Quantifying the economic impact of disasters on businesses using human mobility data: a Bayesian causal inference approach
Quantifying the economic impact of disasters on businesses using human mobility data: a Bayesian causal inference approach
复制标题
DOI：
10.1140/epjds/s13688-020-00255-6
复制
发表时间：
2020-04
期刊：
影响因子：
3.6
通讯作者：
T. Yabe;Yunchang Zhang;S. Ukkusuri
中科院分区：
文献类型：
--
作者：
T. Yabe;Yunchang Zhang;S. Ukkusuri
近年来，自然灾害等极端冲击发生的频率和强度不断增加，给全球许多城市造成了重大经济损失。量化极端冲击后当地企业的经济成本对于灾后评估和灾前规划非常重要。传统上，调查一直是用于量化灾害对企业造成的损失的主要数据来源。然而，调查往往存在成本高、实施时间长、观测时空稀疏以及可扩展性有限等问题。近年来，大规模的人员流动数据（例如手机GPS）已被用于以前所未有的时空粒度和规模来观察和分析人员流动模式。在这项工作中，我们使用从手机收集的位置数据来估计和分析飓风对业务绩效的因果影响。为了量化灾难的因果影响，我们使用贝叶斯结构时间序列模型来预测受影响企业的反事实绩效（如果灾难没有发生怎么办？），该模型可能会使用灾区以外其他企业的绩效作为协变量。该方法经过测试，可量化波多黎各 9 个类别的 635 家企业在飓风玛丽亚之后的恢复能力。此外，分层贝叶斯模型用于揭示位置和类别等业务特征对企业长期弹性的影响。该研究提出了一种新颖且更有效的方法来量化业务弹性，可以帮助政策制定者进行备灾和救灾过程。
In recent years, extreme shocks, such as natural disasters, are increasing in both frequency and intensity, causing significant economic loss to many cities around the world. Quantifying the economic cost of local businesses after extreme shocks is important for post-disaster assessment and pre-disaster planning. Conventionally, surveys have been the primary source of data used to quantify damages inflicted on businesses by disasters. However, surveys often suffer from high cost and long time for implementation, spatio-temporal sparsity in observations, and limitations in scalability. Recently, large scale human mobility data (e.g. mobile phone GPS) have been used to observe and analyze human mobility patterns in an unprecedented spatio-temporal granularity and scale. In this work, we use location data collected from mobile phones to estimate and analyze the causal impact of hurricanes on business performance. To quantify the causal impact of the disaster, we use a Bayesian structural time series model to predict the counterfactual performances of affected businesses (what if the disaster did not occur?), which may use performances of other businesses outside the disaster areas as covariates. The method is tested to quantify the resilience of 635 businesses across 9 categories in Puerto Rico after Hurricane Maria. Furthermore, hierarchical Bayesian models are used to reveal the effect of business characteristics such as location and category on the long-term resilience of businesses. The study presents a novel and more efficient method to quantify business resilience, which could assist policy makers in disaster preparation and relief processes.