Earthquakes are complex phenomena, and understanding their triggers is crucial for both scientific advancement and public safety. In this article, I'll delve into the multifaceted factors that contribute to seismic activity, particularly focusing on the role of groundwater cycles, stress from neighboring ruptures, and fluid injection underground. These elements often play a pivotal role in nudging a fault past its limit, while lunar tides, though significant, leave little to no mark. Every large earthquake, I argue, is a culmination of two distinct causes: the slow loading that builds strain over centuries and the sudden trigger that tips the balance. This dual nature has been a subject of intense study for seismologists, who have made significant progress in unraveling the complexities of earthquake initiation.
The Loading and the Trigger
Plate tectonics is the primary culprit behind the slow loading of faults. Along the Pacific and North American plate boundary, the crust gradually shifts by a few centimeters annually. This movement creates locked patches that resist movement, and the strain must eventually find an outlet. Researchers at the University of Hawaii at Manoa modeled 1,000 years of history on the southern San Andreas and San Jacinto faults, revealing stress levels that were at or above anything simulated in a millennium. This led them to label the fault system as "critically loaded," with a particular focus on Cajon Pass, a critical juncture that can either halt or facilitate a rupture.
Loading, however, is a necessary but not sufficient condition for an earthquake. It merely indicates that a fault is poised for action. The timing of the event is what truly matters. The Cascadia Subduction Zone, for instance, experiences slow loading as the Juan de Fuca plate creeps beneath North America at a pace akin to a fingernail's growth. Yet, the day an earthquake occurs remains unpredictable.
The Role of Stress and Rupture
When a rupture occurs, it significantly rearranges the stress field in its vicinity. Some patches of the fault become more susceptible to failure, while others experience relaxation. This dynamic is well-documented in aftershock patterns, which are primarily influenced by these stress changes. A notable example is the magnitude 7.7 earthquake in Myanmar on March 28, 2025, captured on camera as the ground tore open. Chinese researchers calculated that this event added three kilopascals of stress to five fault segments in southwestern Yunnan, a seemingly small amount but enough to tip a fault already at its breaking point.
Seismic waves also play a role in triggering earthquakes at a distance. These waves, generated by a significant rupture, can induce smaller earthquakes thousands of miles away as they pass through the Earth. This phenomenon, known as dynamic triggering, was evident after the Myanmar earthquake, where microseismicity increased near the Thailand border and in geothermal areas of southern China, where the static stress change was slightly negative.
The Impact of Water
Water, both on the surface and underground, exerts a substantial influence on seismic activity. Snowpack and groundwater in California exert downward pressure, causing the ground to move vertically by as much as 0.4 inches (1 centimeter) annually. Krittanon Sirorattanakul and Jean-Philippe Avouac at Caltech tracked this cycle against California's earthquake catalog, finding that regions with the most significant groundwater level fluctuations also exhibited the most substantial seasonal variations in seismicity. In Northern California, this effect reached approximately 10 percent.
The timing of this effect was particularly intriguing. Peak seismicity trailed peak stressing by about half a month, indicating that the fault does not break instantaneously upon stress arrival. Models that assumed immediate failure overpredicted the response. Friction models incorporating a delay in the fault's response matched both the magnitude and timing of the effect, offering a more accurate representation of earthquake nucleation.
The Human Factor
Human activities, particularly in the energy sector, can also trigger earthquakes. After 2009, Oklahoma experienced a dramatic increase in seismicity, from a handful of magnitude 3 earthquakes annually to hundreds. This surge was not attributed to drilling itself but to the salty water that accompanied oil extraction. Operators in Oklahoma yielded roughly 10 gallons of salty water for every gallon of oil, which was then pumped back into the deep Arbuckle formation, raising fluid pressure on faults several miles below.
The U.S. Geological Survey attributes only one to two percent of Oklahoma's induced earthquakes to hydraulic fracturing, with disposal wells being the primary culprits due to their extended operation and higher fluid volumes. The problem has since migrated to the Permian Basin of West Texas and southeastern New Mexico, where six earthquakes of magnitude 5 or larger have occurred since 2020.
Forecasting and Implications
None of these factors provide a specific date for an earthquake, as forecasting relies on measuring rates and probabilities over extended periods. However, the seasonal variations in seismicity offer valuable insights. Researchers can determine the stress exerted by the water cycle and observe how seismicity responds, thereby deducing the frictional properties of faults that are otherwise inaccessible. This has led to the identification of a key parameter, with values ranging from one to 10 kilopascals, which directly feeds into hazard models and is crucial for planning new oil or gas fields.
Moreover, the connection between seismic hazard and accelerating groundwater depletion in California's Central Valley and other heavily pumped basins is noteworthy. These changes in the load on the crust appear to influence fault behavior, underscoring the intricate relationship between geological processes and human activities.
