Hi Roger,

If you try the following then it might be interesting.

Change your program so that it plots the quakes for 90 day time windows instead of 30 day increments. That is what my program is doing. It averages together the signals detected over a three month period of time to produce each chart line. Then it moves ahead 10 days and produces another 3 month average based line. That long averaging time is an effective “noise filter.” It makes certain that only the strongest signal generating quakes appear on the chart. With shorter time windows such as a few days the chart peaks can be spread out all over the place.

If you generate a chart like that and it still resembles mine I would be quite interested in seeing it. It could be important.

On other topics, I have an extremely important project that I have been working on for the past year. And it has been taking up most of my free time. Hopefully I will be largely done with it this coming week. Then I can start spending more time on other projects such as this earthquake forecasting one.

You and I are focusing on different goals here. And this is one of the primary reasons that other scientists around the world have made so little progress with forecasting earthquakes over the years.

Scientists in general are trying to find a way to tell when and where earthquakes are going to occur. And they are trying to learn how to predict all earthquakes. That is quite difficult. I am simply trying to learn how to generate forecast data which will make it possible for us to save lives. And that is a radically different goal.

Society does not need to know when even powerful earthquakes are going to occur. They only need to know when one which will destroy a city will occur. So my forecasting program, chart, and tables etc. are focused on providing people with information which will tell them that. I don’t care how well the charts match all earthquakes. I just want to determine if they can provide a warning that a deadly one might be about to occur.

So, what the charts do is show when an earthquake is likely and where it might occur. And if it looks like one could be about to occur near you then I advise you to check the earthquake cloud Web sites etc. on my 141.html Web page to see if there are other indicators that an earthquake could be headed your way.

This type of information can be invaluable for groups which are already collecting precursor data such as groups in China, Turkey, Australia, Japan, and even here in the U.S.

For example, people would like to know when the “Big One” is going to occur here in the U.S. It is that mythical and hypothetical earthquake which is supposed to occur on the U.S. West Coast and destroy a major population center. If you look at the November 15, 2006 Kuril Islands earthquake chart peaks at my Web site you can watch the latest chart for peaks like that occurring near 120W. If they start appearing then that might be an indicator that the “Big One.” is finally on the way. If there are no strong peaks there as is presently the case, then that indicates that my own forecasting method has not detected anything. And I would think that it would be nearly impossible for a major earthquake such as one with a 7.0 or greater magnitude to occur near any large city on the U.S. West Coast without my detecting any strong signals beforehand.

The point is, with that 2003-2007.html Web page chart I am slowly adding markers for when significant earthquakes occurred in the past. They are mostly ones which claimed at least one life. If there are peaks on the chart for the times when a good percentage of them occurred in the past then that shows that the forecasting method is capable of achieving my goal of saving lives. The chances are that when the next deadly earthquake occurs there will be some peaks on the chart to indicate that it is about to occur.

A considerable amount of precursor data of different types is being collected in the People’s Republic of China. And they know about my forecasting program. So if there are peaks on the chart for the China area then they could easily check their data for that longitude and see if something might be on the way.

Whether or not the chart shows that a harmless 8.0 magnitude earthquake occurred out in the ocean somewhere is not usually important to me. I am only interested in the deadly ones. True, a powerful ocean earthquake can create a deadly tsunami. But, I believe that most of the normal problem quakes are occurring inland. I was actually quite surprised that the Kuril Islands quake appeared on my chart. I did not think that I could detect ones occurring out in the ocean like that. I thought that they had to be near cities where there are large collections of radio antenna simulation entities such as electric power lines and metal water pipes.

No. Sublunar points etc. are calculated only once and then stored in a massive text file for future use. I calculate those types of values each time I download the USGS earthquake list file. The quake data are then stored in that file. It takes about 5 seconds to do all of the calculations for each earthquake. And it is so slow because I have to have my Perl program manually feed the original quake data into the U.S. Navy MICA program which generates sublunar data etc.

Remember that we created a True Basic program for generating those types of data. And I would still like to use it. But I have not had time to run the necessary tests and incorporate the code into the Perl program. It would need to be tested extensively. I am generating those data for use by forecasters and disaster mitigation personnel who could be trying to save the lives of people living in an entire city. So, the data have to be accurate and reliable.

As I have said, the USGS has a group of Web site files which list significant earthquakes for the past few decades. To see if my forecasting program is working or not we need to try matching those 2003-2007.html Web page peaks with those past earthquakes. If there are chart peaks where a good percentage of the deadly earthquakes occurred then that shows that the program is working. And so far that correlation appears to be pretty good.

Something that I am also checking on is a possible longitude offset problem. In some cases the peaks might be appearing to the east or to the west of the actual earthquake longitude. The latest chart shows a strong peak around 125E. And since most of the latest warning signal activity stopped after that Solomon Islands earthquake I suspect that the peak was pointing to either that earthquake or perhaps a recent significant quake in the Japan area. As I learn more about the peaks on the 2003-2007 chart I should be able to tell if there is such an offset error. Hopefully, it will then be possible to adjust the computer programs to compensate for the offset.

You might try printing out that 2003-2007.html Web page chart and studying it. The data are quite interesting. With some of the peaks for example you can see them shifting to the east or west as time progress. That might show that the area where strain is building in the fault zone is gradually shifting east or west.

What I did was load the chart into a program like Word for Windows, format the chart so that it was 7 inches by 22 inches, save it, load it into Windows Paint, set the page format to “Landscape” with margins close to 0, and then have Paint generate the chart on 3 pages which I taped together.

I could also send you a 7 by 22 “ready to print” GIF chart if you would like. It is much clearer than ones created by using Word etc. to format the chart size. But it does not contain any quake markers.