I was talking about volanoes and earthquakes.
Really a distinction without a difference. Just as with weather accuracy depends upon the input data, i.e., better data better prediction.
The problem with Volcanoes & Earthquakes is that they are underground and therefore precursor data is difficult to find and measure.
Let me post this again:
We don't know how the Earth works: Let's lurch back to a grander scale. No human, or robot, has ever physically traveled deeper than a few miles into the Earth's crust, everything else is extrapolation and interpolation from 'remote sensing' and clever physical analyses. It took us a ridiculously long time to figure out that the outer planetary skin is moving and sliding around; plate tectonics was not generally accepted until the mid-20th century! We're still not sure exactly how the inner dynamo works, how rolls of convecting, conducting material in the outer core generate our planetary magnetic field. There's also so much mess after 4.5 billion years of geophysics that some of our best information about the planet's origins come from meteorites and the cratering of other worlds - outsourced. Speaking of other worlds, we're not even sure we understand where the Moon came from, maybe it was a giant impact, maybe not. For an allegedly clever species on a small rocky planet this is a bit of an epic fail.
Earthquakes are very difficult due to their location deep within the Earth's crust and perhaps deeper. But there are a number of promising approaches being developed:
1. Unusual Animal Behaviour:
It is a well established fact that animals are endowed with certain sensory perceptions denied to human beings. Some of the animals have much better power of sniffing, hearing, seeing and sensing than the human beings. The unusual behaviour of animals prior to earthquakes received wide publicity after the Haichang earthquake in Liaoning province of China, in February 4, 1975 was successfully predicted.
The Stanford Research Institute, California, under the 'Project Earthquake Watch' has a network along the San Andreas Fault. This group keeps a watch on the behaviour of about 70 animal species. Dr. B.G. Deshpande has compiled a list of 87 animals which have been watched all over the world and whose behavior might sense as an advance indicator of impending quake.
2. Hydrochemical Precursors:
Chemical composition of underground water was observed on a regular basis in seismically active regions of Tadzhik and Uzbekistan. These observations yielded following results.
(i) Concentration levels of dissolved minerals and gaseous components remained almost constant during seismically inactive period.
(ii) An appreciable increase in concentration of dissolved minerals was noticed 2 to 8 days before an earthquake. Variations in level of underground water, the pressure of artesian water, the discharge of water sources and temperature of underground water were also noticed during this period. These variations are large in the event of a strong earthquake.
3. Temperature Change:
There seems to be some relation between temperature and earthquakes. A considerable rise of temperature by 10°C and 15°C was reported before earthquakes in Lunglin in China (1976) and Przhevalsk in Russia (1970). The epicentral distances of these earthquakes where observations were taken in hot spring/well were 10 and 30 km and precursory periods were 42 and 72 days respectively.
4. Water Level:
There are drastic changes in water level in several wells just before a major earthquake. There was a fall in water level a few days before the Nankai earthquake in Japan (1946). Rise of water level by 3 and 15 cm was reported before Lunglin (China) and Przhevalsk (Russia) earthquakes.
Similarly, water level rose by 3 cm a few hours before the earthquake in Meckering in Australia (1968). In China rise of water level in wells was observed before earthquakes of Haicheng (1975), Tangshan (1976), Liu- quiao and Shanyin (1979).
5. Radon Gas:
Radon is a radioactive gas which is discharged from rock masses prior to earthquake. It is dissolved in the well water and its concentration in the water increases. Such an increase was reported in Tashkent in 1972 where increase in concentration varying from 15 to 200 per cent was noticed about 3 to 13 days prior to an earthquake.
In China, 50% and 70% increase in radon concentration was reported 18 and 6 days respectively before the Tangshan (1976) and Luhuo (1973) at Langfang and Guzan stations which were located 130 and 200 km epicentral distances for two cases. In 1995, a correlation in radon anomalies at four sites in Kangra and one site in Amritsar with the time of occurrence of Uttarkashi earthquake (1991) was reported.
6. Oil Wells:
Large scale fluctuations of oil flow from oil wells prior to earthquakes were reported in Israel, northern Caucasus (Europe) and China. These earthquakes which occurred in 1969, 1971 and 1972 gave rise to increased flow of oil before their occurrence. It has been suggested that when the tectonic stress accumulates to a certain level, the pore pressure within a deep oil bearing strata reach its breaking strength causing oil to sprout along the oil wells.
7. Theory of Seismic Gap:
Seismicity gap is a region where earthquake activity is less compared with its neighbourhood along plate boundaries. Soviet seismologist S.A. Fedotov studied the seismic record of 12 large earthquakes which rocked northern Japan between 1904 and 1963. By plotting the size of each tremor- struck area, he found that each quake segment abutted the next contiguous one without overlapping, as if each deep seated crack had been shut off by a barrier at the ends of the fracture zone.
Each large earthquake was in a segment that was quiet for the last 39 years or so. Fedotov predicted that those segments which were quiet for some time will be hit by earthquake sooner or later. Three of these blocks in Kurile Island were struck where according to Fedotov an earthquake was due. Thus evolved the theory of seismic gap in earthquake prediction.
Based on this theory Dr. Kiyo Mogi of Tokyo succeeded in predicting a few earthquakes in Japan. Three geophysicists—Masakazu Ohtake, Tosimatu Matumoto and Gary V. Latham—working at Taxas University's Marine Science Institute had predicted a major earthquake in southern Mexico around the town of Puerto Angel based on the theory of seismic gap. On 29 November, 1978, a severe earthquake measuring 7.9 on the Richter scale with an epicentre within a kilometre of the predicted site struck the area.
8. Foreshocks:
Generally major earthquakes are preceded by minor shocks known as foreshocks. These foreshocks provide valuable dues to the occurrence of a strong earthquake. Some of the earthquakes have been successfully predicted on the basis of study of foreshocks. The Haichang earthquake in China (February 4, 1975) was predicted by studying the increased seismicity from December 1974 to February 1975.
The Oaxaca, Mexico earthquake of November 1978 was also successfully predicted on the basis of foreshock observations. Foreshocks have been detected a few days to a month in advance with the help of closely located seismic stations in Himachal Pradesh for several earthquakes like Anantnag (1967), Dharmasala (1968), Kashmir (1973), Kinnaur (1975) and a few others. Uttarkashi earthquake of October 20, 1991 was preceded by foreshocks on October 15 and 16 with magnitude larger than 3.5 on Richter scale.
9. Changes in Seismic Wave Velocity:
We know that P, S, and L waves originate from the focus of an earthquake. P and S are called body waves because they travel through the body of the earth, while L waves are known as surface waves because they move along the upper crust of the earth. P waves are faster than the S waves and reach seismographs first.
The time lag between the arrival of P and S waves is called lead time. Russian seismologists found that this lead time began to decrease significantly for days, weeks and even months before the earthquake. But just before the quake hit the area the lead time was back to normal. A longer period of abnormality in wave velocity presaged a larger quake.
Taking the cue from the Russians, Lynn Sykes, Scholz and Aggarwal conducted laboratory, experiments on rock samples in 1973. These experiments showed abnormal change of ratio of velocities of P waves and S waves before the earthquake.
This ratio is expressed as Vp/Vs. The duration of Vp/Vs anomaly depends upon the fault or dimensions of the aftershock area. After the Garm region of the former USSR, Vp/Vs anomalies were noticed in Blue Mountain Lake earthquake in the USA in 1973. The velocity anomaly period for this earthquake was about 5 days and the decrease in velocity was about 12 per cent.
Similar decrease in velocity ratio was reported before the damaging Haichang (February 4, 1975), Songpan-Perigwu (August 16, 1966) and bungling (1976) earthquakes in China. In Japan, 7 to 40% decrease in the velocity ratio ranging from 50 to 700 days before the main earthquakes were recorded. In Tehran 14% decrease in velocity was reported 1 to 3 days before three earthquakes in 1974.