In the 1970s, two types of seismometers were installed on the nearside of the Moon. One type is called the Long-Period (LP) seismometer, which is sensitive below 1.5 Hz. The other is called the Short-Period (SP) seismometer, whose sensitivity is high around 2 – 10 Hz. So far, more than 13,000 seismic events have been identified through LP data analyses, which allowed us to investigate lunar seismicity and the internal structure. On the other hand, most of the SP data have remained unanalyzed because they include numerous unnatural signals and/or instrumental noises. This fact leads to the hypotheses that (i) we have missed lots of high-frequency seismic events and (ii) lunar seismicity could be underestimated. To verify these ideas, this study conducted an analysis of the SP data. In the analysis, I denoised the original SP data and performed the event detections by comparing the spectral features between the cataloged high-frequency events (such as shallow moonquakes) and the continuous SP data. Eventually, I discovered 22,000 new seismic events, including thermal moonquakes, impact-induced events, and shallow moonquakes. Among these, I focused on analyzing shallow moonquakes — tectonic-related quakes. Consequently, it turned out that there are nearly three times more tectonic events than considered before. Furthermore, additional detections of shallow moonquakes enabled me to see the regionality in seismicity. Comparing three landing sites (Apollo 14, 15, and 16), I found that the Apollo 15 site is more seismically active than others. These findings can change the conventional views of lunar seismicity.