Swarms in Eastern Washington: are there fewer now than in the past?

Recently, while chatting about swarms, Steve Malone, emeritus PNSN director, pointed out that swarms of earthquakes used to be much more common in Eastern Washington than they are now. To prove us wrong, a swarm started on Saturday morning local time (September 21, 2024)! I took a screenshot of our recent event page (https://www.pnsn.org/earthquakes/recent) to show how localized the new cluster of events is. Also, notice how they are primarily tiny (M<=1) with only a few M2s. The events aren’t popping off as quickly anymore. The total number of located events is around 100 at the time of this writing (early Tuesday evening), but our system could have missed some of the smallest ones, and some may not have been analyzed yet. This is a completely natural phenomenon; although this swarm happens just outside the Hanford Site, it has nothing to do with the radioactive waste that is stored there. The term seismic swarm does not have a well-defined definition; simply, we mean “a bunch of earthquakes, without a clear mainshock, in a particular geographic area, with only a short time between events.” We detect many of these all across the region (OR and WA).

This swarm and our recent conversations prompted me to look into more detail about whether or not swarms in that region are less frequent than in the past. Our instrumental catalog started in 1969. When we say eastern Washington, we mean everything east of the Cascade mountain range. In the map below, I plot all the events labeled earthquakes in our local database colored by the year they occurred (our event information is also in the ANSS Comprehensive Catalog and available from earthquake.usgs.gov and pnsn.org). In the map below, each earthquake location is a dot; the darker the dot, the older the event. It is difficult to tell whether there are more swarms or not, although it does look like they may be in different places. Because all the dots plot on top of each other, a different visualization may be better: density of earthquakes across the map region.

 

Below, I show the number of events per 1x1 km grid cell (smoothed out). The color scale is logarithmic, meaning 0 means one event, and 2 means 100 events. Some clusters have more than 100 events, but I saturated the scale so that clusters with fewer earthquakes would also be visible. Since there are no geographical features, it is difficult to orient ourselves. Still, you can compare these two images (before 1998 and since 1998) to the map above and get the sense that, yes, the locations of some of the swarms are different, whereas others seem to be in the same general location.

 

 

 

We also need to keep in mind that our network has changed significantly over the decades. We had a lot fewer stations in the past. Without sensors nearby, you cannot detect the smallest earthquakes, and most of these earthquakes are very small. Compare the network in 1974 (left), to the one in 1985 (middle) to now, 2024 (right). Clearly, we have a lot more data to work with. For example, for a long time, there were no seismic stations in NE Oregon, so we cannot know whether there were small earthquakes there in the past. The network has been dense near the Hanford Site for a long time. Let’s take a closer look at that area, also because that is the area that Steve would have paid most attention to in the past. 

 

Let’s look at the density plots, zoomed into this region:

So, yep, Steve was right, and wrong. We still have a lot of swarms and small earthquakes in Eastern WA, but fewer distinct swarm areas are active. In particular, the many swarms east and north of the Hanford Site (i.e., near the Saddle Mountains) seem to have been quiet in recent decades. I’m anticipating his next comment: they were bigger! Were they? Yes, they were. See the map and cumulative event count graph below that show only the M2.5+ events. The steeper slope in the early years indicates more earthquakes per time unit than in later years. The little bump in 2009 is from the Wooded Island swarm, a very productive swarm with more than 1500 small events (the bright red cluster in the density plots above). Again, with a caveat: our analysis methods have changed over time, and our current magnitudes may not be perfectly comparable to those of the past. Recalibrating magnitudes is another project we plan to tackle to the best extent possible.

We write blog posts about swarms of small earthquakes regularly on this website. We have so many in our region! Interestingly, they occur at lots of different depths in the crust, as well as in different settings. In Eastern Washington, some are very shallow (many of the old ones were), some, like the current one, occur below the thick flood basalts deposited in that region, and some are in the lower crust.  We have recently started an effort to inventory and describe all the swarms as a first step to learning what causes them in those various settings. Various researchers have further analyzed and described some of the swarms in scientific papers. But most have just been popping off, unseen and unnoticed by anyone but our analysts.