To distinguish debris flow infrasound from other infrasound sources, previous works have used one-dimensional infrasound shapes and parameters. In this study, we converted infrasound signals into two-dimensional signal time–frequency graphs and created a time–frequency graph dataset containing five common kinds of infrasound. We used deep learning to distinguish debris flow infrasound from other infrasound and improved the deep learning model to enhance the accuracy of debris flow infrasound identification. By improving the LeNet-5 network, we obtained an infrasound signal recognition method for debris flows based on deep learning. After signal preprocessing and model training, this method was able to differentiate target infrasound from environmental infrasound, and a debris flow infrasound recognition accuracy of 84.1% was achieved. The method described in this paper can effectively recognize debris flow infrasound and distinguish it from other environmental infrasound. By such means, more accurate and more timely debris flow disaster warnings may be obtained.

Representational state transfer (REST) is a resource-based service architectural style. It abstracts data and services as resources and accesses them through a unique Uniform Resource Identifier (URI). Compared with traditional Simple Object Access Protocol (SOAP) methods, REST is more concise. It takes full advantage of HyperText Transfer Protocol (HTTP) and has better scalability and extensibility. Based on REST services, this article integrates geographic information, real-time disaster monitoring data, and warning services in a three-dimensional (3D) digital Earth infrastructure and establishes a three-dimensional geological disaster monitoring GIS platform with good service compatibility and extensibility. The platform visually displays geographical and geological information and real-time monitoring data in a three-dimensional Earth, accesses warning model services to implement disaster warnings, and realizes comprehensive information management, monitoring, and warnings of multiple types of geological disasters. This can provide decision support for disaster prevention and relief and improve the informatization of geological disaster prevention and control.

Due to the inability of remote sensing satellites to monitor avalanches in real time, this study focuses on the glaciers in the rear edge of Jialongcuo, Tibet, and uses infrasound sensors to conduct real-time monitoring of ice avalanches. The following conclusions are drawn: (1) In terms of waveform, compared to background noise, ice avalanche events have a slight left deviation and a slightly steep shape; compared to wind, rain, and floods events, ice avalanche events have less obvious kurtosis and skewness. (2) In terms of frequency distribution, the infrasound frequency generated by ice avalanche events is mainly distributed in the range of 1.5 Hz to 9.5 Hz; compared to other events, ice avalanche events differ some in frequency characteristics. (3) The model based on information entropy and marginal spectral frequency distribution characteristics of infrasound have higher accuracy in signal classification and recognition, as they can better represent the differences between infrasound signals of different events than other features. (4) Compared with the K-nearest neighbor algorithm and classification tree algorithm, the support vector machine and BP (Back Propagation) neural network algorithm are more suitable for identifying infrasound signals in the Jialongcuo ice avalanche. The research results can provide theoretical support for the application of infrasound-based ice avalanche monitoring technology.