|
|
|
Disasters/Pollutions
|
Flood Damage Mapping in North Korea Using
Multi-Sensor Data
In addition, the 3-second DEMs overlaid on the classified map were used to differentiate the flood-suffered area depending on elevation intervals. Table 2 indicates that 97.83 percent of the flood damage area (covering 5.34 percent of the study area) remained below 100 m.
The main reason for this concentrated damage on the lower region below 100 m could be related to deforestation-induced flooding and sediment disaster.
To define discriminations among the above classes more clearly, the SAR image was used to interpret their surface roughness and soil moistures. Figure 1 shows the SAR image intensity of the respective five land-cover classes.
In order to determine significant separability of the five classes in the SAR image, the plotted mean values in Figure 1 should be compared with each other based on the t-hyphothesis test. The range of their t-test values varied from 6.25 to 365.79. On the basis, the five classes with JERS-1 SAR data could be significantly distinguished form each other
Table 2: Classified land-cover areas(%) depending on elevation intervals
|
Class
|
Elevation interval (m)
|
|
0-100
|
101-150
|
151-200
|
201-300
|
301-400
|
400<
|
|
Flood damage
|
5.340
|
0.094
|
0.020
|
0.0003
|
0.000
|
0.000
|
|
Coniferous forest
|
0.225
|
0.542
|
0.636
|
1.168
|
0.593
|
0.264
|
|
Water body
|
2.126
|
0.011
|
0.000
|
0.000
|
0.000 |
0.000
|
|
Pf (seedbed)
|
2.915
|
0.017
|
0.000
|
0.000
|
0.000
|
0.000
|
|
Pf (fallow)
|
22.783
|
0.132
|
0.021
|
0.005
|
0.000
|
0.000
|
|
Dry field
|
18.399
|
0.661
|
0.072
|
0.030
|
0.005
|
0.000
|
|
Deci. bro.forest
|
1.864
|
0.544
|
0.371
|
0.616
|
0.326
|
0.296
|
|
Forest degradation
|
28.463
|
6.377
|
2.454
|
1.897
|
0.482
|
0.248
|
|
Totals
|
82.115
|
8.378
|
3.574
|
3.719
|
1.406
|
0.808
|
Figure 1. Mean grey level with +_ 1 standard deviation of the radar illumination backscattered by the different surfaces of the five classes : food damage of 224.297 pixels, water body of 89,305 pixels, paddy field (seedbed) of 122, 433 pixels, paddy field ( fallow) of 957,005 pixels, and dry field of 772,409 pixels.
Conclusions
The classified map extracted from JERS-1 OPS data was analyzed in conjunction with the 3-second grid digital elevation data. Classification results out of the study area indicate that none of the three forest-cover classes have show mean NDVI values between -0.2330 and -0.0673. Within the 4 types, the highest mean NDVI values is recorded in the flood-suffered paddy field, which covers 5.46 percent, i.e., 7,470 ha, of the area. This is an unexpected finding, for identification of flood-suffered area without using temporal data is very difficult. In addition, it is possible to identify the flood-suffered crop field in the classified map derived from JERS - OPS data, by using JERS-1 SAR data.
Acknowledgments
The National Space Development Agency (NASDA) of Japan provided the JERS-1 OPS/SAR data, whose ownership belong to MITI/NASDA.
References
- Gamon, J.A., C.B. Field, M.L. Goulden, K.L. Griffin, A.E. Hartley, G. Joel, J. Penuela, and R. Valentini, 1995. Relationship between NDVI, canopy structure, and photosynthesis in three Californian vegetation types. Ecological Applications, 5 : 28-41.
- Huguenin, R.L., M.A. Karaska, D. Van Blaricom and J.R. Jensen, 1997. Subpixel classification of bald cypress and tupelo gum trees in Thematic Mapper imagery. Photogrammetric Engineering & Remote Sensing, 63 : 717-725.
- Jensen, J.R., K. Rutchey, M. Koch, and S. Narumalani, 1995. Inland wetland change detection in the Everglades Water Conservation Area 2A using a time series of normalized remotely sensed data. Photogrammertic Engineering & Remote Sensing, 61 : 199-209.
- Kim, C., and K.H. Joung, 1998. Application of vegetation indices for forest degradation using Landsat Tm data. Proc. of International Symposium on Remote Sensing, Kwangju, Korea, pp. 192-197.
- Lyon, J.G.D. Yuan, R.S. Lunetta, and C.D. Elvidge, 1998. A change detection experiment using vegetation indices. Photogrammetric Engineering & Remote Sensing, 64: 143-150.
- Malingreau, J.P., 1989. The vegetation index and the study of vegetation. A;;lications of Remote Sensing to Agrometeorology, Kluwer Academic Publishers, Dordrecht, Netherlands, pp. 285-303.
- Michener, W.K., and P.F. Houhoulis, 1997. Detection of vegetation changes Michener, W.K., and P.F. Houhoulis, 1997 Detection of vegetation changes associated with extensive flooding inn a forested ecosystem. Photogrammetric Engineering & Remote Sensing, 63 : 1363-1374.
- Musiake, K., T. Oki, M. Koike, T. Nakaegawa, and G.Fuchgami, 1993. Ectraction of soil moisture information using JERS-1 SAR image. JERS-1 Information Exchange Meeting Presentation Materials, Tokyo, Japan, pp. 309-326.
- Okamoto, A.,S. Yamakawa, and H.Kawashima, 1998. Estimation of flood damage to rice production in North Korea in 1995. International Journal of Remote Sensing, 19(2) : 365-371.
- Singh, A., 1989. Digital change detection techniques using remotely-sensed data. International Journal of Remote Sensing, 10: 989-1003.
- Tappen, G.D.,D. Tyler, M. Wehde, and D. Moore, 1992. Monitoring rangeland dynamics in Senegal with Advanced Very High Resolution Radiometer data. Geocarto International, 1 : 87-98.
|
|
|
|
|
|
|
