Ceramic Petrography Laboratory

Detailed Methods

Elizabeth J. Miksa, Carlos Lavayen, and Sergio F. Castro-Reino

Introduction

Petrographic modal analysis, or point counting, is a quantitative technique that provides data for the evaluation of samples on the basis of the quantity and composition of grain types present. It allows samples to be grouped based on the absolute abundances of different grain types. This technique was selected because it allows sand samples to be compared directly to sand-size tempering material found in utilitarian ceramics recovered from study area sites. This section describes how sand samples were collected, processed, and point counted in thin section to provide the basis for an actualistic petrofacies model, or map of modern sand composition zones (Ingersoll 1990). Subsequent sections detail the statistical techniques used in the derivation of the petrofacies model, how the thin-section based petrofacies model was used to define expected characteristics of the sand in hand sample as viewed through a binocular stereomicroscope, and how the hand sample model was applied to the sand temper in the potsherds and verified.

Our analytical procedures can be summed up with five basic steps, best illustrated as a simplified analytical flow chart. For the Rye Creek Mitigation Project, Stark and Heidke (1992:Figure 13.14) developed a detailed analytical methods flow chart as a means of elucidating the procedures used in conducting a quantitative petrofacies analysis of sands as source materials. This has been updated and improved and is presented both as an outline of our analytical process and as a suggested template for future sand source studies.

Defining Petrofacies

Sands derived from similar source rocks under similar conditions will have similar compositions (c.f. Dickinson et al. 1983). When we study sands within a well-defined region and determine that those sands can be broken into subsets on the basis of composition and spatial contiguity, we have defined petrofacies (petrographic facies). The petrofacies concept is used for the study of rock units according to their petrologic characteristics as opposed to their form, boundaries, or relationship to other rock units (Bates and Jackson 1984:379). Petrofacies in ancient sedimentary deposits have both lateral (spatial) and vertical (time) components. In modern sands, we deal only with the spatial component. The time component is effectively held constant, barring major climatic and/or tectonic changes within the time scale of the study. In this context, the petrofacies becomes a sand composition zone. For the purposes of archaeology, we can think of petrofacies as temper resource procurement zones whose sand compositions are distinct from one another at a relevant scale of investigation.

The boundaries of the petrofacies are a construct that we create, in the sense that composition rarely changes abruptly from one drainage to the next, and yet there are modal compositional differences within regions. Compositional changes tend to be clinal, with grains from one source rock being lost as grains from another source are gained. Boundaries must be chosen based on the spatial scale on which the source rocks change and the spatial scale on which sampling is conducted. The problem is similar to that encountered in defining archaeological strata on the basis of changes in the sediment. If the surface between two strata is considered the "contact," some strata are well defined with abrupt contacts and are easily mapped. Other strata may have abrupt contacts on only part of their boundary, and some have only gradational contacts with the surrounding sediment. An abrupt contact leads to easy boundary definitions at fine to coarse scales of resolution. A gradational contact allows boundaries to be drawn coarsely, but not on a fine scale. Usually, increasing the resolution of data across the boundary zone allows the contact to be refined. The same principles apply to sand composition boundaries.

Drainage basins rarely coincide exactly with rock units, yet they are the geomorphological unit in which sands are created and transported. A preliminary petrofacies map is created by comparing bedrock geology to geomorphology and drainage pattern in a region. The preliminary petrofacies map is tested by sampling sands and characterizing their composition. Resolution of the petrofacies boundaries depends on the scale of variability in the source rocks, the scale at which samples are collected, the resolution of the analytical technique, and the statistical techniques employed. In our experience, preliminary petrofacies maps constructed from bedrock geology and geomorphological data rarely convey the true variability of the sand as it exists on the landscape. The preliminary map is a starting point, a tool to guide the initial sampling necessary for construction of a model that has some basis in reality. Petrofacies boundaries change in the course of investigation, as the information used to refine the model grows more detailed.

We infer that intrabasinal production has been documented when sherds tempered with compositions found in the basin are recovered from sites in the basin. We do not know whether or not this inference is wholly correct. The potential for redundant compositions (similar sand compositions found in more than one basin) is highest for the granitic and diabasic sands. Where these sands contain lithic fragments characteristic of restricted areas their origin can be assigned more confidently. When they contain only the products of granite or diabase, the possibility exists that sands from one basin could be confused with sands from another basin. However, our preliminary work in comparing similar compositions from different basins indictes that the sands are easily distinguished petrographically and statistically.

A basic distinction exists between those petrofacies that are mineral-rich and those that are rock-fragment-rich. The mineral-rich sources are those that contain primarily quartz, feldspars, and mafic minerals from coarsely crystalline plutonic rocks such as granite and diabase. Often these petrofacies contain rock fragments such as quartzite and basalt, tuff, or rhyolite, but the rock fragments are not dominant in the samples. The rock-fragment-rich sources have a much higher proportion of rock fragments such as quartzite, siltstone, volcanic rocks, greenstone, hornfels, phyllite, schist, and other metamorphic rocks. They contain quartz and feldspars—these minerals are rarely absent from any petrofacies—but the rock fragments dominate the petrofacies.


Next

Back to General Overview

Ceramic Petrography Lab Home


Index

Ceramic Petrography Lab home
Overview
Detailed Methods
Online Research Results

Detailed Methods Index

Introduction
Defining & Mapping Petrofacies
Collection of Sand Samples
Sand Sample Preparation
Point Count Methodology
Point Counting - How
Point Counting - What
Statistical Analysis
Correspondence Analysis
Petrofacies Refinement
Discriminant Analysis
Hand Sample ID Model
Describe Sand in Hand Sample
Flow Chart
Binocular Microscope Sherd Characterization
Sherd Thin Sections

Analytical procedures flow chart (simplified)
Petrofacies Map & Sands

Home | About Us | Compliance Services | Analysis Services
Publications
| Selected Projects


© 2004 Desert Archaeology Inc.