Abstract
Hazards of sandstorm are increasingly recognized and valued by the general public, scientific researchers, and even government decision-making bodies. This paper proposed an efficient sandstorm prediction method that considered both the effect of atmospheric movement and ground factors on sandstorm occurrence, called improved naive Bayesian-CNN classification algorithm (INB-CNN classification algorithm). Firstly, we established a sandstorm prediction model based on the convolutional neural network algorithm, which considered atmospheric movement factors. Convolutional neural network (CNN) is a deep neural network with convolution structure, which can automatically learn features from massive data. Then, we established a sandstorm prediction model based on the Naive Bayesian algorithm, which considered ground factors. Finally, we established a sandstorm prediction model based on the improved naive Bayesian-CNN classification algorithm. Experimental results showed that the prediction accuracy of the sandstorm prediction model based on INB-CNN classification algorithm is higher than that of others and the model can better reflect the law of sandstorm occurrence. This paper used two algorithms, naive Bayesian algorithm and CNN algorithm, to identify and diagnose the strength of sandstorm in Inner Mongolia and found that combining the two algorithms, INB-CNN classification algorithm had the greatest success in predicting the occurrence of sandstorms.
Highlights
Sandstorms have become an important natural disaster affecting human survival and development
In order to solve this problem, combining with the naive Bayesian algorithm, we proposed an improved sandstorm prediction algorithm, improved naive Bayesian-Convolutional neural network (CNN) classification algorithm (INB-CNN classification algorithm), and established a sandstorm prediction model based on this algorithm and analyzed its prediction accuracy
Taking the test set of infrared satellite cloud image data as the research object, the sandstorm prediction model based on convolutional neural network algorithm was tested, and the prediction accuracy was 0.833. rough many experiments, it was found that according to the normalization rule, when the parameter α is set to 0.3, the accuracy of the sandstorm prediction model based on the INB-CNN classi cation algorithm is highest
Summary
Sandstorms have become an important natural disaster affecting human survival and development. Eir research showed that the method has little value in predicting sandstorms in Northwest China. Lu et al [6] combined the BP neural network with genetic algorithm to establish and implement a sandstorm prediction model. Eir research showed that this hybrid method can effectively detect and predict all types of dust storms. Eir research showed that the model can predict the atmospheric life cycle of dust storm erosion. Eir research showed that the method is more accurate than the classification based on spectral data alone. There are few studies on the occurrence of sandstorms based on satellite cloud images and convolutional neural network algorithm. The sandstorm prediction model based on the convolutional neural network algorithm only considers the influence of atmospheric movement on the occurrence of sandstorms. Convolutional neural network algorithm and naive Bayesian algorithm are very important data mining algorithms. ey are applied to data analysis in the field of meteorology which can explore the internal relations among various meteorological elements and find various potential laws to reveal unknown meteorological theories. ey are important for meteorological science research and useful in enriching weather forecasting methods and improving the level of weather forecasting, which plays an active and important role
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
