Hi Don,

This is a very basic description of how this proposed approach to forecasting earthquakes works. I am of the opinion that it does work because I use it myself. And as I learn how to better evaluate the precursor signal patters I believe my forecasts are becoming more accurate and reliable.

Step 1. You put the times of all of the earthquakes that have occurred in some area into this computer program. And it tells you how the fault zone was being bent, stretched, and compressed when each of the earthquakes occurred. I have already done this for quite a few earthquakes. And what you find is that they often occur according to certain types of patterns. Some Turkey earthquakes appear to be especially reliable regarding the patterns they follow.

Step 2. When you detect the right type of precursor such as certain types of electromagnetic energy field fluctuations you feed the precursor time into the program. It calculates how fault zones around the world were being bent, stretched, and compressed at that time. And it then matches those data with the previously generated data for earthquakes around the world. When there is a good match it indicates that another earthquake might be about to occur in the area where the previous earthquake occurred.

Obviously you are collecting data for a large number of earthquakes. But computers are good at doing just that. And if you use the right filtering techniques to match precursor and earthquake data you can probably get the expected location down to a relatively few areas. Then you look for other signs of an approaching earthquake in those areas such as fresh cracks in building foundations etc.

The problem is that we presently have no computer programs capable of generating that original data. I myself have to use indirect, slow, and cumbersome procedures. But they do work to a certain extent. And what I am presently focusing most of my efforts on is developing more and more effective data filtering techniques.

Regarding placing sensors everywhere, I agree that it is too expensive. And that is why I am recommending that a computer program be developed to produce those types of data. In an early version it would not be able to take things like atmospheric storm effects on ocean tides into account. But it should be possible to fairly easily generate data which would take normal ocean and Solid Earth tides into account. And my data indicate to me that at least initially that should be adequate.

Just guessing, I expect the cost of producing the original version of such a computer program would be somewhere between $10,000 and $100,000. The exact cost would depend on who did the work and how sophisticated that original version was. It might require several computer programmers and a team of geophysicists.

Would it be worth it? I saw one estimate that a major earthquake near any number of our large West Coast cities could result in something like $100 BILLION dollars worth of damage and lost productivity. And some of that loss would result from fires started because people were not prepared for the earthquake and ready to quickly shut down fuel and electric lines.

So, how does even a $100,000 cost compare with a $100,000,000,000 loss? In my book such a program would be worth the expense. Plus, I believe that geophysicists would love to have a program like that even if in the end it did not generate high quality earthquake forecasts.