but not to a seismograph :
Because the Comprehensive Test Ban Treaty forbids all nuclear testing, seismologists have redirected their attention toward the detection of nuclear explosions, regardless of size.
This is a difficult task because each day there are hundreds of naturally occurring earthquakes and large non-nuclear industrial explosions associated, for example, with mining and building demolitions. It is generally possible, however, to distinguish earthquakes caused by explosions from naturally occurring earthquakes along faults. In comparison to naturally occurring earthquakes, earthquakes triggered by explosions are very shallow. Explosions occur in small spaces and, because an explosion causes the rock around it to dilate, produce strong compressional body waves that travel through the Earth. Earthquakes along faults, in contrast, are caused by slip distributed over large areas and tend to produce much larger surface waves that travel along Earth's surface. Each of these produces distinctly different seismograms and ratios of long- to short-period seismic waves. The size of an explosion can be estimated from the magnitude of the earthquake it produces. Current efforts are aimed at the identification of nuclear explosions in the 0.001 to 0.01 kiloton range, which produce earthquakes of magnitude 2 to 3.
One of the most significant events since the signing of the Comprehensive Test Ban Treaty was a series of nuclear tests conducted by India and Pakistan in May 1998. India, which was one of only three countries to oppose the treaty, conducted three nuclear tests in the northwestern part of the country. Neighboring Pakistan, which supported the treaty, but refused to sign it as long as it was opposed by India, conducted five nuclear tests in response. The tests produced earthquakes with magnitudes between 4.8 and 5.2, one of which was preceded by a naturally occurring magnitude 6.9 earthquake in Afghanistan. Seismologists have concluded that both India and Pakistan probably exaggerated the size of the tests in order to present more powerful images to each other.
The use of seismology to detect remote explosions is not limited to nuclear test monitoring. It can also be used to learn about large explosions due to other causes, especially in foreign countries or inaccessible areas. Seismologists using publicly available information, for example, were able to determine that two separate explosions occurred when the Russian submarine Kursk sank in August 2000. A small explosion was followed about two minutes later by a second explosion that released about 16 times as much energy as the first and produced a magnitude 4.2 earthquake that was recorded as far as 5000 km away. It was further determined that the energy released in the second explosion was equivalent to that which would have been released by 2000 to 4000 kg (about 2 to 4 kilotons) of TNT. The depth of the second explosion was estimated from a bubble pulse produced during the explosion, which was caused when a bubble of hot gas oscillates while rising quickly through the water. The calculated depth of 100 m is about the same as the seafloor depth at the location of the Kursk accident, so the second explosion probably occurred when the sinking submarine struck the seafloor. More than 150 earthquakes with magnitudes of 1.4 to 1.6 occurred in the months after the sinking. They were probably caused by depth charges that were detonated by the Russian navy to discourage foreign submarines from visiting the wreckage. Similar studies have shed light on incidents such as the 1995 attack on the Murrah Federal Building in Oklahoma City; the 1998 truck-bombing of the American Embassy in Nairobi, Kenya; and the September 11, 2001 terrorist attacks on the World Trade Center and Pentagon.
http://www.espionageinfo.com/Re-Se/Seismology-for-Monitoring-Explosions.htmlthat's why they are trying to track gases
http://www.llnl.gov/str/Carrigan.html