Using nuclear bombs to detect near earth orbit objects This question is based on an article written some years ago by A. C. Clarke, in which he attempted to solve two problems with one solution.  His idea was to remove the Earth's stockpile of nuclear weapons by launching our current number of nuclear warheads
- 10,144 devices as at 2014 - into space. The next stage of his plan was to manoeuvre them onto an orbital position diametrically opposite Earth.
Then detonate the bombs and use the radiation emitted to act as a source for detecting near earth asteroids and comets by measuring the radiation reflected back.  This plan, if feasible, would increase our chances of early detection of potentially Earth directed NEO's and subsequent possible avoidance of another Tunguska like impact. 
I assume this would require 3 (or more) separate bombs, using triangulation for location purposes.
Ignoring the political, environmental and logistical concerns involved in the transport of this nuclear material into space, my questions are:
Can the power output of a nuclear weapon be "tuned" to provide the maximum intensity at the most efficient frequency for reflection and subsequent detection by satellites?
Given the difference in the material composition of asteroids and comets, which frequency (or frequency range) would make this process most efficient? 
 A: This answer is a summary, based  on what I consider to be the most salient points made in the comments above. It is closely linked to the answer of my later question, Deflection of Earth directed NEOs using nuclear powered laser beams
Below is the answer I received in relation to my second question:
Project Excalibur
 The idea of a nuclear pumped X-ray laser was one which was investigated in detail in the Reagan "Star Wars" program of the 1980s, backed by one Edward Teller. Tests were carried out by surrounding the nuke with bundles of rods to create a one-pass laser. Apparently it was nowhere near efficient enough to be used in a military context. 
So further testing would be required to determine if we could actually deflect an NEO using a laser beam. However, as the energy level needed to merely produce a reflection detectable on Earth would be much less than the deflection energy, I would conclude that detection may well be a viable proposition.
This comment from Count Iblis, above,  
You should just assemble a few devices that are as powerful as you can make them, perhaps 100 Gigaton TNT. Then you get an output that is 20 times the power emitted by the Sun, albeit for a fraction of a second. That will generate enough of a signal that can be easily detected from Earth 
may support my argument that it may be possible to detect NEOs. 
This comment from Dwin is also relevant in establishing the orbital parameters involved:
There's nothing that says that you cannot have multiple detonations. That way you could then "see" reflections at several times and thereby calculate trajectories
