ABSTRACT
The analysis of stored stochastic simulation outputs dates to at least the early 1990s with the Scenario Optimization of Ron Dembo. 1 Similar stochastic libraries have been used in Financial Engineering and Insurance. The idea forms the backbone of the discipline of probability management 2 in which uncertainties are stored as vectors of simulated results (Stochastic Information Packets, or SIPs). Think of these as vectors of Monte Carlo or historical data in a column of Excel or R. SIPs may be used with simple vector arithmetic in subsequent calculations to create new SIPs. This concept is not new but benefitted greatly from the Open SIPmath™ Standards developed by the nonprofit that I direct, https://www.w3.org/1999/xlink" xlink:href="https://www.ProbabilityManagement.org">ProbabilityManagement.org. 3 This allows uncertain quantities to be stored as auditable cross-platform data, which, in turn, allows stochastic simulations to be linked into collaborative networks. For example, stochastic libraries of climate-related hazards developed within large time dynamic systems may be used by decision-makers across multiple locations in multiple environments such as Python, R, and Excel. 4, 5 .
