EARTHQUAKE CLOUDS AND
SHORT TERM PREDICTION

Thoughts After a Successful Prediction

- June 5, 1999 -

The 4.9 ML Calexico, Southern California, earthquake on June 1 announced that our first prediction to public on this site was very successful. We celebrate it now! We thank the USGS for giving the date signature, and many people for visiting our site.

To show the quality of the prediction, we give the following comparison between the prediction and the earthquake:

  Date Latitude (N) Longitude (W) Size (ML)
The prediction 5/5~6/5 27~33 114.5 ~115.5 >= 4.8
The earthquake 6/1 15:18 32.37 115.24 4.9

This earthquake is the only one falling into both the predicted size window, and the predicted area window from October 4, 1994 to today (1702 days). Relying on the world earthquake database of the USGS, we found that there have been only five earthquakes (table 1) within both windows since January 1, 1986, including this one. Therefore, the probability of this prediction is 2.46% or a risk of 1 in 40.

Although the prediction is successful, we should state that since we have a lack of experience and necessary resources, such as an hourly distribution of surface current, or an automatic video camera system to monitor earthquake clouds in the sky continuously, we may miss some earthquakes or make mistakes in the future. We hope that people will understand us if those cases occur.

We also hope that people would like to help us solve those difficulties if it is possible, so we will be able to make more and better predictions. In fact, we have at least 95% confidence on both the time window and the size window of all our predictions, so the main problem for our predictions is how to find the gaps where earthquake vapor goes out exactly.

Knowing that some seismologists and people do not like our work, we have to point out that our work has at least two advantages:

  1. Scientific value The Nature Debates from February 25 to April 8,1999 did not come to a conclusion about whether or not earthquakes can be predicted. Max Wyss, an optimist, even wrote, " I am pessimistic about the near future." However, our work gives a positive answer by the low probabilities and high accuracy of our predictions.

  2. Economic value To predict earthquake, Japan spends $185 million a year 1, China spent $45 million in 19962, and the USA spent $8 million in 19953, but no reliable short-term prediction has been exhibited 4-5. However, all of our tools are just a compass, a point-and-shoot camera, and a personal computer. Thus, our work has great economic value.

We have answered many questions from the seismologists. Now let's ask them two "Yes or No" questions:

  1. Can they explain what the cloud on the cover of this web site is?
  2. Can they make predictions as good as ours?

Since the majority of them are authorities in the seismology field and have excellent research resources, they should do much better. If none of them can say yes to our questions, we hope that they would like to support or cooperate with us to complete this honorable historic task.

References

  1. Swinbanks, D. Quake panel admits prediction is 'difficult'. Nature 388, 4 (1997).
  2. Li, H. China's campaign to predict quakes. Science 273, 1484-1486 (1996).
  3. The USGS. National Earthquake Hazards Reduction Program (FY 1995).
  4. Saegusa, A. Japan tries to understand quakes, not predict them. Nature 397, 284 (1999).
  5. Saegusa, A. China clamps down on inaccurate warnings. Nature 397, 284 (1999).

Table 1:
Past Earthquakes in the predicted area and size from 1/1/1986 to 6/4/1999
No
Date
Time
Latitude
N
Longitude
W
Magnitude
mb
Other Data
1 1987 02 07 3:45 32.39 115.31 5.4 5.4 ML Pas
2 1990 01 13 20:05 27.602 115.091 5.1
3 1992 05 22 14:09 27.359 115.019 5.2 4.8Z
4 1994 10 03 14:00 32.066 114.948 4.5 5.2Mw HRV 4.9MD ECX
5 1999 06 01 15:18 32.37 115.24 4.9
There are 4,900 days from January 1, 1986 to June 1, 1999, in which 5 earthquakes fell into both the area window (27~33 N, and 114.5~115.5 W) and size window (>= 4.8ML). Considering two additional data of No. 4, we do not wipe it off even though 4.5 is smaller than 4.8. Thus, the probability is 2.46%.


Original Prediction

5/5/1999 Pasadeda

According to my analysis, I predict that there will be an earthquake of magnitude >= 4 ML in the area 27~33 N and 113~117W within 47 days.

  • More likely time is 5/5~6/5.
  • The most likely time is 5/6~5/26.
  • More likely size is >= 4.5 ML.
  • The most likely size is >= 4.8 ML.
  • More likely latitude is 28~32 N.
  • The most likely latitude is 29.5~30.5 N.
  • More likely longitude is 114~116 W.
  • The most likely longitude is 114.5~115.5W.

Zhonghao Shou

Actual Earthquake
USGS Earthquake Bulletin: 1999 06 01 15:18:01.0 32.297N 115.255W Depth: 5G 4.9mb CALIF.-BAJA CALIF.


Home | Introduction | Publication & News | Predictions | New Predictions | Essays | Links | Contact

Sign Our Guestbook Guestbook View Our Guestbook

Updated: June 20, 1999 | Webmaster