This study investigated the effects of wind direction on trace metal concentrations in particulate matter in southeast Kansas. Air quality monitoring was conducted from March to October 2000 at nine sites. Concentrations of six metals (beryllium, chromium, arsenic, cadmium, barium, and lead) were analyzed. The results showed that wind direction significantly impacted chromium, barium, and lead concentrations, with calm/variable winds associated with higher chromium and lead levels and north winds linked to higher barium. However, upwind/downwind comparisons found no significant differences near targeted industrial sources. Mean levels of chromium, arsenic, and cadmium exceeded EPA risk levels at some sites.

1. Effects of Wind Direction on
Trace Metal Concentration
in Southeast Kansas
Sergio A. Guerra, Dennis D. Lane, Glen A. Marotz,
Ray E. Carter, Richard W. Baldauf, Carrie M. Hohl
Department of Civil, Environmental, and Architectural
Engineering, University of Kansas

2. Introduction
 Southeast Kansas supports the highest concentration of
hazardous waste burners in the country
 3 cement kilns
 1 commercial hazardous waste incinerator
 U.S. EPA sponsored the Southeast Kansas Health Study
to investigate air quality and potential health effects from
ambient air in the area
 (report available at http://www2.kumc.edu/ceoh/skhs/)
 This study was a joint effort between the Department of
Civil and Environmental Engineering at KU and the KU
Medical Center

3. Scope
The environmental sampling element of
the project included the determination of
trace metal concentrations from PM2.5
collected in Teflon filters in Southeast
Kansas.
Effects of spatial and temporal factors on
these concentrations were investigated
Wind direction effects are of particular
interest

4. Air Quality Monitoring
March to October, 2000
Sampling Sites in the cities of Chanute,
Coffeyville, Fredonia, and Independence
Additional Site in Labette County
Sampling Site Design, Sampling and
Analysis Protocols According to EPA
Guidelines

5. Sampling Area in Southeast Kansas

6. Fredonia
Site Name
Site
Code
Fredonia- Lincoln School
FLS
Fredonia- South Farm
FSF
Fredonia- South Mound
FSM
Independence- Eisenhower School
IES
Independence- Radio Tower
IRT
Independence- East of River
IER
Chanute- North Ash Grove
CNA
Chanute- South Ash Grove
CSA
Chanute- Elderly Care Center
CEC
Labette County- Big Hill Lake
LCL

7. Source-specific sites
 Wichita windrose
(1984-1992)

9. Data Collection- MiniVOL

10.  Filters of PM2.5
samples from
each sampling
day were
selected
 Filters were
digested and
analyzed by
ICP/MS for 20
metals

11. Data Analysis
Descriptive statistics were calculated for
all trace metal concentrations
Analyses of variance (anovas) were
performed on:
Possible effects of sampling site
Possible effects of sampling date
Analogous non-parametric tests also
performed to confirm anova results

12. Data Analysis
 The effect of wind direction on PM
concentrations was further investigated by
performing a General Linear Model (GLM)
Data was divided in four wind direction categories (from
NCDC);
 South
 North
 Calm/variable
 Other
 The effect of the targeted sources was also
analyzed by using the paired t-test and the
analogous non-parametric test

13. Integrated Risk Information System (IRIS)
Metal
Concentration (µg/m3) producing risk levels
1 in 10,000
1 in 100,000
1 in 1,000,000
Be
0.04
0.004
0.0004
Cr(VI)
0.006
0.0006
0.00006
As
0.02
0.002
0.0002
Cd
0.06
0.006
0.0006
Al, V, Fe, Co, Rb
Mn, Ni, Cu, Zn,
Se, Mo, Ag, Sb, Ba,
Tl
Pb
Not listed as hazardous air pollutants
Not assessed under IRIS program
Carcinogen Assessment Group recommends
that numerical estimate not be used
 Mean Chromium
conc. was found to
exceed the 1 in
100,000 risk level
 Mean and median
arsenic conc.
exceeded the 1 in
1,000,000 risk
level
 Mean cadmium
conc. exceeded
the 1 in 1,000,000

14. Six metals were chosen for thorough
analysis
Beryllium
Chromium
Arsenic
Cadmium
Barium
Lead

15. Results
 Anova and Kruskal-Wallis analysis shows:
Site is not a statistically significant factor for any metals
(95% confidence)
Temporal factor was found to be statistically significant
for chromium, arsenic, cadmium, barium and lead
 Temporal factor was further investigated
through temperature, wind speed and
atmospheric variables

16. Sampling dates were divided into North,
South, calm/variable and other categories
GLM was run to determine the effect of
site and wind direction; site and wind
direction interaction
Wind direction was significant for chromium,
barium, and lead at 95% confidence level
Wind direction was significant for arsenic at
90% confidence level

17.  Anova and Kruskal-Wallis tests were run on the
database used for the GLM
Wind direction was not statistically significant for
beryllium, arsenic and cadmium
Wind direction was statistically significant for chromium,
barium and lead
 Calm/variable winds produced highest chromium
concentrations
 North winds produced highest barium
concentrations
 South winds and calm/variable winds produced
the highest lead concentrations

18. Upwind versus downwind samples
The effect of selected sources was tested
by applying the t-test and analogous nonparametric test
Data included upwind and downwind
samples for cities of Chanute,
Independence and Fredonia
Differences between upwind and
downwind samples were not found
statistically significant at the 95% level

19. Conclusions
 Mean chromium concentration exceeded the 1
in 100,000 risk level
Though risk level is for Cr(VI)
 Mean arsenic and cadmium concentrations
exceeded the 1 in 1,000,000 risk level
 Sampling site was not significant for beryllium,
chromium, cadmium, arsenic, barium and lead
 Sampling day was significant for all metals
except beryllium

20. Conclusions…
 Calm/ variable winds produced the highest
chromium concentrations
 Predominantly north winds produced the highest
barium concentrations
 Predominantly south and calm/variable winds
produced the highest lead concentrations
 No statistically significant differences were found
between samples upwind and downwind from
targeted sources