TABLE 4,1 - Scaling Exponent, m, for Several Soils Soil Most Probable Nevada Minimum Maximum 344 Dry Clay 2.9 Wet Clay 2.5 Sandstone 3.6 Dry Sand 2.7 “Tn:‘Fig. 4.11 the results for surface charges in various soils are shown. .Far each soil the line drawn is that for the most probable value of m. On.this curve also are shown the nuclear craters at Nevada and in the Marshalls. In plotting the results of the nuclear explosions on this figure, the value of efficiency found for the JANGLE surface shot for the scaling exponent m * 3.4, namely 6per cent, has been assumed to be applicable to the explosions in the Pacific. The logarithmic grid has been adjusted in the region of 1 KT to include this efficiency for all ‘larger yields. Hence the graph can be entered di- rectly with the value of radiochemical yield. This graph gives a real- istic indication of the uncertainty in crater prediction depending on the properties of the soil. All data that have: been used in the development of the extrapola- tion method presented here. are summarized in Appendix A. This appendix also includes data for ‘some TNT. shots, namely those in wet sand, as well as some nuclear charges,> such as Trinity, which were not used in the actual analyses presented,héere.* 4.3 COMMENTS ON THE EXTRAPOLATToftHETHOD Ma It should be noted explicit ihat the extrapolation method de- scribed here is based on an empir ead qquation of the form R = f(W, m). (4g) 1 or Bo ~ R = (WE)™ . £(,) where E is an efficiency which depends on nédiun, scaled charge depth, and type of explosive. As mentioned in section 1.5; this is not the only form of equation which can be postulated, and:defended. The available data are so meager, and their scatter aroundthe curve representing any specific equation is so great, that it is not possible at present to establish unequivocally the relative validity of alternative forms of io the empirical equation. * The wet sand TNT results were not used because data on only ‘one charge size was found and hence a value of slope could not be established. A value for Trinity was not used because the scaled height is greater than that of interest in this report. 56