416 
cess of plant maturation, and a surface plant 
cover sketch with .maximum or minimum biomass 
during whole growing season can be portrayed. 
An average ground biomass map during the 
plant grouwing season are also illustrated. 
Based upon perennial biomass of plant, a pe 
rennial maximum and minimum biomass will be 
produced, then an average Perennial biomass 
can be obtained, and a dynamic map of plant 
cover will be able to be portrayed. 
Such advantages promote the estimation of 
pasture biomass and crop production by using 
NOAA’s data, instead of LANDSAT’ S ones. 
1.3 DATA COLLECTION 
Collection of the NOAA data took place on 
the 29th of July 1984 and the 13th June. 1985 
a.nd to compare with these data, collection 
of 11 representatives and their correspond 
ing spectrural reflection data from 11 va 
rious types of pasture during June to August 
1983 was made (Tal.2). Meteologic data(pre- 
ciptation, temperature, etc.) in 1983 and 
in 1985 were collected to fill in the time- 
non synchronous hap between NOAA’s data on 
29th of July, 1984, and ground sampling data 
in June to August, 1983, and to provide ba 
sis for the establishment of correlation mo 
del and error analyses. 
According to various types of pastures or 
ranges, a representative sampling plot with 
an abea of 1 was collected. The clipped 
sample wet weight was immediately measured 
after clipping. Three subsamples location 
were averaged as a biomass for a sampling 
location. By visual plant cover, the total 
biomass per mu or per ha. can be calculated. 
2 METHODS AND PROCEDURES 
2.1 APPLIED MODELS 
As well known, by amount of ground mass spe 
ctral measurements, the 0.68 urn region cor 
responds strongly to the in vivo red region 
of chlorophyll absorption and is inversely 
related to the chlorophyll density. The 
0.725-1.10 urn region corresponds to tne re 
gion of the spectrum where reflectance is 
proportional to the green leaf density. Ra 
tio combination of these two wavelength re- 
tions are thus related to the chlorophyll 
green leaf interraction (Gates, 1965; Wool- 
ley, 1971 ; Knipling, 1970). Recent years, 
using these two bands for estimating biomass 
has been confimed by many cases, such as u- 
ing LANDSAT’s and NOAA’s data to estimate 
soybean and winter grain yield, to predict 
aqricultureal crop production and pasture 
or range biomass. It should be pointed out 
that the two bands situated at 0.76-0.78 urn 
and 0.92-0.98 um are avoided,because the 
former is not sensitive to vegetation, and 
the latter is suitable for atmospheric water 
vapor absorption. 
To compare results from green leaf area 
response to absorption, reflection, and ra 
diation for the red and the near infrared 
with that of hand clipping method, the 101W 
field spectral radiator was used to integ 
rate the reflective spectrum situated at 
0.55-0.68 vim, and 0.725-1 .10 um which is 
fully corresponded to CH1 and CH2 of NOAA, 
respectively. A regressional analysis(Fig.2) 
between measured spectrum and sampling fra- 
sh grass biomass corresponded in this area 
was made. It is illustrated that the corre 
lation coefficient is 0.77. 
For estimating pasture biomass with the 
AVHRR data of NOAA, an applied model, green 
leaf Normolized Difference(ND) (Rouse et al., 
1973; Tucker et al., 1983 a , 1983) will be 
accepted, namely: ND=CH2 _ CHi/CHjj + GHi . 
It is clear that the ND reflects the dif 
ference between absorption of green leaf 
matter for red and reflection for near inf 
rared. Themore the chlorophyll density, the 
bigger this difference is. Therefore, it is 
refered to as a specific value to estimate 
green leaf biomass. It is noted that the ND 
value is effected by t 56° field of view be 
cause of the atmospheric path- length effec 
ts of solar rediace(Tucker, 1983), and if 
the angle of the sun’s altitude is high and 
sky is clear, this effect will be reduced to 
a minimum extent(Holben et al., 1984). 
2.2 METHOD AND DATA ANALYSIS 
W ith NOAA magnetic types of the Tarim River 
Basin, Xinjiang of China received on the 29th 
of July, 1984, and the 13th of June, 1985, 
by the Beijing Receiving Station of NOAA Sa 
tellite in the Meterological Satellite Cen 
ter, National Meterological Bureau, China, 
by using the model ND = CH^-CH-]/CH2+CH1 , and 
I'able l. 
Chnracto 
inclinati 
ileinht at 
Number o£ 
Times of 
orbital F 
Latitudir 
Cycle dui 
Ground cc 
Field of 
Instantar 
( 1 FuV ) 
Groud res 
h 
Gro^d res 
(maximi 
Samples j 
Number ol 
uata prêt 
♦LANDSAT- 
'of 16 da} 
.of prévit 
matic Ma| 
♦•This is 
than the 
*** un Ut 
Table! 
wet b: 
Plots 
1 
2 
3 
4 
5 
6 
7 
8 
9 
10 
11 
* In L, 
**One 
thro 
in t: 
trib 
illu 
both 
ver 
dese 
To 
biom 
area 
gras 
and