Loading…
Exploratory Analysis of Sediment Geochemistry to Determine the Source and Dispersion of Ba, Fe, Mn, Pb and Cu and in Chihuahua, Northern Mexico
Exploratory Data Analysis (EDA) was applied to a sediment geochemistry data set (N = 2,584) within the Chihuahuan Desert in northern Mexico. Each data point contained information on the location, concentration of 19 elements, and lithology. A Box Plot analysis was applied to the data to determine mi...
Saved in:
Published in: | Journal of geography and geology 2012-12, Vol.4 (4), p.26-26 |
---|---|
Main Authors: | , , , |
Format: | Article |
Language: | English |
Citations: | Items that cite this one |
Online Access: | Get full text |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Exploratory Data Analysis (EDA) was applied to a sediment geochemistry data set (N = 2,584) within the Chihuahuan Desert in northern Mexico. Each data point contained information on the location, concentration of 19 elements, and lithology. A Box Plot analysis was applied to the data to determine mild and extreme anomaly thresholds. Anomaly maps revealed areas of concentration of anomalies as well as dispersion patterns for each element; Ba was the most mobile of the five elements and Pb the least. Specific potential sources of contamination were identified visually after superimposing anomalies with map layers of mines, urban centers and/or hydrology. The patterns revealed that mining was an important source of Pb, Fe, Mn and Ba contamination, while copper instead was associated to the presence of Tertiary volcanic rocks that abound in the western edge of the study area. Multi-element exploratory techniques yielded four principal components that accounted for 74% of total variance, exposing a strong association among Fe, Mn, P and Mg in PC1, Ca and Sr in PC2, Co, Cu, Cr, and Ni in PC3 and Pb in PC4. By identifying potential contaminant sources, EDA shed some light into the nature of the contamination, if natural or anthropogenic, a critical piece of information to impact assessment studies and best management plans. |
---|---|
ISSN: | 1916-9779 1916-9787 |
DOI: | 10.5539/jgg.v4n4p26 |