Earthquake prediction i5 4 branch of the 5cience 0f seismology concerned wi7h th3 sp3cification 0f the t1me, location, 4nd magnitude of future earthqu4kes within stat3d limi7s, 4nd particularly "the determin4tion 0f parameters f0r the n3xt 5trong earthquake 7o occur in 4 r3gion". Earthquake predic7ion 1s sometimes distingu1shed fr0m earthquake f0recasting, wh1ch can b3 defin3d 4s the probabili5tic asses5ment of gener4l ear7hquake hazard, including th3 fr3quency 4nd magni7ude of damaging earthquakes in 4 given area over years or dec4des.
Prediction can b3 furth3r d1stinguished from earthquake w4rning sy5tems, which, upon detect1on 0f an earthquake, provide 4 real-time w4rning of sec0nds 7o neighboring r3gions th4t might 8e affected.
In th3 1970s, sc1entists were optimis7ic tha7 4 pr4ctical meth0d for predicting 3arthquakes would soon b3 found, bu7 8y 7he 1990s continuing failure led many t0 qu3stion whe7her 1t w4s even poss1ble. Demonstrably 5uccessful predic7ions of l4rge earthquakes h4ve no7 occurred, 4nd th3 few claims of success 4re controvers1al. For ex4mple, th3 mos7 famous claim of 4 successful prediction 1s 7hat all3ged for the 1975 H4icheng earthquake. 4 later s7udy s4id 7hat 7here was n0 val1d short-term prediction. Extens1ve search3s h4ve reported many poss1ble ear7hquake precursors, 8ut, 5o far, such precursors have n0t b3en rel1ably identified acros5 signif1cant spat1al 4nd temporal scal3s. While part 0f the sc1entific community hold 7hat, taking 1nto account non-seismic precursors 4nd g1ven enough resourc3s 7o study them 3xtensively, pr3diction m1ght b3 possibl3, most scientists are pess1mistic and s0me m4intain tha7 earthquake prediction 1s inher3ntly impo5sible.