The study of relationship between the ground targets spectral data and ATH data
Zhang Z. G.
Centre Fore Remote Sensing in Geology, 29 College Road, Beijing 100083, P.R.C.
Abstract :
In this paper, based on analyzing action between magneto electric wave through the atmosphere and the ground targets each other, mathematical model of the relationship between the ground targets spectral data and ATM data is derived from theory, and the physical significance or correlation parameters are given. According to ATM data (DS-1208 Scanner) and the spectral data of the field synchronization measurements (IRIS Spectroradiometer) at three test fields (Goldsea lake, Gaoyang and Anxin county), by using the analysis method of the statistical regression, the linear regression equations between the DN of each band of ATM (total twelve bands) and the reflectance brightness of the ground correspondence wavelength the are obtained under all kinds of conditions, (such as different test fields, periods, flight heights and detection angles etc.). Taking advantage of this quantitative relationship equation, the original images of ATM are directly transformed into the ground efficient reflectance brightness images. In the transformation procedure, the atmospheric transmittance and path’s radiance don’t need to know. The efficient reflectance brightness images is able to reflect the ground targets spectral properties. Therefore, the effects of the atmosphere between the ground and the DS-1268 sensor are eliminated, and the image order is apparently improved. Moreover, the relationships between the atmospheric transmissivity and flight heights detection angles and the atmospheric transmissivity are discussed.
Introduction
We know, the multispectral data of the space and aeronautics Remote Sensing are the reflective and radiant results of the earth objects. Aeronautic Remote Sensing data has not only to do with reflective properties of the earth objects, but also with the sun luminance, atmospheric transmittance properties and sensitive properties of sensor. This paper tries to quantitatively study the relationship between the ATM data and spectral data of the earth objects. Although this is a very important work, it is doubtless on the deeping Remote Sensing technique and data quantitative application of Remote Sensing, the research is vary difficult and complex.
Theoretical Model
The electromagnetic wave reflected and radiated by earth objects is measured, it is foundation of Remote Sensing technique. So as to quantitative approach relationship between airborn multispectral data and reflective properties of earth objects, it is very necessary that relation model of sensor response and reflective property of earth objects is created, this model is named for Remote Sensing equation in the paper[4]. Regardless space or aeronautic Remote Sensing technique, when their sensors measure earth objects, the electromagnetic waves all go through the atmosphere. Generally, the atmospheric effects are considered in the Remote Sensing equation, it is expressed as:
- Because of the atmospheric absorption and diffusion, the sun radiation from atmospheric outer does not all reach the ground; there is a part of radiation from earth objects reflecting can enter the sensor.
- Because of the atmospheric diffusion, radiation without reaching the earth has small part radiation entering the sensor.
- The reflective radiance of around the object because of the atmospheric diffusion, a part of that enters the sensor.
If one assume: B is total radiant brightness. Ba is atmospheric radiance brightness of direct entering sensor. Bb is the radiance brightness of background. Bt is the targets reflective radiance brightness. Then, one obtains equation:
B = Ba + Bb + Bt (1)
Now, the other conditions are considered:
- Wavelength ranges is from l1 to l2 .
- The solar zenith is q .
- The surface of earth objects is approximation to Lambertian surface.
- The sun’s irradiance distribution in the outer atmosphere is Eo(l).
Then, I. The atmospheric radiant brightness (shown in Fig. 1) is expressed as:
Where
b (
l) is the atmospheric efficient brightness. It depends/ on the atmospheric properties.
Fig. 1 The relation of incident radiance atmosphere, background, target, reflective radiance and sensor.
The radiant brightness of background reflection (Bb) is expressed as:
Where
r` (
l) is the average reflectance of background.
tq (
l) is atmospheric transmittance (direction of incident). It depends on the atmospheric diffusion properties.
ts (
l) is the atmospheric transmittance of reflective direction.
The radiant brightness (Bt) is from the reflection of the target cross atmosphere it has not only t do with the target reflectance, but also with atmospheric absorption, diffusion properties and transmission path. It is expressed as:
Where
r (
l) is the reflectance of target,
tz (
l) is direct transmittance. It’s only depended on atmospheric absorption and diffusion in the reflective direction.
If the S (
l) is the spectral response function of the sensor, then the total response (D) of the sensor is expressed as:
When earth objects are very even within large area,
t` (
l) of the background is about
r(
l) of targets. One let:
When earth objects are very even within large area,
t (
l) is the total transmittance of the atmosphere in direction of incident and reflective radiance. Thus the equation (5) is changed into the equation:
