Geoscience Reference
In-Depth Information
Tabl e 2. 1
The average bulk chemistry of low, medium, and high-Ti basalts of various grain sizes
such as <45, 20-45, 10-20, and <10 m
Bulk chemistry
(oxides wt %)
Low-Ti basalt
(15071)-52
Medium-Ti basalt
(12030)-14
High-Ti basalt
(71501)-35
SiO
2
46.07
46.25
31.87
TiO
2
1.89
3.32
9.52
Al2O
3
13.87
11.70
11.83
Cr2O
3
0.44
0.43
0.43
MgO
10.88
9.42
9.49
CaO
10.52
9.78
10.36
MnO
0.19
0.20
0.22
FeO
13.87
16.27
16.05
Na
2
O
0.40
0.46
0.38
K
2
O
0.16
0.29
0.09
P
2
O
5
0.15
0.25
0.06
SO
2
0.11
0.12
0.19
Source: RELAB
Tabl e 2. 2
The average modal abundance of minerals for low-, medium-, and high-Ti basalts of
various grain sizes such as <45, 20-45, 10-20, and <10 m
Modal abundance of
minerals (wt %)
Low-Ti basalt
(15071)-52
Medium-Ti basalt
(12030)-14
High-Ti basalt
(71501)-35
Ilmenite
1.63
2.93
9.86
Plagioclase
19.10
15.76
18.76
Pyroxene
16.56
23.50
14.60
Olivine
2.86
3.50
3.40
Agglutinitic glass
52.16
48.06
45.40
Volcanic glass
3.90
1.43
6.70
Others
3.76
4.80
1.30
Source: RELAB
A specific problem of LSU is the determination of the end-members; to this end
we employ two approaches: the Convex Cone Analysis and another one based on
the detection of morphological independence. The planetary surface materials have
specific spectral features, which are related to their composition. Each image pixel is
assumed to be some mixture of various component materials, but regular algorithms
for classification cannot identify more than one class within one pixel. Hence, the
spectral unmixing is used to deal with this case.
Quantitatively retrieving mineral abundances from hyperspectral data is one of
the promising and challenging geological application fields of hyperspectral remote
sensing. However, its mixture characteristic of mineral spectra and deconvolution
method of mixture spectra are the most basic obstacles in using hyperspectral data
Search WWH ::
Custom Search