To address the problems of the text of earthquake emergency information keeps changing incrementally with the time of an earthquake’s occurrence and there being more and more information categories, thus making it difficult to identify earthquake emergency key information, this paper proposes an intelligent recognition algorithm of earthquake emergency information based on the optimized BERT-BiLSTM-CRF algorithm. Based on the historical seismic emergency information dataset of the past 10 years, first, the BIO sequence labeling method is used to classify the seismic entities, and the BERT pretraining model is constructed to represent the seismic emergency text with sentence-level feature vectors. The BiLSTM algorithm is used to obtain the contextual information of the bidirectional seismic emergency text, and we introduce the attention mechanism to enhance the recognition effect of the seismic emergency key information in the statements. Finally, we use conditional randomization to enhance the recognition of earthquake emergency key information in the utterance. The conditional randomization algorithm is applied to extract the dependency relationship between adjacent vectors and improve the accuracy identification to realize the intelligent recognition of earthquake emergency information. The experimental results show that our model can extract earthquake emergency information from online media efficiently and accurately, with better performance than other baseline models.

This paper proposes a novel ship detection method based on analyzing the sea state in optical images. This method is composed of three phases. First, the image is segmented with the improved region splitting and merging method, which divides the sea into separated regions. Then, the sea state of each divided region of sea is analyzed by extracting texture roughness and ripple density of a modified differential box counting (DBC) method. Finally, an appropriate algorithm is applied to detect ships for each region of sea. Experimental results test on 36 real remote sensing images and 133 images obtained from Google earth demonstrate that the method is free of image resolution and has little limitation of sea conditions.

Using an iterative numerical approach, we have obtained the self-consistent thermal expansion, heat capacity, and Grüneisen parameters of diopside (MgCaSi₂O₆) over wide pressure and temperature ranges based on experimental data from the literature. Our results agree well with the published experimental and theoretical data. The determined thermodynamic parameters exhibit nonlinear dependences with increasing pressure. Compared with other minerals in the upper mantle, we found that the adiabatic temperature gradient obtained using the thermodynamic data of diopside is larger than that of garnet while lower than that of olivine, when ignoring the Fe incorporation. Combining our results with thermodynamic parameters of garnet obtained in previous studies, we have estimated the adiabatic temperature gradient and geotherm of an eclogitic upper mantle in a depth range of 200–450 km. The results show that the estimated adiabatic temperature gradient of the eclogite model is ~16% and ~3% lower than that of the pyrolite model at a depth of 200 km and 410 km, respectively. However, the high mantle potential temperature of the eclogite model leads to a similar temperature as the pyrolite model in a depth range of 200–410 km.