Home Articles Multi – sensor radiometric correction: a case study from Malaysia

Multi – sensor radiometric correction: a case study from Malaysia

ACRS 1997

Poster Session 2

Multi-Sensor Radiometric Correction: A Case Study from Malaysia

Muhamad Radzali Mispan* and Paul M. Mather
Department of Geography, University of Nottingham
University Park NG7 2RD Nottingham
United Kingdom
*Present address: Senr, Mardi,
P.O. Box 12301, 50774 Kuala Lumpur,

E-mail: [email protected]


Abstract

Remote Sensing offers practical benefits in our understanding about the nature and characteristics of land surface features. However, cloud cover normally impedes the use of optical remotely sensed data in studies of low-latitude regions (tropical areas). Therefore, data from different sensor such as Landsat TM and SPOT HRV are required to ensure as complete a temporal coverage as possible. These however, increase the complexity of the analysis as the SPOT-HRV and Landsat TM sensors have different characteristics and acquire data under different atmospheric, viewing and illumination conditions. In this study, the radiometric correction required to ensure compatibility of data over time and between sensors are described. Calibration to adjust sensor drift was performed followed by atmospheric correction based on combination 6S model and Dark Dense Vegetation (DDV) method. Next, data from different sensors (TM and SPOT HRV) were adjusted for sensor characteristics (band pass) and viewing conditions. The radiometric correction process will not only, correct the image data but also convert them to radiance or reflectance factor for quantitative analysis. The procedure is illustrated using a case study from Malaysia.


Introduction

Ideally, a comparison of several satellite scenes over a period of time, covering the same area, would show changes in the intrinsic properties of land surface features. Such changes in spectral response, however, may also be due to effects related to sensor performance, atmospheric condition at the time of over pass, and viewing and illumination geometry of the sensor. Hence, the use of data from different sensors to provide better temporal coverage results in an increase in the complexity of data analysis as each sensor has different characteristics. Therefore, there is a need to calibrate and correct these data spectrally and spatially and convert to a common scale or datum, so that they are internally (within scene) and externally (between scene) consistent. This paper demonstrates a radiometric correction technique applied to data acquired by two different sensor systems and subsequently makes sensible comparison between images. It places emphasis on establishing a spectral datum for operational, use of this technique.


Material and Methods

Radiometric correction of multi-sensor data involves three steps: sensor calibration, atmospheric correction, and sensor inter-calibration. This study used a combination of 6S radiative transfer code and pseudo-invariant objects. For sensor calibration and atmospheric correction, dark dense vegetation (DDV) was used as a spectral datum. For sensor inter-calibration, DDV and
other pseudo-invariant objects available in the image data were used to compensate for the effect of different of different sensor characteristics and view geometry between Landsat-TM and SPOT-HRV sensors.


Resources

The study area is located at the central western coast of Peninsular Malaysia (Latitude 30 25′ N and longitude 1010 45′ E) about 20km south of Kuala Lumpur. This study used Landsat -5 Thematic Mapper (TM) and SPOT HRV-2 data acquired on the 6 March 1990 and 26 December 1990 respectively (Table 1). The image processing and analysis were carried out using ERDAS-Imagine software at the Department of Geography, University of Nottingham.


Sensor calibration

A linear model relates the digital value (DN) of an image pixel to the intensity of reflected radiant energy (L)
(Wm-2 sr-1mm-1).
In order to calculate the radiance for a given pixel (i,j) in a spectral band
(l), the calibration gain (A) and offset (B) must be known. However, the accuracy of the calibration coefficients of Landsat TM data stored in the tape header are limited by the undocumented performance of the sensor and the ground receiving station (Price, 1087). Thus the calibration gain and offset values must be updated if accurate results are to be achieved. This study uses a method proposed by Olsson (1995) to derive calibration coefficients for Landsat TM data. This method gives a better estimate of the coefficients (Mispan, 1997). However for SPOT-2 HRV data, the study uses the absolute calibration gain provided in the tape header. Table 1 shows the calibration gain and offset of the Landsat TM images. The relationship can be expressed using equation (1) for Landsat-5 TM and equation (2) for SPOT HRV-2 data:

Lli = Ai (DNli)+Bi (1)

Lli = DNliAi (2)


Table 1: Calibration coefficients for Landsat-5 TM and SPOT HRV1 data

Band
Green
Red
NIR

TM-90
Ai
1.174
0.806
0.816

Bi
6.05
3.36
3.25

SPOT2-HRV1
Ai
1.067
1.177
1.289

Bi
0
0
0

ACRS 1997

Poster Session 2

Multi-Sensor Radiometric Correction: A Case Study from Malaysia


Atmospheric correction

The basic philosophy of atmospheric correction is to obtain information about the atmospheric optical characteristics and to apply this information in a correction scheme (Kaufman, 1989). It is a process of surface reflectance retrieval,
rs,
from the corresponding reflectance at the top of atmosphere, or simply apparent reflectance,
r*. The relationship between the radiance, L, of spectral band
li , and the apparent reflectance can be expressed using equation (3):

r* (li) =
L(li)
————
Ex.d.Cos(q)
(3)

Where Es is the exo-atmospheric solar irradiance at the top of the atmosphere
(Wm-2 sr-1mm-1), d is the distance multiplicative factor (unit less) and q is the solar zenith angle (degree). Subsequently, the relationship between apparent reflectance and surface in the present of atmospheric constituents can be expressed as follows:

Where ra is atmospheric reflectance, D