Sn

1
Comparison of gravimetric PM data from the
Harvard Impactors and Gent Stacked Unit PM10
Samplers in Prague 2004
M. CIVIŠ1, J. HOVORKA1 and J. SCHWARZ2 1Institute
for Environmental Studies, Charles University,
Benátská 2, 128 01, Prague, Czech
Republic 2Institute of Chemical Process
Fundamentals, Academy of Sciences of the Czech
Republic, Rozvojová 2, 165 02, Prague, Czech
Republic

• Motivation
• Systematic sampling (every third day) and during
  four intensive seasonal campaigns (14 days per
  campaign) of aerosol fractions PM10, PM2.5 , PM1
  using Harvard impactors (HI) and PM10, PM2.2
  using Gent Stacked Unit PM10 Samplers (SFU) on
  two locations in Prague within the year 2004.
• Evaluation of the both manual PMx sampling
  methods regarding to meteorology parameters.
• Comparison of HI method with SFU method on
  station 1 and comparison of SFU method on both
  stations.

N
Station 2 (suburban) N 50º07.643 E
014º23.064 285 m ASL
Harvard Impactor HI HI for PM10 and PM2.5 are
single staged with flow rate 10 l/min while HI
PM1 has two stages and flow rate about 23 l/min.
PM10 was sampled on Quartz filters (37 mm, pore
size 0.1-4 µm). PM2.5 and PM1 fractions were
caught on PTFE filters (37 mm, PE holding ring,
pore size 2 µm)
Gent Stacked Unit PM10 - SFU At the inlet of
SFU sampler are single staged, PM10 impactor with
flow rate 16 l/min. Then aerosol passes through
filter cassette, consisting in-serial two PTFE
Nuclepore filters. Pore size of the first filter
is 8 µm. Pore size of the second filter is 0.4
µm. Coarse of particles from 10-2.2 µm is
collected on the first filter, particles smaller
than 2.2 µm are collected on the second filter
Station 1 (urban) N 50º04.280 E 014º25.253
225 m ASL
Pearsont Correlation of the methods
Relations between PM and average day temperature
    Station 1 Station 1 Station 1 Station 1 Station 1 Station 2 Station 2
    PM10 PM2,5 PM1 SFU10 SFU2,2 SFU10 SFU2,2
St. 1 PM10 1            
St. 1   81/0.22            
St. 1 PM2.5 0.44 1          
St. 1   81/0.22 80/0.22          
St. 1 PM1 0.44 0.62 1        
St. 1   81/0.22 80/0.22 80/0.22        
St. 1 SFU10 0.2 0.63 0.51 1      
St. 1   71/0.23 73/0.23 68/0.25 52/0.27      
St. 1 SFU2.2 0.04 0.33 0.34 0.88 1    
St. 1   58/0.27 57/0.27 57/0.27 52/0.27 52/0.27    
St. 2 SFU10 0.24 0.46 0.29 0.52 0.42 1  
St. 2   81/0.22 80/0.22 80/0.22 52/0.27 52/0.27 81/0.22  
St. 2 SFU2.2 0.43 0.56 0.74 0.43 0.44 0.93 1
St. 2   81/0.22 80/0.22 80/0.22 52/0.27 52/0.27 81/0.22 81/0.22
Observed time period Method and location Regression equation Correlation coefficient, number of pairs/critical values
Spring campaign 20.5.-5.6. PM10 St.1 y0.02x1.3 0.4 (14/0.48)
Summer campaign 6.6.-20.6. PM2.5 St. 1 y0.09x1.5 0.76 (15/0.48)
  PM1 St. 1 y0.07x1.3 0.49 (15/0.48)
  SFU10 St. 2 y0.09x1.3 0.78 (13/0.51)
Autumn campaign 9.10.-25.10. PM2.5 St. 1 y0.11x2.2 0.57 (15/0.48)
  PM1 St. 1 y0.12x1.7 0.62 (15/0.48)
  SFU10 St. 2 y0.08x2.3 0.50 (14/0.5)
Year 2004 SFU2.2 St.1 y-0.04x2.7 -0.35 (58/0.27)
  SFU2.2 St. 2 y-0.02x2.8 -0.27 (95/0.21)
PM10 St. 1 y-0.02x1.7 -0.29 (89/0.22)
Relations between PM and average day wind speed
Observed time period Method and location Regression equation Correlation coefficient, number of pairs/ critical values
Winter campaign 16.2.-27.2. PM2.5 St. 1 y-0.21x4.5 -0.59 (12/0.53)
Spring campaign 20.5.-5.6. PM2.5 St. 1 y-0.09x3.5 -0.51 (15/0.48)
  SFU2.2 St. 2 y-0.53x3.4 -0.49 (15/048)
Year 2004 SFU10 St. 1 y-0.13x3.2 -0.39 (74/0.23)
  SFU10 St. 2 y-0.11x3.3 -0.22 (94/0.21)
HI coarse St. 1 y-5.79x52.7 -0.31(89/0.22)
SFU coarse St. 1 y-2.17x16.7 -0.28 (76/0.23)

• Conclusion
• Sampling with HI method measured ( a 0.05)
  higher concentrations of PM10 and PM2.5 than SFU
  PM10 and PM2.2 on both stations. Results show
  that air masses are better mixed in Prague for
  PM2.5 which does not apply for PM10.
• SFU2.2 on both stations correlated well with
  average day temperature during the whole year
  2004. SFU10 on both stations correlated with
  average day wind speed during the whole year
  2004. Correlation of aerosol fractions with
  relative humidity was not found. Negative
  correlations between coarse fractions and average
  wind speed measured by both methods on station 1
  was observed. Highest concentrations of HI coarse
  occurred with wind blowing 2-3 m/s. Highest SFU
  coarse concentrations occurred with wind speed
  1-2 m/s.

The study was conducted within the project
Comprehensive size resolved characterization of
atmospheric particulate matter in Prague
supported by GACR grant No. 205/03/1560