Probably yes.

There are two parts to your question. First is to see if the signal data is detecting any earth quake anomalies and the second part is predicting time to earth quakes.

For real time you need to build the model first and use real-time signal data to do scoring live. In python flask apps are built just to do that.

Yes, deep learning can be used to forecast earthquakes.

But first we need to understand that there are two types of uncertainties when it comes to dealing with earthquakes : 1.) Date and time 2.) Magnitude

To predict the two accurately, scientists require tons of historical data (in order to generate patterns) and terabytes of real time data analysing the current state of the earth (in order to match the patterns)

#############

Infact, a company called Terra Seismic says that earthquakes can be predicted 20-30 days before they occur. Using Big Data and satellite technology, the firm processes large volumes of satellite data which is taken each day from regions where the probability of an earthquake is huge and ground based sensors. This data is then combined with a huge number of earthquake precursors such as ground water level changes, sudden clouds, bizarre behavior of animals, birds and fishes, changes of the ground conductivity, geomagnetic and gravity anomalies, electromagnetic emissions, anomalous atmospheric electric field, geochemical aberrations such as excessive emissions of radon, hydrogen, helium, carbon dioxide, methane and other gases and fluids, and variations in seismic waves velocities. Algorithms built by the firm then analyze this combined data to judge the probability of an earthquake.

With the appropiate learning examples, it is possible