GEOMETRICCONSIDERATIONOF IMAGE COMPOSITE FOR VEGETAION MONITORING 
—— | NOAA/AVHRR, ADEOS/OCTS, ADEOS-II//GLI  — — 
Toshiaki Hashimoto 
Earth Observation Research Center(EORC), 
National Space Development Agency(NASDA) 
Japan 
Commission Vll, Working Group 5 
KEY WORDS : vegetation, image composite, sensor 
ABSTRACT: 
NOAA/AVHRR data are widely used for vegetation monitoring, especially in a global scale. National Space 
Development Agency of Japan (NASDA) is planning to develop newearth observation satellites like ADEOS (Advanced 
Earth Observing Satellite), ADEOS-II, etc. Some sensors on these satellites willbe expected to be applied to vegetation 
monitoring. Ocean Color and Temperature Scanner (OCTS) on ADEOS and Global Imager (GLI) on ADEOS-II will be 
useful for this purpose. The former has the spatial resolution of 700m at nadir and the latter has of 250m in visible and 
near infrared regions. These sensors willbe able to play the same as or better role than AVHRR because of their higher 
resolutions. 
For vegetation monitoring, some images observed sequentially are composed to one image for eliminating clouds. The 
author already proposed the method of image composite for NOAA/AVHRR with the considering on the solar zenith 
angle effect on NDVI, the distortions in both geometry and radiometry, the revolution of satellite. 
If OCTS and GLI images are of integrated use with AVHRR image, the cross calibration will have to be examined, 
because they have differences in geometry like spatial resolution, revolution, orbit, solar zenith angle at observation, etc. 
That results in the radiometric characteristics of composite image of AVHRR, OCTS and GLI are different. 
In this paper, the attentions to pay for image composite are pointed out with respect to geometric properties, andthe 
method of image composite is proposed. 
1. INTRODUCTION 
2. SATELLITE/SENSORCHARACTERISTICS 
Since the discussion on the global environment was 
begun in 1980's, NOAA/AVHRR data have been widely 2.1 NOAA/AVHRR 
utilized in the field of vegetation monitoring. For example, 
GVI data have been produced from AVHRR data by Table 1 shows the satellite/sensor characteristics of 
NOAA/NESDIS and used for global vegetation monitoring.  NOAA/AVHRR, ADEOS/OCTS and  JADEOSI/GLI. 
Vegetation map of 1km resolution is also prepared. NOAA-9 and 11 are indicated as the examples. 
ADEOS/OCTS and ADEOS-II/GLI developed by NASDA 
will be expected to be applied to this field. The spatial NOAA satellite series has been operational since 1978 for 
resolution and spectral property (number of bands) of weather monitoring after the missions of TIROS and 
these sensors are better than those of NOAA/AVHRR, ITOS series. It has been utilized for vegetation monitoring 
while the observation of the same area is less frequent. in large scale because of the worldwde coverage, the 
It seems that the best way to utilize the OCTS and GLI frequent observation and the capability of extracting 
data to vegetation monitoring is the integrated use of NDVI (Normalized Difference Vegetation Index) since the 
these data and AVHRR data so as to take advantages of global environmental problems arose. 
each sensor data. It means to make use of the high 
resolutions and high spectral properties of new sensor In general, two NOAA satellites numbered with both an 
data and the frequent observation of AVHRR data. odd and an even number are operational at a time. The 
satellite withan odd, an even number crosses the equator 
It is supposed in this paper that image composite willbe in ascending mode in the afternoon, in the morning, 
generated from sequential images of each sensor for a respectively. The same area on the ground can be 
certain period. These composite images from three kinds observed from 4 to 6 times in a day using the tw 
of sensors are integratedly used for vegetation satellites system. It was reported that the local time à 
monitoring. In order to conduct such an application, the descending node was delayed gradually (Kajiwara, et al. 
cross calibration of each sensor data has to be examined. 1992). Furthermore, the rate of delay of local time 
The radiometric properties of 3 kinds of composites are became bigger year by year. The daily change rate of 
different from each other because of the differences of mean anomaly of NOAA satellite became faster. 
the scan geometry of the sun-ground-sensor. Consequently, the number of passes in one day became 
less and the recurrent period 
276 
International Archives of Photogrammetry and Remote Sensing. Vol. XXXI, Part B7. Vienna 1996 
Ó—, 09 ai DC *OX I3 UY] 
N 
Aa 
NS 
— OC. C. I» 
PD zz O Oo 
o