobe earthquake, the stricken areas were observed by airborne MSS, which has twelve bands between visible and thermal infrared. In this study, after some training areas were selected in Nada Ward using GIS data based on field damage survey, spectral characteristics of damaged and non-damaged buildings were investigated. Then, the distribution corresponding to the damage level of buildings in the hard-hit area of Kobe City was estimated by the maximum likelihood classifier. The estimated result of areas with burned and severely damaged buildings was relatively in good agreement with the field survey data. An application of this method, based solely on the post-event image, to early damage assessment systems can be expected. : MSS Key words: Airborne MSS image, Spectral characteristics, Building damage areas, The Kobe earthquake 1 4)-) ), ) 1) 2) GIS 1 3) ), ) GIS
Table 1: Specification of the airborne MSS image Band Visible (blue) Visible (green) Visible (red) Wavelength (μm) b02 0.41-0.4 b04 0.4-0.3 b0 0.3-0. b0 0. - 0.0 b0 0.0-0. b0 0. - 0.0 b0 0. - 0.2 Infrared (near) Infrared (middle) infrared (thermal) Acquisition date IFOV Altitude Observation width b 0. - 0.0 b 0.2-0.0 b 1.2-1.2 b1 2.0-2.4 b1.00 -.00 Jan. 24, 2. mrad 2,00 m 4,000 m R,G,B NASDA ), ) MSS ) ), ) Landsat/TM ) ) 2 MSS 1) MSS Fig. 1. Observed (light) and studied (dark) areas in this study. Hatched area and red polygon indicate Nada Ward and seismic intensity VII in the JMA scale, respectively. Fig.1 MSS Landsat 2 2 1 2.mrad 1,000m IFOV 2.m 1 24 2,00m IFOV m m/pixel CCT MSS 24 x 24 3 b1 3 2 MSS NASDA MSS Table 1 1 1 24 2 2 1 24 MSS 400 GIS 1)
Table 2: Digital number of MSS bands for training data Category CCT b02 b04 b0 b0 b0 b0 b0 b b b b1 b1 Burned Severely damaged Slightly damaged Liquefaction Bare ground Railway Vegetation 4 4 0 3 3 4 0 0 1 0 3 3 4 2 4 1 1 4 2 3 4 3 3 0 1 0 1 1 4 3 0 1 1 2 4 1 1 1 4 1 0 4 1 2 0 0 1 4 1 1 21 1 1 2 1 22 22 32 2 1 20 m 1 1 30% Hamada et al.() 1) 1 400 200 4 Table 2 Fig.2 b02-b0 b02 ) ( CCT) - ( CCT for "burned") 0 40 20 0-20 -40-0 Burned Severely damaged Slightly damaged Liquefaction Bare ground Railway Vegetation b02 b04 b0 b0 b0 b0 b0 b b b b1 b1 Fig. 2. Relative spectral characteristics for training data b0-b1 b0-b1 20 b1 20 40 b02-b0b-b ) Landsat/TM 30m
Fig. 3. Estimated distribution of the areas with burned (orange), severely damaged buildings (yellow), and slightly damaged buildings (blue) using maximum likelihood classification. Red polygon indicates the area of seismic intensity VII in the JMA scale. Fig. 4. Result of the field survey on building damage 1). Red and yellow represent the areas with burned and severely damaged buildings, respectively. Red polygon indicates the area of seismic intensity VII in the JMA scale. CCT 1 )-) MSS MSS S/N m 1) c1c2 c3c4 cc Fig.2 b1 ) 0 CCT 40 b0 CCT 0 c1-c3 Fig.3 1 30% Fig.4 c2table 3 c4-c 0%c1-c3 3.% 0.0%.3% c1-c3
Table 3: Classification accuracy (%) Class c1 c2 c3 c4 c c c1: Burned 3. 1.0 2.0 2..2 0.2 c2: Severely damaged.3 0.0 1.0 1. 1. 0. c3: Slightly damaged 2. 1.3.2 3.3 1. 1. c4: Liquefaction / Ground 0.2 0.2 0.0 1.0 0.0 0.0 c: Railway. 2.3 0.3 0.0 1.0 0.0 c: Vegetation 1.0 0.2 0. 1.3 0..3 Cloud - - 0. 0.2 - - Total 0.0 0.0 0.0 0.0 0.0 0.0 30% Fig.2 c1 10m x 10m 21x21 c1 Rpx RpxFig.c1 Rpx30%c2 c Rpx 20% 30% 0% Rpx% 0% c2 % Rpx% c1 % Rpx30% c1-c3 Fig. Fig.4 Cumulative relative frequency 0% 0% 0% 40% 20% MSS c1: Burned c2: Severely damaged c3: Slightly damaged c4: Liq. / Ground c: Railway c: Vegetation 0% 0% 20% 40% 0% 0% Rpx Fig.. Cumulative relative frequency of Rpx for each set of class for maximum likelihood classification 1 MSS MSS m CCT 20 Fig.. Estimated result of the areas with heavy building damage. Orange and yellow represent the area with Rpx30% and Rpx %, respectively in texture analysis. Red polygon indicates the area of seismic intensity VII in the JMA scale.
MSS 000 CD-ROM 2 1) No.2/I-pp.2-22001. 2) Vol.2pp.-2000. 3) Tchimev, P. pp.1-232000. 4) Vol.40No.4pp.2-32001. ) 31 pp.-2001. ) pp.-02001. ) No.2/I-pp.-22001. ) pp.22001. ) Vol.44pp.2-2000. ) 2001. ). ) 1 pp.-. ) MSS 1 pp.-. ) 1. ) No./I-4pp.1-2001. 1) 1. 1) Hamada, M., Isoyama, R. and Wakamatsu, K.: The Hyogoken-nanbu (Kobe) Earthquake, Liquefaction, Ground Displacement, and Soil Condition in Hanshin Area,,. 1) :.