1.Introduction
Flood is one of the most severe nature disaster. It is estimated that the loss owing to flood accounts for 30% of all the loss from the nature disasters[1]. In China, flood makes a loss so high as 15 to 20 billion yuan each year.

It is impractical to acquire the flood area through field investigating. The methods to extract flood extent from remotely sensed images such as TM, FY-IB and NOAA AVHRR are explored by a few persons[2][3][4]. Microwave remotes Sensing is very useful for monitoring flood because it can obtain useful image especially in bad weather. Flood extent can be automatically extracted from JERS SAR image in plain area[5], but it is difficult to automatically extract flood extent from SAR image in mountain area because of mislabeling shade of mountain as flood. Therefore, in mountain area, flood extent is mostly extracted from SAR images by visual interpretation. It is accurate but very labor intensive. The method proposed in this paper focuses on how to solve this problem with TM and SAR images.

2. Study area and principle
The study area of our research is the Qianshan country in Anhui province of china. The data sets used in this research includes: 91) LANDSAT TM 131/39 acquired on December 7,1995 (figure 1). (2)RADARSAT SWA SAR image acquired on July 28,1998 (Figure 2).

Figure 1 TM2 image of the study area


Figure 2 RADARSAT SWA SAR image of the study area

A single radar image is usually displayed as a gray scale image. The intensity of each pixel represents the proportion of microwave backscattered from that area on the ground, which depends on a variety of factors: types, sizes, shapes and orientations of the scatter in the target area: moisture content of the target area; frequency and polarization of the radar pulses; as well as the incident angles of the radar beam. The pixel intensity values are often converted to a physical quantity called the backscattering coefficient or normalized radar cross-section measured in decibel (dB) units with values ranging from +5 dB for very dark surfaces.

The SAR backscattered intensity generally increase with the roughness. Roughness’ is a relative quantity. For SAR images, the reference length scale for surface roughness is the wavelength of the mocrowave. It the surface fluctuation is less than the microwave wavelength, then the surface is considered smooth. For example, little radiation is backscattered from a surface with a functuation of the order of 5 cm if a fluctuation of the order of 5 cm if a L-band (15 to 30 cm wavelength) SAR is used and the surface will appear dark. However, the same surface will appear bright due to increased backscattering in a X-band (2.4 to 3.8cm wavelength) SAR image. Flat surface such as paved roads, runways or calm water normally appear as dark areas in a radar image since most of the incident radar pulses are reflected away. However, rough water surfaces may appear bright especially when the incidence angle is small. The urban area appears bright in the SAR image while the vegetated area has intermediate while the vegetated area has intermediate tone.

In mountain area, when slope is perpendicular to the radar look direction, the radar echoes will be accenturated and show bright in image. Otherwise, the radar echoes will be low and show dark tone such as shade in image, which is confused with the water body. It is the crux of deciding the water body. It is the crux of deciding the flood extent from the SAR image in mountain area.

Remotely sensed image fusion has been proved to be a good method in feature extraction. There are several methods of image fusion such as HIS conversion fusion and principle conversion fusion, which are based on pixel,, with aim at improving visual interpretation results or classification results. We utilize the object fusion technologies in the paper. The shades from LANDSAT TM can be used to eliminate the shades from RADARSAT SAR, if the shades from LANDSAT TM have the same extent with shades from RADARSAT SAR. The water bodies in LANDSAT TM can be used as base water body during disaster evaluation. During imaging, wind may make water surface very rough so that water bodies appear bright in the images, which is confused with land surface especially in the center of large lakes. The water body extent on TM images can be used to compensate the flood extent on SAR images especially for large lake with rough surface. Therefore, the flood extent can be extracted through the compensation of the shades and water body from LANDSAT TM and RADARSAT SAR.