Three Examples Illustrating the Formation Process of an Archaeological Site

An important theme in archaeology is to study the formation process of an archaeology record—or to put it in more general words, to understand how things are formed. To me, a particularly interesting aspect of studying the formation process is when there are more than one artifact or feature in an archaeological site, how do we determine the relationship between them, e.g., how did they grow together and what is the their relative timing order?

This is an intriguing and practical problem because in archaeological sites, we most often will see more than one artifacts or features mixed together. Most likely there is a stark distinction between their appearances. Things sitting currently together doesn’t necessarily mean that they have always been together, especially if they look so different. The appearance distinction makes people wonder how they were formed together.

I recently visited a series of archaeological sites on my road trip from bay area to Austin. I came across of several examples that illustrate this interesting problem.

Example 1: Joshua Tree National Park

At the Joshua Tree National Park, I saw this interesting scenario (in the picture below) where a group of light rocks and a mountain of dark rocks are contacting each other. I went up to take a closer look at the junction point, and it just seemed like that the two parts are naturally grown together. Clearly this is not influenced by human beings intentionally piling the light rocks to the dark mountain. So what happened here?



In fact, geoarchaeologists did find that the two parts were two different materials. The dark rock is gneiss, specially the Pinto gneiss, while the light rock is granite, specially the monzogranite. The monzogranite was formed by magma cooling down underneath the surface, and was originally covered by the Pinto gneiss. During a long period of time, ground water eroded the overlaying gneiss (dark rock), and flood cut through it. Eventually, the monzogranite (light rock) got exposed and even uplifted to the surface due to earth activities, forming the “contacting zone” above the ground with the originally covering gneiss, as we see today [1].


You might think that this formation process so dramatic and hard to believe. Well if you know that it took nature about 100 millions years to form this, you probably wouldn’t be as surprised. In fact, monzogranite is everywhere in the park. Some are contacting the gneiss, some are standing alone as rock piles, such as the Split Rock as shown in the picture to the right. Regardless, the formation process of monzogranite rocks all the follow the same pattern as described above: magma cooling down followed by water flowing through to carve out vertical or horizontal lines. The surrounding materials are flushed away by flood for some rocks (e.g., Split Rock); some are still touching the rock (e.g., the “contact zone” example).

Example 2: Petrified Forest National Park

The second (and a more beautiful!) example of different archaeological artifacts co-exist at the same location is the Painted Desert at the Petrified Forest National Park. In the left figure below, there is a clear color distinction between the black top of the mountain and the colorful body of it. The right figure below examines the color separation more closely.

IMG_3336  IMG_3342

The colorful body is what geologists would call Chinle formation, which by itself has many layers (or members) with each layer has a different color. All members were formed about 225 millions years ago through various deposition processes. More about different layers within the Chinle formation can be found at [2].

What I am really interested in is the black layer on top of the Chinle formation. It just looks so different, and its color is not found in any Chinle formation members. So what is it? Did it form in the same way as the Chinle formation?

It turns out that the top black layer is made of basalt! Basalt is a volcanic rock formed from rapid cooling of lava. About 4-8 million years ago, volcanos erupted at the surrounding area and produced basalts, which were then blew to the top of the Chinle formation. This is why many mountains in the Painted Desert have two distinctive layers as we look at them today. So even though the basalt and the Chinle formation are closely touching together, they are separately by over 200 million years of geologic history!

An interesting fact about the basalt layer is that they actually protect the Chinle formation from being eroded by harsh weather like the monsoon. Geologists believed that the over 200 million years of gap between the Chinle formation and the basalts used to be about 1,000 feet thick, and were eroded before the basalts covered the Chinle formation. Similarly, vegetations covering the mountains can also reduce their erosion rate [3][4].

Example 3: Sunset Crater National Monument

The last example is a little bit different from the previous two in that it is formed by human activities instead of natural forces. The figure below is the Sunset volcano crater as the Sunset Crater National Monument. By the first glance, what immediately caught my attention was the strikingly barren south side of the crater (right side of the picture) in contrast to the north side which had much richer coverage of trees and vegetation.


The question is, why is the south side so barren? Well there could only be two possibilities. First, it could be because the nature of the rocks/soil just isn’t good for trees/plants to grow. Second, the south side did have trees some time in the past, but somehow lost time over time. Which one is it?

The answer is the latter. The geographic formation of both sides are exactly the same — they were both formed by volcano eruptions and depositions. However, people used to hike, among other activities on the south side of the crater a lot, which created wide swaths that removed cinders and made the crater impossible to grow plants. You can actually clearly see the “scars” resulted from the hikers. In fact, people did not only hike. In 1929, a hollywood movie company wanted to dynamite a portion of the Sunset crater as part of a movie [5]! This would lead to devastating results of the crater’s ecology and geology. Fortunately, this plan did not get passed.

Why Does the Formation Process Matter?

Different archaeologists can give you all kinds of professional reasons. However, for archaeological enthusiasts who happened to have an engineering background, being able to explain things is a huge satisfaction. The “oh that’s why” moment is irreplaceable.


[1] Geologic formation of Joshua Tree National Park:
[2] Geologic formation of Petrified Forest National Park:
[3] Geology and the painted desert:
[5] America’s best idea: Sunset crater nearly destroyed by Hollywood:

How Did Researchers Predict the End of the Road for Conventional Microarchitectures 14 Years Ago

Over the past a few decades, we have been witnessing drastic changes in microprocessors. If I have to summarize, there are two milestone transitions: from single core to multicore, and from multicore toward more specialized accelerators. While it’s clear that the so-called dark silicon is the reason that we are gradually shifting from the multicore era to the specialization regime, it is not as clean and easy to explain why we moved from single core to multicore. Continue reading

How to Explain a Scientific Figure

A figure sometimes is more powerful than a hundred words in explaining scientific findings. However, putting figures in a paper without proper explanation might lead to confusion, or at the very least waste precious space real estate. Note that this article is not about how to display data in a figure, but about how to explain a figure, assuming it is elegantly crafted. There are many articles about the former, on which Dr. Edward Tufte has many great articles and books that I highly recommend. Continue reading

Having a “Professional” Personal Webpages

My advisor was (co)charing CGO’2014, and he asked me to compile an email list of all the authors who published at ISCA, MICRO, HPCA, and ASPLOS over the last three years, so that the publicity chair and he could send out the “spam bomb” to publicize CGO.

It took me 10 mins to compile a list of all the names; the tedious and intimidating part was to find the email address for each of them. I could not think of an automatic way other than to google the name, go to the webpage, and look for the email address. It was amazing to me how difficult it was on some webpages to find a single email address, or in general, to quickly locate important information. Continue reading

A Framework For Understanding Cache Coherency


The correctness of shared-memory processor means each processor, or a process/thread from a programmer’s perspective, has a coherent/consistent view of the memory, i.e., they agree upon the value of a particular memory location, i.e., it is not possible that there are two “correct” values of an address at any given point. Whatever the cache coherency protocol does, it has to guarantee a consistent memory view across all the processors and memory. Continue reading

What is the Optimal Logic Depth Per Pipeline Stage?

There are two things that affect the performance: cycles per second (clock rate, frequency), and instruction per cycle (IPC).

The way to increase the frequency is: 1) faster transistor through process generations, and 2) less logic in each pipeline stage (evaluated by #FO4 delays).

This seminal paper assumes that the context is within a particular process generation, that is the FO4 delay stays the same. Continue reading

The Significance of the Recent Discovery of the Voynich Manuscript

Recently Stephen Bax from Bedfordshire University claimed that he had been able to decipher 10 words of the famous Voynich manuscript, BBC news.

The Voynich manuscript is 246 pages of handwritten manuscript discovered by an Italian antique book dealer Wilfrid Voynich in 1912. He purchased it among a collection of second-hand books. The Voynich manuscript largely consists of two major parts: texts and illustrations. Both are yet to to crack and understand. Continue reading

Token Coherence: Decoupling Performance and Correctness

I really like the idea of Token Coherency. It makes so much sense to me, especially if I think about it using my framework of understanding cache coherency.

Workload trend suggests that we want to enable cache to cache transfer to minimize latency. So we need to use direct communication scheme. However, due to the distributed nature of the direct communication, it is hard to enforce order, and traditionally we can only enforce a total order (the strongest assumption) using an ordered interconnect (e.g., bus). For example, destination-set predictor uses the direct communication scheme but has to use an ordered network.That limits the performance. The technology trend suggests that we want to use unordered interconnects. Continue reading