In addition to directly characterizing the extent and level of contamination at the release site, assessments typically define the area in which people, animals, and the environment may be affected by a chemical release, and help estimate the population at risk – that is, the number of people within a region where adverse health effects are possible. These estimates are often generated by modeling tools such as those listed in the Planning, Decision Support, and Modeling Resources for Chemical Incidents section of this document. Tools used each have strengths and weaknesses that are useful to keep in mind when planning response and recovery activities.
Modeling tools can be useful in estimating contamination plumes. However, airborne toxic clouds may meander due to changes in wind direction as they move. This characteristic simultaneously reduces the extent of the direct downwind hazard area while widening the potential hazard area. In some instances, modeling assessments based on current local meteorological conditions can account for uncertainty in the wind direction and estimate the resulting larger area that may be at risk. Modeling can also provide a picture of the downstream movement of chemicals released into a body of water. However, modeling tool outputs are likely to overestimate the extent of the hazard area because they don’t completely account for losses of the chemical in a real-world environment. These losses may be due to reaction with environmental surfaces such as buildings, trees, the ground, water, or a riverbed.
Moreover, buildings can protect their occupants from exposure via the open air, depending upon the speed of the plume, the volatility of the substance, the ability to close building ventilation systems, the integrity of the building envelope, the time of year, location of air intakes, etc. However, standard modeling assessments often do not account for this protection. The fact that more than 75% of people are indoors at any given moment in a typical day has the potential to substantially reduce exposure, especially in urban settings.
Buildings channel the toxic cloud along open streets in urban settings, resulting in hazardous conditions that extend much farther than predicted by typical models, which do not include specific buildings as part of downwind hazard calculations. In addition, congested spaces within the urban environment slow the toxic cloud’s passage and cause the airborne material to linger. When lingering effects are not considered during modeling, the toxic cloud may be predicted to leave the area faster than is the actual case.
A major source of uncertainty when using modeling in the immediate aftermath of a disaster is knowing how much material was released. Response modeling may assume the entire available volume has been released – potentially overestimating the amount of toxic material released and the overall areas impacted. Once again, this assumption will lead to an overestimate of the possible area affected by the hazard, which is probably the most prudent assumption to use until data can be gathered to demonstrate that other areas are truly safe. For all these reasons, sharing of new data regarding the incident as it becomes available is critical to ensuring modeling estimates are refined and updated, and are providing the best possible data to support decision-making.