An earthquake prediction is a prediction that an earthquake in a specific magnitude range will occur in a specific region and time window. Predictions are considered as such to the extent that they are reliable for practical, as well as scientific, purposes. Although there is evidence that at least some earthquakes in some tectonic regimes are predictable with useful accuracy of time and space, the reliability and reproducibility of prediction techniques have not been established and are therefore generally not accepted by seismologists. For practical purposes, seismologists bring forth seismic hazard assessment programs by estimating the probabilities that a given earthquake or suite of earthquakes will occur.

In the effort to predict earthquakes, people have tried to associate an impending earthquake with such varied phenomena as seismicity patterns, electromagnetic fields, weather conditions and unusual clouds, radon or hydrogen gas content of soil or ground water, water level in wells, animal behavior, and the phases of the moon.

Thus far, earthquake prediction is controversial because data are sparse and there is little evidence or verified physical theory to link observable phenomena to subsequent seismicity. The frequent practice of publishing predictions after the fact further complicates matters. Also, given enough predictions, it is virtually inevitable that some will succeed "by chance." Assessing whether a successful prediction is a fluke is challenging. Most assessments rely on chance models for earthquake occurrence, models that are difficult to test or validate, because large earthquakes are so rare, and because earthquake activity is naturally clustered in space and time.

Chinese earthquake prediction research is largely based on unusual events before earthquakes, such as change of ground water levels, strange animal behavior and foreshocks. The Chinese successfully predicted the February 4, 1975 M7.3 Haicheng earthquake and the China State Seismological Bureau ordered an evacuation of 1 million people the day before the earthquake, but failed to predict the July 28, 1976 M7.8 Tangshan earthquake. This failure put Chinese earthquake prediction research in doubt for several years. However, there are messages showing that the Tangshan earthquake was predicted successfully. Chinese research has now merged with Western research, but traditional techniques are still common. Another successful prediction of the November 29, 1999, M5.4 Gushan-Pianling Earthquake in Haicheng city and Xiuyan city, Liaoning Province, China was made a week before the earthquake. No fatalities or injuries were reported

Prediction of volcanic activity (also: volcanic eruption forecasting) is an interdisciplinary scientific and engineering approach to natural catastrophic event forecasting. Volcanic activity prediction has not been perfected, but significant progress has been made in recent decades. Significant resources are spent to monitoring and prediction of volcanic activity by the Italian government through the Istituto Nazionale di Geofisica e Vulcanologia INGV, by the United States Geological Survey (USGS), and by the Geological Survey of Japan. These are the largest institutions that invest significant resources monitoring and researching volcanos (as well as other geological phenomena). Many countries operate volcano observatories at a lesser level of funding, all of which are members of the World Organization of Volcano Observatories (WOVO).

A tropical cyclone (hurricane) forecast model is a computer program that uses meteorological data to forecast the motion and intensity of tropical cyclones. Such models utilize powerful supercomputers with sophisticated mathematical modeling software and meteorological data to calculate paths and intensities. There are three types of models: statistical, dynamical, or combined statistical-dynamic, and two primary types of forecasts, track and intensity.

For More Information:
Earthquake Prediction
Prediction of Volcanic Activity
Tropical Cyclone Forecast Model

Source: Wikipedia (All text is available under the terms of the GNU Free Documentation License)