A combination of machine learning methods was used to create UrbanDenoiser. For training application
The team applied their filtering system toseismic data recorded in Long Beach, California to see how well it performs. They found that the desired signal level improved over background noise by about 15 dB. UrbanDenoiser was then used to analyze data from an earthquake that hit a nearby area in 2014. As a result, the app was able to detect four times more data than unfiltered sensors.
The study was conducted by a team of researchers from Stanford University in conjunction with the Chinese Academy of Sciences.
The developers claim that UrbanDenoiseris able to determine which seismic data are natural for the urban environment and which are artificial, that is, caused by earthquakes, and filters out unnatural ones. The app can also detect tremors of varying magnitudes.
UrbanDenoiser is able to effectively suppresshigh noise levels, although false positives and false negatives are still possible in denoised data, the researchers acknowledge. False negatives occur when the seismic signal is too weak or when the target seismic phases and samples of the training signal are not sufficiently similar to the waveforms of the real earthquake.
Earthquakes cause the most damage and harmespecially in densely populated cities. At the same time, seismologists face difficulties in separating seismic data associated with natural ground movement caused by urban life. They note that human activities in cities, such as vehicles and trains, produce a lot of seismic noise.
The researchers hope that the new application will help to get clear data on earthquakes in cities and thereby minimize damage.
Read more:
MIT builds stationary heat engine that outperforms turbines
After ten years of work, scientists questioned the standard model of physics
See what sunrise looks like on Mars