A new approach for hydrothermal alteration mapping by selecting and interpreting principal components in Landsat ETM+ images
Abstract:
Introduction
In remote sensing studies, especially those in which multi-spectral image data are used, (i.e., Landsat-7 Enhanced Thematic Mapper), various statistical methods are often applied for image enhancement and feature extraction (Reddy, 2008). Principal component analysis is a multivariate statistical technique which is frequently used in multidimensional data analysis. This method attempts to extract and place the spectral information into a smaller set of new components that are more interpretable. However, the results obtained from this method are not so straightforward and require somewhat sophisticated techniques to interpret (Drury, 2001). In this paper we present a new approach for mapping of hydrothermal alteration by analyzing and selecting the principal components extracted through processing of Landsat ETM images.
The study area is located in a mountainous region of southern Kerman. Geologically, it lies in the volcanic belt of central Iran adjacent to the Gogher-Baft ophiolite zone. The region is highly altered with sericitic, propyliticand argillic alterationwell developed, and argillic alteration is limited (Jafari, 2009; Masumi and Ranjbar, 2011).
Multispectral data of Landsat ETM was acquired (path 181, row 34) in this study. In these images the color composites of Band 7, Band 4 and Band 1 in RGB indicate the lithology outcropping in the study area. The principal component analysis (PCA) ofimage data is often implemented computationally using three steps: (1) Calculation of the variance, covariance matrix or correlation matrix of the satellite sensor data. (2) Computation of the eigenvalues and eigenvectors of the variance-covariance matrix or correlation matrix, and (3) Linear transformation of the image data using the coefficients of the eigenvector matrix.
Results
By applying PCA to the spectral data, according to the eigenvectors obtained, 6 principal components were extracted from the data set. In the PCA matrix, theeigen vector differences between the means of the level of significance between two bands (or spectral significance of the PC). The higher value is regarded as the Target Value of the bands which show a lower correlation. The components having maximum spectral significance of PCs, in bands 1 and 3, 5 and 7 and 5 and 3, were selected for enhancement of iron oxides, clay minerals and carbonate minerals, respectively. In each PC matrix, the sum of the significances is regarded as the spectral weight of that PC.
The spectral weight of the extracted PCs, was found to be as follows: PC5> PC4>PC2>PC3>PC7>PC1
The inverse PC4 and –PC3 provide valuable information on vegetation mapping. In order to map the alteration zones and igneous rocks outcropped in the study area, the color composites of the PC5, -PC4 and average of each PC are included in RGB, respectively. The spectral proportion of each PC pertaining to each mineral was calculated as spectral significance in the two bands (e.g. Bands 5 and 7 for clay minerals and Bands 3 and 1 for Fe oxide minerals) divided by spectral weight of that PC. Based on the obtained results, the selectivity of the extracted components for enhancement of clay minerals and Fe oxide minerals was calculated and images of these minerals were produced using the following expressions: Fe oxide minerals: Clay minerals: For carbonate minerals, the proportion of each PC is calculated in terms of the eigenvectors of bands 5 and 3. The selectivity of the PCs used in enhancing of spectral data of carbonate minerals is as follows: PC5>PC2>PC1>PC3>PC4>PC7
In the remotely sensed image, PC5 with high spectral weight was selected as the informative PC for clay minerals, iron oxides and carbonate minerals. This is becausepropylitic alteration and the formation of carbonate minerals can be easily enhanced in the processed images. Eventually, overlapping of the processed images provides patterns of hydrothermal alterationwhich indicate the areas to be prospected. In order to validate the obtained results of the image processing with geological evidence,petrographic studies of rock samples collected from major outcrops in the study area were made. It was found that quartz, calcite, epidote, sericite and chlorite are the main constituents of sericiticand propylitic alteration assemblages in the study area. The minerals are virtually enhanced in Landsat ETM using the proposed methods and confirm the results obtained from multispectral data analysis.
Conclusion
This study provides a new and improved approach to obtain the most meaningful spectral data for oxides, carbonates and clay minerals in multispectral images. As these minerals are typically found in hydrothermal alteration, the method presented in this article can be used for enhancement of such mineral spectral data, which can be very helpful in prospecting and exploration for hidden mineral deposits.
Language:
Persian
Published:
Journal of Economic Geology, Volume:8 Issue: 1, 2016
Pages:
181 to 199
magiran.com/p1572839  
دانلود و مطالعه متن این مقاله با یکی از روشهای زیر امکان پذیر است:
اشتراک شخصی
با عضویت و پرداخت آنلاین حق اشتراک یک‌ساله به مبلغ 990,000ريال می‌توانید 70 عنوان مطلب دانلود کنید!
اشتراک سازمانی
به کتابخانه دانشگاه یا محل کار خود پیشنهاد کنید تا اشتراک سازمانی این پایگاه را برای دسترسی نامحدود همه کاربران به متن مطالب تهیه نمایند!
توجه!
  • حق عضویت دریافتی صرف حمایت از نشریات عضو و نگهداری، تکمیل و توسعه مگیران می‌شود.
  • پرداخت حق اشتراک و دانلود مقالات اجازه بازنشر آن در سایر رسانه‌های چاپی و دیجیتال را به کاربر نمی‌دهد.
دسترسی سراسری کاربران دانشگاه پیام نور!
اعضای هیئت علمی و دانشجویان دانشگاه پیام نور در سراسر کشور، در صورت ثبت نام با ایمیل دانشگاهی، تا پایان فروردین ماه 1403 به مقالات سایت دسترسی خواهند داشت!
In order to view content subscription is required

Personal subscription
Subscribe magiran.com for 50 € euros via PayPal and download 70 articles during a year.
Organization subscription
Please contact us to subscribe your university or library for unlimited access!