Short-Term PM2.5 Mortality

Are there any well-regarded short-term mortality relationships for PM2.5 that have been used in BenMAP or would be well-suited for BenMAP? What would be the best way to handle a PM2.5 episode lasting only a couple of weeks?

Hi David,

I think it depends in part on the location in which you are estimating impacts and the type of event. Can you tell me more about the characteristics of the project?

Thanks for the response, Neal.

It would be for Hawaiian volcano eruptions. I know there are a lot of ways to potentially look at the impacts, but I’m curious what you think might be most appropriate? For now we are just looking at elevated PM2.5 over about 2 weeks from model data. What I’ve done in the past for projects with only summertime data is to input realistic daily values for the rest of the year, but have no change in concentrations between the base and control grids. Then applying long-term mortality HIFs for PM2.5 gives a picture of the long-term impacts of increased summertime exposure. However, am I correct in assuming that using this method would produce the same results if changes in annual average concentrations were due to a large increase over a short period of time or a small increase or a long period of time as long as the annual average works out to be the same? If PM2.5 levels were only altered over a 2-week period would this yield strange results using a long-term mortality HIF?



Hi David,

That’s a tricky one. Here are a few quick thoughts:

  • I’d suggest using an effect coefficient from a short-term time series study. I’m concerned that using a LT function wouldn’t be an appropriate way to characterize impacts from a 2-week episode.

  • However, I’m not familiar with any ST PM mortality studies performed in HI. I’ve seen one study assessing Volcano-related SO2 impacts. I’m not sure if they looked at mortality or not. You can transfer a function from a study performed elsewhere, but you’d of course need to make some strong assumptions about time-activity patterns, particle composition, etc.

  • I’d recommend looking first at the 2009 Particulate Matter Integrated Science Assessment. The 2019 ISA, while still draft and subject to change, is another valuable resource.

I hope this helps. Best of luck with your analysis!