Introduction

Dams are defined as an artificial obstruction to natural flows, which helps to accumulate water for agricultural reasons and energy generation, potable water supply, flood mitigation, and also recreational reasons (ICOLD, 2008). On the other hand, dams are mentioned as one of the key assets in the Critical Infrastructures and Key Resources (CIKRs) by the Department of Homeland Security (DHS) in the United States. There was a huge initiative from the federal government to coordinate a set of resilience and protection plans series, started the process of understanding what features create resilience in CIKRs (Gopalakrishnan & Peeta, 2010). Due to this mission, dams and their environment should be designed for a safer, more secure, and more resilient domination by preventing, deterring, neutralizing, or mitigating the effects of deliberate efforts by terrorists to destroy, incapacitates, or exploits their elements (Fisher et al., 2010). There are also other programs for the protection of critical infrastructures such as the Swiss Federal program for critical infrastructure protection (CIP) to analyze its threats and identify vulnerabilities and assessing risks.
In principle, to assess the risk and provide the protection plan for dams and their downstream areas all relevant hazards and threats must be considered. The risks or vulnerabilities related to dams often refer to natural hazards such as earthquakes or floods or other technological failures. On the other hand, the spectrum of hazards and threats can be of many-sided nature, either technology- or human-related, natural and operational as well as contextual, which should be considered while the vulnerability of infrastructures should be assessed. Most of the researches in the last decades focuses on the natural hazards threats spectrum. However, the emergence of terrorist groups such as Al-Qaeda and ISIS revived the idea of protection of human-related threats. The spectrum of human-related threats ranges from unintended errors to targeted malicious attacks, either physical (e.g., explosive devices) or cyber, which deal with sophisticated models capable of describing thought and ideology transposed to terror and destruction (Kröger & Zio, 2011). Thus, risk assessment and vulnerability assessment should expand to all properties related to the dams from farms, transportation routes, energy lines, pipes, business areas, industries, cities or villages, and even biosphere assets. There are much research works on the loss of life and flood mortality highlighted by different models and frameworks especially developed by Jonkman and his teams (S. Jonkman & Vrijling, 2008; S. Jonkman, Vrijling, & Vrouwenvelder, 2008; S. B. Jonkman, Maaskant, Kolen, & Needham, 2016; S. N. Jonkman, Maaskant, Boyd, & Levitan, 2009). He divided different methods in loss estimation into three categories as macro-level by characterizing the event, Meso-level by characterizing the location, and finally micro-level by investigating the individuals (S. B. Jonkman et al., 2016). In the developed equation, the important property is people, and any risk number is calculated based on the number of people at risk and two reducing factors as sheltering fraction (Roozbahani, Zahraie, & Tabesh) and evacuation fraction (FE) which is illustrated in eq.1. The mortality fraction is the other factor (Fd) that has been used in other risk assessment equations (eq.2) where N is a number of people at risk and NPAR is the number of people at risk.