Figure 2 - uploaded by Xilei Zhao
Content may be subject to copyright.
1: CoPE-WELL systems dynamics model 

1: CoPE-WELL systems dynamics model 

Source publication
Thesis
Full-text available
In an increasingly urbanizing world with growing threats of climate change and terrorism, hazards occur more frequently with more severe consequences, bringing significant long-term impacts and requiring years for a community to recover. In order to be better prepared and reduce the impacts of adverse events, communities should conduct effective em...

Contexts in source publication

Context 1
... [Zhao and Spall, 2016], we discuss the modeling process for a full system with multiple subsystems (as shown in Figure 4.2). However, for a general network, mul- tiple full systems and subsystems need to be considered for analysis. Therefore, we extend this full-system/subsystem concept to a general network as shown in Figure 4.3. Taking a small general transportation network as an example (see Figure ...
Context 2
... this section, we present the methodology for developing a hazard-specific weight- ing scheme using linear regression model. One major reason is that most resilience indices provide users a composite with linear aggregation of all indicators from differ- ent domains, so linear regression is a natural and good choice to estimate the weights of indicators, which can be directly used to aggregate indicators to compute the re- silience indices. Another reason is that linear regression has simple model structure, making it easy to interpret and draw inferences. In the following analysis, we will follow the regime shown in Figure 2.2 to deter- mine the relative weights of indicators in the engineered systems. So, first, we need to identify the indicators required to represent engineered systems in the community resilience model and acquire the data for these indicators. Specifically, the engineered systems, also referred to the infrastructure system, contains multiple subsystems in- cluding buildings, communications/cyber, transportation, water, wastewater, power, natural gas, etc. Note that we only choose indicators with publicly available data, so all the stakeholders, including emergency managers, urban planners, policy makers, and researchers can easily gain access to the data and reproduce the work. Since we could not find publicly available data (for all U.S. counties) for appropriate indica- tors to represent wastewater, power, and natural gas, we finalize seven indicators in the aspects of buildings, communications/cyber, transportation and water to serve as proxies to measure the resilience of the engineered ...
Context 3
... Figure 2. Table 2.4, we show the relative weights and direction of indicators for the best model, where the direction ...
Context 4
... Figure 2.6, we show the values of engineered systems domain (Z-score) com- puted by two weighting methods (i.e., unit weighting and data-driven weighting) in the hurricane case. The results are interesting: 1) we find that many counties share similar values under two weighting schemes, so when disaster damage data is not available, the index computed from unit weighting scheme could still be meaningful for many counties; 2) some counties do experience changes when we use data-driven weighting scheme, for example, Cape May County tends to have lower resilience of engineered systems while Middlesex County and Fairfield County show higher values of the engineered systems domain. Therefore, the hazard-specific weighting scheme is a scientific and sophisticated approach to helping people evaluate and quantify community resilience by using the "index" ...