Volume XXIX - 1998

Crater Lake partially fills a caldera within what was once Mount Mazama, one of the greatest volcanoes in the Cascade Range. Around 7,700 years ago it awoke with great fury and power. Roughly 13 cubic miles of magma erupted from the volcano, covering 500,000 square miles to the north and northeast. This eruption is considered by many volcanologists to be the most violent and devastating that the world has seen in the past 10,000 years.

This climactic eruption occurred in two phases. The first was a "single vent phase" in which ash and pumice were erupted from a single vent forming an eruptive cloud estimated to be 25 miles high. This disturbance emptied the uppermost levels of the chamber beneath Mazama, leaving it weak and unstable. As the eruption continued, it drained the chamber underlying the mountain so that Mazama began to collapse inward. This brought on the "ring vent phase," whereby the remaining magma was pushed out of the mountain along the multiple areas where the upper volcano was cracking and falling. This occurred along circular ring fractures which is now where most of Rim Drive is located. This sequence of events brought about formation of a caldera, much of which we see today.

Perhaps the most unusual aspect of the climactic eruption is that it exhibited two distinct chemistries. The purpose of this article is to describe some field examples of those chemistries, because their appearance from the same volcano during one eruption is rare. Those examples should also help to illuminate how long differentiated magma compositions have characterized what we see around Crater Lake.

Volcanoes are commonly classified by their composition and their structure. Mount Mazama is an andesitic stratovolcano. Andesite is the most common lava type found at the volcano, and you can get a great look at andesite along West Rim Drive around the Watchman. The mountain's structure was that of a sharp faced or angular peak, making it a stratovolcano. (Mounts Hood, Rainier, and Shasta are all stratovolcanoes as well). Mount Mazama is believed to have been a cluster of several overlapping volcanoes that began erupting at least 420,000 years ago. With each new eruption, more and more bulk to the mountain would be added. The Crater Lake volcanic system consisted of three main components: 1) several peaks like Phantom cone, Hillman peak, and Mount Scott that piled on one another to form a conglomerate main cone, which the lake partially fills at present; 2) about twenty smaller volcanoes called cinder cones located within the caldera and on its perimeter (of which Wizard Island is one of the more recognized examples); and 3) volcanic domes, the only ones of which are now beneath the water line. All three of these surface structure groups are believed to have been fed by a large magma chamber. Figure 1 illustrates the locality of that chamber in relation to the volcano.

Figure 1. This is an illustration that modifies a figure drawn by C.R. Bacon of the U.S. Geological Survey. It shows how the chamber may have been layered into these two main zones during the time of the climatic eruption. Notice that the lighter, more silicic magma floats above the heavier, mafic magma. The magma chamber is less than four miles from the surface, and it is in the general shape of a lens.

Each type of volcano is usually made up of a specific type of lava. This is due to different lavas having characteristics which result in distinct types of volcanoes. A few characteristics like viscosity, temperature, and water content vary widely when you look at different types of lavas. These variations that are observed in lavas play an important role in the way that the molten rock will eventually behave when they reach the surface. For example, if you were to drop cookie dough on one side of a plate, and pancake batter on the other side, they would behave differently. This is mainly because the thick and pasty cookie dough has much higher viscosity than does the runny pancake batter. When the dough and batter "cool" and harden, you would see two different shapes (one flat and one tall).

The cookie dough would be analogous to "silicic" lavas , while the pancake batter would be called "mafic" lavas. Silicic lavas usually build stratovolcanoes and domes, and mafic lavas make up the large and broad shield volcanoes and cinder cones. Figure 2a shows how lavas are broken-up into specific classes based mainly on compositional changes in the silica and oxygen.

Figure 2a. There are six main lava and ash types that are found in the Cascade Range. These types are classified as to the amount of silica they contain. Rhyolites have a high amount of silica, while basalts are at the low end of the scale. Lava and ash types (such as numbers 2, 3, 4, and 5) are intermediate.

Figure 2a. There are six main lava and ash types that are found throughout the Cascade Range. These different types are mainly classified in terms of the amount of silica that is in them. Rhyolites have a high amount of silica, and the basalts have a lower amount. Lava and ash types like numbers 2, 3, 4, and 5, are intermediate.

Figure 2b. Let's say that after a volcanic eruption, you go out into the field to gather up a sample of each type of lava and ash that was deposited. In the majority of eruptions, several types of lava and ash are found. If we were to plot each type found from a common eruption with a dot, the distribution would look like the figure above. Notice how all of the samples that you collected are randomly dispersed along a wide range of different types. There are no clumped groups.

Figure 2c. Now, if you were to do the same thing that you did in 2b, the types of ash and lava that you would find after the Crater Lake climactic eruption would only be from the basaltic-andesite (2.) and the rhyodacite (5.) groups. Let's plot them on the same kind of graph that we did before. The dots would only be in those two regions. When the products of an eruption are limited to two distinct, clumped groups, the system that fed this rare eruption is said to be bi-modal. This is one of the aspects of the climactic eruption that makes Crater Lake so unique. ** I should note that the 4 lava flows associated with the climactic eruption occurred up to 4,000 years prior to the collapse. Some geologists who studied Crater Lake do not even consider them to be part of the climactic eruption, but rather lava flows that simply preceded it. So, the dots in figure 2c are the ash deposits from the climactic eruption.

It is important to realize that the majority of volcanoes in the Cascade Range, as well as those around the world, are composed of closely related lavas and other eruptive material. In other words, it is uncommon for volcanoes to display material of one composition as well as large amounts of a completely different type. There is commonly some diversity in the lavas and ash in most eruptions, but their compositions are usually very similar to one another as shown in figure 2b.

Figure 2b. After most eruptions in the Cascades, field samples would show that a range of different lava and ash types (shown in Figure 2a) were deposited. The samples collected would be randomly dispersed among the various types, with their distribution similar to the above figure, so that no clumping of samples were evident in any group.

At Crater Lake, the volcanic system exhibits a pattern where there are only two very distinct groups of eruptive material (figure 2c). This is called a bimodal system. In the climactic eruption of the Crater Lake volcanic system, materials of two distinct compositions were erupted almost right after the other. In the first phase, an enormous (up to 30 miles high) eruptive cloud was produced. This sent large amounts of ash and pumice into the atmosphere and was eventually deposited in an area of nearly 500,000 square miles.

Now, in Figure 2c, if you were to do the same thing that you did in 2b. the types of ash and lava that you would find after the Crater Lake climactic eruption would only be from the basaltic-andesite (#2) and the rhyodacite (#5) groups. Let's plot them on the same kind of graph that we did before. The dots would only be in those two regions. When the products of an eruption are limited to two distinct, clumped groups, the system that fed this rare eruption is said to be bi-modal. This is one of the aspects of the climactic eruption that makes Crater Lake so unique.

As the eruption continued, the cloud grew so large and heavy that the released gases could not hold it all up and some parts fell back down on the volcano. These super hot flows of ash and rock (called pyroclastic flows) rushed down the flanks of Mount Mazama, and deposited pumice with such heat and pressure that the pumice was compressed and welded. A deposit of welded pumice brought about the Wineglass Formation. This is the long orange colored deposit of ash and pumice that stretches from Llao Rock to Red Cloud on the north and east side of the inner caldera wall. It is illustrated in figure 3.

Figure 3. The Wineglass Formation (indicated by a thick black line) stretches from Llao Rock on the north side of the caldera wall to Red Cloud flow on the east side. It resulted from a pyroclastic flow in the first phase of the climactic eruption. The formation is best seen by hiking about 1/8 mile down the Cleetwood Trail, where a huge orange-colored ash flow tuff is evident. Tuff is a name given to ash when it is deposited so hot that it partially melts and recrystallizes, forming a hardened mass. Look closely at the tuff and you should see narrow lens-shaped pieces which are squished pumice from the hot flow.

Both the airfall ash and pumice and the pyroclastic flows forming the Wineglass were from the single vent phase. The single vent phase of the climactic eruption drained a large portion of the magma chamber, which left the volcano without a sturdy foundation. As it began to collapse inward, the force of the volcano was so great that it pushed the remaining contents in the magma chamber out along the fractures and cracks that the volcano produced as it broke inward. This activity sent out even larger pyroclastic flows than did the first phase. These rushed past the flanks and out into the once glaciated valleys and deposited rhyodacite and andesite. It is significant that all of the deposits from the first phase are high silica rhyodacite, while the pyroclastic deposits from the second (or ring vent) phase of that eruption began with rhyodacite but ended with basaltic-andesite. There were no intermediate or transitional compositions during the two phases of this eruption.

It is highly improbable that a magma chamber would show two very different types of chemistries unless the compositions were separated from each other into layers within the magma chamber. Layered magma chambers are not uncommon and, in fact, they are thought to be quite ordinary. Most of the time, however, these chambers exhibit a spectrum of layered compositions--not just two main ones like Mount Mazama, How one of these layered magma chambers forms requires some knowledge of the magmas themselves. Rhyolites, rhyodacites, and dacites are considered to be high in silica, explosive, and extremely viscous--hence the categorization "silicic". Basalts, basaltic-andesites, and andesites are considered to be lower in silica, less viscous, erupted at higher temperatures (less than 1200 degrees Centigrade), more dense than the silicic types, and are classified as "mafic" magma. If we could, hypothetically, pour these two opposite types into a big bowl, the less dense silicic type would rise above the more dense mafic type. The same thing, in essence, happens in a magma chamber where the less dense silica-rich magma rides upon the mafic type in the general shape of a lens.

A layered, lens-shaped magma chamber is believed to have existed beneath Mazama when it catastrophically erupted about 7,700 years ago. The most impressive surface manifestation of the chamber can be seen at the Pinnacles, which are found in the once glaciated valleys of Wheeler and Sand Creeks just south of Lost Creek Campground. During the second phase of the climactic eruption, pyroclastic flows rushed down these valleys and filled them. These violent flows may have rushed for miles with speeds over 100 mph and then suddenly terminated. In stopping so abruptly, the flows trapped large amounts of hot gases at the lower levels. As these gases rose toward the surface, they heated the ash and pumice so that they partially melted and recrystallized. This recrystallization process changed the soft ash into a hardened material. The hardened ash and pumice formed around the path of the escaping gas and began to act as a chimney. Over time, streams flowed through these pumice and ash filled valleys. This left the resistant pinnacles standing and eroded the softer, unaltered ash. When you look at these pinnacles, you are actually seeing the subsurface structure that escaping volcanic gases created.

By observing closely, you will see that the upper regions of the pinnacles are darker mafic material, and the lower regions are lighter silicic material. Notice how sharp the contact is between the two (shown in figure 4). Remember how the upper silicic magma rests upon the denser mafic magma in the magma chamber. When erupted, this material would be in reversed sequence (just like if we were to erupt an "n", it would land as a "u"). That sharp contact shows just how unmixed these compositions actually were, as if they were separated from each other in the chamber by a giant wall that disallowed any mixing to occur. The single vent phase had already removed a large portion of the chamber's upper level, which was high silica rhyodacite. When the volcano began to collapse in on itself and erupt in the ring vent phase, the remaining rhyodacite was pushed out, as well as the underlying basaltic-andesite zone (you may wish to consult Figure 1 again to see the general structure of the magma chamber). When both of these layers erupted, the clouds of gas and ash were violently deposited in these valleys with the rhyodacite on the bottom, and the basaltic-andesite above it.

Figure 4. Distinct layers in the ash flow deposits at the Pinnacles. The "A" layer is rhyodacite that was extruded at the start of the ring vent phase, and the "B" layer of basaltic-andesite followed it. Note the sharp contrast between the two layers, and how resistant the pinnacles are to the stream erosion that has taken place since Mazama's climactic eruption.

As stated previously, the intermingling of silicic and mafic magma is the most important piece of evidence that the chamber feeding the mountain during the climactic eruption was layered with a lens of mostly gaseous, silicic magmas separated from the dense basaltic material beneath it. This evidence is only representative of the climactic eruption, and it is worth asking whether the chamber was becoming bimodal before that time. If so, when? Fortunately the rocks at Williams Crater can help to answer that question.

Williams Crater, once known as Forgotten Crater, is named after Howel Williams who wrote the Geology of Crater Lake National Park. Published in 1942, the work is considered a classic--even though Williams did not have the benefit of modern dating techniques. His interpretation of the park's geology has been modified only slightly in the past half century by Charles Bacon of the U.S. Geological Survey. It is therefore fitting that this namesake feature represents a very crucial and critical aspect in understanding the volcanic system which created Crater Lake.

Williams Crater is roughly between 22,000 and 30,000 years old, and is located about one kilometer west of Hillman Peak. It is a basaltic cone that is aligned on a fissure (a linear crack in the earth's crust) that is radiating outward from the rim. Unlike any other cones in the park, there are bands and inclusions of intermediate and silicic pieces in the basaltic lavas and the volcanic bombs that surround the cone of Williams Crater. These pieces of higher silica lava were most likely entrapped in the magma as blobs and crystal mush prior to being erupted from the basaltic cone. Some high silica magma was eventually made in the chamber and began to separate from the mafic magmas. The higher silica lava found its way up this particular vent to the west and was erupted as entrapments or inclusions in the basalt, due to this cone being close enough to Mount Mazama, Other cinder cones in the park, such as Crater Peak and Red Cone for example, were too far away from the growing chamber for this to occur. The Williams Crater complex, in other words, shows that there was some development of differentiated magma compositions in the chamber beneath Mount Mazama at least 30,000 years ago. It is uncertain whether these inclusions offer sufficient evidence of a bimodal system extending that far back in time, but they do give us some information about when this separation may have begun to take place. Many geologists believe that separation between these compositions continues into the present, but it is also worth asking about the characteristics of volcanic activity since the climactic eruption.

On the bottom of Crater Lake, the Dacite Dome to the east of Wizard Island is made of high silica dacite. Prior to its eruption, however, the formation of Merriam Cone took place just south of Cleetwood Cove from lower silica basaltic-andesite. Once again, two different compositions in the same general vicinity. This contrast beneath the lake is not really enough evidence to conclude that the Crater Lake volcanic system is still bimodal, but samples around the park appear to suggest it still has that capability.

Brandon Browne served as a volunteer-in-parks during 1997 and is presently studying geology at Oregon State University in Corvallis.

Example of "spheroidal weathering" along the Garfield Peak Trail, Nature Notes from Crater Lake, 7:2, August 1934.