Technical Support for Exceptional Event Analysis

for Volcano Impacts on PM2.5 in Hawaii

using theExceptional Event Decision Support System (EE DSS)

Informal background presentation to EPA Region 9 and Hawaii Det. Health

September 19, 2013 Telecon

Rudolf HusarWashington University, St. Louis, MO rhusar@wustl.edu

Response to the Bohnecamp memoRudy Husar rhusar@wustl.edu, Kari Hoijarvi, Washington University, St. Louis

• EE Detection – Which monitors are candidate violators of PM2.5 NAAQS? – What days have values above the PM2.5 NAAQS?

• EE Documentation– What are the other sources of PM2.5 near the monitor(s)

• Do they impact the monitors on exceedance days?• Can one illustrate the effect of other sources vs. the volcano?

– What are the conditions when volcano impacts on EE monitors? • What is the volcanic source/emission strength?• What are the met. conditions/pathways of volcanic impacts on EE monitors?• Are these consistent for most events?

– Are the tools and methods adequate to document the above?• Are existing surface & satellite observations and modeling tools available? • Can those be combined/integrated for robust EE documentation?

• How/when is this to be done? (C. Bohnencamp memo)

Candidate for violating Daily PM2.5 NAAQS

None of the Hawaii monitors are candidates of daily PM2.5 NAAQS violations

Candidates for violating Yearly PM2.5 NAAQS

Kona and Hove(?) on Hawaii are candidates of yearly PM2.5 NAAQS violations

Candidate stations for violating daily or yearly PM2.5 NAAQS

What days have values above the PM2.5 NAAQS?Time series for Kona on standard DataFed Browser

For 2011-2012 the average PM2.5 = 14.2 ug/m3Potentially, any sample over 12 ug/m3 is violating and a candidate for EE flag

Average = 14.2

What days have values above the PM2.5 NAAQS?Based on AQS PM2.5 data for 2011 and 2012

There are 704 daily samples. Of these 477 samples or 68% are > 12 ug/m3 (black dots). If all 477 were successfully flagged, the monitor average would be 8.5 ug/m3 be 8.5 ug/m3 – too many flagged

Compliance could be reached if samples with concentration above 17.5 ug/m3 (168 samples) were successfully flaggedExtrapolated to three years, 2011-13, there will be at least ~ 250 days to be flagged at Kona.

What days have values above the PM2.5 NAAQS? For Kona (2011-12), determined by the Hawaii EE DSS Screening Tool

There are other strategies for flag-day selection from the pool of 477 that is above 12 ug/m3. Another approach is based on spatial anomaly, i.e. deviation from the neighboring sites. Using that criteria would require flagging 174 days, compared to 168 days.

Given the many (over ~250) days to be flagged, choosing a good flag sampling and flag documentation strategy is a key task for the collaborating workgroup.

Other sources of PM2.5 impacting the EE monitor(s)Background PM2.5 – oceanic aerosol

A possible source of background spikes is Asian dust reaching HI – suggested by the NRL NAAPS modelThese should be considered as ‘other sources’.

The Hawaii background PM2.5 is quite uniform at ~ 6 ug/m3But spikes up to 12 ug/m3 can occur near-simultaneously at all monitors.

Aggregate of 6 background monitors

Distinguishing volcano vs background PM2.5 sourcesSpatial anomaly: Excess over the fluctuating background PM2.5 concentration

Red line is the spatial anomaly: Excess PM2.5 at KONA over the spatially uniform but time-varying background

Spatial analysis could help distinguishing background and volcanic PM2.5

Background ~ 6 ug/m3

OMI Satellite Columnar SO2 Seasonal averages, 2004-2010

The SO2 concentration is highest at the volcano – in the vicinity of the EE sourcesThe transport is mostly toward the West with the trade winds

MODIS satellite measurements of volcano aerosol (sulfate)Seasonal averages of AOD, 2004-2010

High AOD downwind of Hawaii, the volcano source. No evidence for other major sourcesSeasonally the volcano aerosol plume is more in the Spring and SummerPlume transport is with the trade winds

OMI satellite measurements of columnar NO2Seasonal averages, 2004-2010

Indicates the high anthropogenic emissions in Oahu. No NO2 signal over Hawaii

Why is the PM2.5 impact so far away from the SO2 source?SO2-Sulfate conversion takes time and distance

OMI SO2 Column Concentration

MODIS AOD PM Column

Satellite measurements of SO2 and AODAverage 2004-2010, June, Jul August

SO2 highest near the source

Aerosol Optical Depth (AOD) highest downwind

Possibly accelerated in fog/clouds -> VOG

Aspects of volcanic emission transport: Conceptual

Stagnating, mixing ‘pool’ in the wake of the island

Routine wind data in EE DSS Surface Winds: Measurements (hourly)

NOGAPS model (6 hourly, vertical profiles

Hawaii EE Documentation Tool

Surface Winds OMI satellite SO2

MODIS AQUA Image MODIS Terra

MODIS AOT

Bad data

All the analysis was performed using different modules in the Hawaii version of the EE Decision Support System. The Hawaii system documentation is marginal at best .

EE DSS Integration with VMAP: Vog Measurement and Predication System

and other systems?

Collaboration with other science teams would also be desirable