Yellowstone Science2

Dr. Robert B. “Bob” Smith has beenassociated with Yellowstone geology forfour decades. Bob is a professor of geol-ogy and geophysics at the University ofUtah. He has conducted research in thepark since 1959 and has operated theYellowstone seismic and GPS networkssince 1982. He is a former president of theSeismology Section and a fellow of boththe American Geophysical Union and theGeological Society of America. A livelyspeaker who talks about the many con-nections of features and resources in whathe calls a greater Yellowstone“geoecosystem,” Bob graciously spokewith senior editor Sue Consolo Murphy in1999 during one of his many trips to thepark. Windows into the Earth: The Geo-logic Story of Yellowstone and GrandTeton National Parks is his new book withco-author Lee J. Siegel (Oxford Univer-sity Press 2000; 240 pages, 69 illustra-tions).

Yellowstone Science (YS): How didyou get interested in geology?

Robert Smith (RS): I actually got startedhere in Yellowstone; I worked in 1956 asa GS-0. I think that’s the truth—maybe itwas a 1.

YS: No pay?RS: Very little. It was a great year

because my job was the lowest GS levelthey had. I was stationed at Lake workingfor the U.S. Fish and Wildlife Service.They brought us on in late February; wedrove “weasels” across Hayden Valley.These were the first snowmobiles, theselittle army weasels, horrid things.

There used to be a grayling fish hatch-ery at Grebe Lake, west of Canyon. Myfirst job was to ski in and open up thisbuilding and get the water flowing andthen install fish traps and wait for the fish

to spawn and be captured for study.When we were taking graylings, griz-zlies would come to our cabin becausethey could smell fish eggs inside thebuilding. I was sitting in my bed onenight, and I heard this roar and poundingon the cabin. After that I slept with atwo-bitted ax across my bed the rest ofthe time. I figured they were going tocome right through the door.

I then helped map the tributaries ofYellowstone Lake that could supportfish spawn. I did surveys of water chem-istry, salinity, and sediment conditions.I think I walked every mile of the drain-age that summer. Monday they put apack on my back and said, “See youFriday.” There were no radios, no GPS(Global Positioning Systems), old maps,nothing, you just went. I would go upevery stream, every tributary. I lived thatsummer at Fern Lake, upper Pelican,and we had cabins at Clear Creek, downat Trail Creek, and at Peale Island. Weworked our way around YellowstoneLake. That was really a fantastic experi-ence.

They also had me assist with surveyinglake bathymetry and limnology. We hadan old surplus navy boat with a depthbottom sounder on it from which we didseismic profiling of the lake. We alsolowered water and bottom sampling de-vices down the water column. All the wayalong, the sounder recorded data frombeneath the lake bed with echoes of rocksediments beneath it. “Hey,” I’d look atmy boss, “what is all this?” He said,“Mind your own business. You’re sup-posed to worry about fish, not about rocks.”But I thought it was pretty neat. That was1956.

I didn’t finish high school, actually. Iwas admitted to college early, but I leftthat year after the opportunity came forme to work in Yellowstone. I ended up atMadison Junction that fall doing streamchemistry and creel censuses, all thesethings about fishing. Then the HebgenLake earthquake ripped off in 1959, and Iswitched into geology. That really got meinterested. We students went up to theHebgen Lake area and saw the aftermathof this major earthquake, including faultmapping and scarp measurements.

YS: You weren’t here at the time? Youdidn’t experience the quake?

RS: No. I was just finishing a summergeology field course in southern Idaho. Ataround midnight the ground started shak-ing as we said, “It’s a big earthquake.” It’swhat really got me interested in this mix-ture of geophysics—a combination ofphysics and geology. I also like the bio-logical side of things because I started outdoing that in Yellowstone. I went on andgot degrees in geology, a Ph.D. in geo-physics, and I started doing lots of otherthings; I went to pilot training in the AirForce, I put seismographs all over Europeto snoop on Russian nuclear testing, and

Windows into YellowstoneAn Interview with Geologist and GeophysicistRobert B. Smith

Bob Smith

Fall 2000 3

Figure 1. Space view of Grand Teton andYellowstone national parks from satelliteimages overlaid on digital elevation maps.The 8,000-foot-high Yellowstone calderawas produced by a giant volcanic eruption630,000 years ago. The caldera occupies a45-by-30-mile-wide area of centralYellowstone. The Teton fault bounds theeast side of the Teton Range and raised themountains high above Jackson Hole’svalley floor. (Image by E. V. Wingert.)

Figures 1, 2, 3, 5, and 7 for this article arefrom Windows into the Earth: TheGeologic Story of Yellowstone and GrandTeton National Parks by Robert B. Smithand Lee J. Siegel, copyright 2000 by RobertB. Smith and Lee J. Siegel. Used bypermission of Oxford University Press, Inc.

Figure 2. Path of the Yellowstone hotspot. Yellow ovals show volcanic centers where the hotspotproduced one or more caldera eruptions—essentially “ancient Yellowstones”—during the timeperiods indicated. As North America drifted southwest over the hotspot, the volcanism progressednortheast, beginning in northern Nevada and southeast Oregon 16.5 million years ago andreaching Yellowstone National Park two million years ago. A bow-wave or parabola-shaped zoneof mountains (browns and tans) and earthquakes (red dots) surrounds the low elevations (greens)of the seismically quiet Snake River Plain. The greater Yellowstone “geoecosystem” is outlined inblue.

Yellowstone Science4

was chosen as the American exchangescientist to the British Antarctic Surveyand went to the Antarctic—lots of reallykind of wild things.

YS: You spent a large portion of yourcareer working with other folks to ex-pand the state of the knowledge fromwhat was just vaguely recognized as avolcano, a volcanic caldera, in Yellow-stone, to what is now recognized as nearlythe largest one in the world.

RS: The largest active hotspot on thecontinents and maybe in the oceans. Wegeophysically mapped the third dimen-sion, that is, the subsurface geology withdepth, and we studied “extinct” volcaniccenters all the way from Boise, Idaho, toYellowstone.

YS: When you first came to work onYellowstone geology, what was the levelof knowledge?

RS: It was somewhat limited, espe-cially in terms of understanding the vol-canic and tectonic processes in a platetectonic framework. I finished my Ph.D.in 1967 and went to Columbia Universityto do post-doctorate research. There Iwent back through all of their old seismicrecords for western U.S. earthquakes, butfocused on learning more about theHebgen Lake earthquake. I also went tothe University of California at Berkeley,because that was a famous seismologicalinstitute, and said, “I want all your data onthis great earthquake at Hebgen Lake.”We went down in the basement and thisguy said, “They used to be here in boxes.”Didn’t find one…

About a month later a technician calledme up and he said, “Dr. Smith, there’s apile of stuff here under dust and garbage,is this what you want?” It was like findinga gold mine! I now had all the world’srecords for the Hebgen Lake earthquakeand its aftershocks. I started workingwith those. This really heightened myinterest. I started coming up to the HebgenLake area, putting out seismographs andstudying Yellowstone’s fault and volca-nic features in the 1960s.

Mapping of geology in Yellowstonewas initiated in the 1870s by FerdinandV. Hayden, the famous naturalist-geolo-gist whose pioneering work helped getYellowstone named as the first nationalpark. Hayden made a prophetic statementfrom on top of Mount Washburn looking

out across the Yellowstone Plateau say-ing, “This basin has been called by sometravelers the vast crater of an ancientvolcano…” But not many paid attentionto his writings in the sense of the extent ofthe system or the youthfulness ofYellowstone’s volcanism.

In 1922, Professor Jagger at Massachu-setts Institute of Technology rode throughYellowstone on horseback on his way toHawaii, where he founded the HawaiianVolcano Observatory. He observedYellowstone’s geology and topographyand made a famous statement, “Anyonewho has spent summers with pack-trainin a place like Yellowstone comes toknow the land to be leaping…The moun-tains are falling all the time and by mil-lions of tons. Something underground isshoving them up.” He recognized thatYellowstone was a dynamic geologic sys-tem. Then after another long hiatus, aPh.D. student named Joe Boyd of Harvardin the late 1950s mapped and outlined thedetail of the Yellowstone caldera.

But it was in the mid-1960s that amodern and major effort to studyYellowstone’s volcanic system was initi-ated by the USGS and funded by NASA,which was training astronauts to go to themoon. They were searching for placesthat had moonish, volcanic rocks—Cra-ters of the Moon, deserts. They funded theUSGS research on geology of Yellow-stone because it was a big volcanic center.In the mid ‘60s, they were doing mappinghere and we were starting our first instal-lations of portable seismographs for earth-quake studies. That’s when I met up withBob Christiansen of the USGS VolcanoHazards Branch. He and I became friendsand close colleagues because our researchreally dovetailed together. I would de-velop some new information that wouldfit his ideas and vice versa. The integra-tion really paid off. The sum of the two ofus was much greater than individualsworking alone. Also, Dave Love of theUSGS, who had mapped in and aroundYellowstone, collaborated with me, start-ing on fault and earthquake studies in thelate ‘60s (Figure 1).

YS: You were at the University of Utahby then?

RS: Yes. I had started doing earth-quake installations and detailed faultmapping in Yellowstone about 1967. The

USGS installed the first permanent seis-mographs in 1973. It was also then thatwe got our first research grants. We putportable seismographs all over Yellow-stone and the Tetons. We started our firstsurvey at Norris, then studied the HebgenLake fault zone near West Yellowstone,Yellowstone Lake, and the BeartoothPlateau; then we went down and did theTeton fault. These were part of a long-term plan to analyze the Quaternary faultand volcanic history of the region.

We also built a boat to do seismicprofiling, bottom-sediment coring, andheat flow measurements. From this ves-sel we ran seismic profiles of Yellow-stone Lake and Jackson Lake. The pistoncores allowed us to determine the com-position and ages of the lake sedimentsfrom which we subsequently determinedthe first estimates of the past 7,000-yearhistory of Yellowstone Lake. BobChristiansen and his colleagues were alsoputting together the volcanic frameworkof Yellowstone at the same time, and BobFournier, also of the USGS, was doinghis hydrothermal work along with DonWhite. Our data and ideas all came to-gether roughly at the same time in theearly ’70s.

I wrote a couple of papers in 1974 thatdescribed the properties of Yellowstoneas a “hotspot.” Remember, plate tecton-ics didn’t come into vogue until the late‘60s, so there was no framework to eventhink about a hotspot until 1972 when aPrinceton geologist plotted all the Earth’svolcanic centers and recognized their pat-tern relating to plate motions.

YS: Is the hotspot considered to becontiguous with the caldera?

RS: “Caldera” is a Spanish word for acooking pot called a caldron. When alarge volume of magma is removed frombeneath a volcano, the ground subsidesor collapses into the emptied space, toform a depression called a caldera. Theycan range in size from a kilometer to tensof kilometers long, like Yellowstone’s.“Hotspot” is a term used to denote an areaof concentrated volcanism on the earth’ssurface with a deep mantle source ofmagma and heat. As the ascending mol-ten rock migrates through the earth’smantle, some of the magma gets en-trained on the base of the overlying plate,while part of the magma leaks upward

Fall 2000 5

into the crust, melting surrounding rocksand creating a shallow heat source. Thatmagma feeds Yellowstone’s magmachambers whose tops are located at depthsof about 8 to 10 km and extend to depthsof about 16 km.

This magma in turn providesYellowstone’s immense heat flow. It isnot so much that Yellowstone’s groundtemperatures are high, but it is the flow ofheat coming out the earth’s surface that is30 to 40 times higher than the heat flow-ing anywhere else in North America.Yellowstone is like an immense heat ra-diator.

There aren’t a lot of big calderas around.Toba in the southwest Pacific is an ex-ample of a large caldera about the size ofYellowstone’s. However, it’s poorlyknown because it’s so remote. As relatedto the giant eruptions and the calderasthat occur elsewhere in the world, theYellowstone caldera is a giant—50 kmlong by nearly 40 km across. This is thedimension of the roof that collapsed intothe magma system.

YS: Tell me about mapping the youngYellowstone caldera and pursuing thebigger picture of how it relates to oldervolcanic activity across the western U.S.

RS: You can’t just study Yellowstonein the context of nothing else. You haveto do it in terms of how it fits into theworld. I have prepared a map of thelocations of the Yellowstone and theSnake River Plain calderas, the oldercalderas along the track of the hotspot(e.g., the shift in the relative position ofthe hotspot as a result of continental drift).The map also shows how the topography,earthquakes, and faults were related tothe Yellowstone hotspot track. That’swhen we first began thinking about theoverall pattern of the effects of the hotspoton the surrounding area and its evolution,but more importantly, how it created thevolcanism, earthquakes, and how it tiedto the faulting—the energetics of a hotspot(Figure 2).

The USGS had mapped most of thepieces of Yellowstone by 1970. And apaper on a global hotspot and plume waspublished by a professor at Princeton in1973. I published two papers in 1974 inwhich I described the effects of the Yel-lowstone hotspot and its volcanism. Aprofessor at Yale, Dick Armstrong, first

noted “old Yellowstone” volcanic cen-ters along the Snake River Plain. Hedated volcanic rocks, rhyolites, scatteredalong the floor of the Snake River Can-yon; the rocks get older and older downthe Snake River Plain from Yellowstone.But they were buried beneath the youngbasalts. Now, the Snake River Plain is abroad topographic depression, and wereasoned that there had to be mountainsthere before. You just don’t blow awayRocky Mountains. Something destroyedthem. Destruction is a product of explo-sion plus foundering of the mountainroots back into the magma system.Armstrong showed the progressive ageof the volcanic rocks, oldest in south-western Idaho and northwestern Nevada,and youngest in Yellowstone.

Plate tectonics had just hit. So in 1972I calculated the North American plate’sinteraction with the Pacific plate. I said,“Let’s assume that the source of Hawaiivolcanism is fixed deep in the Earth andcompare how Yellowstone relates to Ha-waii.” The model predicted the south-west motion of the North American plateat Yellowstone. Its trace, the 800-kmtrack of the progressively younger volca-nic centers of the Snake River Plain to-wards Yellowstone, fit this model of aplate overriding a magma source anchoreddeep in the earth, sometimes called aplume. Hawaii is over a hotspot beneaththe Pacific plate, and we are over a hotspotbeneath the North American plate.

Armstrong dated the rocks by potas-sium argon methods and showed that theoldest were to the southwest and theyoungest to the northeast. That gave aplate velocity of 4 1/2 centimeters peryear of movement to the southwest. Icalculated the motion of the North Ameri-can plate, using Hawaii as a referenceframe, and I added in some extension,and it fit Armstrong’s data within themargin of error. The light went on: thevolcanic activity at Yellowstone is froma fixed Earth’s mantle, and the progres-sive volcanic ages are just the record ofthe plate motion across this source. Youcan’t see old Yellowstone calderas toowell in the Snake River Plain becausethey got covered by younger basalts. Butyou can infer roughly where they arebecause of the ages of the rhyolites thatare mapped. For the same reason, most of

the Yellowstone caldera has all been cov-ered up. Lisa Morgan and Ken Pierce ofthe USGS and Mike Perkins of the Uni-versity of Utah have detailed the volcanichistory of the Snake River Plain anddelineated the details of many of its vol-canic centers.

Thus we determined the track of theplate over the hotspot, and from seismicdata we determined the size and locationof its deep magma system. The magmathat makes up the spread-out hotspot onthe base of the plate is only 3 to 6 percentmelt; the rest is solid rock. By integratingall available data from geophysics, geol-ogy, mapping, and dating, it all started tofall together.

YS: And the plate—where Yellow-stone currently is—moves relative to thishotspot of liquid rock under the Earth’ssurface.

RS: You have to think about the frame-work. The mantle is fixed. The magmacomes up and interacts with the overlyingEarth’s plates that are moving. It’s likemoving your hand across a burning candle.The flame leaves a line of burns on yourhand and, if you leave it there long enough,the candle flame burns a hole throughyour hand.

Beneath southern Idaho, we’ve had acandle flame made up of magma burningupwards into the plate moving over it,starting down in the Boise area 16 millionyears ago. And the plate has moved fromnortheast to southwest since then. Theyoungest rocks are at Yellowstone (Fig-ure 3).

YS: And so what is cold now, the rockthat you map, was once a hot piece ofrock.

RS: There were old Yellowstones allthe way from Boise up here, but they arenow inactive, cold and buried beneathbasalt. Of course, the surface rocks inYellowstone cool rapidly after exposureto the Earth’s atmosphere.

YS: And it all became rhyolite, the rockwe often see on the surface of the park?

RS: At the Idaho National EngineeringLaboratory near Idaho Falls, 15 or soyears ago, they wanted to learn about thesubsurface geology beneath the site. Theydrilled a deep borehole, and low andbehold, they drilled through surfacebasalts into rocks that we call rhyodacites.These are similar in composition to

Yellowstone Science6

Figure 3. A cross section of Yellowstone reveals molten rock under the caldera at depths of about three to eight miles. Heatemitted by the molten rock powers Yellowstone’s geysers and hot springs.

Figure 4. Crustal deforma-tion of Yellowstone fromGPS measurements. Three-dimensional GPS stationvelocities for Yellowstonefrom 1987–95. Arrows showhorizontal velocity vectors atstations; color contoursrepresent vertical velocities.Large arrows indicatedirection of regionalextension across theYellowstone volcanic field.Courtesy Bob Smith,University of Utah, Yellow-stone Hotspot Project.

-20

-15

-10

-5

0

5

10

-10

-10

-10

-5

-5

-5

-5

-5

-5

-5

-5

-5

-5

-5

-5

-5

0

0

0

00

0

0

0

0

0

00

0

0

5

111˚ 110˚

44˚

45˚

Vertical Velocity (mm/yr),Contours 5 mm/yr

Yellowstone National Park

10 km5 mm/yr

Regional extensiondirection, ~5 mm/yr

5 mm/yrRegionalExtension

Fall 2000 7

Yellowstone’s rhyolites. The rhyoliteshave been covered by younger basaltsthat you see when you drive south fromAshton to Idaho Falls to Pocatello. Theyare ragged black rocks that make up thesurface. The inference then, from theseand our geophysical data, is that the rhyo-lites actually make up a much thickercomponent of the Snake River Plain butthey are buried by the basalts, which arejust a thin layer at the top.

YS: You moved on, to monitoring themovement of Yellowstone—the breath-ing of the caldera, so to speak.

RS: We got into crustal deformation. Irecognized from a return visit to the southend of Yellowstone Lake in the early ‘70sthat there was something strange goingon here; things didn’t look right. Thetrees at the shoreline appeared to be inun-dated by rising lake water, and parts ofPeale Island, where I had worked in 1956,were under water. I reasoned that the lakewas tilting to the south, inundating itssouthern reaches and uplifting its north-ern parts. This effect would also increasethe height of land at the north end of thelake, rising and expanding the beach be-hind the Fishing Bridge Visitor Center.We were witnessing the effect of a tiltingtoward the south of a bathtub ring, itsshoreline, around Yellowstone Lake.

It was then that I realized that if we didprecise measurements of the elevation ofbenchmarks originally established whenroads were built in Yellowstone in 1923and 1934, and we went back and re-observed those marks, we could see ifthey had moved vertically or not. Wewere contracted by the USGS and, withone of their crews, we surveyed and com-pared the data for three summers. Ourfirst year we went across the caldera fromCanyon to Lake. Our surveyor had theoriginal surveyors’ notes from 1923, andhe said to me, “There’s something reallywrong here—we’re way off from theirelevations, I mean, we’re like a foot anda half off.”

I thought, boy, we’ve done somethingwrong; we’ve got to go back and redo oursurvey. No, we were doing even moreprecise surveying than the 1923 surveys.Excluding the errors that could have beenin the 1923 survey, we showed that thiswhole portion of the Hayden Valley wasgoing up. We ran a survey line from West

Thumb to Old Faithful to Madison Junc-tion, and we also went across the oldroad, from Nez Perce Creek over the topof Mary Mountain to Hayden Valley.That is how we connected the two pro-files together. Altogether these measure-ments revealed that the Yellowstonecaldera was rising, like a giant bulgingstomach of a breathing creature.

This unprecedented discovery revealedwhat I called a living caldera. It had risen75 cm—3/4 of a meter over a calderathat’s 50 kilometers long. It really wasunprecedented, seeing deformation thisbig, greater than most anywhere within acontinent that we knew of with the excep-tion of active volcanoes such as Rabual inthe southwest Pacific and the Phlegreanvolcanic field near Mt. Vesuvius vol-cano, Italy. Continued leveling of thepoints by Dan Dzurisin of the USGS andour new GPS measurements, however,showed a cessation of the caldera uplift,returning to subsidence about 1985 (Fig-ure 4).

It was at the same time that we receivedan NSF grant to employ the new technol-ogy of Global Positioning Systems (GPS)to study Yellowstone. With this newmethod we didn’t have to be on roads andwere able to go all over the backcountry,essentially putting a grid of GPS bench-marks across Yellowstone. And we re-observed them every other year, from ‘87to ‘95. These measurements revealed thatthe caldera had indeed reversed motionand began moving down at about 1.5 cmper year, at nearly the same rate as theuplift and over the same uplifted area.What, is this thing breathing? We wereall excited about that. In looking at our‘95 survey data, we noticed things werestarting to bottom out. A couple of GPSstations around Old Faithful and LeHardyhad come back up a little bit. But wedidn’t have any more money to continueour study. Starting in the mid ‘90s, WayneThatcher and Charles Wicks of the USGSused Interferometric Synthetic ApertureRadar (InSAR) that suggested the calderabegan rising in 1994—another majorchange in the caldera dynamics, althoughour new continuous GPS data up to sum-mer 2000 do not corroborate this uplift(Figure 5).

Nonetheless, we were lucky to haveseen a giant caldera change from a period

of uplift to subsidence, and perhaps an-other uplift in our lifetime. In a paralleleffort, we studied the most intense earth-quake swarm in Yellowstone’s recordedhistory and found that the earthquakesoccurred at the greatest rate during thechange from caldera uplift to subsidencein late 1985 and continued into 1986. Sowe have this great correlation of earth-quakes and changes in crustal deforma-tion. I then coined the phrase a “living,breathing caldera.”

YS: Do you have any idea, from yourwork done here or elsewhere, whetherdeformation has anything to do with thepredictability of volcanic eruptions?

RS: We’d like to think it does…inHawaii, where they have predicted erup-tions on the basis of earthquakes, theycan see the correlation of earthquakesrelated to migrating magmas which even-tually erupt to the surface. But those arebasaltic magmas—they flow much faster,they’re not as explosive as Yellowstone’smuch more viscous rhyolitic magma andthere’s no precedent, no historic exampleto understand this behavior.

Eruptions in Rabaul, New Guinea, werepreceded by uplift and subsidence andunusual periods of seismicity. On theother hand, there was no eruption in ornear the Bay of Naples, Italy, during theperiod that land rose and subsided severalfeet, so that at one time some beautifulItalian buildings were buried under thewater. Now they’re back out of the water.They came up in the 1950s and ‘60s,when the ground did a lot of huffing andpuffing—you know, it’s also a caldera.It’s erupted before. And of course,Vesuvius is nearby with three millionpeople living in the area.

We’ve had to envision analog modelsfrom the basaltic cases as a working modelfor Yellowstone’s rhyolitic eruptions. Therhyolitic magma would be more viscousand retain fluids and gases, causing upliftand subsidence with changes in pressure.Another reasonable model is one withlarge volumes of hydrothermal fluids un-derlying the caldera—the ones that feedits geysers, hot springs, fumeroles. Thesemore easily running fluids pressurize theirchambers, uplifting the ground then drain-ing out its sides, and dropping theground—a mechanism that also explainsmechanics and numbers of earthquakes

Yellowstone Science8

that we observed in the 1985 northwestcaldera swarm. Above all we are dealingwith a large rhyolitic volcanic system fedby a hotspot. Nowhere else does such afeature exist on a continent.

YS: So right now we’re in a period ofuplift again?

RS: Perhaps we are back into uplift.That does not necessarily mean a pendingvolcanic eruption. But remember, we’vehad 30 or so smaller but still explosiveeruptions since the last giant eruption630,000 years ago, the youngest only70,000 years ago that occurred on thePitchstone Plateau. It was however a gi-ant, catastrophic eruption that created theYellowstone caldera and blew ash allover much of the West. These post-calderaeruptions were smaller, but were tens tohundreds of times bigger than the Mt. St.Helens eruptions. And, based on BobChristiansen’s USGS work, they in-creased in frequency around 125,000years ago. But there have been no volca-nic eruptions for 70,000 years.

We’ve also looked carefully at the align-ment of Yellowstone’s post-caldera vol-canic vents that line up northwest-south-east. These are smaller volcanoes, alongwith our new earthquake epicenters in thecaldera, and they both line up; they’resitting there parallel one to another. Wethink the vents are along vertical dikes, ifyou wish. These are active magma sys-tems just below the surface, and theycreate earthquakes, and they create vol-canoes.

At the time of the big earthquake swarmin 1985, we called the situation to theattention of the Park Service. The earth-quakes were coming at a very high rate.This was on the northwest side of thepark, just beyond the caldera. It began inOctober, peaked about the first week ofDecember, and continued through March1986. We studied the sequence very care-fully. The earthquakes progressed fromthe caldera outward to the northwest andgoing deeper as the sequence progressed.We interpreted the earthquakes as relatedto motion of fluid along a vertical dike,propagating fluid to the northwest.

It doesn’t have to be magma to causethis effect; hydrothermal fluids may havebeen the responsible mechanism. My im-pression is that it could have been watermoving outward from the caldera, leav-

ing it to subside as supporting fluids wereremoved. As the caldera is subsiding it’sgot to get rid of the volume at about 0.02cubic km per year. That means you’retaking a volume roughly the size of Mam-moth—the hot springs terraces, or theentire Mammoth developed area, wideand high. Interestingly, that is about therate that magma would need to be in-jected into the crust to create the calderauplift from 1923 to 1985 and to sustain itshigh heat flow.

And then the earth’s surface startedgoing down. The earth doesn’t let youpush fluid back down in it, so we surmisethat it is being squeezed out the sides.We’ve hypothesized that caldera fluids,either hydrothermal or magma, could bemigrating radially outwards from thecaldera along dikes or vertical sheets offluid. But Yellowstone’s a big place. Thestuff could leak out, especially if there isa lot of gas in it, and you may never seethem if they are hydrothermal fluids. Thatmechanism is something reverse to up-lift.

One of the nice things about our obser-vations is the synchronicity of the upliftand subsidence between Hayden Valleyand Old Faithful. They’re 20 miles apart,yet they go up and down together. Thatimplies you have a connected plumbingsystem. A “pipe” from Hayden Valleymust be connected to the Old Faithfularea at depth. That’s probably the top ofthe magma system. So where we mappedthe magma, which is actually a partialmelt, that’s the magma chamber. Thisbody gives off heat that’s coming up andcreating the high heat flux, and it is what’sheating the groundwater that makes thegeysers. This system must extend undermost of the caldera. But it shallows underthe southeastern corner and the north-eastern caldera. And one place where itseems to come shallowest, northeast ofSour Creek, is north of the Hot SpringsBasin country, where it looks like there’sa connection of these low-velocity magmabodies above the surface and a shallowhydrothermal system (Figure 6).

YS: Does the shallowness of the magmanecessarily relate to where a next erup-tion might be?

RS: Oh, I think it would. You’ve gotthe two main pods in the middle, thesoutheast and northeast beneath the

domes; if I’d be laying bets, I’d be think-ing it probably wants to come up in thenortheast. On the other hand, the geo-logic mapping shows the oldest post-caldera flows in the northeast, and theyget progressively younger toward theMadison Plateau. So if you count on thepast 150,000 years of volcanic history,you’d say the biggest potential is in thesouthwest plateau, along the caldera’ssouthwest rim.

But I would look to the geophysicalevidence, such as seismic images ofmagma and of where the earth’s surfaceis moving from GPS measurements, andsee if magma or hydrothermal fluids maybe coming up on the northeast side. Wejust don’t know the physics of these flu-ids well enough to predict that.

YS: You’re talking about the shallowmagma under the Mirror Plateau, thenortheast edge of the caldera, and yet thehottest spot is under Norris Geyser Basin.

RS: Norris Geyser Basin has the hot-test water reservoir temperature, but it isonly probably 1 to 3 km deep. You haveto differentiate between temperature andheat. You can have a concentrated blob ofhot, hot water that’s 450° to 600° Fahren-heit in a geyser reservoir, but it’s anisolated body with a high temperature.Whereas the area of the caldera, on aver-age, has a higher release of heat per unitarea—that’s heat flux.

YS: Now, describe how you map thismagma. How long does it take to do that?And theoretically, can you retake a snap-shot over time?

RS: We use the methodology used inmedicine called CAT or MRI scans to dothe same for the Earth, but using earth-quake recordings of seismic waves pass-ing through the earth. CAT scans are justa way of sending x-rays into the body,and they get reflected back from differentparts of the body to produce an image ofthe body. X-rays transmit easily throughsoft tissues, but harder material like bonesmore easily reflect the rays. You’ve hadCAT scans, right? A radiologist takes hislittle device and puts the gel on your bellyand moves it around. He’s sending rays inso he gets coverage. Lots of rays gothrough the whole volume and you getgood coverage, then they are broughttogether in a computer to create a pictureof your internal organs.

Fall 2000 9

Figure 5. Earthquakes of the Yellowstone-Teton region. Epicenters of earthquakesfrom 1973 to 1996 are shown by red dots.Most of the quakes were under magnitude 5.The most intense earthquake activity is inthe northwest corner of Yellowstonebetween Norris Geyser Basin and theHebgen Lake fault. The Teton fault now isseismically quiet. Active faults are shown asblack lines and post-caldera volcanic ventsas orange stars.

Figure 6. The Yellowstone magmachamber. Cross-section of theYellowstone caldera from seismicimages of the P-wave velocity usinglocal earthquake tomography. Itreveals the location of magmachambers beneath Yellowstone. Themagma chambers are composed ofpartially molten rock containing 10–30 percent melted rocks. Warm colorat depths of 8 to 16 km are hot rocks,blue colors are cold rock (from Millerand Smith, 1999).

Saturated andoverpressuredfracture zone

4.0

5.0

6.0

3.4-

6.8+

Vp (km/s)

74 0

16

080km

km

0

37

8

40

Depth (km

)

N

Partial melts?

?Outline ofgranitic body

16

8Thermally

undisturbedbasement

??

Yellowstone caldera boundary

Dome

Dome

Yellowstone Science10

Geologists apply the same method thatwe call tomography—seismologists de-veloped it before the medical professiondid. We use seismic rays that go into theearth. When the seismic wave encountersa hot rock, its speed of propagation slowsdown; when it encounters a cold rock, itspeeds up. So if you have enough earth-quakes in a region, recorded on enoughseismographs, then you can reconstructthe ray paths of where they’re fast andwhere they’re slow. That’s what we’vedone for Yellowstone’s upper 15 km. Wemade a three-dimensional image of itsstructure. This is the method that we usedto prepare the figure of Yellowstone’smagma system.

We found magma, here 10 to 30 per-cent of melted rock, in a porous space ofsolid rock at depths as shallow as 8 to 10km beneath the Earth’s surface. It ex-tends the length of the caldera with aconduit aimed toward the surface at thenortheast side of the caldera. We do nothave data that will provide the same kindof images deeper into the lower crust. Butwith our new NSF project focused on thedynamics and detailed mapping of theYellowstone hotspot, we hope to probe asdeep as 1,000 km beneath the surface andmap the magma conduit all the way fromthe hotspot to the surface.

YS: And you do this taking advantageof the natural seismicity?

RS: We use the naturally occurringearthquakes. In 1978–1980, we recordedseismic waves generated by explosivesin drill holes along the Snake River Plainto study the track of the hotspot all theway from Twin Falls to the BeartoothPlateau. We used our own earthquakes(we made them very small magnitude).And we also relied upon natural earth-quakes. Natural earthquakes pose theproblem that you don’t know exactlywhere they are. So we have to calculatethe location of the earthquake plus thevelocity field. It’s a more difficult math-ematical problem. But if you have a fewplaces where you’ve got a controlledsource, like an explosion, and you knowexactly where it is, that helps us calibrateit. We just put slices through the velocitymodel, and the low velocities map out thehydrothermal and magma systems com-pared to colder, higher velocity earth.

YS: So your mapping can show not

only the spread of the magma horizon-tally, but the depth. Does it vary a lot?

RS: Not within the chamber, now, that’sthe very-near surface. It seems to be nor-mal down to about 50 miles (80 km). Thatis where magma associated with the realhotspot begins. We do not know if thehotspot magma originates at the core-mantle boundary (at 2,700 km deep) or isthe result of decompression melting orrock at much shallower depths of 100 to200 km deep. Our new experiment shoulddiscern that model.

Regardless, magma is generated in theearth’s mantle. Part of it leaks through theoverlying lithosphere into the crust, melt-ing surrounding rocks and producing amelt that resides in upper crustal magmachambers. This is what feeds Yellow-stone’s volcanism and enormous heatflow. However, most of the hotspotmagma is sheared off on the bottom of themoving plate spreading out to the south-west beneath the Snake River Plain. Weimaged that just recently, all the wayfrom southern Idaho to Yellowstone, andit’s about 100 km deep.

You asked an important question be-fore: do things change with time? Well,geologists say that they do. I proposed anidea to the USGS volcano hazards group.If you went to an active volcano andmagma was moving, as the magma cameup it would heat the rocks around it andslow down the velocities of the seismicwaves traveling through them, so youcould do the tomography in real-time,like a doctor does. As magma ascends itslows down the seismic velocity and cre-ates earthquakes. After the magma passes,the seismicity ceases and the velocityincreases as the rock cools. I couldn’t doit on hourly scales, but I could do it ondaily to monthly intervals and see howthe rocks are affected by heat creatingdifferent types of earthquakes and chang-ing the rock velocity. I’m talking aboutan active volcano like Hawaii or some ofthe Alaskan or the Aleutian ones. Yel-lowstone would not be so practical be-cause it is just so big and does not have thehigher extent of rapidly moving magma.Remember, this technology is brand new.The ideas are brand new. It takes a lot ofcomputing power. It takes modern andreliable real-time data.

YS: Up until now you just mapped it

once?RS: Once. We took one snapshot in

time.YS: But with this new work you hope

to do it on repeat intervals?RS: Right. That’s why adding corrobo-

rating data from such methods as GPS isso important. GPS tells you how fast theground is going up and down or side-ways, due to magma or hydrothermalfluid migration. This motion must bedifferentiated from the overall global platemotions to ascertain how fast the groundmight be moving as it builds up energy onfaults or in its magma chambers. Seismicdata also tells you the geometry of themagma body, so you can actually workout the dynamics. Because we map pres-sures and we can infer from pressures ofthe magma, we can say how fast it’sactually deforming.

We run the Yellowstone seismic andGPS network. It contains 22 seismic sta-tions—18 in Yellowstone and four out-side the park along the Hebgen Lakefault, because they are integral to theinteraction between the caldera and thefault mechanism. The seismic stationscontinuously transmit data by radio linksvia Mt. Washburn, to Sawtelle Peak, andfrom there on an FAA line to our SaltLake City recording laboratory. We alsoemploy satellite telemetry from our co-operative university-USGS Lake stationthat is sent to Golden, Colorado, then onto Salt Lake City via the Internet.

YS: So when you’re down there inUtah, and something is happening here inYellowstone, it instantaneously gives youa picture of what’s going on.

RS: Yes. If there’s an earthquake, ittakes about 10 seconds to calculate themagnitude and location. That is broad-cast to me and others via automatic tele-phone paging systems and sent to ouronline web site.

YS: I felt the Borah Peak, Idaho, earth-quake when I was at Old Faithful in 1983.Why do I recall it took geologists a whileto figure out exactly where the epicenterwas?

RS: Because we didn’t have the fastcomputers then, nor did we expect such alarge quake in central Idaho and did nothave an array of seismographs there. Wedidn’t have the data coming in real time.We have no excuse not to do it now.

Fall 2000 11

YS: Is it a triangulation process?RS: Exactly. It’s just surveying with

seismic waves. But it’s a much tougherproblem, because the land surveyor trans-mits his signals through the air electroni-cally. We have to transmit through thiscrummy earth. There are fast rocks andslow rocks. The surveyor points his eye atsomething and he assumes it’s along astraight line. In the earth, it bends. Wehave to calculate all the bends. This newmethod of tomography allows us to cal-culate the earthquakes in this very hetero-geneous earth.

YS: So, you’re down there in Salt LakeCity, and there’s been an earthquake inYellowstone, and it is “x” magnitude,and here’s the epicenter. What does theGPS network add?

RS: The Yellowstone GPS network ismade up of receivers over benchmarks onthe ground. They continuously recordtransit times of radio waves from GPSsatellites. It transmits these data back toour lab in the same way as the seismicdata. Every thousandth of a second theseismographs are all transmitting data.We sample the GPS every 15 seconds.This type of recording provides an accu-racy at the centimeter level.

All these data are dumped into fullydedicated computers that calculate thecoordinate of that benchmark on the sur-face. We compare the coordinates of thepoint with time to see how fast it ismoving. Now, the majority of the earth’smotion isn’t associated with earthquakes;only about 1 percent or less of the earth’smotion is released as the energy in earth-quakes. Earthquakes are just the creakingand groaning. But the earth is movingacross the hotspot continuously. The restof the motion, we call aseismic motion. Itreflects the slowly deforming earth thatmoves more like silly putty, it is plastic.Plastically it’s not going to create earth-quakes, but it is what records the slowmotions of earth’s processes—such asmagma movement, bending the rock be-fore an earthquake, or uplifting the groundover magma. We subtract out the amountof movement related to earthquakes andget the total amount of deformation that’sdue purely to the plasticity, the volcanicmechanism. We can measure both upliftand horizontal movement with GPS toaccuracies of millimeters now.

NSF just awarded us a multiple-year collaborative research grant,“Geodynamics of the YellowstoneHotspot,” between the University of Utahand the University of Oregon. The objec-tive is to understand how the Yellow-stone hotspot works, how magma getsfrom hotspot to the surface, and how thiseffects the topography as well as how itchanges the pressure on its faults. To dothis, we will conduct a GPS and seismicsurvey of the whole Yellowstone system.We’re going to look at the effects of thehotspot across a broad region from Casperto Boise to Helena to Salt Lake. We’reputting in permanent and portable GPSstations. There’s a permanent GPS sta-tion right over here at the baseball field inMammoth. They are also at Lake Junc-tion and Old Faithful. And we’ve in-stalled two in the backcountry in coop-eration with the USGS, one in the lowerHayden Valley and one on the Sour CreekDome. We’re going to have about a dozeneventually, just like the seismic network.We will operate this network in continu-ous recording, just like our seismographnetwork.

Also in 2000, we’re going to bring inabout 80 portable seismographs and placethem around the Yellowstone hotspotfrom as far away as 200 miles on a 30-mile grid. Then we will do tomography ofthe much deeper earth, just like we did itfor the crust of Yellowstone, and we willbe able to resolve the source and depth ofthe Yellowstone hotspot.

I feel like I’m just an earth internistdoctor who’s running his CAT scan—Ijust do it a little bit slower. We’re goingto record all the earthquakes, record allthe GPS, find out what the structure is todepths of about 1,000 miles. Then we willadd in the data on fault movements andinformation on Yellowstone’s magmasystems and determine from computermodels what to expect on the Earth’ssurface, and perhaps what to expect in thefuture. So we’ll really be able to definethe form of the hotspot. We will use theGPS to measure the ground motion, howfast it’s moving over a big region, not justYellowstone park. Then we will put allthese new observations together with Yel-lowstone volcanic history in a mathemati-cal model and create a mathematical im-age of the hotspot.

We’re going to put in all the faults andlet the them rupture at the rate they wantto. We’re going to let this thing stepthrough time. Probably in an hour ofcomputing we can simulate 10,000 yearsand let things move according to the rateswe see today. We can then try and predictwhat’s happening, where the magma is,how big it is. We’re going to try to calcu-late the magma reservoir sizes, the tem-peratures. We’ll get all the physical char-acteristics we want out of this body. We’rejust doing internal medicine. Same thing.That tells us how active it is. The Yellow-stone hotspot is a global community—remember it’s the biggest one on thecontinents. It’s affected 20 percent of thenorthwestern U.S. in its 16-million-yearhistory. It’s a big feature. It’s much big-ger than the National Park Service.

YS: Tell your story about Peale Island.RS: In the summer of 1956, we had

fish-research stations at Chipmunk andGrouse creeks. I lived at the cabin on thatisland in the South Arm of YellowstoneLake. One incident I remember is that aswe ran out of food—we had a few poundsof cheddar cheese and nothing else; wewere eating fish and cheese. We werecatching about 1,000 fish a day in the fishtraps. We were so sick of eating fish. Inever wanted to see another one.

I said to my partner, “This is enough.I’m through with this.” So we took ourlittle boat over to the shore, then walkedwest in the melting snow and muddyground. It must have been in early June.We hiked about 14 miles to Heart Lake.We were really post-holing in the snow.We thought a grizzly bear would chew usup. And we finally worked our way to theroad over by Shoshone Lake. Someonepicked us up and took us back to Lake.There our boss said, “Here, take somefood and get back to work.”

I’ve been back at Peale Island severaltimes since. I went back with Ken Diem[of the University of Wyoming] in about1974, and he said, “Hey, we can’t parkour boat at the dock.” Well, the boat dockwas partly under water. And many of thetrees around the south shore looked likethey were being inundated. I reasonedthat the only way to do that is to “tilt thebathtub back.” That’s why at FishingBridge you have this emergent beach allthe way over to Storm Point; it’s an

Yellowstone Science12

emergent beach in oceanographic terms.It’s a beach that’s rising because the up-lift of the Yellowstone caldera is centeredto the north, and that process of uplift andsubsidence has no doubt been going onfor thousands of years. The net effect isuplift of the Sour Creek Dome and thesurrounding area, producing the dam-ming of Lake Yellowstone at Le Hardyrapids.

YS: In the short-term, you don’t expectto see the beach disappear and the dockcome back up?

RS: No, I do not. That’s the other thingabout Yellowstone—the caldera is apimple on the overall deformation of theentire Yellowstone Plateau. The calderaitself is moving up and down, but thewhole region up to 300 miles wide hasbeen uplifted 500 meters. The Yellow-stone Plateau goes well beyond the bound-ary of Yellowstone Park—it goes out for200 kilometers or more; it encompassesthe greater Yellowstone ecosystem.That’s a whole region of uplifted topog-raphy that probably wouldn’t be there ifthe hotspot wasn’t there. So you have thebroad uplift of the hotspot that’s veryslow, and it’s different from this littlepimple that goes up and down.

YS: What is the relative rate of seismic-ity compared to other places in the coun-try?

RS: Very high. Yellowstone seismic-ity, including the Hebgen Lake earth-quake, is certainly the highest in the RockyMountains in historic time. If you calcu-late the amount of energy per squarekilometer, it’s higher than anywhere elsein the lower 48 states except the SanAndreas fault and related faults in Cali-fornia. Certainly within the interior of thecontinent it has the highest rate of energyuse.

YS: And yet, the rate of “felt” earth-quakes varies quite a bit from year toyear?

RS: It varies, but when it’s active thereare a lot of felt earthquakes.

YS: How many did we have in 1998,for example?

RS: I think there were 11 or so. Butback in 1985, there were 30 or moreearthquakes over magnitude 3.5. In 1995,they were being felt pretty routinely.When we had the swarm on July 3rd, Ithought, “Wow, July 4th is going to be real

fireworks.”YS: In your book you get into questions

of emergency preparedness.RS: We point out the need for pre-

paredness planning in the sense of theawareness of its potential volcanic andearthquake hazards. I have suggestedpeople prepare emergency response plansaccordingly.

YS: Is that based on projected trends ofan increasing rate of seismicity?

RS: No. We’re saying that all the agen-cies, the Park Service, Forest Service, thesurrounding communities, should beaware of potential geologic disasters thatcan happen in time frames that they’reresponsible for and should be planningfor. Most people in emergency manage-ment deal with a 24-hour clock or, at best,about a year ahead, as far as budgets areestimated. But remember, we have hadthe largest historic earthquake in the In-termountain West, the magnitude 7.5Hebgen Lake earthquake that killed 28 in1959. This gives us an idea of what toexpect in the future.

FEMA (the Federal Emergency Man-agement Agency) considers both shortand long-term effects. The volcanic orthe earthquake threat for Yellowstone is

very low, in a human time frame. Thepublic has got more important things toworry about, like getting creamed on theroad or having the stock market fall. Butthe agencies ought to take, I think, a muchlonger-term view, that says we realizethere’s a much lower probability, butwhen it does happen it can be catastrophic,beyond things you’ve even thought about.

YS: This long-term uplift wouldn’t nec-essarily be associated with a greater like-lihood of a more serious event?

RS: We just don’t know. The Yellow-stone deformation field is a situation likethat of a blind man coming up to theelephant. He’s never seen an elephantbefore. He touches this thing and he feelsit breathing and he says, “What is this? Isit an organism? Is it a tree that’s mov-ing?” We (the scientific community) havenever seen an eruption or a major earth-quake inside of a caldera in historic time.So we cannot say what to expect, but wecan wisely estimate its effect by extrapo-lating observations from other volcanoesand earthquakes and using the geologicrecord to estimate the rates of occurrence.These data, along with real-time seismicand GPS observations, will provide uswith a good working model and ideas of

Figure 7. Depths of volcanic ash that could be deposited by future calderaeruptions (gray) and by smaller eruptions (white oval) at Yellowstone. Prevailingwinds would determine actual ashfall patterns. Contour lines show ash depths infeet. (Michael Perkins.)

00

Estim

ated

ash

thic

knes

s (fe

et)

Giant, cataclysmic eruption

0.1

10

1

0.1

3

0.3

0.03

Downwind distance (miles)

3

9

6

15

125 250 375 500

Small eruption

12

18

Area of observed ash fallout

Estimated ashfall thicknesses from giant volcanic eruption

Estimated ashfall thickness from small volcanic eruption

20

1 0.3

Fall 2000

the expectations about precursory earthactivity.

When we first discovered the uplift,people said, “Oh, boy, Yellowstone’s inuplift and if it keeps uplifting it’s gonnablow away.” I’m very careful, and Ithought, well, we don’t know. We saw a10-year period of uplift, subsidence, anduplift. We’ve seen a complete cycle ofsomething. We don’t know what the some-thing is yet.

YS: You’ve said there was a 0.01 per-cent chance on an annual basis of either avolcanic eruption or a 7.5 earthquake.

RS: The actual probability is even lowerthan that. I was calculating the groundmotion. People want us to predict things.Well, we can’t predict things, we canpredict the effects of things. And theeffect of things that is most easy to predictis how the ground is going to move. So Ipredict the ground motion by predictingthe acceleration of the ground. I can’tpredict when the fault’s going to go off.But I can predict that if the fault goes offit’s going to shake the ground over here acertain amount. Volcano prediction hereis so far in its infancy no one knows whatto predict. If you look at Hawaii, you cansee that preceding so many eruptions theground was slowly moving. There theyhave nice, runny basalts. And they have alot of seismographs. They can actuallysee the earthquakes coming up with themagma and the ground rising. When theseismologists see anything unusual orstarting to change, they radio the scien-tists working in the field to get them out.And they get people out. You can’t dothat in a rhyolite system because themotion is far too slow.

YS: In terms of emergency prepared-ness, then, you can’t really tell us what’sgoing to happen.

RS: We can tell you what will probablyhappen in a time frame of, at best, days,but mostly in months to years. We cangive you a deterministic view—a sce-nario of the worst thing to expect.

YS: And how soon in advance of anevent do you think you could do that?

RS: Oh, I could give you a scenariotoday—here’s what could happen with abig eruption, a little eruption, and a tinyeruption. And I could say, “Give theseideas to the emergency management folksand plan around these scenarios.” I ask

13

the questions such as: Do you have built-in road escape? What about when some-thing happens in the middle of Yellow-stone and all the roads/canyons are closed?How likely are accompanying landslides?How vulnerable are medical facilities?Are outside groups prepared to assist?What are you going to do with 30,000people on Sunday night during a busysummer season?

YS: When you say a “tiny” eruption,you’re not really talking little, are you?

RS: I’m talking the size of a Mt. St.Helens eruption at the smallest, to maybean eruption 1,000 times bigger. Or per-haps it may be a phreatic or a pure steameruption. These do not have magma;phreatic eruptions are hot water and steameruptions that, for example, blew out MaryBay and Indian Pond on the north side ofYellowstone Lake.

YS: If we were to have even one ofthose little eruptions, would we have no-tice in terms of hours? Weeks? Years?

RS: I think we’d have notice in termsof weeks, if they’re rhyolite. Perhapsshorter for basalt or phreatic eruptions,with a context of a modern seismic array,modern GPS, and bringing in thegeochemists who can study the chemis-try of the fluids. The USGS was doingchemical monitoring here, and theystopped it because of budget cuts. But acombination of monitoring would prob-ably give you reasonable lead-time, onthe order of days, weeks, months, be-cause these things are slow. They’re big;they’re catastrophic in the sense thatthey’re this gooey stuff. They build up somuch pressure that when they finally gothey’re really explosive.

YS: If it were one of the big ones,wouldn’t the scale of it be so large thatone could argue that you couldn’t beprepared anyway? You’d have to evacu-ate the entire western U.S. It’s the end ofthe world as we know it (Figure 7).

RS: You’re right. If it was a cata-strophic caldera-forming eruption, yeah,like, who cares? Well, it would certainlycreate a globally significant change.You’d have pyroclastic flows from thevolcanic vents destroying and cookingeverything in their way for tens of milesfrom the volcano. In the surrounding area,you’d have 10 to 20 feet of ashfall thatcould decrease in thickness but could

extend for hundreds of miles. What doyou do with 10 feet of snow? Imagineturning it into ash—it ain’t going to melt!

YS: Even in Salt Lake you’d get a footof ash.

RS: A foot. Imagine a foot down inthose clogged freeways.

YS: Why won’t any of you even specu-late on the next giant earthquake or volca-nic eruption?

RS: Because we don’t have a basis fortheir understanding yet. This whole sci-ence is so new, remember I’m the blindman coming up to the elephant. I finallyfigured out that the elephant is alive. I’vekind of got its dimensions. I walked fromone side to the other. And I’ve probablyfigured out it’s an elephant. But I don’tknow if it’s standing up ready to fall onme, or if it’s laying down breathing, or ifit’s a rogue or what. I don’t know if it’strained or if it’s wild. So, we’re justlearning. Yellowstone and rhyolitic vol-canism and the relationship to big earth-quakes are so unique that we don’t havea basis of experience to build on.

YS: Someday, when the hotspot is un-der Billings or wherever, Red Lodge,what’s Yellowstone going to look likethen?

RS: I would guess first it’ll look likeIsland Park: lower elevation, much lesshydrothermal activity, no geysers. It’lldie away. Then it’ll look like Ashton,Idaho. Then you’ll start growing potatoeson it! See how the topography of theSnake River Plain falls away to the south-west? The hotspot has raised the groundup here; it’s moving north, but behind itthe land is collapsing in, creating a lowerelevation and a depression. That fills inwith basalts. And the basalts then pro-duce the soils and the soils produce pota-toes.

YS: Where do the basalts come from?RS: The basalts are derived from the

hotspot. They are the last thing that comesout of it. They’re going to be more likeHawaii eruptions. They’ll be exciting andthey’ll be on television, but they’re notgoing to kill a lot of people.

YS: We don’t have to quite worry aboutmoving the Old Faithful Visitor Centeryet.

RS: No, it’ll move itself eventually.You’ve got to get a new one anyway.