Japan Meteorological Agency seismic intensity scale

The Japan Meteorological Agency (JMA) Seismic Intensity Scale (known in Japan as the Shindo seismic scale) is a seismic intensity scale used in Japan to categorize the intensity of local ground shaking caused by earthquakes.



The JMA intensity scale should not be confused or conflated with magnitude measurements like the moment magnitude (Mw) and the earlier Richter scales, which represent how much energy an earthquake releases. Much like the Mercalli scale, the JMA scheme quantifies how much ground-surface shaking takes place at measurement sites distributed throughout an affected area. Intensities are expressed as numerical values called "seismic intensity" (震度); the higher the value, the more intense the shaking. Values are derived from peak ground acceleration and duration of the shaking, which are themselves influenced by factors such as distance to and depth of the hypocenter (focus), local soil conditions, and nature of the geology in between, as well as the event's magnitude; every quake thus entails numerous intensities.

The data needed for calculating intensity are obtained from a network of 670 observation stations using "Model 95" strong ground motion accelerometers. The agency provides the public with real-time reports through the media and Internet giving event time, epicenter (location), magnitude, and depth followed by intensity readings at affected localities.

History
The Tokyo Meteorological Observatory, which in 1887 became the Central Meteorological Observatory first defined a four-increment intensity scale in 1884 with the levels faint (微), weak (弱), strong (強), and violent (烈). In 1898 the scale was changed to a numerical scheme, assigning earthquakes levels 0–7.

In 1908, descriptive parameters were defined for each level on the scale, and the intensities at particular locales accompanying an earthquake were assigned a level according to perceived effect on people at each observation site. This was widely used during the Meiji period and revised during the Shōwa period with the descriptions seeing an overhaul.

Following the Great Hanshin Earthquake of 1995, the first quake to generate shaking of the scale's strongest intensity (7), intensities 5 and 6 were each redefined into two new levels, reconfiguring the scale into one of 10 increments: 0–4, 5-lower (5–), 5-upper (5+), 6-lower (6–), 6-upper (6+), and 7. This scale has been in use since 1996.

Scale overview
The JMA scale is expressed in levels of seismic intensity from 0 to 7 in a manner similar to that of the Mercalli intensity scale, which is not commonly used in Japan. Real-time earthquake reports are calculated automatically from seismic-intensity-meter measurements of peak ground acceleration throughout an affected area, and the JMA reports the intensities for a given quake according to the ground acceleration at measurement points. Since there is no simple, linear correlation between ground acceleration and intensity (it also depends on the duration of shaking  ), the ground-acceleration values in the following table are approximations.

Intensity 7
The Intensity 7 (震度7, Shindo 7) is the maximum intensity in the Japan Meteorological Agency seismic intensity scale, covering earthquakes with an instrumental intensity (計測震度) of 6.5 and up. At Intensity 7, it becomes impossible to move at will. The intensity was created following the 1948 Fukui earthquake. It was observed for the first time in the 1995 Great Hanshin earthquake.

Observation system
Since April 1997, Japan has been using automated devices known as "seismic intensity meters" to measure and report the strength of earthquakes based on the JMA scale. This replaced the old system that relied on human observation and damage assessment.

The installation of these meters began in 1991 with the "Model 90 seismic intensity meter," which didn't have the capability to record waveforms. In 1994, an upgraded version, the "Model 93 seismic intensity meter," was introduced. This model could record digital waveforms on memory cards. Later, the "Model 95 seismic intensity meter" was introduced, which had several improvements including the ability to observe double the acceleration limit and a higher sampling rate. Today, all of JMA's seismic intensity meters are of this "Model 95" type.

Specifications of the Model 95 Seismic Intensity Meter


 * Observation components: NS (North-South), EW (East-West), UD (Up-Down) - three components (seismic intensity is a composite of the three components)
 * Measurement range: 2048 gal to -2048 gal
 * Sampling: 100Hz rate, 24-bit
 * Recording standard: Seismic intensity of 0.5 or higher (collected in one-minute intervals)
 * Recording medium: IC memory card

By the end of 2009, about 4,200 of these meters were in use for JMA's "seismic intensity information," and by August 2011, this number had grown to 4,313. This was a significant increase from the roughly 600 units in use when the switch to measured seismic intensity was made. This shows that Japan's network for observing seismic activity is one of the most comprehensive in the world. Of these meters, around 600 are managed by the JMA, about 780 by the National Research Institute for Earth Science and Disaster Resilience (NIED), and roughly 2,900 by local government bodies.

The network was designed with the aim of having one seismometer in each municipality before the major municipal mergers of the Heisei era. Additional units were installed in remote islands and areas with low populations to ensure complete coverage.

Besides the seismic intensity meters used for JMA's information, many other meters have been installed by local government bodies that are not used for JMA's information. Public institutions and public transportation organizations have also independently installed meters to ensure the safety of infrastructure like dams, rivers, and railways.

Observation instrument installation
To ensure the accuracy of earthquake intensity measurements, there are specific guidelines for setting up seismic intensity meters. The JMA doesn't use data from meters that are set up in unsuitable locations for their earthquake intensity information.

Firstly, these meters should be placed on a sturdy stand designed for them. Because the ground can shake more on embankments or cliffs, the meters should be set up outside on flat, stable ground with no steps nearby, and at least two-thirds of the stand should be buried in the ground. There are also rules about nearby structures. The meters should be far enough away from trees or fences that could fall over and hit the meter. If the meters are set up inside, they should be placed near the pillars on the ground floor, and they can be set up anywhere from the basement to the second floor. Meters aren't set up in buildings that have earthquake isolation or control construction.

Seismic intensity meters should be securely attached to the stand or, if they're inside, to the floor. It's recommended to follow the setup instructions provided for each type of meter and, if possible, to secure them with anchor bolts.

The JMA rates the setup location of seismic intensity meters used for earthquake intensity information on a scale from A to E. Grades A to C are acceptable, D is generally not used but may be used after careful consideration, and E is not acceptable.

However, there have been cases where earthquake intensity information was used even though the meters were set up in unsuitable locations, and later the accuracy of the information was questioned and corrected. For instance, during the July 2008 Iwate earthquake, an earthquake intensity of 6+ (later changed to 6-) was recorded in Ono, Hirono Town, Iwate Prefecture. This intensity was much higher than in nearby municipalities, which led to an investigation. On October 29 of the same year, the JMA announced that the meter in Ono was in an unsuitable location for earthquake observation and removed it from the earthquake intensity data, correcting the maximum intensity from 6+ to 6-. Since the meter in Ono was originally rated as acceptable, it's been suggested that other meters could also be in deteriorating setup locations.

Density of station placement and maximum seismic intensity
The number of seismic monitoring stations significantly grew in 1996, thanks to the JMA increasing the number of seismic observation points. This growth has made it easier to detect strong earthquakes near their origin point. For example, the 1984 Nagano earthquake, which caused a lot of damage but was only rated as a 4 in terms of seismic intensity, and the 1946 Nankai earthquake, a huge earthquake that was rated as a 5, would have been given lower ratings if there weren't any monitoring stations near their origin points before 1995. After the increase in monitoring stations, even if an earthquake is the same size as before, it's likely to be given a higher seismic intensity rating, and high intensity ratings like 6- are reported more often. The increase in seismic observation points has made it possible to detect earthquake intensities closer to their origin point, and the JMA is studying the differences between the highest earthquake intensities detected at all monitoring stations and the intensities measured at JMA offices, to understand how the increase in monitoring stations has changed the maximum seismic intensities. Here are a few examples: In earthquakes with smaller magnitudes, the range of seismic intensity 6- becomes narrower. Even so, if there are many observation points, some will fall within the range of seismic intensity 6-. However, if there are fewer observation points, there is a high possibility that the maximum seismic intensity will be lower because it will not be captured by the observation points. Before 1995, an earthquake with a maximum seismic intensity of 6 was certainly a "major earthquake" in terms of magnitude. However, since 1996, even very shallow minor earthquakes are more likely to report seismic intensities of 5 or 6, so it is not appropriate to treat "earthquakes with a maximum seismic intensity of 6" on par with those before 1995. It may seem as if there have been more earthquakes since the Great Hanshin-Awaji Earthquake, but this is not because there have been more earthquakes, but because there have been more reports of seismic intensity.

Furthermore, seismic intensity observation points are not uniformly distributed by area. They are often installed in regions with high population density, especially in urban areas. This tendency is particularly strong for observation points set up by local public entities. In these high population density areas, there tends to be a higher amplification rate of seismic intensity in the surface soil layer.

Seismic intensity calculation
The seismometers used by the JMA and others observe shaking through accelerometers. They first measure the three components of motion - vertical, north-south, and east-west - as time-domain signals of acceleration. The instrumental seismic intensity (decimal value) is then calculated through the following process:

Round the instrumental seismic intensity (if negative, it is 0, if 8 or more, it is 7) to determine the seismic intensity level from 0 to 7. In the case of seismic intensity 5 and 6, it is further divided into lower and upper depending on whether it is rounded up or down (refer to the Scale overview section).
 * 1) The time-domain signals of vertical, north-south, and east-west motion are converted into frequency-domain signals through Fourier transform.
 * 2) To correct for the effects of the earthquake wave period, filtering is applied to each of the frequency-domain signals of vertical, north-south, and east-west motion. The filter used here is a product of several filters, each of which is a function of frequency ($$f$$).
 * 3) * Low-cut (low frequency elimination) filter: $$\sqrt{1-\exp \left( -\left(\frac{f}{0.5}\right)^3 \right)}$$
 * 4) * High-cut (high frequency elimination) filter: $$\frac{1}{\sqrt{1+0.694x^2+0.241x^4+0.0557x^6+0.009664x^8+0.00134x^{10}+0.000155x^{12}}}$$  (where $$x=\frac{f}{10}$$)
 * 5) * Periodic effect filter: $$\sqrt{\frac{1}{f}}$$
 * 6) Convert the frequency-domain signals of the vertical, north-south, and east-west movements that have been filtered back into time-domain (acceleration) signals by inverse Fourier transform.
 * 7) Combine the three components of vertical, north-south, and east-west movements to create a single composite acceleration.
 * 8) Find a threshold value $$a$$ such that the total time when the absolute value of the composite acceleration is $$a$$ or more is exactly 0.3 seconds. In other words, let $$A(t)$$ be the composite acceleration signal as a function of time $$t$$. We need to find a threshold value $$a$$ such that $$\int_{T_1}^{T_2} H(|A(t)| - a) dt = 0.3$$, where $$H(x)$$ is the Heaviside step function, and $$T_1$$, $$T_2$$ are the bounds of the time interval being considered. The aim is to standardize the magnitude $$a$$, which is the basis for calculating the seismic intensity, to a shaking that lasts for 0.3 seconds, in order to bring the seismic intensity calculated closer to the actual damage caused by the shaking.
 * 9) Calculate $$I=2 \log_{10}a+0.94$$.
 * 10) Round the third decimal place of $$I$$ and truncate the second decimal place to determine the instrumental seismic intensity.

Comparison with other seismic scales
A 1971 study that collected and compared intensities according to the JMA and the Medvedev–Sponheuer–Karnik (MSK) scales showed that the JMA scale was more suited to smaller earthquakes whereas the MSK scale was more suited to larger earthquakes. The research also suggested that for small earthquakes up to JMA intensity 3, a correlation between the MSK and JMA values could be calculated with the formula MSK = JMA1.5 + 1.5, whereas for larger earthquakes the correlation was MSK = JMA1.5 + 0.75.