Transcript Causes of Haze Assessment Update for Fire Emissions Joint Forum 12

Causes of Haze Assessment
Update for Fire Emissions Joint
Forum -12/9/04 Meeting
Marc Pitchford
Causes of Haze Assessment
(COHA)
• Overview
– Starting the third year of a 4-year contract effort to
use ambient monitoring data & some emissions
information to assess haze influential factors (e.g.
aerosol species, atmospheric processes, source
types & regions)
– Assessment information & products are stored &
distributed on the COHA web site (no paper report)
http://www.coha.dri.edu/
– Primary uses of the COHA products are to support the
technical assessments for SIPs & TIPs, and
specifically the Attribution of Haze Workgroup effort
COHA Tasks
• Completed for each Class I Area
– Emissions mapping & Descriptions (maps of NOx, SO2 & fire
within 20km)
– Monitoring Site Setting Descriptions (maps of terrain, landuse, air quality & met. monitoring sites, urban & industrial
locations)
– Meteorological Site Setting Descriptions (local flow patterns,
site representativeness, climate data, etc.)
– Aerosol Descriptions (figures, tables, & text of typical, best &
worst aerosol components, monthly distribution & composition of
worst days)
– Back Trajectory Analysis (3 years of 8 per day 8-day back
trajectories at 3 starting heights [>3 million trajectories],
summary maps of all days, best & worst haze days, best & worst
for each component days, etc [>5000 summary maps])
– Trajectory Regression Analysis (statistical relationship
between transport time over source regions & air quality
measured at the monitoring site, sufficient data at about 80 sites,
used by the AOH workgroup to compare to WRAP modeling
attribution results
Bandelier Wilderness Area : Nearby Emissions
US Fire Database
US National fire database 1970 to 2002 reported
by USFS, BLM, BIA, NPS, FWS
Key Data Fields:
• Lat/long
• Fire start date (date
discovered & controlled
on USFS land only)
• Area
• Cause
• QA flags (DRI)
Canadian Fire Database
Canadian Large Fire Database 1959 to1999
Large forest fires in Canada, 1959–1997, Stocks et al., 2003 J.G.R.
Key Data Fields:
• Lat/long
• Province
• Region
• Fire start date
• Size
• Cause
MODIS Fire Database
2001 to present from Forest Service, Remote
Sensing Application Center
http://activefiremaps.fs.fed.us/
The fire detections are discerned using the 1-km
thermal bands of MODIS. Detections are
provided as the centroids of the 1-km pixels
2002 fires
shown here
Key Data Fields:
• Lat/long
• Fire start date
• Brightness temp
• Pixels
NOAA NESDID Satellite Services
Division Fire Database
• 2003 to present
• Detections from GOES, AVHRR, DMSP/OLS
and MODIS
• Archived visible smoke plume polygons
Key Data Fields:
• Lat/long
• Time
http://www.firedetect.noaa.gov
Alaska Fire Database
• Alaska Forest Service
• Large wildland fires from 1950 to 2003
• Provides only year not day
Key Data Fields:
• Lat/long
• Year
• Area, perimeter
• Fire ID
http://agdc.usgs.gov/data/blm/fire/index.html
Example Trajectory
Summary Maps
Bandelier National Park –
residence time maps show
fraction of transport hours spent
over each 1o by 1o cell.
• Top map shows the residence
time for the worst 20% haze
days
• Bottom map shows the
residence time for the best 20%
haze days
Trajectory Regression Results
• Bar graphs of concentration & percent
contributions associated with 15 to 20
source regions
• Source regions are based on state
borders
• Includes statistical uncertainty bars &
tabular regression analysis output (not
shown)
• Currently applied to sulfate and aerosol
extinction data
COHA Tasks
• To be completed
– Episode analysis (custom assessment of
causes of worst haze periods using
supplemental information, examples: forest
fires, regional & global dust, eastern sulfate
transported west, west coast nitrates, etc) ~10
episodes completed to date
– Receptor modeling (more transport
regression, CMB, etc.)
– Update previous work with additional data
Additional COHA Activities
• Tribal Causes of Haze (conduct COHA
for tribal Class I Areas, assess regional
haze monitoring needs for tribal lands)
• Dust Causes of Haze (develop methods
to characterize dust by source
type/geographic scale, apply method to
several years of data)