 What do I see when I look at this plot? I see that we have magnitude on the x-axis, and we have mean annual occurrence of earthquakes greater than or equal to some magnitude on the y-axis. So this means this plot is the same as the plot that we sketched earlier, showing how often earthquakes or different magnitudes happen. But this data is not for the whole world. It's just for the new-matterhood seismic zone. And there's two different datasets being plotted on top of each other here. So let's just look at those. One is Otto Nutley's catalogue, and it spans 1816 through 1984. It might be a little hard to see, so I'll just draw over it, but it looks like this, right? It goes down here, and there's some kind of jaggedy parts like that. Okay. The other catalogue is the new-matterhood seismic zone catalogue. This is the digital catalogue, and it's shorter. It only spans 24 years. It looks sort of like this. Great. So the difference between these two catalogs is that one spans a lot longer in time, but since it isn't a digital catalogue, it doesn't have as many small earthquakes in it. The other one has more small earthquakes, but isn't as long in time, so it doesn't have as many big earthquakes. And neither of these looks exactly like the plot we sketched, right? Remember the plot we sketched looks sort of like this, where you had magnitude here and log of the number here, and there was this perfect straight line. And these don't have that perfect straight line. They tip over and kind of go flat at some point. But you can extrapolate if they did have a perfect straight line, they would look like this. So what this is saying is that earthquakes smaller than about magnitude 3 in the Nutley catalogue, or about magnitude 1.5 or so in the digital catalogue, those earthquakes are there, but they're just not getting recorded. So when a plot like this flatlines, that's what it means. At this end, it's not a perfect straight line. It's kind of jaggedy. That's just because the catalogue is limited in time. So there's just not enough big earthquakes to really do useful statistics. In order to get a recurrence interval, these researchers extrapolate from the part where they have data to the part where they don't. And they calculate these time intervals. This is how often you can expect a magnitude 6 to happen in this region based on this earthquake data. Here's how often you can expect a magnitude 7, and here's how often you can expect a magnitude 8. Now is a good time to point out that uncertainty in the magnitude of those historic earthquakes in 1811 and 1812 becomes uncertainty in calculating the recurrence interval because it makes a big difference if you know you're trying to calculate how often a magnitude 7 is going to happen as opposed to how often a magnitude 8 is going to happen, right? See how different these numbers are. So there's kind of two big ways that you can have uncertainty in one of these calculations. One is just the limit of your catalogue, right? The shorter your catalogue is in time, the more you have to extrapolate from what you know to what you don't know. And the more uncertainty you have in the magnitude of the earthquake that you're looking for, like whether it's a 7 or an 8, the more uncertain you're going to be in calculating the recurrence time also.