There is also the problem of which "metric" system to use. The metric system was invented in the Napoleonic era and an international standards body has existed since 1875. Over the last two centuries many metric units had come into use, with many variations by nationality and discipline. In 1960 the SI system (Le Systeme International d'Unites) was adopted - a thoroughgoing effort to clean-up and regularize the mess that had accumulated. Many of the units that came into use early in the atomic age - curies, rems, rads, roentgens, barns, fermis, etc. - were eliminated. These units are outside the SI system, but are recognized by the SI standards as anciliary metric units to be discouraged and eventually phased out. However the original historical and scientific documents regarding the atomic age, and most writings about it since, are completely dominated by these non-SI units. Accordingly I have made no attempt to keep to the SI system.

A continuing source of confusion with nuclear weapons is the meaning of the word "ton". Normally this is used as a unit of mass or weight (a distinction I am going to ignore) in either the Metric, British Imperial, or U.S. Customary measurement systems (the last two both having two types of tons, short and long). In connection with nuclear weapons the term "ton" and its metric extensions (kiloton, megaton, etc.) are also used as units of explosive energy output or yield. The confusion is further heightened by the non-standard convention sometimes employed in the U.S. or Britain of using the abbreviation MT (or Mt, or mt) to distinguish metric tons from short tons, while also using MT (or Mt, or mt) to mean "megaton".

The SI system does not use the ton, but it does recognize a metric unit of 1000 kg called the "tonne" (aka the metric ton). In this FAQ I use tonne to mean metric ton. Ton is used either to refer to the U.S./Imperial short ton, or for the energy yield of small explosions. Which is which should be apparent from context.

Now the units of explosive energy (megatons, kilotons, or even just tons, depending on yield) are derived from attempts to compare the explosive force of a bomb to conventional explosives, the original intention was to equate it with tons of trinitrotoluene (TNT) - a workhorse military explosive. This presented problems very quickly. To which "tons" does the comparison refer? And the explosive force of TNT is not exactly a universal constant. The energy release is affected by such things as charge density, degree of confinement, temperature, and the reference end state of the explosion products. Energy outputs ranging over 980-1100 calories/g are reported.

To clarify the situation kilotons (megatons, etc.) were redefined to be a metric unit equal to exactly 10¹² calories (4.186x10¹² joules). Thus treating kilotons as a metric mass measurement (kilotonnes) of TNT gives a value of 1000 c/g, well within the reported range, while treating it as "kilo short tons of TNT" gives 1102 c/g, at the extreme upper end of the reported range. Thus a kiloton can be called a "kilo metric ton of TNT" and a "kilo short ton of TNT" with about equal validity.

Note that the metric definition of kiloton refers to ALL of the energy immediately released by the device, regardless of form. Although chemical explosives release essentially all of their energy as kinetic or blast energy, only part of the energy in a nuclear explosion is released in this form (though under most conditions, it is the major part). Thus a kiloton nuclear explosion actually has significantly less blast effect than a kiloton chemical explosion.

The proper capitalization of the abbreviations for kiloton and megaton is also non-standard. kt, kt, kT, and KT can all be found in the literature (the authoritative Effects of Nuclear Weapons avoids the issue by never abbreviating these terms). The SI standards guide recognizes the use of a lower case "t" for tonne, but no capitalization rule has ever been adopted for the explosive ton. The official SI capitalization rule for kilo is "k" and for mega it is "M" (I think it would have been more logical to reserve lower case letters for metric scale fractional units - deci, centi, milli, etc., while reserving upper case for multiple units - deka, kilo, mega, etc., but SI doesn't do this).

For no particular reason I use the lower case "t" for the explosive yield measurement ton, and I follow the SI standard in metric scale abbreviations thus we have kt and Mt for kiloton and megaton.

Information provided by: http://www.fas.org