A Study on Natural Disasters Predictive Model Using Periodic Natural Phenomenon Based on Non-homogenous Poisson Process

Wenhui Ju,Guofeng Su,Priscilla Obeng Oforiwaa
2023-01-01
Abstract:Extreme weather and climate events strengthen the frequency and magnitude of natural disasters, the most threatening risks to human life. To understand how climate change affects disasters, an improved predictive model of natural disasters based on the Non-homogenous Poisson process was developed. Using historical events and data from BC180 to 2019 in China, an intensity function curve, expressed by the Fourier series expansion, was fitted with a 95% confidence interval giving a distinct result scale. It showed that natural disasters do fluctuate periodically and are significantly impacted by environmental factors, such as sunspot activities. Currently, the earth is in its peak phase, implying a probabilistic increase in the future occurrence of natural disasters. Although limited by historical disasters data and the analysis of humans’ positive efforts to climate change, the research may provide the expected frequencies of natural disasters to guide in building mitigation and resilience.
What problem does this paper attempt to address?