1. AM broadcast wave propagation

  Figure 1 shows the propagation of MF band broadcast waves.
The ionospheric D layer exists only during the day between 60km and 90km.
It reflects the VLF/LF band electromagnetic waves and attenuates the MF band and above electromagnetic waves.
The E layer is located between 90km and 130km and exists both day and night. Therefore, it reflects electromagnetic waves in the MF band and reaches long distances only at night.

　Radio wave propagation loss is expressed as Loss (dB)=20*log(4*π*r/λ) (λ=wavelength, r=distance).
　The propagation loss at 1000kHz is 98dB at 2000km. Attenuation in the ionosphere is empirically thought to be about 70 dB.
  As an example, find the received power of a broadcast wave emitted with a transmitting power of 50kW.　

Transmission power is +77dBm when converted by [dBm] = 10*log10([watts]*1000).

　If the gain of the transmitting and receiving antenna is 0dB, the received power will be +77dBm-98-70=-91dBm, which is considered to be the reception limit.
  In other words, the propagation of AM broadcast waves is limited to 2000km, and it is thought that only one hop is reflected in the ionosphere.

In fact, Sagamihara can barely receive stations in Sapporo, Kushiro, Kumamoto, and Kagoshima, which are approximately 1,000 km away.

  On the other hand, the VLF/LF band has less propagation loss and attenuation in the ionosphere than the MF band, and it propagates more than halfway around the Earth.
  From this, it can be said that it is easier to identify the epicenter area by using AM broadcast waves in the MF band.
  It is relatively easy to identify transmitting stations in Japan and neighboring countries, and you can check them by frequency using a frequency book.
  Also, it is easy to check their voices, such as occasional call signs, traffic information, etc.
  You can check local news, address of music program requester, etc.

　　　　　　　Fig.1 Radio wave propagation diagram of MF band AM broadcast waves

2. AM broadcast wave path and predicted epicenter area
　As shown in Fig. 2, this method has the advantage of being able to cover the entire Japan's epicenter area with a small number of observation points, and making it easy to identify the predicted epicenter area.
  For M=6 or less, the predicted epicenter area is assumed to be a radius of 50 km from the center between the transmitting point and the receiving point.
  In order to observe the epicenter area throughout Japan without any gaps, it is necessary to conduct observations in Sakhalin, Etorofu Island, Taiwan, and the Philippines.

Fig.2 Nationwide observation using 4 domestic observation points

3. Example of ionospheric disturbance observation using AM broadcast waves
　Fig. 3 shows an example of observation data for one day from Sapporo to Sagamihara.
  Fig. 4 shows an example of the standard deviation of variations in morning and evening reception end and start times from Sapporo to Sagamihara over the last four months.
  Only four times did it exceed σ=2.
  Figure 5 shows the predicted epicenter area on the path from Sapporo to Sagamihara.
  The predicted epicenter area is Akita Prefecture, western Iwate Prefecture, and northern Miyagi Prefecture.
  List-1 shows a list of earthquakes that occurred in the predicted epicenter area within approximately one week from the day when σ exceeded 2.

　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　　Fig.3 Data Graph for one day from Sapporo to Sagamihara

4. How to read the ionospheric disturbance observation statistical graph
　　-Green line graph: Morning Terminator Time (JST, left scale, hereinafter referred to as TT)
　　-Yellow line graph: Evening Terminator Time (JST, left scale)

　　-Green bar graph: Standard deviation of morning TT (σ Sigma, right scale)
　　-Orenge bar graph: Standard deviation of evening TT (σ Sigma, right scale)
　　
　　-EQ: Occured Time of Noto Peninsula Earthquake (2024/1/1 16 : 10)

Fig.4 Standard deviation of morning and evening reception end/start time fluctuations from Seoul to Sagamihara in the latest month

5. About standard deviation
　　(1) An easy-to-understand explanation of standard deviation is here
　　　(2) The probability distribution (normal distribution) of data is expressed as shown below.
　　　　　

　　　　　　　　　　　　　　　　　　　　　　　　　(Image source：wikipedia)

・The probability that the deviation from the average value is within ±1σ is 68.27%, 95.45% when it is ±2σ or less, and 99.73%
when it is±3σ.
・A deviation of ±2σ is an anomaly that occurs only 4.55% of the time, and a deviation of ±3σ is an anomaly that occurs only 0.27%.
・For the time being, the difference in the morning and evening reception end/start times of medium-wave band AM broadcast waves
will be about ±2σ, and in 1 to 2 days it will be M=4 class, about ±3σ, If it continues for 1 to 2 days, it will be considered
as M=5 grade, and if it continues for 3 days or more, it will be judged as M=6 grade or higher.

　　　　　　　　　　　　　

　
Fig.5 Predicted epicenter area on the path from Sapporo to Sagamihara

List-1 List of earthquakes that occurred in the predicted epicenter area within one week from the day when σ exceeded 2

                           Please use all data and descriptions on this website at your own risk.
           Persons affiliated with this center are not responsible for the results of using this product.
          Copyright (c) Certified NPO Earthquake Precursor Observation Center (JEPCOC). All rights reserved.

σ>2 observed days	Earthquakes that occur in the predicted epicenter area within approximately one week
6/30, 7/4	2016/7/16 22:12	Northern inland area of Akita Prefecture	M4.6
8/18	2016/8/27 19:00	Northern Miyagi Prefecture	M3.5
10/1, 2, 3	2016/10/10 8:15	Northern inland area of Akita Prefecture	M3.2
10/12	2016/10/16 15:14	Southern Miyagi Prefecture	M4.1