Air Plume Model Predictive Plot Checklists

Air Plume Model Predictive Plot Checklists

Use the checklists below to help you assess the utility of an air plume model “predictive plot” in supporting decision-making in responding to a release of air toxics.  Start by using the first checklist to help you examine the plot carefully and do a "quality assurance" check.  If you have additional time and access to individuals who can help you answer the questions (e.g., experts who ran the model and developed your plot), continue with the second checklist to help you understand more about the model and the assumptions used to produce the plot.

As you use these checklists, keep in mind the overarching question you are trying to answer:  Does the plot have enough detail to support the decisions I need to make? 

Checklist 1.  Examine the plot carefully (Quality Assurance)

Factors	Questions to Consider
1) Key locations	·         Are important locations (schools, hospitals, roads) clearly identified on the plot?
2) Interpreting the Plot	·         What do the different color schemes mean? Is there a key/legend? ·         What do the different shaded areas mean in term of their impact on human health (e.g., are they dose or concentration levels)?  ·         For radiation releases: How do these levels differ from normal background levels?
3) Time frame	·         How long ago was the plot generated?  Does it still provide a valid picture of the release?  ·         Is the map a forecast of the total cumulative effects of the release, or does it represent a snapshot in time?
4) Field measurements	·         Are there any field measurements from the area (e.g., from survey teams or aerial flights) on the plot? ·         Are the reports from the field consistent with your plot?
5) Released material	·         Was the release a one time “puff” (a minute or less) or did the release continue.  If so for how long (e.g., minutes, hours)?

Checklist 2.  Understand the inputs and capabilities of the computer model that produced the plot

Additional Factors	Questions to Consider
1) Inputs and Assumptions	·         How was the material in the release identified (e.g., knowledge of the material onsite at a facility, real-time monitoring)? Has the accuracy of the material released been confirmed? ·         How was the amount of material released determined? Has the accuracy been confirmed?  If the release amount is an estimate, is it a best case or worst case estimate? ·         What was the height of the release? Has this been included as an input to the model? ·         Do you have any information from witnesses at or near the scene?  For example, if there a visible smoke plume, which direction is it blowing?
2) Meteorology	·         Where did the weather data used in the predictive plot get collected relative to the release location? ·         How variable is the meteorology (i.e. gusty, or no wind at all)?  Is this reflected in the predictive plot?  ·         Is rain predicted in next 24/48 hours? ·         How does the weather/wind interact with the terrain? ·         Is possible upwind dispersal being taken into account and reflected in the map? ·         Has an expert been consulted on meteorology?
3) Model Capabilities	·         What is it able to model (e.g. single chemical, multiple chemicals such as petrochemicals, radiation, fire or explosions)? ·         Is the model able to use real-world meteorological data?  Does the model incorporate upper air winds in addition to ground level winds? ·         How far downwind is the model capable of making predictions for?   Are there any other limitations of the model to be aware of? ·         Does the model take local topography (hills, buildings, trees) into account? Can it model complex urban environments? ·         How far above ground level is the release?  Is the release part of an explosion or fire (which can transport toxic materials higher into the atmosphere)? Is the model able to accommodate these factors? ·         What height is the plot for (e.g., ground level or above)? ·         Has the expert who provided your predictive plot been consulted?