Last Preliminary/Auto Event
This is the preliminary event which produce by automatic system
Magnitude

---

--
--

Tsunami Travel Time Modelling, This map is generated by Latest Auto Earthquake Information.

TTT software developed by Dr. Paul Wessel (Geoware, http://www.geowareonline.com), and is used by the NOAA Pacific Tsunami Warning Center to predict of tsunami travel times on a geographic (lat–lon) grid derived from a supplied bathymetric data grid. This map generated by Generic Mapping Tools (GMT) developed by Dr. Paul Wessel and Walter Smith (https://www.generic-mapping-tools.org/).

Last Final Event
This is the final event revised by seismologist

Last Advisory/Warning
This is the advisory/warning of last event

InaTNT Background

Tectonic earthquakes are the trigger for up to 90% of tsunami events that have ever occurred on this earth. In addition to tectonic (non-tectonic) earthquakes, tsunamis can be caused by several factors, including volcanic eruptions, landslides or meteor strikes. The detection of the causes of non-tectonic tsunamis is quite complex and difficult to model in relation to the tsunami potential in real time. Even now, not a single country in the world has actually implemented a tsunami early warning system to anticipate tsunamis from these non-tectonic generating sources.

The tsunami that hit the Sunda Strait on December 22, 2018 opened our eyes that the threat of a tsunami from non-tectonic sources still exists and can occur at any time. At that time, the collapse of mountain body material covering an area of ​​64 hectares due to the explosive eruption of Mount Anak Krakatau caused tsunami waves that hit a number of beaches on the outskirts of the Sunda Strait (We, 2018). The enormity of this tsunami caused substantial material losses and fatalities. The National Disaster Management Agency (BNPB) noted that the death toll from the tsunami in the Sunda Strait reached 437 people, as many as 2,752 houses were damaged, 92 inns and stalls were damaged, 510 boats and ships were damaged, and 147 vehicles were damaged (Farisa, 2018).

The tsunami event on December 22, 2018 was not preceded by a tectonic earthquake and was purely caused by an avalanche due to volcanism. Practically, the BMKG InaTEWS (Indonesia Tsunami Early Warning System), which is designed to detect tsunamis due to tectonic earthquakes, is not able to detect tsunamis with such characteristics. The BMKG was only able to confirm this tsunami event in the form of an official press release after approximately 1.5 hours from the actual time of the incident. Confirmation is done by manually checking the mareogram data of the Tide Gauge equipment in the Sunda Strait through the website of the Geospatial Information Agency (BIG).

The main problem is the ineffectiveness of the tsunami early warning system, especially in detecting non-tectonic tsunami events in the Sunda Strait and concerns about the potential for subsequent events in the future, becoming a strategic issue in the Deputy for Geophysics which is very urgent and has become a serious concern for the BMKG institution. BMKG has started to pioneer the development of a Non-Tectonic Tsunami Early Warning System or known as InaTNT (Indonesia Tsunami Non-Tectonic) to address this major problem. The Sunda Strait area was used as a pilot project for the development and testing of this system.

Basic Concept

In order for the Non-Tectonic Tsunami Early Warning System (InaTNT) to run effectively, its development is focused on a number of stages or main targets, namely:

a. Integrate sea level observation data from various internal and external BMKG stakeholders into an integrated system and display.

Sea level observation data from a number of external sources in the existing BMKG InaTEWS System, have not been integrated into a system or integrated visual display so that tsunami monitoring efforts are less efficient. Therefore, this is the first major step that is the target to be taken. InaTNT integrates sea level observation data from Tide Gauge, Buoy, HF Radar, IDSL and Automatic Weather Station (AWS) Water Level devices from BMKG internal and external stakeholders. Figure 1 and table 1 show the distribution of sea level observation equipment in the Sunda Strait which is currently integrated into InaTNT.

Figure 1. Distribution of sea level observation equipment in the Sunda Strait

All sea level observation data from various external sources are integrated through a pull data mechanism by the Central BMKG InaTNT server via the internet.

b. Provide an automatic detection facility for sea level anomalies indicating a tsunami through mareogram data from sea level observation devices.

This system mainly acts as a head-up for InaTEWS operators to immediately respond and validate whether the sea level anomaly event that occurs is actually a tidal wave indicating a tsunami and cross-checks with seismic data and immediately issues non-tectonic tsunami early warning dissemination through a special dissemination format. However, if the tectonic tsunami early warning system, which is the main existing system, has been activated first, the SOP for regular Tsunami Early Warning and its dissemination will be carried out. This system is connected to the main alarm alert in the InaTEWS operational room which can sound at any time if an indication of a tsunami is found automatically.

InaTNT, in principle, detects or scans the anatomical presence of a signal that indicates a sea level anomaly that indicates a tsunami, adopting the tested automatic earthquake detection method, namely the STA/LTA method. This method uses the ratio between the signal average and the short period window (STA) which represents the window anomaly with the average signal with the long period window (LTA) which represents the basic noise of the data. The STA/LTA ratio of the data if it exceeds the predetermined threshold value means that there is an indication of a sea level anomaly detected in the signal recording. Based on the simulation results of a number of tsunami events triggered by non-tectonic sources such as Palu (submarine erosion) (figure 3) and Krakatau (sea volcano eruption) (figure 4), the results of the STA/LTA impulsive curve detection are quite promising. The detection results on the InaTNT system are then refined using the AIC Picker method to obtain more precise tsunami arrival times at certain locations. This becomes very important when it will be applied to the search for tsunami sources using the Back Propagation method. The trigger will also activate sound alerts in the InaTEWS operational room and also provide notifications on the web display.