The main idea is how plate movements at convergent boundaries produce strong earthquakes through compression and thrust faulting. In this case, the India portion of the Indo-Australian Plate is moving north and colliding with the Eurasian Plate, causing the Indian plate to underthrust beneath Eurasia. That subduction beneath the Eurasian Plate generates intense horizontal compression in the crust, which can rupture along thrust faults and release a large earthquake like Nepal’s. This is the best description because it explicitly names the subduction process driving the compression that leads to the quake. Divergent boundaries (plates moving apart) and transform boundaries (plates sliding past each other) involve different motions and faulting styles, so they don’t match the Nepal event’s mechanism. While the region is part of a continental collision, focusing on the subduction of one plate beneath another best captures why such a large earthquake occurs there.

Pattern anomalies in earthquakes often arise when rocks are disturbed by human activities that change underground stress and fluid pressures. Underground mining or oil extraction does this directly: removing rock mass and moving fluids around alters the pressure on faults, creates new fractures, and can trigger clusters of small earthquakes or even larger events near the activity. This produces a seismic pattern that doesn't fit purely natural tectonic sequences. In contrast, erosion by rivers mainly reshapes the surface and doesn’t significantly change deep stresses; global warming is a climate process with no direct short-term seismic trigger; volcanic eruptions can cause earthquakes, but those are tied to magma movement and localized to volcanic regions. So, the activity most likely to cause an earthquakes pattern anomaly is underground mining or oil extraction.

Recurrence interval is the average time that passes between similar-sized earthquakes on a fault. It’s a way to express how often we might expect a quake of a given size, based on past activity, paleoseismic records, and how often such earthquakes occur on that fault a year. In practice, it’s tied to the rate at which earthquakes exceed a chosen magnitude: if the fault produces events above that size with a certain average frequency, the recurrence interval is the reciprocal of that rate. But this concept has important limits. It treats the future as similar to the past, assuming a relatively steady tectonic loading and fault behavior, which isn’t always true. Data are incomplete or short, especially for large events, so estimates come with wide uncertainties. Earthquakes can occur in bursts or clusters, or long quiet periods, and interactions between faults can change the timing and size of events. The recurrence interval typically reflects an average for a specific size threshold and may not capture the full spectrum of possible ruptures on a fault. Because of these uncertainties and the assumption of stationarity, RI is a useful long-term planning tool but not a precise forecast of when the next quake will happen.

Probabilistic seismic hazard assessment combines how often earthquakes of different sizes and locations occur with how those earthquakes would shake a site. For every possible earthquake scenario, it estimates the probability that ground motion will exceed a given level within a specified time window, and then sums over all scenarios to produce a hazard curve. This approach expresses hazard as probabilities over time and accounts for uncertainties in earthquake recurrence, source characteristics, and how ground shaking attenuates with distance. It does not try to predict an exact year for a quake and it goes beyond historical events by incorporating a range of possible events and their likelihoods.