|
|
|
AirSAR/MASTER
|
Analysis of Multipolarization and Multi-Band
Radar Data Over Iloilo City and Vicinities
4.2 Image Processing Analysis on Test Areas Observations
Surface roughness is measured in centimeters and determined by textural features comparable in size to the radar wavelength, such as leaves and twigs of vegetation and sand, gravel and cobble particles (Sabins, 1987). The Dingle Limestone (blue outline) in Dumangas test area (Figure 2) shows a bright area in relatively dark areas. Its surrounding plain has agricultural crops and gravel (Lituanas, 1999) with actual size range of 2cm to 15cm of the AirSAR sensors, and show up as bright areas in C- and L-band but darker in the P-band. The limestone produces a backscatter related to its topographic relief. Thus, it shows up as a bright area in all 3 bands. Using Table 1, an object has to be >18.64cm to be rough and exhibit a bright response for the P-band. Anything that is <18.64cm will show up either as intermediate (gray color) or smooth (black). By contrast, if L- and C-bands are used, an object needs only be >1.56cm and 6.85cm to be rough and exhibit the same response. Since the vertical size range of the agricultural crops and the gravel is 2cm and 15cm, they also show up as bright areas in L-band and more so in the C-band. Thus, marking the limestone hill will be difficult in the C- and L-bands but not in the P-band. In the P-band, the hill is outlined by its bright response against the dark response of the alluvial plain with its vegetation and gravel.
In Oton test area (Figure 3), bright areas in the center image are newly developed urban areas. The rectangular-shaped, light toned objects on the right are agricultural areas. The dark areas in the upper part are land submerged in water due to the sea's transgression inland forming tidal flats. The beach area of sand and gravel deposit shows uniform intermediate response to the L-HH band, bright in C-HH band and dark in the P-HH band. Using Table 1, we can deduce the beach sand deposits in the range of 1.21cm and 6.8cm (intermediate). Pebble- sized sands and cobbles have composite size range of 0.4 cm to 25.6cm. These form a series of micro-ridges that lie sub-parallel to the curve of the coastline as outlined. It can be deduced that these are series of deposition of sediments parallel to the coastline caused by the interaction of the land and the sea. The relative uniformity of the size range is due to the winnowing effect of the waves. This explains the relatively uniform intermediate response in the L-HH image. Multi-band radar is useful in estimating surface relief of materials. We can state some of the process that took place in the past by observing closely the patterns of lineation on the coast, and correlating it with field observation.
The Guimaras test area (Figure 4) provides the radar data for lithologic terrain mapping. Geologic structure and erosional characteristics of underlying rocks control the topography. The Buenavista Limestone (50m to 80m el.) in white outline, exhibits a pitted texture of karst topography (Sabins, 1987) and has some bedding that forms asymmetric ridge. The beds dip gently to the northwest. On the lower right of the image is the Guimaras Diorite identified in the field as quartz diorite porphyry. It has a higher relief than the limestone, producing a bright backscatter and shows no definitive pattern. On the upper part is the Quaternary alluvium that exhibits a fine texture, dark gray color and forms as fringe on the northwest side of the limestone.
5.0 Summary and conclusions
In the test areas, we demonstrated that multi-band radar data could help approximate the surface relief of materials in the field by describing the relief in terms of surface roughness derived from the Modified Rayleigh Criterion. Such relief correlates with signatures in the radar image. By observing closely the patterns of lineation in the Oton coast image and by correlating it with field observation, we could deduce some of the process that took place in the past.
We demonstrated that L-band radar is useful for recognizing major lithologic terrain in a tropical environment such as in Guimaras. These are (1) the Buenavista Limestone - karstic, gently dipping and fossiliferrous limestone in north and northwest Guimaras (2) the Guimaras Diorite - as low-lying hills and mounds in northern and central parts; (3) and the Quaternary Alluvium - the sediments along the coast, along river deltas and active tidal flats.
We also showed that AirSAR image is useful for geomorphologic terrain mapping. We identified three landforms namely, the fluvio-denudational, the fluvial and the fluvio-marine landforms. Lastly, by comparing the present image with an existing reference map, we can detect changes in the land cover and in the geomorphologic setting of the area. In terms of impact assessment for flooding hazard of the three terrain, the fluvio-marine environment is prone to frequent flooding mainly due to its low relief and geomorphologic setting. The next area prone to flooding is the fluvial environment.
6.0 Acknowledgment
We thank the DOST-PCASTRD for the generous financial support to this project. We also thank our colleagues at the MGB Central Office and Regional Office No.6, personnel from NAMRIA and UP-TCAGP for joining us in the field. We also appreciate Mr. Bobby Crisostomo of NAMRIA for his leadership in this project.
7.0 References
- Abundo, R.V., Parringit, E., Domingo III, D.S. and Lituanas, M.B., 1998. AirSAR Capability to Geomorphology and Geologic Mapping Application. Proceedings of the 19th Asian Conference on Remote Sensing, pp. J-2-1 - J-2-5, Manila, Philippines.
- Calomarde, R.I., 1987a. Geology and Geomorphological Terrain Classification of the Guimaras Island, Iloilo City. Mines and Geosciences Development Services Regional Office No. 06, Iloilo City, Philippines.
- Evans, D.L., Farr, T.G., Van Zyl, J.J., Zebker, H.A., 1988. Radar Polarimetry: Analysis and Applications. EEE Transactions on Geoscience and Remote Sensing, 26, pp.774-789.
- Geology of the Philippines,1996. Unedited version. Mines and Geosciences Bureau.
- Hewson, R.D and Taylor, G.R., 1995. AirSAR Imagery as an Ancillary Data Set for Geological Mapping and Fowlers Gap, Western New South Wales. Proceedings of the International Workshop on Radar Image Processing and Applications held in the UNSW, Sydney in November 6-8, 1995.
- Javelosa, R.,1994. Report on the Quaternary Geological and Geomorphological Survey of Iloilo City. Mines and Geosciences Bureau, Quezon City, Philippines.
- Sabins, Jr., F.F.,1987. Remote Sensing: Principles and Interpretation, 2nd ed. (New York: W.H. Freeman and Company).
8.0 Table and Figure
Table 1. Summary of the calculation results for the AirSAR bands using the Modified Rayleigh Criterion (Peake and Oliver,1971) at a look angle of 560. Values are in cm. H refers to the vertical height of the object
| Category |
C-band |
L-band |
P-band |
Signature as observed In the image |
| Smooth |
H< 0.275 |
H<1.21 |
H<3.28 |
Dark |
| Intermediate |
1.56<H<.275 |
6.85<H<1.21 |
18.64<H3.28 |
Shades of gray |
| Rough |
H>1.56 |
H>6.85 |
H>18.64 |
Bright white |
Figure 1:
Figure 2:Dingle Limestone
Figure 3:Oton Coastline
Figure 4: Guimaras Diorite
|
|
|
|
|
|
|
