{"id":9849,"date":"2021-08-02T22:33:30","date_gmt":"2021-08-03T06:33:30","guid":{"rendered":"http:\/\/earthjay.com\/?p=9849"},"modified":"2025-07-19T23:37:32","modified_gmt":"2025-07-19T23:37:32","slug":"earthquake-report-m-8-2-near-perryville-alaska","status":"publish","type":"post","link":"https:\/\/earthjay.com\/?p=9849","title":{"rendered":"Earthquake Report: M 8.2 near Perryville, Alaska"},"content":{"rendered":"

A few days ago, I was passed out on my couch (sleep apnea) and for some reason I awoke and noticed that I had gotten a CSEM notification of a large earthquake offshore of Alaska. Well, after looking into that, I sent my boss, Rick, a text message: “8.2.”
\nhttps:\/\/earthquake.usgs.gov\/earthquakes\/eventpage\/us6000f02w\/executive<\/a>
\nRick Wilson runs the tsunami program at the California Geological Survey (CGS)<\/a> and works with the California Governor’s Office of Emergency Services (Cal OES)<\/a> to use official forecasts of tsunami size from the National Tsunami Warning Center (NTWC) to alert coastal emergency managers about the level of potential evacuation that they may want to act upon.
\nMore about this process can be found here<\/a>. Take a look at the CGS Special Report 236 to learn about the Tsunami Playbooks and the “FASTER” approach for tsunami evacuation guidance. Evacuation is something that is done at the local level, so CGS and Cal OES can only provide recommendations.
\nNeedless to say, we were both at the ready to respond. Rick has hourly phone calls with the NTWC and follows up with phone calls and emails to specific interested parties (e.g. the emergency managers). We each went into tsunami response mode. I manage the Tsunami Event Response Team, which may be activated to collect observations of tsunami inundation or ocean currents.
\nI started looking at tide gage and DART Buoy data to see how large the tsunami was in the epicentral region. The M 8.2 was in the region of the 1938 M 8.2 earthquake which generated a transoceanic tsunami. I also looked into the literature about the 1938 tsunami, to see what size that tsunami was. The 1938 tsunami had a decimeter scale wave height (peak to trough) for gages in Alaska and in California (Johnson and Satake, 1994). Jeff Freymueller et al. (2021) had also recently worked on the 1938 earthquake source area and tsunami modeling as well.
\nThe nearest tide gage for this 2021 event is at Sand Point, but the nearest gage in 1938 was in Unalaska. So, in order to get a modest comparison between 1938 and 2021, I felt a need to wait for the Unalaska data to trickle in. This may give us some idea whether the 1938 tsunami recorded in Crescent City and San Francisco might be a decent analogue. Of course, we need to get the official forecast from the NTWC prior to sending out any information. But, that process can take hours (over 3 hours in this case). So, we need to get our minds wrangled around the possibilities in the absence of more information.
\nEarthquake and Tectonic Background:<\/font><\/strong>
\nThe plate boundary in the north Pacific is a convergent (pushing together) plate boundary where the Pacific plate on the south ‘subducts’ northwards beneath the North America plate on the north. The Alaska-Aleutian subduction zone forms a deep sea trench which can be seen in maps of the region. The subduction zone fault dips into the Earth, getting deeper to the north.
\nBetween earthquakes (the interseismic period), the megathrust fault is seismogenically coupled (i.e. ‘locked’) just like velcro has the ability to hold together one’s wallet. The plates are always moving towards each other. Because the fault is locked, the crust surrounding the fault bends elastically to accommodate this convergent motion.
\nAs the crust bends and flexes, it stores energy (i.e. tectonic strain). The part of the fault closest to the seafloor (the southernmost part of this subduction zone fault) gets pulled downwards, while the part of the crust further to the north flexes upwards.
\nThe materials along the earthquake fault have properties that resist motion (like the velcro). But, as the plates converge and increase the amount of energy stored, the forces on the fault may exceed the strength of the fault. At this time, the fault slips, causing an earthquake.
\nThe part of the fault that was being pulled downwards gets pushed upwards during the earthquake (the coseismic period), while the crust that was being flexed upwards between earthquakes thus subsides downwards during the earthquake.
\nThe Alaska-Aleutian subduction zone has a history of subduction zone earthquakes and tsunami, plus there exists a prehistory of earthquakes and tsunami in some parts of this plate boundary. Geologists are often asked to determine the potential hazard of future earthquakes and tsunami and their answers are based on what we know from the past (using both historic and prehistoric data).
\nThe 2021 M 8.2 earthquake happened in the same location as a 1938 M 8.2 earthquake, just to the east of a sequence of earthquakes from last year (22 July and 19 October 2020).
\nTsunami:<\/font><\/strong>
\nWhen the earthquake fault slips, and the upper plate deforms, the vertical motion of the plate can elevate (or lower) the overlying ocean water. After the water changes position, it seeks to return to sea-level (an equipotential surface). If elevated, the water drops downwards and then oscillates up and down. This is the process that generates waves that radiate from the area with seafloor deformed by the earthquake.<\/p>\n

    \nThings that make a tsunami larger are [generally]:<\/p>\n
  1. More vertical land motion (possibly from larger slip on the fault, e.g. from a larger magnitude earthquake)<\/li>\n
  2. Deeper water (deeper water = more volume of water moving = more energy to create larger tsunami waves)<\/li>\n<\/ol>\n

    So let’s take a look at the things that may have affected the size of the tsunami from this 2021 M 8.2 earthquake.
    \nFirst of all, based on the earthquake slip models (estimates of how the earthquake slipped, in meters, and how that slip varied along the fault) suggest that a majority of the largest slip happened beneath the continental shelf. The water depth on the shelf is similar to many shelfs worldwide, shallower than about 200 meters. How does this affect the size of the tsunami?
    \nWell, I guess that is the main point, the ground deformation that generated the tsunami was beneath shallow water.
    \nThese slip models are based on a variety of data and most of the data are seismic data. Some tsunami are generated by slow slip (not generating seismic waves) on the shallow part of the fault. These are called tsunami earthquakes.
    \nBecause tsunami earthquakes may be generated by slip in this way, slip models using seismic data cannot resolve the location of the slip on the fault that created these tsunami. However, the tsunami from this 2021 M 8.2 earthquake were small. Therefore the updip part of the fault probably did not contribute significantly to the tsunamigenic ground deformation.<\/p>\n

    Below is my interpretive poster for this earthquake<\/font><\/h2>\n
      \n
    • I plot the seismicity from the past 3 months, with diameter representing magnitude (see legend). I also include earthquake epicenters from 1921-2021 with magnitudes M \u2265 7.5. <\/li>\n
    • I plot the USGS fault plane solutions (moment tensors in blue and focal mechanisms in orange), possibly in addition to some relevant historic earthquakes. <\/li>\n
    • A review of the basic base map variations and data that I use for the interpretive posters can be found on the Earthquake Reports page<\/a>. I have improved these posters over time and some of this background information applies to the older posters.<\/li>\n
    • Some basic fundamentals of earthquake geology and plate tectonics can be found on the Earthquake Plate Tectonic Fundamentals page<\/a>.<\/li>\n
    • I include outlines of the historic subduction zone earthquakes as prepared by Peter Haeussler from the USGS in Anchorage. He appears in the video about the 1964 earthquake below.<\/li>\n
    • Some of the tide gage and DART buoy locations are labeled.<\/li>\n
    • Note how there are still aftershocks from the 2018 M 7.9 earthquake sequence.<\/li>\n<\/ul>\n
        \n

        I include some inset figures. Some of the same figures are located in different places on the larger scale map below. I present 3 posters, each with slightly different information.<\/font><\/h3>\n
      • This is the first poster I prepared.<\/li>\n
      • In the upper center is a low-angle oblique view of the plate boundary. Note the oceanic Pacific plate is subducting beneath the continental North America plate. As the plate goes down, the water embedded in the rocks and sediment are released into the overlying mantle wedge. This water causes the mantle to melt, which rises, erupts as lava and forms the volcanic chain we call the Aleutian Islands. I place a green star in the “epicentral” location of the 2021 M 8.2 earthquake.<\/li>\n
      • In the upper left corner is part of a figure from Witter et al. (2019) that shows sections of the megathrust fault relative to how much the fault is thought to be locked. This is called the coupling ratio. For a fault that is fully coupled (or locked), the ratio is 1.0. For a fault that is slipping about 50% and accumulating about 50% of the plate motion rate, the coupling ratio is 0.5. Many subduction zones have low coupling ratios of 0.2-0.6. The region of the fault west of the 1938 and 2021 M 8.2 earthquakes is called the Shumagin Gap, thought to be possibly aseismic (with a coupling ratio closer to 0). But the 2020 sequence of M 7.8 and 7.9 earthquakes filled much of this gap.<\/li>\n
      • In the upper right corner is a plot showing the earthquake shaking intensity using the Modified Mercalli Intensity Scale (MMI). This is a USGS model based on observations of intensity from thousands of earthquakes. Read more about MMI here<\/a>.<\/li>\n
      • In the center right is a plot showing the aftershocks within a couple hours of the mainshock<\/li>\n
      • In the lower right corner is the initial record of the tsunami at the Sand Point tide gage (see map for gage location).<\/li>\n
      • I labeled the USGS slab 2.0 slab contours (Hayes et al., 2018). These depth contours represent the depth of the megathrust fault at these locations. The M 8.2 hypocentral depth is 32.2 km and the slab2 depth is about 35 km. Nice!<\/li>\n<\/ul>\n
          \n
        • Here is the map with 3 month’s seismicity plotted. There are 3 posters. The first one is something I put together around 2 hours after I awoke on the couch (abt 2am my time). I prepared the 2nd poster an hour later, which includes some information about tsunami prehistory. I prepared the 3rd poster late Sunday evening, about 3 days after the earthquake.<\/li>\n


          \n\"\"<\/a><\/p>\n

        • This is the second map I prepared and some figures are the same as in the first poster.<\/li>\n
        • Below the low-angle oblique map is a slip model from the USGS. The color represents the amount of slip on the fault. Note that the maximum slip is close to the epicenter. This is not always the case, as for the 1938 event, it appears that the maximum slip was not where the mainshock epicenter was.<\/li>\n
        • In the upper left corner is a map from Nelson et al. (2015). Those authors studied the prehistoric tsunami records at Chrikof Island, an island about 200 km to the east of the 2021 M 8.2 epicenter. The lower map shows GPS derived plate motion rates.<\/li>\n
        • In the lower right corner is also from Nelson et al. (2015). On this plot, the vertical axis represents time with “today” at the top and over 5000 years ago at the bottom. The horizontal axis is space, west to east from left to right. Each colored symbol represents the time of a prehistoric tsunami. The vertical size of these symbols represents the uncertainty (or “error”) associated with those chronologic data. We can take the number of earthquakes or tsunami over a period of time to estimate how frequently those process happen over time.<\/li>\n
        • To the left is a more updated version of the Sand Point tide gage, showing a wave height (peak to trough) of about 45 cm. We cannot compare this to the 1938 tsunami as there was not a tide gage at Sand Point in 1938<\/li>\n


          \n\"\"<\/a><\/p>\n

        • I prepared a 3rd poster, but updated it to this 4th poster.<\/li>\n
        • In the Intensity Data area, I added USGS “Did You Feel It?” data, which come from reports from real people. Learn more about dyfi here<\/a>. The model data are the colored lines labeled in white and the dyfi data are colored polygons labeled in yellow.<\/li>\n
        • In the aftershocks plot, I added epicenters from the several days after the mainshock. I also added a transparent overlay of the USGS finite fault model (the slip model). Compare the overlap, or non-overlap, of the slip region and the aftershocks. Why do you think that they are not completely overlapping?<\/li>\n
        • In the lower right section are tide gage records from gages in the area included in the poster. I plot the tidal forecast (dark blue), the tide gage observed water surface elevation (medium blue), and the difference between these data (in light blue) which is a record of the tsunami (and other waves, like wind waves). I made a rough approximation estimate of the maximum wave height and labeled this in yellow. The San Point tide gage has a mx wave height of about 0.8 m!<\/li>\n
        • I also plot the data from the DART buoy 46403, which is the closest DART buoy to the mainshock epicenter. The DART buoy network<\/a> is used to help calibrate tsunami forecast models during tsunami events. These are basically pressure transducers on the seafloor that measure changes in pressure caused by waves and atmospheric processes. The data plotted here are not tsunami data, but seismic wave data. One reason we know that this is not a tsunami is that the waveform initiated about 3 minutes after the earthquake. A tsunami would take longer to get to the buoy.<\/li>\n
        • In the upper left corner is a pair of maps that show USGS earthquake induced ground failure models. The map on the right shows what areas have likelihood of having landslides triggered by the 2021 M 8.2 earthquake. The panel on the right shows the possibility that areas might experience liquefaction induced by the earthquake.<\/li>\n
        • I added aftershocks associated with the 2020 M 7.8\/7.5 sequence that filled the Shumagin Gap (green circles) and outlined the aftershock region for both 2020 and 2021 sequences. The 2021 sequence is not yet over. The largest aftershock so far has only been M 6.1. The 1938 M 8.2 event had a M~7 event 5 days after the mainshock. Stay tuned?<\/li>\n


          \n\"\"<\/a>\n<\/ul>\n

          Tectonic Overview<\/font><\/strong><\/h2>\n

          Below is an educational video from the USGS that presents material about subduction zones and the 1964 earthquake and tsunami in particular.
          \n           Youtube Source <\/a> IRIS
          \n           mp4 file <\/a> for downloading.
          \n

          \nhttp:\/\/earthjay.com\/earthquakes\/19640327_alaska\/1964_great_alaska_earthquake_USGS.mp4<\/a><\/video><\/div><\/p>\n
            \nCredits:<\/p>\n
          • Animation & graphics by Jenda Johnson, geologist<\/li>\n
          • Directed by Robert F. Butler, University of Portland<\/li>\n
          • U.S. Geological Survey consultants: Robert C. Witter, Alaska Science Center Peter J. Haeussler, Alaska Science Center<\/li>\n
          • Narrated by Roger Groom, Mount Tabor Middle School<\/li>\n<\/ul>\n

            This is a map from Haeussler et al. (2014)<\/a>. The region in red shows the area that subsided and the area in blue shows the region that uplifted during the earthquake. These regions were originally measured in the field by George Plafker and published in several documents, including this USGS Professional Paper<\/a> (Plafker, 1969).<\/p>\n


            \n\"\"<\/a><\/p>\n

            Here is a cross section showing the differences of vertical deformation between the coseismic (during the earthquake) and interseismic (between earthquakes).<\/p>\n


            \n\"\"<\/a><\/p>\n

            This figure, from Atwater et al. (2005) shows the earthquake deformation cycle and includes the aspect that the uplift deformation of the seafloor can cause a tsunami.<\/p>\n


            \n\"\"<\/a><\/p>\n

            Here is a figure recently published in the 5th International Conference of IGCP 588 by the Division of Geological and Geophysical Surveys<\/a>, Dept. of Natural Resources, State of Alaska (State of Alaska, 2015). This is derived from a figure published originally by Plafker (1969). There is a cross section included that shows how the slip was distributed along upper plate faults (e.g. the Patton Bay and Middleton Island faults).<\/p>\n


            \n\"\"<\/a><\/p>\n

            Here is a graphic showing the sediment-stratigraphic evidence of earthquakes in Cascadia, but the analogy works for Alaska also. Atwater et al., 2005. There are 3 panels on the left, showing times of (1) prior to earthquake, (2) several years following the earthquake, and (3) centuries after the earthquake. Before the earthquake, the ground is sufficiently above sea level that trees can grow without fear of being inundated with salt water. During the earthquake, the ground subsides (lowers) so that the area is now inundated during high tides. The salt water kills the trees and other plants. Tidal sediment (like mud) starts to be deposited above the pre-earthquake ground surface. This sediment has organisms within it that reflect the tidal environment. Eventually, the sediment builds up and the crust deforms interseismically until the ground surface is again above sea level. Now plants that can survive in this environment start growing again. There are stumps and tree snags that were rooted in the pre-earthquake soil that can be used to estimate the age of the earthquake using radiocarbon age determinations. The tree snags form “ghost forests.<\/p>\n


            \n\"\"<\/a><\/p>\n

            This is a photo that I took along the Seward HWY 1, that runs east of Anchorage along the Turnagain Arm. I attended the 2014 Seismological Society of America Meeting that was located in Anchorage to commemorate the anniversary of the Good Friday Earthquake. This is a ghost forest of trees that perished as a result of coseismic subsidence during the earthquake. Copyright Jason R. Patton (2014). This region subsided coseismically during the 1964 earthquake. Here are some photos from the paleoseismology field trip<\/a>. (Please contact me for a higher resolution version of this image: quakejay at gmail.com)<\/p>\n


            \n\"\"<\/a><\/p>\n

            This is another video about the 1964 Good Friday Earthquake and how we learned about what happened.<\/p>\n