The USGS earthquake sequence product is a tool that identifies and describes earthquakes that are clustered in space and time as “sequences.” Its overarching goal is to provide simple contextual information regarding the spatial and temporal interconnection of earthquakes. The sequence product provides a description of a sequence, including information on the number and size of aftershocks, as well as visualizations of the sequence.
Summary
The sequence product relies on earthquakes identified as part of the Operational Aftershock Forecast (OAF) product to cluster mainshock-aftershock sequences and provide related information on the mainshock event page. Sequence identification relies on automated processing, including a simplified model of the expected space-time extent of an aftershock sequence. It therefore provides a first-order estimation of the earthquakes that are directly related to an ongoing sequence. Currently, the sequence product is available only for Mainshock-Aftershock type sequences (i.e., not earthquake swarms) and is not exhaustive, as it does not contain all sequences. Check out the OAF Overview page for details on the triggering criteria for domestic sequences. International OAF and sequence products are only occasionally posted, specifically for significant international earthquakes that require an international humanitarian response.
Background
Earthquakes occur frequently across the globe, with hundreds of earthquakes cataloged by the Advanced National Seismic System daily. Most earthquakes are too minor to be felt or occur in remote locations, such as beneath the oceans. The smallest earthquakes people typically begin to feel are about magnitude (M) 2.5. Although such small earthquakes occur frequently, they are felt over a small area and rarely result in damage. Less frequently, moderate magnitude earthquakes may cause felt shaking over larger areas and potentially cause damage. Only rarely do large, destructive earthquakes occur. Between 2015 and 2025, the Advanced National Seismic System cataloged roughly:
1600 M5-M5.9 per year
110 M6-M6.9 per year
13 M7-M7.9 per year
0.7 M8-M8.9 per year (less than one per year).
Earthquakes rarely occur in isolation but instead are usually clustered in space and time as part of sequences. All earthquakes result from some stress change to the Earth, and sequences may behave differently depending on the type of stress change that causes them. Commonly, earthquakes result from stress caused by the movement of tectonic plates, and because of that, most earthquakes occur near plate boundaries. Scientists classify sequences into two main types (though some sequences combine features of both):
Mainshock-Aftershock Sequences: Mainshock-aftershock sequences usually begin with the largest earthquake (the “mainshock”), followed by smaller earthquakes whose rate and size decay over time. Occasionally, a mainshock may be preceded by smaller earthquakes that are retrospectively termed “foreshocks.” Every time an earthquake occurs, it redistributes stress along nearby faults. Although scientists still debate the details, most agree that this stress transfer is a primary driver of mainshock-aftershock sequences. The larger an earthquake is, the larger the stress change it causes and the further the stress change extends in space from the mainshock hypocenter. For this reason, massive earthquakes can cause aftershocks at vast distances. For example, the 2004 M9.1 Sumatra-Andaman Island Earthquake caused aftershocks in an area 800 miles long, roughly in a zone the size of the state of California. Aftershock sequences typically last longer for larger earthquakes than for small earthquakes.
- Swarm Sequences: Earthquake swarms are sequences of elevated earthquake activity lacking a clear mainshock. Swarms often have their largest earthquake later in the sequence, rather than at its beginning. Swarms may also have several earthquakes with magnitudes similar to the largest earthquake. In addition to being driven by stress transfer like mainshock-aftershock sequences, swarms are thought to have an additional “ingredient” that promotes earthquake activity, such as fluid-pressure diffusion, magma movement, or slow fault slip. As a result, swarm intensity (earthquake rate or magnitude) may fluctuate based on this additional driving process, causing the sequence to evolve differently than a typical mainshock-aftershock sequence.
Words like mainshock, aftershock, foreshock, and swarm are just descriptors we place on earthquakes depending on the behavior of a sequence. An aftershock is simply any earthquake in a sequence that occurs after the mainshock. Some sequences have foreshocks, or earthquakes that occur before a mainshock. The mainshock is the largest magnitude earthquake in the mainshock-aftershock sequence. If an aftershock is larger than the previously identified mainshock, that earthquake would become the mainshock, and the earlier earthquakes would be termed foreshocks. Swarms are clusters of seismicity in space and time that do not follow a clear mainshock-aftershock pattern.
Scientists have studied the behavior of earthquakes using earthquake catalogs for a long time and have devised a few simple rules that, on average, describe earthquake behavior. Still, there is a significant variability in the evolution of each mainshock-aftershock sequence, and nature can always surprise us.
The relationship between the magnitude of a mainshock and its largest aftershock is described by Båth's law. Båth's law states that the largest aftershock is on average a little over 1 magnitude unit smaller than the mainshock. For example, if a M7 earthquake occurred, Båth's law would suggest that the largest aftershock would be about M6. This only describes average behavior among many sequences, and this behavior can vary wildly for individual sequences. Take, for example, the 2023 Kahramanmaras earthquake sequence, which had a M7.8 mainshock followed by a M7.5 aftershock. This variability also goes both ways; for instance, the largest aftershock of this M7.6 125-mile deep earthquake in Tonga was only an M5.
Omori’s law describes how the rate of aftershocks decays with time. It explains how we expect to see fewer aftershocks as we get further in time from a mainshock. How fast the aftershock rate decays can vary quite a bit between Mainshock-Aftershock sequences, but we expect that, on average, we would see fewer aftershocks each day following a mainshock. Further details on aftershock decay rates are available on the Aftershock Forecast Product Scientific Background page.
Another behavior we see in mainshock-aftershock sequences is that larger mainshocks cause aftershocks in larger regions and for more extended time periods, as discussed above. To describe this, we often rely on statistical models. These models can be complex because mainshocks aren’t the only earthquakes that change stress in the Earth; all of their aftershocks do as well. The USGS uses these models to forecast aftershocks. The Aftershock Forecast Product Scientific Background page provides details on how these models work.
How do we determine an Earthquake Sequence for the Earthquake Sequence Product?
Earthquake sequences occur often and having humans decide what earthquakes should belong in a sequence is time-consuming and uncertain. Instead, we automatically classify earthquakes as being part of a sequence using the same statistical models used by the Aftershock Forecast Product. This ensures that the earthquakes we display as part of a sequence are the same ones used as input for our Aftershock Forecasts. For most earthquakes, the spatial and temporal extent of the sequence is defined based on the location and magnitude of the mainshock. While this works well for USGS automated systems and broadly describes a sequence, users should be aware that important earthquakes may be inadvertently excluded, or that earthquakes that could be considered outside a sequence may be included after more scrutiny.
Currently, earthquakes that are part of an earthquake sequence contain a blue banner on the mainshock’s event page. This banner links to a map where all earthquakes, foreshocks and aftershocks are plotted. The USGS Sequence Product currently only showcases aftershock-mainshock sequences. Swarm sequences may be included in the future. Click here to view all USGS Sequence Product mainshocks.
The USGS earthquake sequence product is a tool that identifies and describes earthquakes that are clustered in space and time as “sequences.” Its overarching goal is to provide simple contextual information regarding the spatial and temporal interconnection of earthquakes. The sequence product provides a description of a sequence, including information on the number and size of aftershocks, as well as visualizations of the sequence.
Summary
The sequence product relies on earthquakes identified as part of the Operational Aftershock Forecast (OAF) product to cluster mainshock-aftershock sequences and provide related information on the mainshock event page. Sequence identification relies on automated processing, including a simplified model of the expected space-time extent of an aftershock sequence. It therefore provides a first-order estimation of the earthquakes that are directly related to an ongoing sequence. Currently, the sequence product is available only for Mainshock-Aftershock type sequences (i.e., not earthquake swarms) and is not exhaustive, as it does not contain all sequences. Check out the OAF Overview page for details on the triggering criteria for domestic sequences. International OAF and sequence products are only occasionally posted, specifically for significant international earthquakes that require an international humanitarian response.
Background
Earthquakes occur frequently across the globe, with hundreds of earthquakes cataloged by the Advanced National Seismic System daily. Most earthquakes are too minor to be felt or occur in remote locations, such as beneath the oceans. The smallest earthquakes people typically begin to feel are about magnitude (M) 2.5. Although such small earthquakes occur frequently, they are felt over a small area and rarely result in damage. Less frequently, moderate magnitude earthquakes may cause felt shaking over larger areas and potentially cause damage. Only rarely do large, destructive earthquakes occur. Between 2015 and 2025, the Advanced National Seismic System cataloged roughly:
1600 M5-M5.9 per year
110 M6-M6.9 per year
13 M7-M7.9 per year
0.7 M8-M8.9 per year (less than one per year).
Earthquakes rarely occur in isolation but instead are usually clustered in space and time as part of sequences. All earthquakes result from some stress change to the Earth, and sequences may behave differently depending on the type of stress change that causes them. Commonly, earthquakes result from stress caused by the movement of tectonic plates, and because of that, most earthquakes occur near plate boundaries. Scientists classify sequences into two main types (though some sequences combine features of both):
Mainshock-Aftershock Sequences: Mainshock-aftershock sequences usually begin with the largest earthquake (the “mainshock”), followed by smaller earthquakes whose rate and size decay over time. Occasionally, a mainshock may be preceded by smaller earthquakes that are retrospectively termed “foreshocks.” Every time an earthquake occurs, it redistributes stress along nearby faults. Although scientists still debate the details, most agree that this stress transfer is a primary driver of mainshock-aftershock sequences. The larger an earthquake is, the larger the stress change it causes and the further the stress change extends in space from the mainshock hypocenter. For this reason, massive earthquakes can cause aftershocks at vast distances. For example, the 2004 M9.1 Sumatra-Andaman Island Earthquake caused aftershocks in an area 800 miles long, roughly in a zone the size of the state of California. Aftershock sequences typically last longer for larger earthquakes than for small earthquakes.
- Swarm Sequences: Earthquake swarms are sequences of elevated earthquake activity lacking a clear mainshock. Swarms often have their largest earthquake later in the sequence, rather than at its beginning. Swarms may also have several earthquakes with magnitudes similar to the largest earthquake. In addition to being driven by stress transfer like mainshock-aftershock sequences, swarms are thought to have an additional “ingredient” that promotes earthquake activity, such as fluid-pressure diffusion, magma movement, or slow fault slip. As a result, swarm intensity (earthquake rate or magnitude) may fluctuate based on this additional driving process, causing the sequence to evolve differently than a typical mainshock-aftershock sequence.
Words like mainshock, aftershock, foreshock, and swarm are just descriptors we place on earthquakes depending on the behavior of a sequence. An aftershock is simply any earthquake in a sequence that occurs after the mainshock. Some sequences have foreshocks, or earthquakes that occur before a mainshock. The mainshock is the largest magnitude earthquake in the mainshock-aftershock sequence. If an aftershock is larger than the previously identified mainshock, that earthquake would become the mainshock, and the earlier earthquakes would be termed foreshocks. Swarms are clusters of seismicity in space and time that do not follow a clear mainshock-aftershock pattern.
Scientists have studied the behavior of earthquakes using earthquake catalogs for a long time and have devised a few simple rules that, on average, describe earthquake behavior. Still, there is a significant variability in the evolution of each mainshock-aftershock sequence, and nature can always surprise us.
The relationship between the magnitude of a mainshock and its largest aftershock is described by Båth's law. Båth's law states that the largest aftershock is on average a little over 1 magnitude unit smaller than the mainshock. For example, if a M7 earthquake occurred, Båth's law would suggest that the largest aftershock would be about M6. This only describes average behavior among many sequences, and this behavior can vary wildly for individual sequences. Take, for example, the 2023 Kahramanmaras earthquake sequence, which had a M7.8 mainshock followed by a M7.5 aftershock. This variability also goes both ways; for instance, the largest aftershock of this M7.6 125-mile deep earthquake in Tonga was only an M5.
Omori’s law describes how the rate of aftershocks decays with time. It explains how we expect to see fewer aftershocks as we get further in time from a mainshock. How fast the aftershock rate decays can vary quite a bit between Mainshock-Aftershock sequences, but we expect that, on average, we would see fewer aftershocks each day following a mainshock. Further details on aftershock decay rates are available on the Aftershock Forecast Product Scientific Background page.
Another behavior we see in mainshock-aftershock sequences is that larger mainshocks cause aftershocks in larger regions and for more extended time periods, as discussed above. To describe this, we often rely on statistical models. These models can be complex because mainshocks aren’t the only earthquakes that change stress in the Earth; all of their aftershocks do as well. The USGS uses these models to forecast aftershocks. The Aftershock Forecast Product Scientific Background page provides details on how these models work.
How do we determine an Earthquake Sequence for the Earthquake Sequence Product?
Earthquake sequences occur often and having humans decide what earthquakes should belong in a sequence is time-consuming and uncertain. Instead, we automatically classify earthquakes as being part of a sequence using the same statistical models used by the Aftershock Forecast Product. This ensures that the earthquakes we display as part of a sequence are the same ones used as input for our Aftershock Forecasts. For most earthquakes, the spatial and temporal extent of the sequence is defined based on the location and magnitude of the mainshock. While this works well for USGS automated systems and broadly describes a sequence, users should be aware that important earthquakes may be inadvertently excluded, or that earthquakes that could be considered outside a sequence may be included after more scrutiny.
Currently, earthquakes that are part of an earthquake sequence contain a blue banner on the mainshock’s event page. This banner links to a map where all earthquakes, foreshocks and aftershocks are plotted. The USGS Sequence Product currently only showcases aftershock-mainshock sequences. Swarm sequences may be included in the future. Click here to view all USGS Sequence Product mainshocks.