62
P. Lazo et al.
45
40
35
30
25
20
15
10
5
1
0.7
0.6
0.5
0.4
0.3
0.2
0.1
MAPE 41.7559
MAD
0.0967
MSD
0.0157
Accuracy Measures
n
M o
Actual
Fits
Variable
Spatial Analysis Plot for Mo
Linear Trend Model
Mo = 0.2776 - 0.00031×n
45
40
35
30
25
20
15
10
5
1
0.5
0.4
0.3
0.2
0.1
0.0
MAPE 43.1193
MAD
0.0564
MSD
0.0058
Accuracy Measures
n
S e
Actual
Fits
Variable
Spatial Analysis Plot for Se
Linear Trend Model
Se = 0.189 - 0.00149×n
a.
b.
Fig. 4.8 Spatial analysis plots of a Mo and b Se
The spatial distribution of Mo was very stable. It showed a moderate coefficient of
variation (CV% = 45%) and small values of kurtosis, by indicating the concentration
data show a relatively central tendency. Mo concentration data show a lognormal
distribution (p = 0.524 > 0.05) that usually indicates a tendency of crustal origin. Se
is also showed a moderate coefficient of variation (CV% = 52%), but it is followed
by high positive value of kurtosis (aprox. 5.6) by indicating the concentrations data
were shifted right with a positive tendency in their distribution.
The spatial analysis plot of Mo (linear trend model) (Fig. 4.8a) showed a very
small increasing gradient (slope = 0.0003 < intercept = 0.278) from the south to the
north (Mo = 0.278 − 0.0003 × n, n represent the number of sampling sites). Similar
to Mo, the linear spatial analysis of Se (Fig. 4.8b) showed similar increasing gradient
(slope = 0.00168 < intercept = 0.193) from the south to the north (Se = 0.189 −
0.00149 × n, n represent the number of sampling sites). Very high intercept values
compared to the slope of the regression lines indicate a relatively stable tendency
of the concentrations data that was probably linked with the combination of the
long-range transport and small effects of local emission sources.
High Mo contents in moss samples were found in coastal areas (St. 11, 13, 14), in
Memaliaj area that is mentioned for its abandoned lignite mine (St. 19), Kruja area
(St. 27), and Shkodra region (St. 31, 32 and 33), that is probably linked with soil
geochemistry. The highest Mo contents were found in St. 24 (0.606 mg kg
−1 ) and St.
33 (0.628 mg kg
−1 ) (Elbasan and Shkodra), positioned in central part of Albania and
only 30 km far from Tirana, and in the North part of the country. High Mo enrichment
in this area was probably derived by historical deposition and the accumulation of
Mo from iron and steel metallurgy (St. 24) and geogenic factors.
The highest Se contents in moss samples were found in St. 13 and 14. The next
relatively high Se contents in moss samples were found in St. 2 and 4 positioned in
the Western part of the country, that should be affected by oil refine industry and
coastal sea salts effects. All these findings indicate that Mo and Se are mostly derived
by atmospheric deposition supply and only some selected points are affected by local
factors. It can be explained by contributions from long-range transport of pollutants
in a similar way as other elements.
P. Lazo et al.
45
40
35
30
25
20
15
10
5
1
0.7
0.6
0.5
0.4
0.3
0.2
0.1
MAPE 41.7559
MAD
0.0967
MSD
0.0157
Accuracy Measures
n
M o
Actual
Fits
Variable
Spatial Analysis Plot for Mo
Linear Trend Model
Mo = 0.2776 - 0.00031×n
45
40
35
30
25
20
15
10
5
1
0.5
0.4
0.3
0.2
0.1
0.0
MAPE 43.1193
MAD
0.0564
MSD
0.0058
Accuracy Measures
n
S e
Actual
Fits
Variable
Spatial Analysis Plot for Se
Linear Trend Model
Se = 0.189 - 0.00149×n
a.
b.
Fig. 4.8 Spatial analysis plots of a Mo and b Se
The spatial distribution of Mo was very stable. It showed a moderate coefficient of
variation (CV% = 45%) and small values of kurtosis, by indicating the concentration
data show a relatively central tendency. Mo concentration data show a lognormal
distribution (p = 0.524 > 0.05) that usually indicates a tendency of crustal origin. Se
is also showed a moderate coefficient of variation (CV% = 52%), but it is followed
by high positive value of kurtosis (aprox. 5.6) by indicating the concentrations data
were shifted right with a positive tendency in their distribution.
The spatial analysis plot of Mo (linear trend model) (Fig. 4.8a) showed a very
small increasing gradient (slope = 0.0003 < intercept = 0.278) from the south to the
north (Mo = 0.278 − 0.0003 × n, n represent the number of sampling sites). Similar
to Mo, the linear spatial analysis of Se (Fig. 4.8b) showed similar increasing gradient
(slope = 0.00168 < intercept = 0.193) from the south to the north (Se = 0.189 −
0.00149 × n, n represent the number of sampling sites). Very high intercept values
compared to the slope of the regression lines indicate a relatively stable tendency
of the concentrations data that was probably linked with the combination of the
long-range transport and small effects of local emission sources.
High Mo contents in moss samples were found in coastal areas (St. 11, 13, 14), in
Memaliaj area that is mentioned for its abandoned lignite mine (St. 19), Kruja area
(St. 27), and Shkodra region (St. 31, 32 and 33), that is probably linked with soil
geochemistry. The highest Mo contents were found in St. 24 (0.606 mg kg
−1 ) and St.
33 (0.628 mg kg
−1 ) (Elbasan and Shkodra), positioned in central part of Albania and
only 30 km far from Tirana, and in the North part of the country. High Mo enrichment
in this area was probably derived by historical deposition and the accumulation of
Mo from iron and steel metallurgy (St. 24) and geogenic factors.
The highest Se contents in moss samples were found in St. 13 and 14. The next
relatively high Se contents in moss samples were found in St. 2 and 4 positioned in
the Western part of the country, that should be affected by oil refine industry and
coastal sea salts effects. All these findings indicate that Mo and Se are mostly derived
by atmospheric deposition supply and only some selected points are affected by local
factors. It can be explained by contributions from long-range transport of pollutants
in a similar way as other elements.