CHAPTER 4: REVISIONS TO THE GEOLOGY OF THE TRES VÍRGENES-LA
4.3 Methods
4.3.3 Results of Classification
As discussed in Chapter 2, the 28 images of Visible and Near Infrared to Short-wave Infrared (VNIR-SWIR) MASTER data in the TVLRVR were atmospherically corrected and classified. The VNIR-SWIR MASTER data were atmospherically corrected to a reasonable accuracy with the Empirical Line Calibration (ELC) mode in ENVI.
Classifying the atmospherically corrected MASTER data, resized to exclude the bands highly affected by water absorption, was done successfully by choosing endmembers with the highest Pixel Purity Indices (PPITM) values. Endmembers from a select number of field spectra and one image-derived region were added to the list of endmembers to expand the classification. This yielded a classification map that roughly delineated surfaces that vary in their secondary minerals, produced by weathering and alteration, namely ferric oxides, hydrates, hydroxyls, and carbonates. Attempts to identify these minerals using ENVI’s Spectral AnalystTM were only moderately successful. The resulting classification map was overlain with the unit contacts shown in Figures 1 and 2 (Plate 1).
By comparing the delineations of previously mapped units with the classification map, and keeping in mind the likely identity of the various endmembers, a few things can be noted. The recent lavas of the Tres Vírgenes volcanoes, the older Miocene lavas, and many of the central domes of La Reforma and El Aguajito are all mapped as the endmember “Basalt.” In Chapter 2 the similarity of this field spectrum to hematite, a common oxidation mineral of iron-bearing, mafic volcanic rocks, was noted. The
classification clearly does not illuminate the difference between the different lavas in the region, but does distinguish these regions from more silicic material, pyroclastic deposits, sedimentary rocks, and alluvium. It is significant that the central domes of both La Reforma and El Aguajito are primarily classified as this endmember. Because they were previously assigned to Miocene Comondú Formation, one would expect them to match the classification of the Miocene Comondú Formation in the southeastern portion Plate 1, which is classified as “n-d-2.” Endmember “n-d-2” was determined in Chapter 2 to represent weathered detritus. This would be a reasonable classification for the Comondú Formation, which is described as largely tuff breccias and debris flow deposits. Although there is variation within the units of the Comondú Formation [Heim, 1922], the classification of the central domes as endmember “Basalt” further verifies that these central domes are not made up of the Comondú Formation, but rather are stratified domes of welded ash-flow tuffs and lavas.
The abundance of pixels classified as “Basalt” in the central dome of La Reforma also indicates this unit is different from the main ignimbrite emissions mapped as Qi, which are primarily classified as “n-d-2.” This verifies the central dome has a different composition than the pyroclastic layers surrounding the caldera, but it does not establish their age relationship.
As noted in Chapter 2, endmember “n-d-10” is found near the summits of the recent mafic lavas around the La Reforma caldera. Endmember “n-d-10” was determined in Chapter 2 to most likely represent a surface with large amounts of oxidized iron. This
endmember is also found within regions previously mapped as more silicic material:
within a small region mapped as Tc to the north of the La Reforma resurgent dome (Fig.
7) and within a region mapped as Qps, just northwest of El Azufre (Fig. 8).
Figure 8. (Previous page) Revised geologic units and faults on top of a mosaic of two ASTER images (Red: band 3, Green: band 2, Blue: band 1). Line symbology is the same as in Figure 1. Geologic unit abbreviations are described in Table 3. Map datum and projection: NAD27 Mexico, UTM zone 12.
Mesa La Tinaja, described by Garduño-Monroy et al. [1993] as an ignimbrite sheet (Qps, Figs. 1 and 2) that erupted from El Aguajito and is interfingered with sand deposits, is primarily classified as class “Faroash” (Plate 1), similar to other pyroclastic units surrounding El Aguajito. The majority of the ignimbrites attributed to eruptions from La Reforma (Qi, Figs. 1 and 2) are classified as “Farosh” as well, with their detritus primarily identified as “n-d-2.” In the classification, therefore, the ignimbrites of Mesa La Tinaja cannot be distinguished from ignimbrites that belong to La Reforma caldera.
However, it is expected that two ignimbrite deposits, even if erupted from separate volcanic complexes, would be very similar in their VNIR-SWIR spectral signature and weight percent silica.
In El Azufre canyon, within the heavily faulted region, which gives rise to hydrothermal vents, there are pixels classified as “n-d-3.” Endmember “n-d-3” was determined in Chapter 2 to indicate the presence of mineral alunite, a common alteration mineral of hydrothermal reactions. There are also scattered pixels classified as “n-d-3” within the Las Minitas region of La Reforma, another verified region of hydrothermal activity.
Within the central portion of the El Aguajito volcanic complex, however, there are no pixels classified as this endmember, indicating a lack of similar hydrothermal alteration in this volcanic complex. The only other location of this endmember is in the alluvium and Boleo Formation (Tm) in the southeast corner of Figures 1 and 2.
The unit (Qtt) within the southern moat of the La Reforma caldera (Figs. 1 and 2) is largely identified as endmember “n-d-1.” In Chapter 2 it was determined that this endmember likely represents a surface with a large percentage of the mineral montmorillonite, an alteration mineral of volcanic ash. This identification helps to confirm the identity of this unit as altered ash. A large percentage of La Reforma’s pyroclastic material is also identified as this endmember. Additionally, a large percentage of the pixels within the El Aguajito complex is also identified as endmember
“n-d-1.” This identification led to the addition of a unit similar to Qtt within the El Aguajito complex (Qett in Figs. 7 and 8).
A significant percentage of the pixels within the units identified as Tm and Qal in Figures 1 and 2, inland from Punta El Bajito, is classified as endmember “n-d-6.” This endmember was determined to most likely represent a surface with a high gypsum content. Gypsum is found in large quantities within the Boleo Formation [Wilson, 1948].
This region, including the previously mapped alluvium, must therefore primarily contain outcrops of the Boleo Formation and has been updated accordingly in Figures 1 and 2.
Units mapped as Qaa and Qtb in Figures 1 and 2 appear nearly identical in the classification (Plate 1), the weight percent silica map (Plate 2), the ASTER image (Figs. 1 and 2), and the Thematic Mapper image (Fig. 3). These units are largely mapped as endmembers “n-d-2” and “Basalt” and appear to have a mafic composition in the weight percent silica map. Distinguishing these surfaces as two separate units was done based
on previous mapping [Demant, 1984]; however, nothing in the imagery suggests they differ in composition.