Johns Hopkins University, Krieger School of Arts and Sciences (Advanced Academic Programs), Washington DC, 20036, USA.
Sci Rep. 2018 Jul 27;8(1):11342. doi: 10.1038/s41598-018-29680-4.
A series of Poisson distributions are fit to sets of global cost-of-impact data representing large-scale accidents and anthropogenic catastrophes. The fits are used to build a function representing data means and are designated the Inverse Poisson Functional. Climate and environmental data have been used to develop a cost-frequency population distribution and to estimate the expected time between events. On a global scale, we show that expected wait- or reaction- times can be estimated using the Poisson density function. The functional is generated, representing the locus of means (peaks) from the individual Poisson distributions from different impact costs. Past (ex-post) forecasts relate to a range of natural and anthropogenic disasters; future (ex-ante) forecast presents global CO emissions. This paper shows that a substantial reaction to global climate change (CO emissions extremum) will occur in 55 to 120 years (95% CI) with a model prediction of 80 years.
将一系列泊松分布拟合到代表大规模事故和人为灾害的全球影响成本数据集上。这些拟合用于构建表示数据均值的函数，并被指定为逆泊松函数。已经使用气候和环境数据来开发成本-频率人口分布，并估计事件之间的预期时间。在全球范围内，我们表明可以使用泊松密度函数估计等待或反应时间。该函数是通过从不同影响成本的单个泊松分布中生成表示均值（峰值）的轨迹来生成的。过去（事后）的预测涉及一系列自然和人为灾害；未来（事前）的预测呈现全球 CO 排放。本文表明，对全球气候变化（CO 排放极值）的强烈反应将在 55 到 120 年内发生（95%置信区间），模型预测为 80 年。
