United States
Environmental Protection
Agency
Office of Research and
Development
Washington, D.C. 20460
EPA/600/R-01/016
December 2000
Mountain Acid
Deposition Program
(MADPro)
Cloud Deposition to the
Appalachian Mountains
1994 to 1999
r'.*-tv%;fs!i.s?s»im
' . ' ,':fi^:? *?&A^>^f^
U.S. EPA Headquarters Library
Mail cods 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
155LCB01.COV •:• 3/7/01
-------
-------
EPA/600/R-01/016
December 2000
Mountain Acid Deposition Program
(MADPro)
V
H.
V
Cloud Deposition
to the Appalachian Mountains
1994 to 1999
EPA Contract No. 68-D2-0134
Author: Setma Isi!
Contributors: Dr. Volker Mohnen
Dr. Gary Lovett
Dr. Eric Miller
Dr. James Anderson
Thomas F. Lavery
Ralph Baumgardner
Prepared by: Harding ESE, Inc.
Gainesville, Florida
Project Officer: Ralph Baumgardner
U.S. Environmental Protection Agency
Office of Research and Development
Research Triangle Park, NC
-------
-------
The information in this document has been funded wholly by the U.S. Environmental Protection
Agency (EPA) under Contract Nos. 68-02-4451 and 68-D2-0134 to Harding ESE, Inc.
(Harding ESE). It has been subjected to the Agency's reviews, and it has been approved for
publication as an EPA document.
Any mention of trade names or commercial products does not constitute endorsement or
recommendation for use.
-------
-------
MADPro Final Summary Report
Table of Contents
Acknowledgments iv
Abstract v
List of Tables vii
List of Figures ix
List of Acronyms and Abbreviations xiv
Executive Summary xvii
1.0 Introduction 1-1
2.0 Network Description and Methods 2-1
2,1 Network Description 2-1
2.2 Field Operations 2-1
2.3 Laboratory Operations 2-4
2.4 Data Management 2-5
2.5 Quality Assurance 2-6
2.5.1 Field Data Audits 2-6
2.5.2 Laboratory Data Audits 2-7
2.5.3 External Audits 2-7
2.5.4 Precision and Accuracy 2-7
2.5.5 Quality Assurance Experiments 2-8
2.5.5.1 Valente Versus Gerber PVM Field Comparison 2-8
2.5.5.2 Comparison of Two EPA Gerber PVMs 2-8
2.5.5.3 Comparison of EPA versus Colorado State University Gerber PVMs . .. 2-9
2.5.5.4 Evaluation of LWC Instruments 2-9
2.5.6 Whitetop Mountain LWC Data 2-10
p/castnet2/wa02/fnf su m/fnlrpt_5. wpd
Harding ESE, Inc.
-------
MADPro Final Summary Report
Table of Contents (continued)
3.0 Liquid Water Content and Cloudwater Chemistry 3-1
3.1 Cloud Frequency and Mean Liquid Water Content 3-1
3.2 Cloudwater Chemistry 3-1
3.2.1 Cloudwater pH 3-2
3.2.2 Cloudwater Anions 3-3
3.2.3 Cloudwater Cations 3-4
3.2.4 Normalized Cloudwater Concentrations 3-4
3.2.5 Slide Mountain and Hunter Mountain Results 3-4
3.2.6 Discussion of Results 3-5
4.0 Cloud Deposition 4-1
4.1 Cloudwater Deposition Model (CLOUD) 4-1
4.1.1 Model Description 4-1
4.1.1.1 Canopy Structure 4-2
4.1.1.2 Modifications from Original Version of the Model 4-2
4.1.1.3 Parameter Adjustments at Whiteface Mountain 4-4
4.1.2 Model Calculations 4-5
4.1.3 Model Sensitivity 4-6
4.1.4 Results of CLOUD Model Calculations 4-7
4.2 MCLOUD Model Calculations 4-7
4.2.1 MCLOUD Model Structure and Parameterization for MADPro Sites 4-8
4,2.1.1 Vertical Distribution of Leaf Area in the Forest Canopy 4-8
4.2.1.2 Representation of Cloud Droplet Size Distributions 4-8
4.2.1.3 Number of Droplet Size Classes in the Models 4-9
4.2.1.4 Height of the Wind Speed Measurement Above the Whiteface Mountain
Canopy 4-9
4.2.1.5 Representation of Atmospheric Conditions 4-9
4.2.2 MCLOUD Calculations 4-10
4.2.2.1 Data Screening 4-10
4.2.2.2 Data Aggregation and Summary 4-10
4.2.3 Sensitivity Analysis 4-11
4.2.3.1 Model Response to Variation in Forest Canopy Description 4-12
4.2.3.2 Model Response to Variation in Wind Speed and LWC 4-12
4.2.3.3 Model Layer Thickness 4-12
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
11
Harding ESE, Inc.
-------
MADPro Final Summon Repurt
Table of Contents (continued)
4.2.4 MCLOUD Model Results 4-13
4.2.5 Conclusions 4-13
4.3 Comparison of Results Between the CLOUD and MCLOUD Models 4-14
4.3.1 Results and Discussion 4-14
4.4 Best Estimates of Seasonal Deposition Rates 4-15
5.0 Total Deposition 5-1
5.1 Procedures 5-1
5.2 Results and Discussion 5-3
6.0 Comparison with Other Networks 6-1
6.1 Concentration of Pollutant Ions in Clouds 6-1
6.1.1 Comparison of Results with European Studies 6-2
6.2 Deposition 6-3
6.2.1 Comparison of Results with European Studies 6-5
7.0 Conclusions and Recommendations 7-1
7.1 Cloudwater Concentrations 7-1
7.2 Cloudwater Depositions 7-2
7.3 Recommendations 7-2
References REF-1
Appendices
U.S. EPA Headquarters Library
Mai! code 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
111
Harding ESE, Inc.
-------
-------
MADPro Final Summary Report
Acknowledgements
The success of MADPro is due in large part to the dedication and hard work of the individuals
involved with the day-to-day operations at the collection sites. The field operations group and
laboratory analysts of Harding ESE, Paul Casson, Dave Patrick, Joe Beeler, Don Ho, Tom
Davenport, and Donna Foley, are to be commended for their high-quality work and commitment to
the goals of MADPro.
p/castnet2/wa02/fn!sum/fnlrpt_5. wpd
IV
Harding ESE. Inc.
-------
-------
MADPro Final Summary Report
Abstract
The Mountain Acid Deposition Program (MADPro) was initiated in 1993 as part of the research
necessary to support the objectives of the Clean Air Status and Trends Network (CASTNet), which
was created to address the requirements of the Clean Air Act Amendments (CAAA). The two main
objectives of MADPro were to develop cloudwater measurement systems to be used in a network
monitoring environment and to update the cloudwater concentration and deposition data collected in
the Appalachian Mountains during the National Acid Precipitation Assessment Program (NAPAP) in
the 1980s. MADPro measurements were conducted from 1994 through 1999 during the warm season
at three 'permanent' mountaintop sampling stations. These sampling stations were located at
Whiteface Mountain, New York; Clingman's Dome, Tennessee; and Whitetop Mountain, Virginia.
A mobile manual sampling station was also operated at two locations in the Catskill Mountains in
New York during 1995, 1997, and 1998.
MADPro cloudwater concentrations (normalized with respect to liquid water content) of major ions
(SOf, NC>3, NHJ, and H+) showed statistically significant results for all three sites: Clingman's Dome
showed an increase for all four ions for both normalized and non-normalized concentrations;
Whiteface Mountain results showed a significant decrease for NHJ and SO*" for both normalized and
non-normalized concentrations; and Whitetop Mountain normalized concentrations exhibited a
significant increase for SOf only. All results refer to temporal trends and are based on trends
analyses using simple statistical procedures that did not account for variations in meteorology.
Clingman's Dome exhibited the highest mean and median 6-year (1994 to 1999) average
concentrations for the four major ions, whereas Whiteface Mountain consistently had the lowest
mean and median concentrations for these ions.
Cloudwater deposition estimates were made by applying the cloudwater deposition computer model,
CLOUD (Lovett, 1984), parameterized with site-specific cloudwater chemistry and meteorological
data. In addition, a semi-independent model (MCLOUD) was employed to explore alternative
parameterizations and additional model components beyond those offered by the primary CLOUD
model.
Monthly cloudwater deposition estimates (via CLOUD) were highly variable with deposition values
typically peaking in July or August. Seasonal deposition amounts (June through September) were
highest for Whiteface Mountain, opposite of results for cloudwater concentrations, because of the
higher wind speeds and liquid water content (LWC) experienced at this site. No temporal trend is
evident in the deposition estimates.
Dry, wet, and cloud deposition estimates were calculated on a monthly basis for June through
September for 1994 through 1998 at all three sites. Between 80 and 90 percent of sulfur (S)
11
-------
MADPro Final Summary Report
deposition occurs via cloud exposure at all three sites as does 70 to 87 percent of the total H* loading.
Cloud deposition is also responsible for 90 to 95 percent of NHJ deposition at the southern sites. Dry
deposition is a very minor contributor to the total S and NHJ loading, but contributes between 22 and
28 percent of nitrogen (N) deposition and approximately 16 percent of H* deposition at the southern
sites.
In comparison to nearby low-elevation CASTNet sites, total deposition values from MADPro sites
are approximately 6 to 20 times greater for S, N and NHJ while H+ depositions are from 1.3 to
10 times greater. Dry deposition values from MADPro sites for S, N, and H+ fall within the range of
dry deposition values for CASTNet sites. Wet deposition values for all three species are generally 1
to 3 times higher at the CASTNet sites. Thus, the difference in total deposition between MADPro
and CASTNet sites is directly attributable to cloud deposition.
Concentration ranges for the ions reported for MADPro are comparable to concentration ranges
reported for the Canadian High Elevation Fog (CHEF) project and the Mountain Cloud Chemistry
Program (MCCP). However, the MADPro means are higher for all four major ions for Whiteface and
Whitetop Mountains in comparison to MCCP results. In general, the MADPro mean calculated
deposition values (CLOUD model) for water and four major ions, when compared to MCCP values
and several other studies, fal! within the range of those measured previously for Clingman's Dome
and Whitetop Mountain, while those from the MADPro Whiteface Mountain site are slightly above
the range.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
VI
Harding ESE, Inc.
-------
MADPro Final Summary Report
List of Tables
Table 2-1. Schedule of Routine QC Checks, Calibrations, and Audits Performed at the CASTNet
Laboratory (Harding ESE)
Table 2-2. Precision and Accuracy Objectives for CASTNet Laboratory Data
Table 2-3. Data Quality Objectives for Continuous Measurements
Table 2-4. Results From the QA Testing at ECN Facilities in the Netherlands
Table 3-1. MADPro Cloud Frequency Summary
Table 3-2. Summary Statistics for Cloudwater Samples Collected at MADPro Sites from
1994 through 1999
Table 3-3. Number of Records Accepted for Analysis
Table 3-4. Mountain Cloud Linear Regression Results
Table 3-5. Summary Statistics of Major Ion Concentrations for June through September 1994
through 1999
Table 4-1. Monthly Deposition Estimates Produced with the CLOUD Model
Table 4-2. Seasonal Deposition Estimates Produced with the CLOUD Model
Table 4-3. Effect of Data Screening on Sample Retention
Table 4-4. Combinations of Wind Speed and LWC Evaluated
Table 4-5. Alternate Forest Canopy Descriptions
Table 4-6. MCLOUD Seasonal Deposition Estimates
Table 4-7. Summary of CLOUD Model Sensitivity, Potential Bias, and Expected Differences with
MCLOUD Modeling Results
Table 4-8. Comparison of Deposition Velocities and Water Depositions
Table 4-9. Best Estimates of Seasonal Depositions
p/casmet2/wa02/fnlsu m/fnltp t_5. wpd
Vll
Harding ESE, Inc.
-------
MADPro Final Summary Report
List of Tables (continued)
Table 5-1. Summary of Cloud, Precipitation, Dry, and Total Deposition Estimates for
Whiteface Mountain
Table 5-2. Summary of Cloud, Precipitation, Dry, and Total Deposition Estimates for
Whitetop Mountain
Table 5-3. Summary of Cloud, Precipitation, Dry, and Total Deposition Estimates for
Clingman's Dome
Table 5-4. Percent Composition of Total Deposition at the Three MADPro Sites
Table 5-5. Dry, Wet, and Total Seasonal Depositions (June through September) for Whiteface
Mountain and Two Nearby CASTNet Sites for 1995 through 1998
Table 5-6. Dry, Wet, and Total Seasonal Depositions (June through September) for Whitetop
Mountain and Two Nearby CASTNet Sites for 1996
Table 6-1. Warm Season Average Ion Concentrations for the Six MCCP Sites 1986 through 1988
Table 6-2. Comparison (RPD) of MCCP versus MADPro Average Ion Concentrations
Table 6-3. Mean Chemical Composition Including Minima and Maxima of Cloudwater Collected
at Mt. Brocken, Germany
Table 6-4. Statistical Characterization of the Chemical Composition of Fog Samples from
European Investigations
Table 6-5. Comparison of MADPro Cloudwater Deposition Estimates to Previous Studies
Table 6-6. A Summary of Cloudwater Chemical Deposition via Droplet Interception for the
Eastern United States
Table 6-7. Comparison of the Proportion of Total Ion Deposition Delivered by the Dry,
Cloudwater, and Precipitation Deposition Estimates for Whiteface Mountain at an
Elevation of 1,050 m
p/castnet2/wa02/fnlsu m/fnlrpt_5. wpd
V1I1
Harding ESE, Inc.
-------
MADPro Final Summary Report
List of Figures
Figure 2-1. Locations of Mountain Acid Deposition Sites
Figure 2-2. Regional Map for Whiteface Mountain, NY
Figure 2-3. Regional Map for Whitetop Mountain, VA
Figure 2-4. Regional Map for Clingman's Dome, TN
Figure 2-5. MADPro Sampling System
Figure 2-6. Flowchart of Laboratory Operations for Cloudwater and Precipitation
Sample Analyses
Figure 2-7. PVM-100 Intercomparison at Whitetop Mountain, VA, 1998
Figure 3-1. Monthly Cloud Frequency (1994 through 1999), Whiteface Mountain, NY
Figure 3-2. Monthly Cloud Frequency (1994 through 1999), Whitetop Mountain, VA
Figure 3-3. Monthly Cloud Frequency (1994 through 1999), Clingman's Dome, TN
Figure 3-4. Mean Liquid Water Content of Clouds (1994 through 1999), Whiteface
Mountain, NY
Figure 3-5. Mean Liquid Water Content of Clouds (1994 through 1999), Whitetop
Mountain, VA
Figure 3-6. Mean Liquid Water Content of Clouds (1994 through 1999), Clingman's Dome, TN
Figure 3-7. Mean Liquid Water Content of Clouds
Figure 3-8. Mean pH of Cloudwater Samples at MADPro Sites (1994 through 1999)
Figure 3-9. Frequency Distribution for Cloudwater pH at Whiteface Mountain, NY
(1994 through 1999)
Figure 3-10. Frequency Distribution for Cloudwater pH at Whitetop Mountain, VA
(1994 through 1999)
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
IX
Harding ESE, Inc.
-------
JAADPro Final Summary Report
List of Figures (continued)
Figure 3-11. Frequency Distribution for Cloudwater pH at Clingman's Dome, TN
(1994 through 1999)
Figure 3-12. Normalized SO*" Concentrations in Cloudwater at Whitetop Mountain, 1994 through
1996 - Calculated as LWC Weighted Means versus Arrival Sector
Figure 3-13. Normalized SOl~ Concentrations in Cloudwater at Whiteface Mountain, 1994
through 1996 - Calculated as LWC Weighted Means versus Arrival Sector
Figure 3-14. Annual SO2 Emissions for 1995
Figure 3-15. Mean Normalized SO^' Concentrations (ueq/L); Segregated by Back Trajectory
Arrival Sector
Figure 3-16. pH of Cloudwater Samples, Whiteface Mountain, NY (1995)
Figure 3-17. pH of Cloudwater Samples, Whiteface Mountain, NY (1997)
Figure 3-18. pH of Cloudwater Samples, Whiteface Mountain, NY (1999)
Figure 3-19. pH of Cloudwater Samples, Whitetop Mountain, VA (1995)
Figure 3-20. pH of Cloudwater Samples, Whitetop Mountain, VA (1997)
Figure 3-21. pH of Cloudwater Samples, Whitetop Mountain, VA (1999)
Figure 3-22. pH of Cloudwater Samples, Clingman's Dome, TN (1995)
Figure 3-23. pH of Cloudwater Samples, Clingman's Dome, TN (1997)
Figure 3-24. pH of Cloudwater Samples, Clingman's Dome, TN (1999)
Figure 3-25. Mean Hydrogen Ion Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Figure 3-26. Mean Hydrogen Ion Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
Harding BSE, Inc.
-------
MADPro Final Summary Report
List of Figures (continued)
Figure 3-27. Monthly Variation in H+ Concentrations in Cloudwater (Means Across All Years,
1994 through 1999)
Figure 3-28. Mean SO^' Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Figure 3-29. Mean SO^" Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Figure 3-30. Monthly Variation in SO*' Concentrations in Cloudwater (Means Across All Years,
1994 through 1999)
Figure 3-31. Mean NO^ Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Figure 3-32. Mean NOj Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Figure 3-33. Monthly Variation in NOj Concentrations in Cloudwater (Means Across All Years,
1994 through 1999)
Figure 3-34. Mean NHJ Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Figure 3-35. Mean NHJ Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Figure 3-36. Average Minor Ion Concentrations from 1994 to 1999
Figure 3-37. Normalized Mean Hydrogen Ion Concentrations of Cloudwater Samples at MADPro
Sites (1994 through 1999)
Figure 3-38. Normalized Mean SO^' Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Figure 3-39. Normalized Mean NOj Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
XI
Harding ESE. Inc.
-------
MADPro Final Summary Report
List of Figures (continued)
Figure 3-40. Normalized Mean NHJ Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Figure 3-41. Concentrations for SLD/HUN301 (1995,1997, 1998)
Figure 3-42. CLD303 1994 through 1999 Regressions Using All Data Points - Non-Normalized
Concentrations
Figure 3-43. CLD303 1994 through 1999 Regressions Using All Data Points - Normalized
Concentrations
Figure 3-44. WFM300 1994 through 1999 Regressions Using All Data Points - Non-Normalized
Concentrations
Figure 3-45. WFM300 1994 through 1999 Regressions Using All Data Points - Normalized
Concentrations
Figure 3-46. WTM302 1994 through 1999 Regressions Using All Data Points - Normalized
Concentrations
Figure 4-1. Relationship Between Square of the Mean Droplet Diameter (D) and LWC for
Clouds on Whitetop Mountain
Figure 4-2. Relative Percent Difference Versus Total Anion Concentration for Each Sample
Figure 4-3. Sensitivity of Deposition Velocity to Wind Speed
Figure 4-4. Sensitivity of Deposition Velocity to LWC
Figure 4-5. Results of Numerical Experiments with Model Layer Thickness and Shifts in the
Vertical Distribution of Leaf Area
Figure 4-6. Difference in Cloudwater Deposition Rate between Model Runs Using the Whitetop
Mountain Cloud Droplet Size Distribution and the Whiteface Mountain Cloud
Droplet Size Distribution for a Range of LWC and Wind Speed
Figure 4-7. Frequency Distribution of Sample Water Deposition for Clingman's Dome 1997
Data Set Calculated with Either 500 or 20 Droplet Size Classes Representing the
Continuous Droplet Size Distribution
p/castnet2Ava02/fnlsum/fnlrpt_5. wpd
Xll
Harding ESE. Inc.
-------
MADPro Final Summary Renon
List of Figures (continued)
Figure 4-8. Comparison of Estimated Cioudwater Fluxes Using Four Potential Model Scenarios
at Whiteface Mountain, 1997
Figure 4-9. Frequency Distributions of LWC and Wind Speed Scaled to Values Representative
of 1,225-m Elevation on Whiteface Mountain
Figure 4-10. Percent Deviation of Model Response to a Hypothetical Pure Balsam Fir Canopy
from Model Response to an Observed Average Canopy at 1,225-m Elevation on
Whiteface Mountain over a Range of Wind Speed and LWC Values
Figure 4-11. Percent Deviation of Model Response to Several Canopy Height Specifications from
Model Response to the Observed Height of 17 m at 1,225-m Elevation
Figure 4-12. Percent Deviation of Model Response to Variations in LAI with a Constant Canopy
Height of 10m
Figure 4-13. Model Response (Hourly Water Flux) to Variation in Wind Speed for a Range of
Cloud LWC
Figure 5-1. Areas in the Eastern United States with Elevations Above 800 Meters
Figure 5-2. Percent Composition of Total Deposition at MADPro Sites
Figure A-l. Mean Liquid Water Content of Clouds with Scaled 1998 WTM LWC
Figure A-2. Normalized Mean Hydrogen Ion Concentrations of Cioudwater Samples at MADPro
Sites (1994 through 1999) with Scaled 1998 WTM LWC
Figure A-3. Normalized Mean NHJ Concentrations of Cioudwater Samples at MADPro Sites
(1994 through 1999) with 1998 WTM LWC
Figure A-4. Normalized Mean NOj Concentrations of Cioudwater Samples at MADPro Sites
(1994 through 1999) with 1998 WTM LWC
Figure A-5. Normalized Mean SO*' Concentrations of Cioudwater Samples at MADPro Sites
(1994 through 1999) with 1998 WTM LWC
p/castnet2AvaQ2/fn Isu m/fnlrpt_5. wpd
XI11
Harding ESE, Inc.
-------
-------
MADPra Final Summary Report
List of Acronyms and Abbreviations
AIRMoN
AIRS
ARS
ASRC
°C
Ca2+
CAAA
CASTNet
CDM-D
CDM-S
CE
CHEF
cr
CLASS™
CLD303
cm
cm/sec
cond.
CSU
CVS
CWP
DAS
DEC
ECEB
ECN
EPA
EPRI-IFS
ft
g/cm2/min
g/m3
H+
Harding ESE
HN03
HUN301
1C
ICAP-AE
K+
K2CO3
Atmospheric Integrated Research Monitoring Network
Aerometric Information Retrieval System
Air Resource Specialists
Atmospheric Sciences Research Center
degrees Celsius
calcium
Clean Air Act Amendments
Clean Air Status and Trends Network
cloud deposition model - deciduous
cloud deposition model - spruce
collection efficiency
Chemistry of High Elevation Fog
chloride
Chemistry Laboratory Analysis and Scheduling System
Clingman's Dome, Tennessee Sampling Station
centimeter
centimeters per second
conductivity
Colorado State University
continuing verification sample
Cloud water Project
data acquisition system
New York Department of Environmental Conservation
eddy covariance-energy budget
Energy Center for the Netherlands
U.S. Environmental Protection Agency
Electric Power Research Institute-Integrated Forest Study
foot
grams per square centimeter per minute
grams per cubic meter
hydrogen
Harding ESE, Inc.
nitric acid
Hunter Mountain, New York Mobile Sampling Station
ion chromatography
inductively coupled argon plasma - atomic emission
potassium
potassium carbonate
p/castnet2/wa02/ftilsum/fhlrpt_5.wpd
XIV
Harding ESE, Inc.
-------
MADPro Final Summary Report
List of Acronyms and Abbreviations (continued)
kg/ha/mo
kg/ha/yr
km
L
LAI
Lpm
LWC
m2/m2
m
m/sec
MADPro
MCCP
MFC
tnL
MLM
mm
mtn/hr
MmaD
N
Na*
NADP/NTN
NAPAP
NH;
NIST
nm
NOj
NO;
NOAA
NPS
03
Ogden
ppb
PVM
QA
QC
%RD
Rh
kilograms per hectare per month
kilograms per hectare per year
kilometer
liter
leaf area index
liters per minute
liquid water content
square meter per square meter
meter
meters per second
Mountain Acid Deposition Program
Mountain Cloud Chemistry Program
mass flow controller
magnesium
milliliter
multi-layer model
millimeter
millimeters per hour
mass median aerodynamic diameter
nitrogen
sodium
National Atmospheric Deposition Program/National Trends Network
National Acid Precipitation Assessment Program
ammonium
National Institute for Standards and Technology
nanometer
nitrite
nitrate
National Oceanic and Atmospheric Administration
National Park Service
ozone
Ogden Environmental and Energy Services, Inc.
parts per billion
particle volume monitor
quality assurance
quality control
percent relative deficit
relative humidity
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
XV
Harding ESE, Inc.
-------
MADPro Final Summary Report
List of Acronyms and Abbreviations (continued)
RPD relative percent difference
S sulfur
SLD301 Slide Mountain, New York Mobile Sampling Station
SOf sulfate
SO, sulfur dioxide
SOP standard operating procedure
TRAACS Technicon Random Access Automated Chemistry System
ueq/L microequivalent per liter
ug/filter microgram per filter
urn micrometer
UV ultraviolet
WFM Whiteface Mountain, New York Sampling Station
WFML Whiteface Mountain lower
WFMS Whiteface Mountain summit
WTM302 Whitetop Mountain, Virginia Sampling Station
p/eastnet2/wa02/fnls um/fhlrpt_5. wpd
XVI
Harding ESE, Inc.
-------
-------
MADPrn Final Summary Report
Executive Summary
The 1990 Clean Air Act Amendments (CAAA) required a reduction in sulfur dioxide (SO2)
emissions by approximately 10 million tons annually to occur in two phases. The first phase was
implemented in 1995 when large electric generating facilities reduced emissions. The second phase is
scheduled for 2000 and targets other power plants. Title IX of the CAAA also mandated the
deployment of a comprehensive research and monitoring program. This program would track the
effectiveness of emission reduction programs with respect to deposition, air quality, and changes to
affected ecosystems. The Clean Air Status and Trends Network (CASTNet) was implemented in
1991 in response to this mandate.
The Mountain Acid Deposition Program (MADPro) was initiated in 1993 as part of the research
necessary to support the objectives of CASTNet. The two main objectives of MADPro were to
develop cloudwater measurement systems to be used in a network monitoring environment and to
update the cloudwater concentration and deposition data collected in the Appalachian Mountains
during the National Acid Precipitation Assessment Program (NAPAP) in the 1980s. MADPro
measurements were conducted from 1994 through 1999 during the warm season at three 'permanent'
mountaintop sampling stations. These sampling stations were located at Whiteface Mountain, New
York; Clingman's Dome, Tennessee; and Whitetop Mountain, Virginia. A mobile manual sampling
station was also operated at two locations in the Catskill Mountains in New York during 1995,1997,
and 1998. The two main objectives of MADPro have been successfully met. The results of the first
objective are summarized in Baumgardner et al. (1997) and Anderson et al (1999). The results of the
second objective are provided in this report.
This report summarizes the analysis and interpretation of MADPro measurements collected from
1994 through 1999. Summary statistics, analyses of cloud frequency, liquid water content (LWC),
cloud chemistry, and cloud and total deposition are presented for the three permanent sites. A
summary of European and other cloud studies, their comparison to MADPro results, and
comparisons to the Mountain Cloud Chemistry Program (MCCP) results are included along with
conclusions and recommendations for future studies.
MADPro cloudwater concentrations (normalized with respect to liquid water content) of major ions
(SOf, NOj, NHJ, and H+) showed statistically significant trend results for all three sites: Clingman's
Dome showed an increase for all four ions for both normalized and non-normalized concentrations;
Whiteface Mountain results showed a significant decrease for NHJ and SOf for both normalized and
non-normalized concentrations; and Whitetop Mountain normalized concentrations exhibited a
significant increase for SO^ only. All results refer to temporal trends and are based on trends
analyses using simple statistical procedures that did not account for variations in meteorology.
Clingman's Dome exhibited the highest mean and median 6-year (1994 and 1999) average
p/qastnet2/wa02/fnlsum/fnlipt_5,wpd
XV11
Harding ESE. Inc.
-------
MADPro Final Summary Report
concentrations for the four major ions, whereas Whiteface Mountain consistently had the lowest
mean and median concentrations for these ions.
The lack of any discernible temporal trends for some of the major ions and the increasing SOf
temporal trend exhibited at Whitetop Mountain is inconsistent with the downward trend in SO2
emission levels in the eastern United States [U.S. Environmental Protection Agency (EPA), 2000].
Given the magnitude of the decrease in the SO2 emission levels, stronger decreasing temporal trends
might also be expected at Whiteface Mountain. These results indicate that cloudwater concentrations
of SOl' mav not be linearly related to SO2 emission reductions because SOf production in clouds is
oxidant-limited. More research and monitoring is necessary before any solid conclusions can be
made as the lack of trends, weak trends, or increasing trends, within this period may also be a
function of, or influenced by, year-to-year variations in air mass trajectories and regional and local
meteorology. Depending on the year, variations in LWC are also shown to have an effect, although
not large, on the results.
Cloudwater deposition estimates were made by applying the cloudwater deposition computer model,
CLOUD (Lovett, 1984), parameterized with site-specific cloudwater chemistry and meteorological
data. In addition, a semi-independent model (MCLOUD) was employed to explore alternative
parameterizations and additional model components beyond those offered by the primary CLOUD
model. MCLOUD was used to produce an alternative set of cloudwater deposition estimates for
eight site-by-year combinations as well as deposition estimates for all 1999 data.
Monthly cloudwater deposition estimates (via CLOUD) were highly variable with depositions
typically peaking in July or August. Seasonal deposition values (June through September) were
highest for Whiteface Mountain, opposite of results for cloudwater concentrations, because of the
higher wind speeds and LWC experienced at this site. No temporal trend is evident in the deposition
estimates.
Deposition velocities and cloudwater deposition estimates produced by the CLOUD model were
approximately 45 percent higher than the MCLOUD runs for the same site-year combinations.
Differences in model formulation and parameterizations, input data, and data screening and data
aggregation procedures all contribute to the differences seen in the results from both models. The
1999 deposition estimates produced by MCLOUD were, therefore, scaled up by 1.45 to account for
these differences.
Total deposition consists of three components: 1) dry, 2) wet, and 3) cloud depositions. Estimates
for all three components were calculated on a monthly basis for June through September for 1994
through 1998 at all three sites. The multi-layer model (MLM) was used to estimate dry deposition
(Meyers et al., 1998; Finkelstein et al, 2000). Estimates of wet deposition were made using onsite or
nearby precipitation chemistry measurements. These results show that clouds are the largest source
p/castnet2/wa02/fnlsum/fnlrpt_5, wpd
XV111
Harding ESE. Inc.
-------
MADPro Final Summary Report
for deposition of pollutants to high-elevation ecosystems. Between 80 and 90 percent of S deposition
occurs via cloud exposure at all three sites as does 70 to 87 percent of the total H+ loading. Cloud
deposition is also responsible for 90 to 95 percent of NHJ deposition at the southern sites. Dry
deposition is a very minor contributor to the total S and NH} loading, but contributes between 22 and
28 percent of N deposition and approximately 15 to 16 percent of H+ deposition at the southern sites.
In comparison to nearby low-elevation CASTNet sites, total deposition values from MADPro sites
are approximately 6 to 20 times greater for S, N and NH}, while H* depositions are usually 1.3 to
10 times greater. Dry deposition values from MADPro sites for S, N, and H+ fall within the range of
dry deposition values for CASTNet sites. Wet deposition values for all three species are generally 1
to 3 times higher at the CASTNet sites. Thus, the difference in total deposition between MADPro
and CASTNet sites is directly attributable to cloud deposition.
The uncertainties in the calculations of cloud deposition values are estimated to range from at least
509c to 100% or more based on the comparison of the results from the CLOUD and MCLOUD
models and on an analysis of evaluation studies for resistance type models. The model comparisons
suggest that the CLOUD model overestimates cloud deposition.
The uncertainties in the calculations of dry and wet deposition values are more variable. The
uncertainties in the dry deposition values generated using the MLM model for Whitetop Mountain
and Clingman's Dome are estimated as less than 100% for seasonal fluxes based on model evaluation
studies. These model calculations are considered underestimates (Finkelstein etai, 2000). The
uncertainty in the estimate of wet deposition for Clingman's Dome is unknown. Wet deposition data
from nearby Mt. Mitchell were used for this site. The estimates of both dry and wet deposition for
Whiteface Mountain are also considered underestimates due to the difference in elevation of the
sampling locations. These underestimates may be as high as 400% for dry deposition while the
amount of underestimation for wet deposition is unknown.
Concentration ranges for the ions reported for MADPro are comparable to concentration ranges
reported for the Canadian High Elevation Fog (CHEF) project and the MCCP. However, the
MADPro means are higher for all four major ions for Whiteface and Whitetop mountains in
comparison to MCCP results. In general, the MADPro mean calculated deposition values (CLOUD
model) for water and four major ions, when compared to MCCP values and several other studies, fall
within the range of those measured previously for Clingman's Dome and Whitetop Mountain, while
those from the MADPro Whiteface Mountain site are slightly above the range.
Since the commencement of operation in 1994, MADPro has produced a 6-year data set that is
comparable to data produced by past networks. Although not designed for ecological studies, results
from this program will aid ecologists in damage assessment of high-elevation ecosystems through the
provision of a data set of known uncertainty.
p/castnet2/wa02/fn!sum/fnlipt_5.wpd
XIX
Harding ESE, Inc.
-------
-------
MADPrn Final Summary Report
1.0 Introduction
Cloudwater droplets have been shown to contain high concentrations of acidic and other dissolved
ions depending on back or past trajectories (Mohnen and Vong, 1993). Cloudwater is typically 5 to
20 times more acidic than rain water (Mohnen et al, 1990; Vong et ai, 1991). Thus, clouds can be
the primary pathway for exposure and deposition of pollutants to high-elevation ecosystems (Aneja
and Kim, 1993). Subalpine, high-elevation forests of the eastern United States receive some of the
largest air pollution loadings of any rural environment in North America (Miller et al., 1993b;
Lindberg, 1992; Lovett and Kinsman, 1990). This large loading of pollutants is due to a combination
of factors including high frequency of cloud immersion, high wind speeds, orographic enhancement
of precipitation, and the large leaf areas of tree species typical of these environments (Miller and
Friedland, 1999). There has been consistent evidence that exposure to acidic Cloudwater, which
reduces cold tolerance, is one of the causal factors leading to the recent decline of red spruce (Picea
rubens) in the northeastern United States (Eager and Adams, 1992). Other possible mechanisms of
forest damage from acidic deposition include extensive leaching of cations and amino acids from the
foliage (Scherbatskoy and Klein, 1983), nitrogen overload, aluminum toxicity, and the combined
effect of acid rain, acid fog, oxidants, and heavy metals (Schemenauer, 1986; McLaughlin etal,
1990, 1991).
Previous cloud monitoring efforts that have characterized cloudwater chemistry in North America
include:
1. The Cloudwater Project (CWP), Weathers et al. (\ 986, 1988,1995) - Studied rain and
cloudwater chemistry on mountain ranges from Puerto Rico to Alaska;
2. The U.S. Environmental Protection Agency's (EPA's) Mountain Cloud Chemistry Program
(MCCP) - Collected and analyzed cloudwater samples and estimates of cloud deposition
from the mountains of the eastern United States from 1986 to 1989 (Mohnen and Kadlecek,
1989; Saxena etal., 1989; Vong etal., 1991; Li and Aneja, 1992; Mohnen and Vong, 1993);
and
3, Canada's Chemistry of High Elevation Fog (CHEF) Project - Measured cloud chemistry on
three mountains in southern Quebec from 1985 to 1991 (Schemenauer, 1986; Schemenauer
and Winston, 1988; Schemenauer etal., 1995).
Data collected by these projects were used by the National Acid Precipitation Assessment Program
(NAPAP) to evaluate the role of airborne chemicals on the changing conditions of forests. The
NAPAP Integrated Assessment of 1990 concluded that a limited number of forest ecosystems were at
risk from acidic deposition and additional ecosystems would be at risk in the future. The 1990 Clean
Air Act Amendments (CAAA) required a 10 million ton reduction of sulfur dioxide (SO2) emissions
to occur in two phases. The first phase was to be completed by the year 2000 and the second phase
by 2010. Title EX of the CAAA also mandated the deployment of a national monitoring network to
p/castnet2/wa02/fn!sum/fnlrpt_5.wpd
1-1
Harding ESE. Inc.
-------
MADPro Final Summary Report
track the effectiveness of emission reduction programs with respect to deposition, air quality, and
changes to affected ecosystems. The Clean Air Status and Trends Network (CASTNet) was
implemented in 1991 in response to this mandate.
The Mountain Acid Deposition Program (MADPro) was initiated in 1993 as part of the research
necessary to support the objectives of CASTNet. The two main objectives of the program were to
develop cloudwater measurement systems to be used in a network monitoring environment and to
update the cloudwater concentration and deposition data collected in the Appalachian Mountains
during NAPAP in the 1980s. MADPro measurements were conducted from 1994 through 1999
during the warm season at three 'permanent' mountaintop sampling stations. These sampling
stations were located at Whiteface Mountain, New York; Clingman's Dome, Tennessee; and
Whitetop Mountain, Virginia. A mobile manual sampling station was also operated at two locations
in the Catskill Mountains in New York during 1995, 1997, and 1998.
The two main objectives of MADPro have been met. The results of the first objective are presented
in Baumgardner et al. (1997) and Anderson et al. (1999). The results of the second objective are
provided in this report.
The report includes a detailed network description in Chapter 2.0. Summary statistics and analyses
of cloud frequency, liquid water content (LWC), and cloud chemistry are presented in Chapter 3.0.
Modeled cloud deposition estimates are discussed in Chapter 4.0. Estimates of cloud, wet, dry, and
total depositions are discussed in Chapter 5.0. A summary of European and other cloud studies, their
comparison to MADPro results, and comparisons to MCCP results are presented in Chapter 6.0.
Conclusions and recommendations for future programs are presented in Chapter 7.0.
p/castnet2/wa02/fnlsum/fnlrpt_5. wpd
1-2
Harding ESE, Inc.
-------
MADPro Final Summar\ Report
2.0 Network Description and Methods
2.1 Network Description
The three permanent MADPro sites were widely dispersed along the Appalachian Mountains
(Figure 2-1). Whiteface Mountain (44'21'58"N, 73'54'10"W) is located in the northeastern
Adirondacks in upstate New York (Figure 2-2). The cloudwater sampling research station was above
the tree line at 1,483 meters (m). The instrumentation was located on the top floor of a four-story
observatory at the summit, with the cloudwater collector, particle volume monitor (PVM), and
meteorological sensors mounted on the flat circular roof.
Whitetop Mountain (36'38'20"N, 81'36'19"W) is located in the Mount Rogers National Recreation
Area of the Jefferson National Forest in southwestern Virginia (Figure 2-3). It is 6 kilometers (km)
southwest of Mount Rogers, the highest peak in Virginia. The MADPro research station was at
1,686 m on the main ridge line of the Appalachian range.
Clingman's Dome (35'33'47"N, 83'29'55"W) is the highest mountain in the Great Smoky Mountains
National Park (Figure 2-4). The solar-powered MADPro site was situated at an elevation of 2,014 m
approximately 50 m southwest of the summit spiral tourist tower. Electronic instrumentation was
housed in a small National Park Service (NPS) building and the cloudwater collector, PVM, and
meteorological sensors were positioned on top of a 50-foot (ft) scaffold tower.
Collection at the sites was initiated each spring, as soon as local weather conditions would allow, and
continued through autumn. Collection ran June through September for Whiteface Mountain, and
May through October for Whitetop Mountain and Clingman's Dome.
A mobile site was located on Slide Mountain in the Catskill Mountains of New York in 1995. A
semi-automated system was relocated to Hunter Mountain in 1997 and collected samples during 1997
and 1998. The Hunter Mountain site also collected wind speed and wind direction data measured
with a portable Davis anemometer.
2.2 Field Operations
The MADPro cloud collection system consisted of an automated cloudwater collector for hourly
cloudwater sampling, a PVM for continuous determination of cloud LWC, a meteorological station
for continuous measurements of wind speed, wind direction, temperature, relative humidity, solar
radiation, and precipitation, and a data acquisition system (DAS) for collection and storage of
electronic information from the various monitors and sensors (Figure 2-5). The Clingman's Dome
and Whitetop Mountain sites also had a filter pack system for dry deposition estimation, with an
p/castnet2/wa02/fnlsum/fnlrp(_5.wpd
2-1
Harding ESE. Inc.
-------
MADPro Final Summary Report
additional wet deposition system consisting of a precipitation collector and rain gauge at Whitetop
Mountain. The filter pack and additional wet deposition systems were not used at Whiteface
Mountain because the National Oceanic and Atmospheric Administration (NOAA) and the National
Atmospheric Deposition Program/National Trends Network (NADP/NTN) were measuring these
parameters in the vicinity of the cloudwater collection site. Likewise, the Cingman's Dome site had
access to the data collected by a nearby wet deposition system operated by the NFS.
The Whitetop Mountain site was also equipped with a continuous ozone (O3) monitor. In addition,
continuous O3 data [through Aerometric Information Retrieval System (AIRS)] and SO2 data
[through the New York Department of Environmental Conservation (DEC)] are available for
Whiteface Mountain.
The core of the automated cloud collection system is a passive string collector [also known as the
Mohnen collector or the Atmospheric Sciences Research Center (ASRC) design] previously used in
the MCCP study. Collection occurs when ambient winds transport cloudwater droplets onto 0.4-
millimeter (mm) Teflon® wires strung between two circular disks (Falconer and Falconer, 1980;
Mohnen and Kadlecek, 1989). Once impacted, the droplets slide down the strings into a funnel and
through Teflon® tubing into sample bottles in a refrigerated carousel. The development and design of
this system is described in detail in Baumgardner et al. (1997).
The PVM-100 by Gerber Scientific (Gerber, 1984) measures LWC and effective droplet radius of
ambient clouds by directing a narrow laser beam from a 780-nanometer (nm) diode along a
40-centimeter (cm) path. The forward scatter of the cloud droplets in the open air along the path is
measured, translated, and expressed as grams of water per cubic rneter (g/rn3) of air. This system was
programmed so that the collector would be activated and projected out of the protective housing
when threshold levels for LWC (0.05 g/m3), wind speed [>2.5 meters per second (m/sec)], and
ambient air temperature [2 degrees Celsius (°C)] were reached. In addition, the system was activated
only when no precipitation was measured. Within the context of MADPro, therefore, a cloud was
defined by a LWC of 0.05 g/m3 or higher, as measured by the PVM. This threshold was established
to have comparability with the MCCP measurements, which were made for the most part with
Mallant Optical Cloud Detectors set at a threshold of approximately 0.04 g/m3 (Mohnen et al., 1990).
The wind speed threshold of 2.5 m/sec was established because cloudwater collection is erratic and
inefficient at lower wind speeds. Higher wind speeds were necessary to yield the minimum
30 milliliters (mL) of cloudwater required for sample analysis. The temperature limit served to
protect against damage from rime ice formation. The absence of rainfall was required because within
the objectives of this study, as well as MCCP, only samples from non-precipitating clouds were
collected. If a rain detector was activated, the string collector retracted into the protective case and
collection was suspended.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
2-2
Harding ESE, Inc.
-------
MADPro Final Summary Report
Collection of cloud samples only when these criteria were met did not result in loss of cloud
frequency and cloud duration information due to continuous data collection by the PVM. All LWC
values of 0.05 g/m3 or greater, independent of the type of cloud (i.e. precipitating or
non-precipitating), were used to calculate cloud frequency and cloud duration information. It is
possible that cloud deposition estimates presented later on in Chapter 4 may be biased by not
sampling for the cloud deposition that occurs during precipitating clouds. Collection of cloud water
samples during precipitating clouds with an ETH cloud impactor as deployed by Collett et al. (1993)
was not within the scope of this project. However, the bias due to this lack of sampling during
precipitating clouds is offset by the fact that cloud deposition totals were estimated by multiplying
the duration-weighted mean chemical fluxes by the cloud-hours for the month. The cloud-hours were
calculated as the cloud frequency times the total hours in the month. See Chapter 4 for a detailed
discussion on estimation of cloud depositions.
Hourly samples of cloudwater were gathered from the collector cooler by the site operator within
24 hours of each cloud event. The time, date, and volume of each sample were recorded on a report
form by the site operator. The site operator also measured pH and conductivity and decanted
samples into 250-mL bottles for shipment in coolers to the Harding ESE, Inc. (Harding ESE)
laboratory. A rinse/sample blank was also included with each shipment to the Harding ESE
laboratory. According to MADPro protocols, the cloud collector was rinsed with deionized water
after each cloud event until the conductivity of the rinsate measured < 10 uS (micro Siemens). The
rinse/sampler blank consisted of a portion of this 'clean' rinsate.
Filter packs were prepared and shipped to the field on a weekly basis and exchanged at the
Clingman's Dome and Whitetop Mountain sites every Tuesday. For a description of the filter pack
set-up, types of filters used and the fraction collected on each filter, refer to the draft CASTNet
Quality Assurance Project Plan (CASTNet QAPP) (Harding ESE, 1999) and/or the CASTNet
Deposition Summary Report (EPA, 1998). A discussion on filter pack sampling artifacts can be
found in Anlaulf etal. (1986).
Filter pack flow was maintained at 1.50 liters per minute (Lpm) with a mass flow controller (MFC)
through the middle of the 1998 season, at which time the flow was increased to 3.0 Lpm for the
duration of the project. During 1994 and 1995, a continuous flow was drawn through the filter pack.
In 1996, the flow was programmed to shut off during a cloud or rain event to allow for determination
of truly dry deposition. Since the total hours of flow and, hence, volume were substantially reduced
depending on the weekly weather conditions, an increase in flow to 3.0 Lpm was implemented in
1998 to increase the volume of flow through the filter pack to better detect the lower concentrations
of analytes.
A wet deposition sample from Whitetop Mountain was also collected on a weekly basis (according to
NADP/NTN protocols) in precleaned polyethylene buckets using an Anderson Model precipitation
p/castnet2/wa02/fnlsum/fnlipt_5.wpd
2-3
Harding ESE, Inc.
-------
MADPro Final Summary Report
sampler. Buckets were placed on the sampler on Tuesday and removed, whether or not rainfall had
occurred, the following Tuesday. Buckets were weighed in the field, decanted to a 1-liter (L)
polyethylene bottle, sealed, and shipped to Harding ESE for chemical analysis. In addition,
precipitation amount (depth) was monitored at Whitetop Mountain with a Belfort rain gauge.
O3 was measured at Whitetop Mountain (as well as Whiteface Mountain via the DEC) via ultraviolet
(UV) absorbance with a Thermo-Environmental Model 49-103 analyzer operating in the 0- to
500-parts per billion (ppb) range. Ambient air was drawn from the air quality tower through a
3/8-inch TFE Teflon® sampling line. The analyzer and Teflon® filters housed at the tower inlet
prevented particle deposition within the system. Periodic checks indicated that online losses through
the inlet system were consistently less than 3 percent. Zero, precision (60 ppb), and span (400 ppb)
checks of the O3 analyzer were performed every third day using an internal O3 generator. The O3
data collected from Whitetop Mountain will not be presented in this report. The data have been
validated and are available, however, through the EPA.
All field equipment received start-up calibrations during site initialization each field season and end-
of-season calibrations. Calibration checks were performed on the PVM throughout the field season
and the results were used to adjust the instrument immediately after the calibration check.
Calibrations on the remaining instruments were conducted using standards traceable to the National
Institute for Standards and Technology (NIST). In addition, independent equipment audits were
performed in 1997 by Ogden Environmental and Energy Services, Inc. (Ogden), and by Air
Resources Specialists (ARS) in 1999. Results of the end-of-season calibrations were used to assess
sensor accuracy and to flag, adjust, or invalidate data.
2.3 Laboratory Operations
Cloudwater and wet deposition samples were analyzed for pH, conductivity, sodium (Na+), potassium
(K+), ammonium (NHJ), calcium (Ca2"1"), magnesium (Mg2*), chloride (Cl"), nitrite (NO^), nitrate
(NOs), and sulfate ion (SOf) in the Harding ESE laboratory. Wet deposition samples were filtered
before analysis and also analyzed for acidity. During the first 2 years of the project, all cloudwater
samples were analyzed for pH and conductivity. Starting in 1996, every tenth sample was analyzed
in the Harding ESE laboratory for these two parameters to reduce redundancy but to still provide
quality control (QC) data.
Cloud water and wet deposition samples were stored at 4° C until analysis. All analyses were
performed within 30 days of sample receipt at the laboratory. The effects of storage on wet
deposition samples have been addressed in NAPAP Report #6 (Sisterson et ai, 1991). This
discussion applies, for the most part, to cloud water samples as well.
p/castnet2/wa02/fnisum/fnlrpt_5. wpd
2-4
Harding ESE, Inc.
-------
MADPro Final Summary Report
Concentrations of the four anions were determined by micromembrane-suppressed ion
chromatography (1C). Analysis of Na+, Mg2*, and Ca2+ was performed with a Perkin-Elmer P-2
inductively coupled argon plasma atomic emission (ICAP-AE) spectrometer, whereas K+ was
analyzed via atomic emission in 1994 and subsequently via ICAP. NHJ concentrations were
determined by the automated indophenol method using a Technicon II or Technicon Random Access
Automated Chemistry System (TRAACS)-SOO autoanalyzer system. Titration to a pH of
approximately 8.3 was used for acidity measurements. Figure 2-6 depicts the sequence of laboratory
operations for cloudwater and wet deposition samples.
Filter pack samples were loaded, shipped, received, extracted, and analyzed by Harding ESE
personnel at the Harding ESE laboratory. For specific extraction procedures refer to Anlauf et al.
(1986) and the draft CASTNet QAPP (Harding ESE, 1999). Filter packs contained three filter types
in sequence: a Teflon® filter for collection of aerosols, a nylon filter for collection of nitric acid
(HNOj), and dual potassium carbonate (K2CO3)-impregnated cellulose filters for collection of sulfur
dioxide (SO2). Following receipt from the field, exposed filters and blanks were extracted and
analyzed for anions and NHJ, as described previously for cloudwater and wet deposition samples.
Refer to the draft CASTNet QAPP (Harding ESE, 1999) for detailed descriptions of laboratory
receipt, breakdown, storage, extraction and analytical procedures for all three sample types.
Results of all valid analyses were stored in the laboratory data management system [Chemistry
Laboratory Analysis and Scheduling System (CLASS™)]. Atmospheric concentrations were
calculated based on the volume of air sampled following validation of the hourly flow data.
Atmospheric concentrations of paniculate SOf, NO3) and NHJ were calculated based on analysis of
Teflon® filter extracts; HNO3 was calculated based on the NO3 found in the nylon filter extracts; and
SO, was calculated based on the sum of SOJ found in nylon and cellulose filter extracts.
2.4 Data Management
Continuous data (meteorological, LWC, flow, and O3) were collected from the three permanent sites
in hourly and 5-minute averages. Hourly data were collected by nightly polling via telephone
modem, and 5-minute data were downloaded to diskettes from the DAS cartridge at least once
weekly. The polling software also recovered status files, power failure logs, and automated
calibration results (O3 only) from the previous 7 days. The hourly data and associated status flags
were ingested into Microsoft Excel™ spreadsheets. The continuous data were validated (flagged,
adjusted, or invalidated) based on the end-of-season calibration results, periodic calibration check
results (PVM only), and information provided by status flags and logbook entries.
p/castnet2/wa02/fnlsu m/fnlrpt_5. wpd
2-5
Harding ESE, Inc.
-------
MADPro Final Summary Report
Discrete data (filter pack, wet deposition, and cloudwater sample results) were managed by
CLASS™. In CLASS™, the analytical batches were processed through an automated QC check
routine. For each analytical batch, an alarm flag was generated if any of the following occurred:
1. Insufficient QC data were run for the batch;
2. The correlation coefficient of the standard curve was less than 0.995;
3. The 95-percent confidence limit of the Y-intercept exceeded the limit of quantitation;
4. Sample response exceeded the maximum standard response in the standard curve (i.e.,
sample required dilution);
5. Continuing verification samples (CVSs) exceeded recovery limits; or
6. Reference samples exceeded accuracy acceptance limits.
A batch of one or more flags was accepted only if written justification was provided by the
Laboratory Operations Manager.
Atmospheric concentrations for filter pack samples were calculated by merging validated continuous
flow data with the laboratory data [micrograms per filter (ug/filter)].
For cloudwater and wet deposition samples, a second laboratory check involved three interparameter
consistency checks:
1. Percent difference of cations versus anions (ion balance),
2. Percent difference of predicted versus measured conductivity, and
3. pH versus conductivity relationship of the sample compared to the expected relationship
when rainfall is assumed to be controlled by strong inorganic acid.
Evaluation of these interparameter consistency checks provided a method for determining whether
the analysis should be repeated or verified.
2.5 Quality Assurance
The quality assurance (QA) program for MADPro consisted of routine audits performed for
CASTNet, if applicable, and testing/comparisons of instruments unique to MADPro.
2.5.1 Field Data Audits
The following audits were conducted for field data:
1. Review of all reported problems with sensors and equipment at the sites and the actions
taken to solve such problems.
p/castnet2/wa02/fnlsum/fnlrpt_5. wpd
2-6
Harding ESE. Inc.
-------
MADPro Final Summarv Kepon
2. Review of calibration files for completeness and adherence to standard operating procedures
(SOPs). Certification results for transfer standards were also reviewed, and transfer sensor
serial numbers were cross-referenced with the transfer sensor serial numbers on the
calibration forms.
3. Comparison of final validated data tables to the raw data tables for identification and
verification of all changes made to the data. Summary statistics and results of diagnostic
tests for assessment of data accuracy were also reviewed.
2.5.2 Laboratory Data Audits
Laboratory data audits consisted of:
I. Review of all media acceptance test results,
2. Review of chain-of-custody documentation, and
3. Review of all QC sample results associated with analytical batches.
Table 2-1 lists the routine QC checks, calibrations, and audits performed at the Harding ESE
laboratory for CASTNet analyses. Table 2-2 lists the laboratory precision and accuracy objectives
for CASTNet and MADPro.
2.5.3 External Audits
As discussed in Section 2.2, external audits were performed at all three sites in 1997 by Ogden. Spot
reports summarizing sensor and system performance were submitted to Harding ESE within 72 hours
of a site audit. In 1999, ARS performed an external audit on the Whiteface Mountain site only, and
results were reviewed on a real-time basis with an official final report submitted within 1 month.
2.5.4 Precision and Accuracy
Accuracy of field measurements was determined by challenging instruments, with the exception of
the automated cloud sampler and Gerber PVM, with standards that were traceable to NIST.
Continuing accuracy was verified by end-of-season calibrations by Harding ESE personnel and by
external audits. No certified standards are currently available to determine the accuracy of the cloud
sampler and the PVM on a routine basis. Accuracy objectives for the rest of the field measurements
are listed in Table 2-3.
Overall precision of field measurements is best determined by collocating instruments and assessing
the difference between simultaneous measurements. Even though collocated sampling was not
conducted at MADPro sites, it was conducted at CASTNet sites. Since the meteorological
instrumentation and O3 analyzers at MADPro sites were identical to those used at CASTNet sites,
precision of these instruments can be inferred from the CASTNet Deposition Summary Report (EPA,
1998), and the CASTNet Annual Reports may be referenced for precision and accuracy results for
1994 through 1999. Precision objectives for field measurements are listed in Table 2-3.
p/castnet2/wa02/fnlsu m/fnlrp!_5. wpd
2-7
Harding ESE, Inc.
-------
MADPro Fined Summary Report
Accuracy of laboratory measurements was determined by analyzing an independently prepared
reference sample in each batch and calculating the percent recovery relative to the target value. The
percent recovery was expected to meet or exceed the acceptance criteria listed in Table 2-2. When
possible, the references were traceable to NIST or obtained directly from NIST. On occasion,
references were ordered from other laboratories.
Analytical precision within sample batches was assessed by calculating the relative percent
difference (RPD) and percent recovery of CVSs run within that batch. CVSs are independently
produced standards which approximate the midpoint of the analytical range for an analyte and are run
after every tenth environmental sample. Precision within a batch was also assessed by replicating
5 percent of the samples within a run. Replicated samples were selected randomly.
2.5.5 Quality Assurance Experiments
During the course of MADPro, several field and laboratory tests were conducted for QA purposes
and for comparability to other studies. The following sections briefly describe these comparisons.
2.5.5.1 Valente Versus Gerber PVM Field Comparison
The Valente Method, described in Valente, et al. (1989), was used in MCCP (Mohnen et al, 1990) to
estimate cloud LWC. During the MCCP program, a field intercomparison of a number of
instruments was conducted. The results of the tests indicated that the Valente Method provided
accurate estimates of cloud LWC. The Gerber PVM was under development during MCCP. A
comparison of the Valente Method for LWC and the current Gerber PVM for LWC provides an
additional check of the comparability of the two studies.
During the 1997 field season, a comparison of the Valente Method and PVM-100 was conducted at
the Whiteface Mountain and Whitetop Mountain sites. Hourly LWC values measured during cloud
events from both sites yielded differences of approximately 50 percent, with the Gerber PVM
recording the higher readings. The differences between the two methods were not affected by
variation in LWC values. The same Valente instrument was used at both sites whereas the Gerber
PVM was the instrument assigned to the specific site. Since the differences between the two
methods were the same at two different sites using different Gerber PVMs, the differences were
attributed to differences between the two methods and not specific instruments.
2.5.5.2 Comparison of Two EPA Gerber PVMs
In May 1998, the PVM assigned to Whiteface Mountain was collocated with the PVM at Whitetop
Mountain for approximately 3 weeks. The PVMs measured LWC values within 5 percent of each
other during all cloud events within this time period (Figure 2-7). This collocation was conducted to
check the precision between the PVMs used on the project. Similar precision (^5 percent) between
PVM-100s was obtained in a European study (Pahl and Winkler, 1995).
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
2-8
Harding ESE, Inc.
-------
MADPro Final Summary Report
2.5.5.3 Comparison of EPA Versus Colorado State University (CSU) Gerber PVMs
In January of 1999, three Gerber PVMs were destroyed in a fire at the Harding ESE Calibration
Laboratory, leaving only two PVMs for the three sites for the upcoming 1999 field season. A Gerber
PVM was leased from CSU for use at the Whiteface Mountain site. Prior to deployment at the site,
the CSU PVM was collocated with the Whitetop Mountain EPA PVM during May 1999 for
approximately 3 weeks. During the first half of this time period, the CSU PVM was reading
approximately 45 percent lower with respect to the EPA PVM. The LWC values measured with the
EPA PVM ranged from 0.04 to 0.63 g/m3. The PVMs were collocated 5 m apart, at the same height
and alignment. However, to rule out any differences due to location, the two PVMs were switched
around. The same difference (45 percent) was observed.
Since the difference between the EPA and CSU PVMs was very similar to the difference observed
between the Valente and EPA PVMs, the accuracy of the EPA PVMs was questioned.
2.5,5.4 Evaluation of LWC Instruments
The Energy Center for the Netherlands (ECN) was contacted to verify the accuracy of the PVMs
used in MADPro. The cloud chamber at ECN has the capability to generate clouds of known LWC
and was used to calibrate the Mallant cloud detector used during MCCP. The facility is described in
detail in Gerber etal. (1993) and is summarized below.
The fogs are characterized in the test section (or chamber) by measuring the LWC with a filter
method and measuring the droplet spectra with an FSSP-100. Two filter systems are run side by side
as a check of the measurement procedure. The filters consist of hydrophobic Pall filters placed in
housings that face the flow. Air is drawn through the filters isokinetically at a flow rate of
2.15 m/sec. The filters are conditioned in the operating chamber for at least 2 hours before LWC is
measured by weighing the filter both before and after a specified time interval, generally about
1 hour. The constant high relative humidity (Rh) in the chamber and filter-sampling procedures
minimizes LWC measurement errors due to evaporation or growth of droplets collected in the filter.
Such errors may be important when the filter method is used under ambient conditions where Rh can
be subsaturated or supersaturated (e.g., see Valente et al., 1989). The estimated accuracy of filter
measurements of LWC in the chamber is 10 percent. The FSSP-100 was located near the filters in
the test section.
The MADPro QA testing was conducted in February 2000 and included the two EPA PVMs and the
Valente Method used in the 1997 comparison. The CSU PVM was unavailable at the time of the
test. The tests included a series of cloud LWC from 0.1 to 0.9 g/m3 at cloud droplet sizes of JO and
25 mass median aerodynamic diameter (MmaD). The test concentrations were replicated on
different test days and different runs. A summary of the results is provided in Table 2-4. The two
EPA PVMs measured within 5 to 8.4 percent of the calibration standard over the entire test range of
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
2-9
Harding ESE, Inc.
-------
MADPro Final Summary Report
LWCs and droplet sizes. The Valente instrument measured within 18.2 percent of the Pall filters.
This result is contrary to the performance of the Valente during field tests when it consistently
measured the LWC approximately 50 percent lower with respect to the PVM-100.
Based on the ECN results and results obtained from field testing (Section 2.5.5.2), the precision
between the EPA PVMs is within ±5 percent. This precision value also agrees with the results found
by Pahl and Winkler (1995) in a German inter-comparison study of PVM-lOOs. Due to the
consistency of test results for the EPA PVMs, the LWC values collected with the CSU PVM at
Whiteface Mountain during 1999 were adjusted by +35.0 percent.
The ECN results for the Valente, however, did not agree with results of field testing between the
PVM-lOOs and the Valente. The difference is most likely caused by evaporation of water from the
instrument when deployed in the field and/or faulty technique during removal and taring of the
cartridge. Before any conclusions can be made with any certainty, however, further field testing by
collocation of the Valente and the PVM-100 is essential. Once the field accuracy of the Vaiente
instrument is quantified, the MCCP LWC data may require adjustment and cloud deposition
estimates may need to be recalculated based on the adjusted LWC to provide a database that would
provide better comparison with MADPro results.
2.5.6 Whitetop Mountain LWC Data
During validation of the 1998 MADPro data, it was noticed that Whitetop Mountain LWC values
never rose above 0.41 ug/m3 after mid-June. Before this time period the LWC data exhibited the
normal range of values for this parameter (0.0 to 0.98 ug/m3). Review of the site logbook entries and
of the LWC data showed that the PVM began to malfunction on June 13, 1998. Data were
invalidated from June 13 through 16 when repairs were conducted on the PVM. However, from this
point on, the LWC values did not rise above 0.41 ug/m3 as stated above. Based on analysis of LWC
data (range, average, time series plots, etc.) from Whitetop Mountain from previous and subsequent
years, the accuracy of the LWC data from June 17 until the end of the season on October 7 is
suspect. The documentation describing the instrument repair procedures does not provide enough
information to confidently adjust the data to correct for this suspected error. What can be confidently
inferred from the documentation and review of the data is that the zero reading was accurate as well
as readings at the calibration span value of approximately 0.16 ug/m3. The site operator had
performed monthly calibrations which checked the zero and span values. No problems were
reported with calibration results. Review of the data also indicate that around the 0.05 ug/m3 value,
indicative of the existence of a cloud, the PVM results agree with logbook comments and other
meteorological parameters. Values below 0.05 ug/m3 also correlate with recorded weather conditions
(clear, no clouds) in the logbook as well as weather conditions indicated by the meteorological data.
Since there was no reliable information to adjust data points above the calibration span value of 0.16
ug/m3, the analyses contained within the main body of this report use the 1998 LWC data from
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd 2-10 Harding ESE, Inc.
-------
MADPrtr Final Summary Report
Whitetop Mountain as reported. The accuracy of cloud frequency and duration information is not
affected since for calculation of these values, accuracy of the PVM at 0.05 ug/m3 is most important.
The accuracy of the values above this point is not relevant as the individual values are not used. Only
the number of cloud hours are important and useful. The analyses which may be affected are those in
which the Whitetop Mountain hourly LWC values are used: time series plot of LWC, normalized
(with respect to LWC) concentration values, and linear regressions of normalized concentrations.
Review of maximum LWC values from previous and subsequent years indicate that at Whitetop
Mountain the maximum values range from 0.66 ug/m3 (1995) to 1.01 ug/m3 (1997). Given that the
zero and span values were accurate, several slope/intercept adjustments were applied to the data
collected after June 16. The observed maximum (after June 16) of 0.41 ug/m3 was set to equal
values of 0.75, 0.80, 0.85, 0.90, and 0.95 ug/m3 to generate slope/intercept values with which to
adjust the data. Analysis of the data after application of each of these adjustment factors showed the
0.85 ug/m3 maximum set-point to yield the most reasonable results based on comparison to data
from other years as well as data from the Clingman's Dome site. As can be seen in Figure 3-7, the
LWC data from Whitetop Mountain and Clingman's Dome track each other closely, with a few
exceptions, up until July of 1998 and then again in 1999.
Even though this report uses the unadjusted data, the slope/intercept adjustment, using 0.85 ug/m3 as
the maximum set point, was applied (for investigative purposes) to the 1998 LWC data collected
after June 16. The LWC time series plot and the normalized concentration bar charts (Figures 3-37
through 3-40) were replotted with this adjusted data for comparison purposes. These results are
presented in Appendix A as Figures A-l through A-5.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
2-11
Harding ESE, Inc.
-------
-------
MADPro Final Summary Report
3.0 Liquid Water Content and Cloudwater Chemistry
3.1 Cloud Frequency and Mean Liquid Water Content
Mean monthly cloud frequencies by year for the three sites are summarized in Table 3-1. The
monthly cloud frequencies by month and year are also depicted as bar charts in Figures 3-1
through 3-3. Monthly cloud frequencies were determined by calculating the relative percent of all
hourly LWC values equal to or greater than 0.05 g/m3, or:
CF =
100 * (# of valid hourly LWC values *> 0.05 g/m3)
(3-D
where n is the number of valid hourly LWC values per month.
Any month with less than 70 percent valid LWC data was not considered representative of the
monthly weather conditions for that month and was not used in any analyses. Cloud frequencies vary
from month-to-month, year-to-year, and from site-to-site. Generally, Whitetop Mountain exhibits
slightly lower cloud frequencies with respect to Whiteface Mountain and Clingman's Dome.
Monthly mean LWC values for each year and site are shown in Figures 3-4 through 3-6. Mean LWC
was calculated by taking the average of all hourly LWC values equal to or greater than 0.05 g/m3
during the month. Only those monthly values that passed the 70 percent completeness criteria were
plotted. Figure 3-7 depicts the same information as a line graph with all three sites plotted together.
In this figure it is apparent that Whiteface Mountain experiences clouds with significantly higher
LWCs than the two southern sites, usually 1.5 to 2 times greater.
Height above cloud base could be a significant factor in determining the LWC values at the
collection location. Even though the height above cloud base in not quantitatively known for the
three sites, it is estimated that cloud base is lower at Whiteface Mountain with respect to the two
southern sites (Mohnen et.al, 1990). Since LWC increases vertically from cloud base up through the
cloud, height above cloud base could account for the higher LWC values at Whiteface Mountain.
3.2 Cloudwater Chemistry
Annual summary Cloudwater chemistry and LWC statistics are presented in Table 3-2. Table 3-3
lists the total number of samples or 'records' that were collected each season of operation at the three
sites. Samples were accepted and used for all subsequent analyses if they met an acceptance criteria
p/castnet2Ava02/fnlsum/fnlrpt_5.wpd
3-1
Harding ESE, Inc.
-------
MADPro Final Summary Report
based on the cation-to-anion ratio. Samples were eliminated if:
1. Both the anion sum and the cation sum were a 100 microequivalents per liter
(ueq/L) and the absolute value of the RPD was >100 percent; or
2. Either the anion sum or the cation sum was > 100 ueq/L and the absolute value of
the RPD was >25 percent.
The RPD was calculated from the following formula:
RPD = 200* (cations-anions)/(cations+anions)
Applying this acceptance criteria eliminated 694 out of 6,186 samples, or 11.8 percent.
3.2.1 CloudwaterpH
The mean annual pH of cloudwater samples for 1994 through 1999 for all three sites is shown in
Figure 3-8. Mean pH was obtained by first calculating the mean of the hydrogen ion (H*)
concentration and then converting this value back to pH units by taking the negative log. A steady
decrease in mean annual pH is evident for Ciingman's Dome.
The pH values are also depicted as frequency distributions averaged over all years in Figures 3-9
through 3-11. The frequency distribution for Whiteface Mountain (Figure 3-9) shows a greater
spread of pH values over the range of pH, 3.4 to 5.0, than either Whitetop Mountain or Ciingman's
Dome. Whitetop Mountain (Figure 3-10) exhibits a relatively normal distribution with most pH
values falling in the 3.6 to 3.8 range; whereas, the majority of pH values for Ciingman's Dome are in
the 3.4 to 3.6 range with very few samples above a pH of 4.0 (Figure 3-11).
The difference in the distribution of pH values may be related to the prevailing wind directions and
upwind sources. Figures 3-12 and 3-13 show cloudwater SOf concentrations versus arrival sector
for each of the three years 1994 through 1996 for Whitetop Mountain and Whiteface Mountain,
respectively. Arrival sector is defined as one of eight upwind directions estimated from 36-hour back
trajectories (Heffler, 1983). Arrival sector 1 represents winds from approximately north-northeast.
Figure 3-12 shows that the highest cloudwater SO^" concentrations at Whitetop Mountain are related
to winds from the north. Similarly, Figure 3-13 shows that the highest cloudwater sulfate
concentrations at Whiteface Mountain are associated with winds from the southwest. Figure 3-14
provides information on annual SO2 emissions by state for the eastern U.S. for 1995, the mid-year of
the three-year period.
The information in Figures 3-12 and 3-13 is shown as pollution roses in Figure 3-15. The pollution
roses were constructed by plotting cloudwater concentrations by arrival sector for 1994 first, then for
1995, and finally for 1996. This procedure obscures 1994 and 1995 data if, for example, 1996
observed the highest concentration. This is the situation for arrival sector 8 for Whitetop Mountain.
p/castnet2/wa02/fhlsum/fnlrpt_5.wpd 3-2 Harding ESE, Inc.
-------
MADPro Final Summary Report
In any event, Figure 3-15 relates cloudwater concentrations to upwind source regions. Whiteface
Mountain experiences relatively high concentrations with trajectories originating from Pennsylvania,
western New York, and the Ohio Valley, areas with high SO2 emissions (Figure 3-14). Relatively
low concentrations are observed with trajectories passing over Vermont and northern New England.
Whitetop Mountain experiences high cloudwater concentrations with trajectories from the Ohio
Valley.
The data in Figures 3-12 and 3-13 are relatively consistent over the three years. Consequently, the
concentrations reported for 1997 through 1999 are probably associated with polluted air masses from
the southwest in the case of Whiteface Mountain and from the north for Whitetop Mountain. Annual
variability in cloudwater concentrations can be at least partially explained by backward trajectories
and upwind source regions.
Figures 3-16 through 3-24 show the pH measurements from all accepted samples as a time series for
the years 1995, 1997, and 1999 for the three sites. These time series plots are useful for illustrating
seasonal differences in pH values within a site and differences between sites within the same year.
The pH values are also plotted as mean annual hydrogen concentrations in Figure 3-25 and as a line
chart with the values for all three sites plotted together for easier comparison in Figure 3-26. From
these figures it can be seen that the H* concentrations at Clingman's Dome have steadily increased
since 1996. A definite pattern is not discernible for either Whiteface Mountain or Whitetop
Mountain.
In Figure 3-27, H* concentrations are plotted as monthly means averaged across all years (1994
through 1999) to show monthly variations: Whitetop Mountain shows a definite peak in August;
Clingman's Dome exhibits higher monthly H* concentrations earlier in the season (June and July);
and there is no obvious peak for Whiteface Mountain.
3.2.2 Cloudwater Anions
Figures 3-28 through 3-30 show annual means and monthly variations for SOf and Figures 3-31
through 3-33 show the annual and monthly statistics for NOj. A slow but steady increase in
SO^" means can be seen for Clingman's Dome with no evident pattern at the other two sites. It is
interesting to note the divergence between Clingman's Dome, Whitetop Mountain, and Whiteface
Mountain in 1999. Clingman's Dome continues the pattern of steady increase, but the other two sites
show a decline, which is rather sharp for Whitetop Mountain, in the annual cloudwater SO^
concentrations. Back trajectory analyses for 1999 would be useful in determining causes for the
difference observed at Clingman's Dome.
A similar pattern is observed for NOj except that, in this case, Clingman's Dome also exhibits a
decline in NOj values in 1999 (Figure 3-32).
p/castnetZ'wa02/fnlsum/fnlrpt_5.wpd
3-3
Harding ESE. inc.
-------
MADPro Final Summary Report
An early season peak during June and July, like the results for H*, is seen for SOf and NC>3 at
Clingman's Dome. Whitetop Mountain shows a peak in August for SOf and NOj which mirrors the
results for H*. The number of months sampled and the smaller magnitude of differences from month
to month at Whiteface Mountain does not allow for any such observations.
3.2.3 Cloudwater Cations
Mean annual NHJ concentrations for 1994 through 1999 for the three sites are presented in
Figures 3-34 and 3-35. The Whiteface Mountain and Whitetop Mountain sites follow the same
annual pattern from 1995 through 1999. Clingman's Dome shows a steady increase until a decline in
1999, similar to the decline exhibited by the other two sites.
The minor cation annual mean concentrations (Ca2+, Na% Mg2*, and K+) are shown in Figure 3-36.
Cl" annual means are plotted in this figure as well. The most notable observation is the steady
increase in annual Ca2* concentrations at Clingman's Dome.
3.2.4 Normalized Cloudwater Concentrations
Since cloudwater analyte concentrations may be influenced by year-to-year variations in LWC, the
concentrations presented in sections 3.2.1 through 3.2.3 were normalized for LWC by multiplying
each sample concentration (for H+, SOf, NO^ and NHJ) by its respective LWC value. These results
are presented in Figures 3-37 through 3-40. Normalizing the concentrations with respect to LWC had
no effect on results from Whiteface Mountain except for NHJ in 1998 (Figure 3-40) where the
normalized value is slightly Jower than the 1997 value instead of higher as in Figure 3-34 (non-
normalized NHJ concentrations). For Clingman's Dome, the 1996 normalized values for H+, SOf
and NO^ exhibit a different pattern (a decrease with respect to 1995) than the non-normalized
concentrations. Also, the 1998 SOf value (Figure 3-38) is lower with respect to the 1997 value
instead of slightly higher as in the non-normalized concentrations (Figure 3-28). There was no
change in the pattern of normalized concentrations compared to the non-normalized concentrations
for H+ and SOf at Whitetop Mountain. The NO, and NHJ results (Figures 3-39 and 3-40) however,
exhibit the same change in pattern for 1996 and 1998 with respect to the non-normalized
concentrations. The 1996 normalized values are higher than the 1995 values rather than a little lower
as in the non-normalized concentrations, and the 1998 normalized results are lower than the 1997
results rather than higher as exhibited by the non-normalized values (Figures 3-31 and 3-34).
3.2.5 Slide Mountain and Hunter Mountain Results
Figure 3-41 presents results from the 3 years of operation of the mobile site in the Catskill Mountains
of New York. In 1995, collection was conducted on Slide Mountain, and during 1997 and 1998 on
Hunter Mountain. The two locations are within approximately 30 miles of each other and experience
the same weather regime. Therefore, data from the 3 years are comparable.
p/castnet2/wa02/fn!sum/fnlrpt_5.wpd
3-4
Harding ESE, Inc.
-------
MADPro Final Summary Report
A large increase in SOf values is evident between 1997 and 1998. In 1995 and 1997, the SOf
means were lower than the seasonal SOf means for the three permanent sites. In 1998, however, the
SOf value is noticeably higher than the Whiteface Mountain mean, slightly higher than the Whitetop
Mountain mean, and just slightly lower than the Clingman's Dome mean for 1998. The NOj values
are lower with respect to the other sites in 1995 and 1997, but are similar to the Clingman's Dome
seasonal mean and surpass the Whiteface Mountain and Whitetop Mountain means in 1998. The
NHJ means are lower than the other three sites across all 3 years.
These results, especially the SOf mean concentrations, indicate that the Slide Mountain and Hunter
Mountain sites experience somewhat different weather patterns or different types of clouds than
Whiteface Mountain.
3.2.6 Discussion of Results
Linear regression analyses were performed for the three sites for the four major ions from 1994
through 1999 to check for statistically significant increases or decreases in concentrations over time.
The data were analyzed for both the normalized and non-normalized data sets using all sample
concentrations rather than monthly or seasonal averages. The p-values from these analyses are
presented in Table 3-4. For Clingman's Dome, the p-values for all four analytes for both data sets
(normalized and non-normalized) show a statistically significant increase. The p-values for
normalized NHJ and SOf concentrations exhibit greater significance. Whiteface Mountain p-values
were significant for NHJ and SOf for both data sets and show a decrease in these concentrations.
The p-values for only normalized SOf concentrations show a statistically significant increase at
Whitetop Mountain.
The regression plots for those sites and analytes with significant p-values are shown in Figures 3-42
through 3-46. The individual sample points have been removed from these plots since the large
number of points (>5000 in some cases) when plotted, visually obscure the information. The very
low R2 values for these regressions indicate that, even though the p-values show high statistical
significance (in most cases), the increases or decreases are not linear. The low R2 statistics are also a
result of the scatter from the large number of samples analyzed. Linear regression may not be the
most appropriate analytical tool for this particular data set. However, for the purposes of providing a
preliminary look at statistical significance in temporal trends, the linear regression results are useful.
More appropriate statistical analyses (e.g., Mann-Kendall trend test, etc.) should be performed to
further elucidate the nature of the temporal trends exhibited by this data.
Even though there were statistically significant temporal trends discernible for the major ions from
1994 through 1999, 6-year average concentrations were calculated and are presented in Table 3-5 for
information purposes only. For all four major ions (SOf, NOj, H+, and NH^), Clingman's Dome
exhibits the highest mean and median values. Whiteface Mountain, on the other hand, consistently
has the lowest mean and median concentrations for all four ions.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
3-5
Harding ESE, Inc.
-------
MADPro Final Summary Report
Clingman's Dome also exhibited statistically significant upward trends in seasonal SOf, NOj, H*,
and to a lesser extent, NHJ values as depicted in Figures 3-42 and 3-43. To fully understand the
consistently higher concentrations at this site, back trajectory analyses are recommended. Back
trajectory analyses for Whiteface Mountain and Whitetop Mountain from 1994 through 1996 (see
Figure 3-15) clearly show the different air masses experienced at these two sites. More local
emission sources and the possible resulting effect on the air quality of Clingman's Dome should also
be investigated (see Mueller, 2000). Butler et al. (in press) suggest that the lack of improvement, and
even an increase in the vicinity nearby Clingman's Dome, may be attributed to several large coal-
fired electric generating facilities that were not required to reduce SO2 or NO^ emissions under
CAAA.
Since SO2 emission levels in the eastern United States have decreased (EPA, 2000), a stronger
decreasing correlation might be expected at Whiteface Mountain. The increasing temporal trend for
SOf and the lack of discernible trends for the other three analytes at Whitetop Mountain, combined
with the results from Whiteface Mountain, indicate that cloudwater concentrations may not be
linearly related to SO2 emission reductions. Clouds may not contain sufficient amounts of the
necessary oxidants for a complete conversion of all available SO2. At the cloud water pH levels
usually observed at the MADPro sites, hydrogen peroxide (H2O2) is the primary, if not the only,
oxidant of significance (Seinfeld, 1986). The research of Dutkiewicz et al. (1995) at Whiteface
Mountain, NY indicates that, at SO2 concentrations above 1 ppb, H2O2 is not present in sufficient
quantities to oxidize all of the SO2 present. The situation is further complicated by the fact that
sources of SO2 and H,O2 in the atmosphere are unrelated and large variations in atmospheric
concentrations of H2O2 are observed (Guns and Hoffmann, 1990). However, more research and
monitoring must be conducted before any solid conclusions can be made since the lack of temporal
trends within this period may also be a function of, or influenced by, year-to-year variations in air
mass trajectories and regional and local meteorology.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
3-6
Harding ESE, Inc.
-------
MADPro Final Summary Report
4.0 Cloud Deposition
This chapter summarizes the modeled cloudwater deposition estimates for the three MADPro sites
for the period 1994 through 1998. The deposition estimates were made by applying the cloudwater
deposition computer model (CLOUD) (Lovett, 1984), parameterized with site-specific cloudwater
chemistry and meteorological data. The structure of the model, the input parameters, and the
sensitivity of the mode! to several parameters and assumptions are described in Section 4.1.
Deposition estimates for the three sites are also reported for monthly, seasonal, yearly, and multi-year
periods.
In addition, a semi-independent model (MCLOUD) was employed to explore alternative
parameterizations and additional model components beyond those offered by the primary model.
MCLOUD was used to produce an alternative set of cloudwater deposition estimates for eight site-
by-year combinations as well as deposition estimates for all 1999 data. These findings are discussed
in Section 4.2. A comparison of the results between the CLOUD and MCLOUD models is presented
in Section 4.3.
4.1 Cloudwater Deposition Model (CLOUD)
4.1.1 Model Description
The CLOUD model was originally described by Lovett (1984). Briefly, it uses an electrical
resistance network analogy to model the deposition of cloudwater to forest canopies. The model is
one-dimensional, assuming vertical mixing of droplet-laden air to the canopy from the top.
Turbulence mixes the droplets into the canopy space, where they cross the boundary layers of canopy
tissues by impaction and sedimentation. Sedimentation rates are strictly a function of droplet size.
Impaction efficiencies are a function of the Stokes number, which integrates droplet size, obstacle
size, and wind speed (Lovett, 1984). The impaction efficiency as a function of the Stokes number is
based on wind tunnel measurements by Thorne et al. (1982).
A resistance or deposition velocity type model does not solely represent the important atmospheric
dynamics controlling deposition. It also simulates the properties of the receptor surface and the
depositing substance. This model is the most common type found in the literature for simulating
atmospheric deposition. A resistance model is used because it simulates the relationship between
atmospheric concentration and flux to a surface.
The forest canopy is modeled as stacked 1-m layers containing specified amounts of various canopy
tissues such as leaves, twigs, and trunks. Wind speed at any height within the canopy space is
determined based on the above-canopy wind speed and an exponential decline of wind speed as a
function of downward-cumulated canopy surface area. The wind speed determines the efficiency of
p/castnet2/wa02/fnlsum/fnlrpt_5. wpd
4-1
Harding ESE, Inc.
-------
MADPro Final Summary Report
mixing of air and droplets into the canopy and also the efficiency with which droplets impact onto
canopy surfaces. The model is deterministic and assumes a steady state, so that for one set of
above-canopy conditions it calculates one deposition rate. The model requires the following as
input data:
1. Leaf area index (LAI) of canopy tissues in each height layer in the canopy,
2. Zero-plane displacement height and roughness length of the canopy,
3. Wind speed at the canopy top,
4. LWC of the cloud above the canopy, and
5. Mode of the droplet diameter distribution in the cloud.
From these input parameters, the model calculates the deposition of cloudwater, expressed both as a
water flux rate [grams per square centimeter per minute (g/cm2/min)] and as a deposition velocity
[flux rate/LWC, in units of centimeters per second (cm/sec)]. Deposition rates of ions are calculated
by multiplying the water flux rate by the ion concentration in cloudwater above the canopy. In the
original version of the model, a calculation of the evaporation rate from the canopy was also included
to estimate net deposition of cloudwater. For this report, only gross deposition rate was estimated, so
the evaporation routine was not invoked.
4.1.1.1 Canopy Structure
Comparing cloud deposition rates across sites is difficult because canopy structure (e.g., height and
LAI), which has an effect on deposition rates, can vary between locations even within a single site.
To focus on meteorological and chemical controls on the deposition process, a "standard" canopy
structure was used for all model runs at all sites. The structure chosen was described by
Lovett (1984). It is a monospecific Balsam Fir canopy with the following parameters: maximum
structure height of 1061 cm, zero-plane displacement of 837 cm, roughness length of 97 cm, and LAI
of 7.6 square meters per square meter (m2/m2). While this canopy parameterization was originally
developed for subalpine forests of New Hampshire, it is similar to forests that can be found at high
elevations on Whiteface Mountain (Miller et al., 1993a). Forests near the summits of Whitetop
Mountain and Clingman's Dome are similar in structure but may be somewhat taller (e.g., Mueller
et al., 1991). Nevertheless, the CLOUD model was executed with the same parameters at all sites.
Furthermore, the deposition model calculations represent the deposition expected to a forest stand
with the canopy structure specified as homogeneous throughout the stand. If that stand were at a
forest edge, the deposition would be greater. However, no information is available on the gap and
edge distribution in the forests at the three MADPro sites. Consequently, deposition estimates were
made only for homogeneous canopies.
4.1.1.2 Modifications from Original Version of the Model
Several changes in the model were made to improve its performance and tailor it to the sites under
study. The model calculates the droplet size distribution associated with each measured LWC
according to the formula of Best (1951). The calculation requires specifying the mode of the
p/castnet2/wa02/fnlsum/fnlrpt_5 .wpd
4-2
Harding ESE. Inc.
-------
MADPro Final Summary Report
distribution. Because droplet size information was not collected at the sites, the droplet size at all
sites was assumed to be a function of LWC as reported by Joslin et al. (1990) for the Whitetop
Mountain site. A regression was run based on the data in Figure 2 of the Joslin et al. paper
(see Figure 4-1), and the following equation was calculated:
Z)2 = 49.9874 + 308.1664X LWC
(4-1)
where D is the mean droplet diameter (adjusted r2 = 0.94, n = 6). The mean diameter, calculated by
taking the square root of both sides of Equation 4-1, was assumed to be identical to the mode
diameter used in the Best (1951) distribution, which is a good assumption because of the roughly
normal-shaped distribution of cloud droplet diameters. Equation 4-1 was used for all sites.
In summary, the Best (1951) paper provides a method for calculating the droplet size distribution
when the modal droplet diameter is known. Data from Whitetop Mountain were used to estimate the
mode of the size distribution as a function of LWC. The Best function was then used to calculate the
full droplet size distributions.
Another issue concerning the droplet size distribution is the discretization of the distribution in the
model simulations. In the original model, three droplet diameter classes were used: 0 to 10, 10 to 20,
and 20 to 30 micrometers (um). Miller et al. (1993a) suggested that as many as 500 droplet size
classes may be necessary to simulate the deposition rate to within 0.1 percent. Simulations were run
with different numbers of droplet size classes and it was found that the deposition velocity using
20 classes was within <1 percent of the deposition velocity using 500 classes. The accuracy of
<1 percent was considered sufficient, given the error inherent in calculating or measuring a droplet
size distribution; and 20 droplet size classes (in the range of 0 to 100 um) were used for all
simulations.
The model uses canopy-top wind speed as an input parameter; however, in practice wind speed is
usually measured some distance above the canopy to avoid shadowing by surrounding tree crowns.
Assuming a logarithmic wind profile above the canopy, the canopy-top wind speed (uh) can be
calculated from the wind speed at any height z (u2) by the following equation:
uh =
In[(z-<0/z0]
(4-2)
where h is the height above the canopy where wind speed was measured, d is the zero-plane
displacement height, and ZQ is the roughness length. The anemometers at Clingman's Dome and
Whitetop Mountain were 3.2 m above the canopy, and that information was used to calculate
canopy-top wind speed using Equation 4-2. The anemometer at Whiteface Mountain was on the roof
of a building above tree line, so this site required a more complicated adjustment, described below.
p/casinet2/wa02/fnlsum/fn!rpt_5.wpd
4-3
Harding ESE, inc.
-------
MADPro Final Summary Report
4.1.13 Parameter Adjustments at Whiteface Mountain
The cloud chemistry, LWC, and wind speed measurements at Whiteface Mountain were made on the
roof of the summit observatory building at an elevation of 1,483 m. This elevation is above the tree
line. Miller et al. (1993a) report an elevational gradient of canopy structure on Whiteface Mountain,
and their highest site, at an elevation of 1,350 m, closely resembles the height and species
composition of the forest used in the deposition calculations. Therefore, 1,350 m was used as the
target elevation for the model application at Whiteface Mountain.
Wind speed, LWC, and cloud chemistry all change with elevation and, hence, the measured values at
the top must be scaled if they are used for modeling at lower elevations. Miller et al. (1993a)
describe the elevational wind profile u(z) with the function:
u(z)=ufa/(l+[(ufa-u0je-r(z-Zi))
(4-3)
where u0 is the above-canopy wind speed at the 1,050-m measurement site, ufa = 3.2 u0 and is the
projected free-air wind speed above the mountain, z> = 1,000 m and is the elevation of the inflection
point of the wind profile, and r = 0.003675. Evaluating Equation 4-3 with z = 1,350 (for the chosen
elevation of 1,350 m), and again with z = 1,483 (the summit elevation), and then taking the ratio of
the two, yields:
= 0.854«(l,483)
(4-4)
Equation 4-4 was used to calculate the above-canopy wind speed for the 1,350 m site. Equation 4-2
was then used to scale that to the canopy-top wind speed, assuming the above-canopy wind speed
estimated by Equation 4-4 corresponds to a point 3.2 m above the canopy.
Based on data from the summit and a site at 1,050 m on Whiteface Mountain, Miller et al. (1993a)
estimated that cloud LWC decreases linearly with elevation on Whiteface Mountain, from a
growing-season average of 0.50 g/m3 at the summit to 0.39 g/m3 at the 1,050-m site. This represents
a 22-percent decrease in LWC over a 433-m elevational drop. A linear function was calculated to
describe the proportional decrease in LWC with elevation as follows:
LWC(z) = LWC(s)[l - (0.22/433)(1,483 -
(4-5)
where LWC(s) is the LWC at the summit. For z = 1,350 m, Equation 4-5 reduces to:
(4-6)
p/castnet2/waQ2/fnlsuin/fnlrpt_5. wpd
4-4
Harding ESE, Inc.
-------
MAD Pro Final Summary Report
All LWC values reported for Whiteface Mountain were adjusted by Equation 4-6 to calculate the
LWC at the modeled site.
Following Mohnen et al. (1990) and Miller et ai (1993a), the elevational change in cloudwater
chemistry was also modeled as a dilution process in which higher values of LWC result in lower ion
concentrations in cloudwater. Consequently, in correspondence to Equation 4-6, chemical
concentrations C in cloudwater at the 1,350-m target site were calculated as:
C(l,350) = C(j)/0.932
(4-7)
where C(s) is the ion concentration at the summit (1,483 m) site.
In reality, cloud pollutant concentrations vary as a function of droplet size. However, MADPro did
not collect size-resolved concentration data. Instead, cloud water was deposited onto a passive
collector thereby measuring deposition-weighted pollutant concentrations (see Section 2.2). These
concentration measurements are precisely the type of concentrations required by the CLOUD model
for simulating deposition to forest canopies. Consequently, not using size-resolved concentration
data will produce no bias in the calculations.
Chemical and meteorological parameters were estimated for two separate sites on Whiteface
Mountain:
• Whiteface Mountain Summit (WFMS) which represents the summit data; and
• Whiteface Mountain Lower (WFML), which represents the 1,350-m site for which
deposition was modeled.
4.1.2 Model Calculations
The CLOUD model was run for all samples from the three MADPro sites for which wind speed and
LWC data were available. The modeled cloudwater flux data were merged with the chemical data
and were used to make the following calculations:
1. The Whiteface Mountain data were processed to create results for two sites, WFMS
(summit site) and WFML (1,350-m site). Deposition fluxes were calculated only for
WFML.
2. Chemical fluxes were calculated as the product of modeled water fluxes and
chemical concentrations for all observations for which chemical data were available.
3. The data set was screened for ion balance in the chemical data. The chemical data
were screened based on the anion and cation sums and the RPD between anions and
cations presented in Section 3.2. Figure 4-2 shows the RPD plotted against the anion
sum and illustrates the screening criteria as lines on the plot. All calculations on
Figure 4-2 were made on the screened data set.
p/castnel2/wa02/fnlsum/fn lrpt_5. wpd
4-5
Harding ESE. Inc.
-------
MADPro Final Summary Report
4. Monthly means were calculated for each site and year, weighted by duration of the
sample to give the most appropriate mean for calculating deposition. These data
were merged with monthly data on cloud frequency from each site.
5. Deposition totals were calculated by site, year, and month by multiplying the
duration-weighted mean chemical fluxes by the cloud-hours for the month. The
cloud-hours were calculated as the cloud frequency times the total hours in the
month.
6. Deposition fluxes for each sampling season were calculated. To compare equitably
across sites, the deposition season was considered to be June through September.
Data outside that period were almost never available for the Whiteface Mountain site
and were spotty at the other sites. Even within the June through September period,
the monthly deposition for each site could not be summed to calculate the seasonal
totals because some months had no available data due to equipment failures or data
screening. Therefore, the seasonal deposition totals were calculated as the
duration-weighted mean chemical fluxes for all observations in the 4-month period
multiplied by the total cloud-hours for the 4-month period. Seasonal deposition
totals were not calculated unless at least 3 months of valid cloud frequency data were
available for that season.
4.1.3 Model Sensitivity
Before the production runs for the three MADPro sites are presented, the sensitivity of the CLOUD
model is discussed. The sensitivity of the CLOUD model to various input parameters, including
wind speed, droplet size, and canopy structure, has been reported in several publications (Lovett,
1984; Lovett and Reiners, 1986; Mueller, 1991; Miller etal, 1993a). The modeled deposition
velocity shows a nearly linear response to changes in canopy-top wind speed in the range of wind
speeds commonly encountered (Figure 4-3). This indicates that the model estimates of cloudwater
deposition are very sensitive to the correct specification of wind speed at the canopy top.
The slight flattening of the slope of the curve at low wind speeds in Figure 4-3 results from the
increased importance of sedimentation flux of droplets when the wind speed decreases. The
relationship shown in the figure has little to do with turbulence. Deposition velocity increases with
wind speed primarily because increasing wind speeds increase the momentum of the cloud droplets,
causing more efficient impaction on vegetation surfaces.
In principle, there should be no sensitivity of the cloud deposition velocity to LWC, if all other
parameters are held constant, because the deposition velocity is the deposition flux normalized by the
LWC. (Another way of expressing this is that the deposition flux is directly proportional to the
LWC, and the deposition velocity is the constant of proportionality.) However, in reality, other
parameters are not always held constant. It has been shown many times that the droplet size
distribution varies with the LWC [e.g., Mohnen etal. (1990) and Mueller (1991)], and the deposition
p/castnet2/wa02/fnlsum/fnlrpt_5. wpd
4-6
Harding ESE. Inc.
-------
MADPro Final Summary Report
velocity does depend on the droplet size distribution. This creates a de facto dependence of the
deposition velocity on the LWC as shown in Figure 4-4. In Figure 4-4, the data in parentheses are
the modes of the droplet size distributions corresponding to each LWC, calculated from
Equation 4-1.
With less than a two-fold increase in deposition velocity resulting from a five-fold increase in LWC,
the deposition velocity is much less sensitive to the LWC than to wind speed. The cloudwater flux
rate, however, is the product of the deposition velocity and the LWC, and, as such, has a greater than
1 : 1 sensitivity to LWC. In other words, doubling the LWC will result in more than doubling the
modeled water flux rate.
The model sensitivity to canopy structure has been examined in detail by Lovett and Reiners (1986)
and Mueller (1991). However, canopy structure was not varied in this study. This version of the
model does not depend on temperature, Rh, solar radiation, or evaporation.
Two other parameters, although not used in the calculations of instantaneous deposition rates in the
model, are important in the final calculations of cloud deposition values. These include cloud
chemistry and cloud frequency (the percent of hours that the site is immersed in cloud). Calculated
cloud deposition totals are directly proportional to both of these factors. For example, if other
parameters are held constant, doubling the cloud SOJ" concentration will double the calculated cloud
f deposition, and doubling the cloud frequency will double the calculated deposition of all ions.
4.1.4 Results of CLOUD Model Calculations
Monthly deposition estimates are given in Table 4-1. Monthly variability is significant [e.g., suifate
deposition varies from approximately 16 kilograms per hectare per month (kg/ha/mo) to less than
1 .0 kg/ha/mo]. Monthly deposition values typically peak in July or August. The highest fluxes were
modeled for Whiteface Mountain and the lowest for Clingman's Dome. Depositions of pollutants are
generally related to cloudwater deposition.
Seasonal deposition estimates are provided in Table 4-2. For this purpose, a season is defined as the
period from June through September. At least three valid months were required to calculate the
seasonal deposition value. The highest seasonal fluxes were modeled for Whiteface Mountain. The
highest seasonal deposition amount was modeled for Whiteface Mountain because of higher wind
speeds and LWC (e.g., see Figure 3-7). No temporal trend is evident in the data.
4.2 MCLOUD Model Calculations
In addition to the model results discussed in Section 4.1, a second model (MCLOUD) was run to
explore uncertainties related to the structure and parameterization of inferential deposition models
and to assess how these uncertainties affect estimates of water and chemical fluxes. The strategy for
p/castnet2Ava02/fnlsum/fnfrpt_5.wpd 4-7 Harding ESE. Inc.
-------
MADPro Final Summary Report
MADPro cloudwater deposition modeling is to use this general sensitivity analysis to better interpret
the results of the CLOUD model used to provide the deposition estimates. Specifically, this analysis
investigated differences in results between MCLOUD (Miller et al, 1993a) and CLOUD (Lovett,
1984). To maintain a partially blind comparison of model calculations, MCLOUD used most (but
not all) of the model parameterization information employed in CLOUD. The scope of the
MCLOUD modeling effort was limited to the following site-year data sets: Clingman's Dome, TN
(1997); Whitetop Mountain, VA (1995 and 1997); and Whiteface Mountain, NY (1994
through 1998).
The MCLOUD deposition model used for these additional analyses is described in detail by
Miller et al. (1993a). MCLOUD is an expansion of the Lovett (1984) CLOUD model and is,
therefore, only semi-independent of the primary CASTNet model. Differences between the two
models are discussed below. The MCLOUD model permits alternative representations of the
physical processes involved in cloudwater deposition as well as alternative "dimensions" for
computational analysis. Potential differences in model results that might occur because of
differences in model parameter systems are discussed in the following subsections.
4.2.1 MCLOUD Model Structure and Parameterization for MADPro Sites
4.2.1.1 Vertical Distribution of Leaf Area in the Forest Canopy
The LAI varies vertically within a forest canopy. Since the actual variation is not known for the
three sites, the vertical distribution of leaf area from a similar Balsam Fir forest at 1,350-m elevation
on Whiteface Mountain (Miller et al., 1993a) was scaled to the LAI of the CLOUD standard canopy
(Section 4.1.1.1) to parameterize MCLOUD. Consequently, the two models use different vertical
distributions of leaf area; and model results appear to be fairly sensitive to the vertical distribution of
leaf area (Figure 4-5). Therefore, it is expected that differences between the modeling results of
>±10 percent will exist on this basis alone.
4.2.1.2 Representation of Cloud Droplet Size Distributions
For the MCLOUD deposition modeling, the CLOUD assumptions regarding droplet-size
distributions were used for Whitetop Mountain and Clingman's Dome sites, and the Miller et al
(1993a,b) assumptions were used for Whiteface Mountain. The CLOUD model used the distribution
of droplet sizes using the Whitetop Mountain distribution parameters for ali three sites.
The consequences of the use of the Whitetop Mountain droplet-size distribution for the Whiteface
Mountain site were evaluated by both sensitivity analysis and by running MCLOUD for the WFM-97
data set under both sets of assumptions. Sensitivity analyses suggested a complicated pattern of
model response to different combinations of LWC and wind speed using the two different dropsize
distributions (see Figure 4-6). The model runs using the Whitetop Mountain distribution
underpredicted the results obtained with the Whiteface Mountain distribution by 1 to 24 percent for
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
4-8
Harding ESE. Inc.
-------
MADPro Final Summarv Kenan
all LWCs at wind speeds below 10 m/sec. At wind speeds > 10 m/sec, use of the Whitetop Mountain
distribution increased deposition rates from 1 to 6 percent over the results obtained with the
Whiteface Mountain distribution. This complicated model response tended to average out over the
broad range of combinations of wind speed and LWC encountered at Whiteface Mountain. Seasonal
average cloudwater fluxes (millimeters of water) differed by ~3 percent.
4.2.1.3 Number of Droplet Size Classes in the Models
Five-hundred droplet size class "bins" were used to represent the continuous droplet size distribution
in a cloud for MCLOUD. This is a practical compromise between computational efficiency and
effective description of the continuous droplet size distribution (see Miller et ai, 1993a, Figure 1, for
more explanation). CLOUD employed 20 droplet size classes, so comparative runs with 20 droplet
size classes were also performed.
The consequences of the number of size classes chosen to represent the continuous size distribution
was previously discussed by Miller et ai (1993a, Figure 1). Miller et al. (1993a) report that
500 droplet size classes are required to arrive at a deposition velocity within 0.1 percent of the
limiting value. The consequences of these two different numerical parameterizations (20 versus 500)
were explored with MCLOUD using data from Clingman's Dome for 1997. Droplet fluxes were
~10 percent less using 20 droplet size classes compared to 500 classes (Figure 4-7). This degree of
deviation is consistent with the previous analysis of Miller et al. (1993a) for 20 droplet classes.
4.2.1,4 Height of the Wind Speed Measurement Above the Whiteface Mountain Canopy
The CLOUD model was assigned the reference height of the wind speed versus elevation function
presented by Miller et al. (1993a) as 3.2 m above the canopy. The actual reference height for this
function is 1.0 m above the canopy (Miller et ai, 1993a, Figure 3). The 3.2-m reference height is the
height of the wind speed measurement above the forest canopies at Whitetop Mountain and
Clingman's Dome. The effect of choosing a 3.2-m wind speed reference height versus a 1.0-m height
at Whiteface Mountain was explored.
CLOUD uses a reference height of the wind speed versus elevation function as 3.2 m above the
canopy. Using a 3.2-m wind speed reference height (versus 1.0-m) at Whiteface Mountain results in
a lower cloudwater deposition of -20 percent (Figure 4-8).
4.2.1.5 Representation of Atmospheric Conditions
Atmospheric pressure, temperature, and the non-condensed water vapor content of the air have subtle
effects on droplet deposition due to their effects on the density and kinematic viscosity of the air.
Variation in the acceleration due to gravity with distance affects large droplet sedimentation
velocities and particle rebound effects. The MCLOUD model includes these subtle effects. CLOUD
p/castnet2/wa02/fnlsum/fnlrpl_5.wpd
4-9
Harding ESE, Inc.
-------
MADPro Final Summary Report
does not include these effects (Lovett, 1984). Evaporation is not being considered by either model.
The more accurate representation of atmospheric conditions such as the density and kinematic
viscosity of air in the model has < 1 percent effect on modeled fluxes.
4.2.2 MCLOUD Calculations
4.2.2.1 Data Screening
Data provided for each site and year were screened according to the following criteria:
1. If LWC was missing, the sample was discarded because water deposition could not
be determined.
2. If wind speed was missing, the sample was discarded because water deposition could
not be determined.
3. If no sample duration information was available, the sample was discarded because
deposition could not be calculated.
4. If a sample duration of greater than 3 hours was required to obtain a satisfactory
sample volume for chemical analysis, the sample was discarded because of the
potential for evaporation of droplets on the cloudwater collector.
5. If the total measured ionic content was > 200 ueq/L and the percent relative ion
deficit (%RD) was > 25 percent, the sample was discarded due to suspect chemical
analysis [%RD = (sumcations-sumanions)/total ions * 200].
6. If the total measured ionic content was < 200 ueq/L and the %RD was > 100 percent,
the sample was discarded due to suspect chemical analysis.
Table 4-3 illustrates the effect of data screening on sample retention in the data sets. The
predominant cause for sample rejection was lack of a valid wind speed measurement. In the
WFM-95 data set, unacceptable ion balance was the primary reason for sample rejection, but missing
wind speed observations were still significant. Sample retention rates ranged from 47 percent for
WFM-95 to 83 percent for WTM-97.
Water flux to the forest canopy for individual samples was calculated using the cloudwater
deposition model, parameterized as described above, and using the LWC, wind speed, and duration
values for each sample that passed the data screening process. Water flux was combined with the
chemical concentrations to calculate individual sample ion fluxes to the forest canopy. Ion
concentrations in cloudwater were adjusted to account for LWC scaling for the 1,350-m site on
Whiteface Mountain.
4.2.2.2 Data Aggregation and Summary
Individual sample water and ion fluxes were summed for each month. Monthly sums were divided
by the amount of time represented by all samples from that month to arrive at an estimate of the
p/castnet2/wa02/fnlsum/fnlrpt_5. wpd
4-10
Harding ESE, Inc.
-------
MADPra Final Summary Report
average flux per hour of cloud duration for a given month. The average flux rate was multiplied by
the cloud frequency (Table 3-1) for each month to estimate monthly deposition. Monthly deposition
fluxes were summed over the June to September growing season.
4.2.3 Sensitivity Analysis
A sensitivity analysis was also performed using only Whiteface Mountain data. Due to the lack of
vegetation at the Whiteface Mountain summit measurement site, it was necessary to choose a
"representative" elevation on Whiteface Mountain to develop a forest canopy description and to
establish a range of environmental conditions that would be meaningful with respect to modeling
such a canopy. Placing a simulated forest canopy in the atmospheric conditions characteristic of the
summit measurement location would result in calculated deposition rates many times the rates likely
to be experienced by actual forest canopies in the northeastern United States. Consequently, a
representative elevation of 1,225 m was selected to define the conditions for the model sensitivity
analysis. The CLOUD modeling was based on an elevation of 1,350 m. The 1,225-m canopy is
considered similar enough to the CLOUD specification that the sensitivity analysis remains very
informative.
The atmospheric conditions observed at the Whiteface Mountain summit station during the 1996
measurement campaign were scaled to the 1,225-m elevation. For the purpose of evaluating
variations in the model canopy description, the 10th percentile, median, and 90th percentile values of
the scaled wind speed distribution were selected to represent model response over a range of wind
conditions (Figure 4-9). Similarly, the 25th percentile, median, and 90th percentile values of the scaled
LWC were selected to test the range of model response to varying cloud conditions. Model
simulations of cloudwater deposition rates were conducted using the nine combinations of wind
speed and LWC (Table 4-4) with a range of possible canopy descriptions (Table 4-5). To simplify
the interpretation of the simulation results, the instantaneous cloud droplet deposition velocities
simulated by the model were extrapolated to a growing season water deposition value (centimeters of
water deposited). This extrapolation assumed that the simulation conditions (LWC and wind speed)
were constant during cloud immersion over a growing season of 3 months. Cloud immersion
frequency was fixed at 25 percent of the period (typical for Whiteface Mountain, see Miller et ai,
1993a), except for one analysis which explicitly explored the effect of scaling cloud frequency with
elevation.
The range of wind speeds and LWC selected from Figure 4-9 were used to investigate the sensitivity
of 3-month, growing-season water depositions to changes in canopy structure and meteorological
conditions. Generally, the cloud frequency was assumed constant at 25 percent. Several figures
were prepared to illustrate MCLOUD model sensitivity; they are discussed in the following
subsections.
p/castnet2/wa02/fnlsum/fnlrpt_5. wpd
4-11
Harding ESE. Inc.
-------
MADPro Final Summary Report
4.2.3.1 Model Response to Variation in Forest Canopy Description
Observed Versus Pure Fir Canopy
Replacing 16 percent of the leaf area attributed to broad-leaf vegetation in the observed canopy with
a needle-leaf representation produced a minor effect (Figure 4-10). Generally, the pure needle-leaf
canopy had a 1.5 to 2.5 percent higher collection efficiency (CE) than the observed canopy with the
exception of high wind speed combined with moderate and high LWC conditions. The slightly
increased "openness" of a canopy with some broad-leaf foliage allows both momentum and droplets
to penetrate to deeper layers of the canopy. Under lower wind speed (lower turbulence) regimes, CE
is reduced rapidly as wind speed decreases in the upper part of the canopy. In lower LWC
conditions, a majority of droplets are scavenged in the upper canopy, therefore reducing the effective
LWC in the lower canopy space to near zero and reducing deposition, so the increased "openness"
has little effect.
Canopy Height
For a canopy with 10 m2/m2 LAI, decreasing the canopy height from the observed value of 17 m
resulted in decreasing cloudwater deposition (Figure 4-11). At a height of 10 m, the reduction was
generally less than 2.5 percent, except in low LWC conditions where the reduction approached
3.5 percent.
Leaf Area Index
For a 10-m canopy, variation in LAI from the observed value of 10 m2/m2 resulted in a mixed
response, depending on the combination of LWC and wind speed (Figure 4-12). Generally,
deviations were in the range of +2 to —4 percent. However, conditions of low wind speed, combined
with medium and high LWC, produced deviations of -6.5 and -9.8 percent, respectively, at
6 nr/nr LAI.
4.2.3.2 Model Response to Variation in Wind Speed and LWC
The MCLOUD deposition model, parameterized with the standard MADPro canopy (described
previously) and 1-m thick layers, has a moderately non-linear response to both increasing wind speed
and increasing LWC (Figure 4-13). For example, a 50 percent reduction in wind speed produces a
30 to 40 percent reduction in flux for constant LWC. Similarly, doubling the LWC produces almost
twice the level of the cloudwater flux for a constant wind speed.
4.2.3.3 Model Layer Thickness
Previous applications of 1-dimensional turbulent diffusion models for cloud droplet flux have used a
1-m layer thickness to describe the vertical structure of the forest canopy and meteorological
parameters within the forest canopy (Lovett, 1984; Lovett and Reiners, 1986; Mohnen, 1988;
Mueller and Weatherford, 1988; Saxena et al, 1989; Saxena and Lin, 1990; Joslin et al., 1990;
Mueller et al., 1991; Mueller, 1991; Miller et al., 1993a and b). Natural forest canopies are
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
4-12
Harding ESE. Inc.
-------
MADPrn Final Summary Report
heterogeneous enough, however, for a 1-m layer thickness description of the vertical distribution of
leaf area to introduce significant uncertainty for the value of any given layer. Regressions of model
results with selected layer thicknesses against results from the 1-m thickness yielded a maximum bias
of 0.92 (Figure 4-5). The use of 1-m layers to describe the standard canopy may result in a systematic
underestimation of cloudwater flux by 8 percent.
4.2.4 MCLOUD Model Results
MCLOUD seasonal deposition estimates are provided in Table 4-6. Again, a season was defined as
the period from June through September and at least three valid months were required to calculate
the seasonal deposition value. Except for a few instances, ion deposition was higher at Whiteface
Mountain. Total ion deposition estimates could only be compared among the three sites for 1997,
due to the limited scope of the MCLOUD runs. Monthly and seasonal totals of cloudwater
deposition estimates tended to be highest at Whiteface Mountain due to the higher wind speed and
LWC regime of that site. Ion deposition was also higher at Whiteface Mountain than at the southern
sites. Whitetop Mountain had somewhat higher ion deposition rates than Clingman's Dome.
4.2.5 Conclusions
The relationship between general model sensitivity and potential bias and uncertainty in CASTNet
cloudwater deposition estimates has been assessed for several key model parameters and input data
streams. Table 4-7 provides a summary of this assessment.
Uncertainty in estimates of cloudwater deposition is nearly directly proportional to uncertainty in
wind speed and LWC measurements, both of which are thought to be small for most site-year data
sets. Deposition estimates are directly proportional to cloud frequency, a parameter that is also well
known at each site for the majority of the months of the study. Whiteface Mountain is an exception
because scaling of wind speed, LWC, and cloud frequency from the summit observations to canopy
elevations is required. Therefore, selection of the "representative elevation" for deposition
calculations at Whiteface Mountain has a significant effect on how the results from that site will
compare to the other observation sites with forested summits.
The response of the cloudwater model appears to be the least sensitive to the least well-constrained
parameters such as canopy species composition, height, LAI, and vertical distribution of leaf area.
Model response is ± 12 percent or less for 10- to 40-percent changes in these parameters.
Estimates at all three sites are subject to a constant -8 percent bias resulting from numerical error
introduced into the models by the choice of a 1-m layer thickness. There is some potential
uncertainty (generally < ±10 percent in LWC, wind speed, and ion concentrations) introduced in a
small number of years from the rejection of observations in the data screening process. There is
p/castnet2/wa02/fnlsum/fnlipt_5.wpd
4-13
Harding ESE, Inc.
-------
MADPro Final Summary Report
some potential for an unknown but significant uncertainty in the estimates arising from limited
representation of meteorological and chemical conditions during cloud periods. Generally, less than
50 percent of observed cloud periods were represented in each site-year final data set.
4.3 Comparison of Results Between the CLOUD and MCLOUD Models
Monthly deposition velocities and cloudwater deposition estimates generated by the CLOUD model
were compared with those generated by MCLOUD. Estimates from all three CASTNet MADPro
sites were evaluated. Comparisons were made using monthly aggregation of the data.
Monthly average cloudwater and chemical deposition rates were highly correlated between the two
semi-independent estimates. Correlations, as measured by the r2 value, were generally greater
than 0.8. Despite the high correlations, differences between estimates were often as large as
50 percent. Most of the differences (including outlier months) appear to be the result of differences
in data screening or the aggregation of results to monthly averages. Other important sources of
differences between the estimates include the effect of different vertical distributions of leaf areas
used to parameterize the models as well as different approaches to scale cloud frequency from the
summit to the lower model canopy elevation at Whiteface Mountain.
4.3.1 Results and Discussion
Overall, there was good general agreement between the two semi-independent cloudwater deposition
estimates and underlying data aggregated at monthly intervals. After the removal of outliers, all
parameters were highly correlated between the two data sets with adjusted r2 values greater than 0.8.
Table 4-8 presents estimates of deposition velocities and cloudwater fluxes for the 13 months that
both models were run. The table shows significant differences between results from Whiteface
Mountain and results from Whitetop Mountain and Clingman's Dome. Deposition velocities from
the two models are comparable for Whiteface Mountain, i.e., within 15%. Differences in deposition
velocities from the two models for Whitetop Mountain and Clingman's Dome range from about 20%
to 50%. Cloudwater deposition estimates were approximately 35 to 55 percent higher for the
CLOUD runs for all three sites.
Model formulation and parameterizations, input data, and data screening and data aggregation
procedures all contribute to the differences between the two sets of model calculations. The
approximate 45-percent difference between the model results characterizes the uncertainty inherent
in modeling procedures (including data screening and data aggregation procedures). The CLOUD
and MCLOUD results have not been evaluated with field measurements. This type of evaluation
could help explain model uncertainties and improve overall model performance.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
4-14
Harding ESE, lac.
-------
MADPra Final Summary Report
Resistance type models that simulate dry deposition have been evaluated (e.g., Meyers et al., 1998;
Finkelstein etal., 2000) using field measurements. These evaluation studies have shown that the
resistance models have little bias for relatively flat, open settings and tend to underestimate
deposition velocities for forested, complex settings. Resistance models also have considerable scatter
based on modeling individual periods (i.e., '/2-hour flux measurements). The scatter shows
differences as high as a factor of 2.0 for individual simulations. However, the differences and scatter
decrease for aggregated periods - monthly, seasonal, and annual events. Assuming these
uncertainties apply to the CLOUD and MCLOUD models, typical uncertainties of at least 50% to
100% or more can be assigned to monthly and seasonal simulations.
4.4 Best Estimates of Seasonal Deposition Rates
Table 4-9 provides the best estimates of seasonal deposition rates for the three MADPro sites. The
CLOUD model is considered the primary model and was used to provide most of the information
shown in the table. However, because CLOUD model results were not available for 1999 and for
1994 for Whiteface Mountain, MCLOUD results were used for these periods. The MCLOUD results
were multiplied by 1.45 to account for the uncertainties discussed in the previous section.
The results show generally higher deposition rates at Whiteface Mountain. The 1997 estimates show
a south - north gradient in depositions. The 1999 data are different, however, in that Clingman's
Dome has the highest deposition values.
U.S. EPA Headquarters Library
Mai! cods 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
p/casmet2/wa02/fn lsum/fnlrpt_5. wpd
4-15
Harding ESE, inc.
-------
-------
MADPro Final Summarv Report
5.0 Total Deposition
Total deposition, as the term is used most of the time, is defined as dry and wet (or precipitation)
deposition summed together. In the eastern United States, frequent cloud exposure usually occurs at
elevations above 800 m (Mohnen et al., 1990). Land area above this elevation has been determined to
total approximately 25 million hectares. These locations, shown in Figure 5-1, cover a considerable
portion of the Appalachian chain, illustrating that the impact of cloud deposition reaches beyond
state boundaries and is not unique to a few isolated areas. Total deposition in such areas of frequent
cloud exposure then is the sum of cloud, wet, and dry deposition. The extra deposition of pollutants
via cloud exposure in these areas is often the cause of environmental damage, especially to red
spruce (Eager and Adams, 1992). An analysis of total deposition shows the contribution of each
fraction (i.e., cloud, wet, dry) to the total loading at a location. Total deposition from high-elevation
sites can be compared to locations where total loading results mostly from dry and wet depositions.
Such analyses may result in a better understanding of the processes involved that lead to ecological
damage.
5.1 Procedures
Cloud, wet, and dry deposition estimates were calculated on a monthly basis from 1994 through 1999
for the months of June through September at all three sites. Cloud deposition values at the three sites
for 1994 through 1998 were estimated by the CLOUD model and for 1999 by the MCLOUD model.
Total deposition amounts were calculated for the years 1994 through 1998.
The multi-layer model (MLM) was used to estimate dry depositions from filter pack concentration
data (Meyers etaL, 1998; Finkelstein etal. 2000) for Whitetop Mountain and Clingman's Dome. At
both sites, the filter packs were collocated with the automated cloud sampler. The dry deposition
values for Whiteface Mountain were obtained from NOAA Atmospheric Integrated Monitoring
Network (AIRMoN) and were also estimated by the MLM. The MLM calculations are considered
reasonable and representative for Clingman's Dome and Whitetop Mountain because onsite
meteorological measurements were used directly in the model. Although the MLM was developed
and evaluated using measurements from flat terrain settings, the model evaluation results are
considered roughly applicable to these two sites. The data from Meyers et al. (1998) show little
overall bias and up to 100% differences for individual V£-hr simulations. More recent data
(Finkelstein et al., 2000) suggest the MLM underestimates deposition velocities for SO2 for complex,
forested sites. The differences are expected to be lower for longer averaging times, i.e., monthly and
seasonal periods. Consequently, the uncertainty in the dry deposition estimates is approximately
100% or lower, and the MLM calculations probably underestimate the dry fluxes.
Wet deposition values for Whitetop Mountain were calculated by Harding ESE from data collected at
the site. Wet deposition data for Whiteface Mountain and Clingman's Dome were obtained from
p/castnet2/wa02/fnisum/fnlrpt_5.wpd
5-1
Harding ESE, Inc.
-------
MADPro Final Summary Report
NADP/NTN. Data from Mt. Mitchell were used for Clingman's Dome since the wet deposition data
collected at Clingman's Dome by NFS were not available.
The location of the filter pack and wet/dry bucket at Whiteface Mountain differed by approximately
750 m in elevation from the location of the cloud deposition estimates. The difference in elevation
also results in a difference in forest canopies and wind speed regimes which are very different from
the 1,350-m cloud-modeling site. Higher wind speeds and greater LAI's tend to increase dry
deposition of HNO3 at higher elevations on Whiteface (MiHer et al, 1993a). Also, the orographic
increase in precipitation over 750 m is significant (approximately 20 percent). It is estimated,
therefore, that wet and dry deposition values may be underestimated significantly in each case with
respect to cloud deposition values at the 1,350-m cloud-modeling site. CASTNet estimates (EPA,
1998) of dry deposition fluxes for two high elevation sites near Coweeta, NC show differences up to
a factor of 4.4 for weekly averages. Again, differences in seasonal averages are expected to be
lower. In short, dry and wet deposition calculations for Whiteface Mountain are considered
underestimates, maybe as high as 400% for dry deposition.
Monthly total deposition values were calculated by summing the monthly cloud, wet, and dry
deposition values for June through September. Tables 5-1 through 5-3 present the monthly cloud,
wet, and dry deposition values, when available or when data completeness permitted, for 1994
through-1999 for Whiteface Mountain, Whitetop Mountain, and Clingman's Dome. Monthly total
deposition amounts were calculated and presented only if all three fractions for a month were
available. As presented in these tables, data completeness for all three fractions for a month was not
high for Whitetop Mountain and Clingman's Dome. To determine the composition of total
deposition from its three constituents in a statistically meaningful way, the following procedures
were employed:
1. For Whiteface Mountain and Whitetop Mountain, the percent that each fraction
contributed to the total deposition was calculated for those months when data for all
three fractions were available. An overall mean percent value was then calculated
from the monthly percentages of each fraction. For Whiteface Mountain, overall
mean percentages were calculated for sulfur (S), nitrogen (N), and H+. Dry
deposition estimates for NHJ were not available from NOAA. For Whitetop
Mountain, overall mean percentages were calculated for S, N, NHJ, and H+.
2. The above procedure was not suitable for Clingman's Dome because there was only
1 month in 1998 when data from all three fractions were available. Therefore,
deposition values of each fraction for all months when data were available from
1994 through 1999 were averaged to yield a fractional mean. The overall fractional
means were then summed to yield an overall total deposition value. The percent
each fraction contributed to the total deposition was then calculated.
p/castnet2/wa02/fnlsum/fnlipt_5.wpd
5-2
Harding ESE, Inc.
-------
MADPro Final Summary Report
Procedure 2 was tested by calculating the overall percent composition for Whiteface Mountain and
Whitetop Mountain using this same procedure. The percent composition values produced by
procedures 1 and 2 for the two sites were almost identical for cloud deposition and only differed by 1
to 2 percent for the dry and wet fractions. It was deemed suitable, therefore, to use procedure 2 to
produce fractional mean percentages of S, N, NHJ, and H+ for Clingman's Dome. The percent
composition results determined by procedure 1 for Whitetop Mountain and Whiteface Mountain, and
by procedure 2 for Clingman's Dome are presented in Table 5-4.
Seasonal deposition values (June through September) were calculated by summing the monthly
values for each fraction when data for all 4 months for a given year were available. If data for only 3
of the 4 months were available for a given fraction, then the fractional value was estimated by scaling
up from the average of the three available monthly values. A seasonal fractional deposition amount
was not calculated if data were not available for at least 3 months in the season. The fractional
seasonal deposition values were then summed to yield a seasonal total deposition value. Seasonal
total deposition estimates for Whiteface Mountain and Whitetop Mountain are presented in Tables 5-
5 and 5-6, respectively. The lack of data completeness at Clingman's Dome did not allow for
calculation of seasonal deposition estimates for any ions.
5.2 Results and Discussion
Tables 5-1 through 5-4 show that clouds are, by far, the largest source for deposition of pollutants to
high-elevation ecosystems. Between 80 and 90 percent of S deposition occurs via cloud exposure at
all three sites as does 70 to 87 percent of the total H+ loading. Cloud deposition is also responsible
for 90 to 95 percent of NHJ deposition at the southern sites. Dry deposition is a very minor
contributor to the total S and NHJ loading, but contributes between 22 and 28 percent of N
deposition and approximately 15 to 16 percent of H* deposition at the southern sites. The percent
composition of total deposition at the MADPro sites is depicted in Figure 5-1.
It is important to reiterate here that the dry and wet deposition values for Whiteface Mountain are
probably underestimated due to the elevation difference in sampling locations. The uncertainties in
the mean estimates of dry deposition for Whitetop Mountain and Clingman's Dome are estimated to
be less than 100% for seasonal fluxes based on model evaluation studies. These calculations
probably underestimate fluxes.
Seasonal dry, wet, and total deposition estimates for the major ions for Whiteface Mountain and two
nearby CASTNet sites are presented in Table 5-5. The same information is presented in Table 5-6
for Whitetop Mountain and two nearby CASTNet sites. Total deposition values from the MADPro
sites are approximately 6 to 20 times greater for S, N, and NH} (NHJ results refer to Whitetop
Mountain only) when compared to the CASTNet sites. With one exception, Whiteface Mountain H*
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
5-3
Harding ESE. Inc.
-------
MADPro Final Summary Report
deposition values are 1.3 to 2.4 times greater than the CASTNet H+ depositions. Whitetop Mountain
H* deposition values for 1996 are 6.5 to 10 times greater than H* deposition amounts at the CASTNet
sites.
Dry deposition values at Whiteface Mountain fall within the range of dry deposition values for the
CASTNet sites for S, N, and H+. Wet deposition values for all three species are 1 to 3 times higher at
the CASTNet sites, except for 1998 when the Whiteface Mountain values were slightly higher for S
and nearly equal for N and H+.
Even though total seasonal deposition estimates could only be calculated for 1996 for Whitetop
Mountain, similar relationships can be ascertained for dry and wet deposition estimates when
compared to the closest CASTNet sites.
It is evident from the data presented in Tables 5-5 and 5-6 that the total deposition at MADPro sites
is much greater due to cloud exposure than total deposition at lower elevation sites that experience
mostly dry and wet depositions.
p/castnet2/wa02/fnlsum/fnlrpl_5.wpd
5-4
Harding ESE, Inc.
-------
MADPm Final Summary Report
6.0 Comparison with Other Networks
6.1 Concentration of Pollutant Ions in Clouds
Two major networks for collecting long-term data on cloud chemical composition and concentration
have been operated in eastern North America: the CHEF project, which began with field
measurements in southern Quebec in 1985; and MCCP, which began in the eastern mountains of the
United States in 1986. Protocols for the two programs were closely linked, with the objective of
having comparable databases covering eastern North America. Results from CHEF have been
published by Schemenauer (1986), Schemenauer and Winston (1988), and Schemenauer et al.
(1995). Results from MCCP projects have been published by Mohnen and Kadiecek (1989) and
Saxena et al. (1989). Other recent reviews of the results from MCCP have been prepared by Vong
etal. (1991), Li and Aneja (1992), and Mohnen and Vong (1993).
For the CHEF sample set as a whole (including the main sites at Roundtop, Mont Tremblant, and
Montmorency, and the subsites at Mont Tremblant mid-level and Roundtop summit), the mean
volume-weighted fog composition for all seasons (n = 599) is as follows (in ueq/L):
sof
NO;
cr
NH;
Ca2+
Mg2*
Na+
K+
H*
279.8
163.3
14.7
210.3
67.9
14.2
15.8
4.5
162.2
The total anion concentration is 457.8 jueq/L and the total cation concentration is 474.9 ueq/L. The
ion balance was calculated to be 1.8 percent, indicating that it is unlikely that other stable ions are
present in significant concentrations.
The CHEF cloud chemistry data are of high quality and are comparable to MCCP data presented in
Table 6-1 for 1986 through 1988. The results from MADPro for the sample set as a whole (warm
season June through September) during the 1994 through 1999 time period are presented in
Table 3-5-
The comparison of regionally and temporally averaged mean ion concentrations indicate that the
concentration of major ions in clouds for northern versus southern sites in eastern North America
p/castnet2Ava02/fnlsum/fnlrpt_5.wpd
6-1
Harding ESE, Inc.
-------
MADPro Final Summary Report
(North Carolina to Quebec and altitudes of 3,000 to 6,000 feet) fall within the minimum and
maximum ranges listed below:
sof
NOj
tr
NH;
Northern Sites
280-300 ueq/L
120-170 ueq/L
160-230 ueq/L
160-220 ueq/L
Southern Sites
380-450 ueq/L
170-1 90 ueq/L
290-370 ueq/L
210-250 ueq/L
There is a distinct north-south gradient in cloud ion concentrations that has already been detected in
all previous studies. Based on the largest combined data set on cloud chemistry now available, this
north-south gradient has been well established.
The MADPro project means are higher for all four major ions for Whiteface Mountain and Whitetop
Mountain in comparison with MGCP project means for these same sites. The MADPro Clingman's
Dome means, when compared to MCCP's Mt. Mitchell means (the closest MCCP site), were higher
for the nitrogen compounds but lower for H+ and SOj'. Table 6-2 presents the relative percent
differences between the MCCP and MADPro overall project means.
Significant variations occur during and between individual cloud events both with altitude (height
above cloud base), cloud LWC, and origin of air masses (proximity of emission sources) as indicated
by the wide spread of maximum and minimum concentrations observed at the monitoring sites.
AU CHEF, MCCP and MADPro concentrations that are discussed in this section are volume-
weighted means, but have not been normalized for LWC.
6.1.1 Comparison of Results with European Studies
A long-term cloud chemistry monitoring program was operated from 1991 through 1993 at Mt.
Brocken, Germany [Moller et al. (1994, 1996) and Acker et al. (1999)]. For the manual cloudwater
collection, the same collector and SOP were used as in MCCP. Therefore, the hourly samples
collected at this mountain site are directly comparable to MCCP and MADPro cloud results. LWC
was measured with the Gerber PVM-100. The overall results of cloud chemical composition are
presented in Table 6-3.
Additional fog collection was performed at other European sites, although for a shorter duration.
These investigations are summarized in Table 6-4.
In general, the concentrations of major and minor ions reported by European investigators are higher
than those measured in eastern North American networks. This is particularly true for NOj and NHJ,
which is indicative of the higher atmospheric pollution burden over Europe and the closer proximity
p/castnet2/wa02/fnlsum/fnlrpt_5. wpd
6-2
Harding ESE, Inc.
-------
MADPrn Final Summary Report
of major emission sources to the European measurement sites. Most sites also show lower LWC due
to the lower elevation of the sampling sites where ground-level fog occurs frequently.
6.2 Deposition
Assessment of atmospheric pollutant transfer is essential for informed ecosystem management of
mountain forests to preserve their biological, recreational, and aesthetic value.
Cloud droplet capture by the forest canopy is the principal transfer mechanism. Its magnitude can be
obtained from cloud deposition models, cloud deposition measurements by eddy correlation, or
estimated from the hydrological mass balance using direct measurement of evapotranspiration by the
eddy covariance-energy-budget (ECEB) method [Herckes et al. (1999), Kalina et al. (1998), and
Wrzesinsky and Klemm (2000)]. The latter two approaches are used mostly for short term field
research projects; results from such studies will not be discussed here.
All three approaches provide as output the flux rate of liquid cloudwater [millimeters per hour
(mm/hr)] and, after adjusting for the time the forest is immersed in clouds (cloud frequency), the
monthly or annual total amount of cloudwater (millimeters of water). Incorporating the cloudwater
chemical concentration subsequently yields the chemical deposition kg/ha/mo or kilograms per
hectare per year (kg/ha/yr).
The best documented estimates of cloudwater chemical deposition prior to MADPro are from MCCP
| Mohnen (1988) and Mohnen et al. (1990)]. The MCCP collected hourly cloudwater data (solute
and LWC concentrations) and meteorological data (winds, cloud frequency, relative humidity, and
solar radiation) for about 5 months each summer (May through September) over 3 years (1986
through 1988). The MADPro mean calculated deposition values (CLOUD model) for water and four
major ions are compared to MCCP values as well as to values ascertained by Lovett et al. (1982) and
Miller et al. (1993a) in TabJe 6-5. In general, the values from the MADPro Clingman's Dome and
Whitetop Mountain sites fall within the range of those measured previously, while those from the
MADPro Whiteface Mountain site are slightly above the range. The calculated values for the
MADPro Whitetop Mountain site are similar to those calculated for the Whitetop Mountain (B) site
in the MCCP project (Mohnen et al. 1990), which had a similar canopy structure to the stand used in
the CLOUD model. The calculated values for the MADPro WFML site are higher than those
modeled for the MCCP Whiteface Mountain site, although it is not clear from the MCCP report what
sort of canopy was modeled for this site or at what elevation. The MCCP report (p. 4-61, Mohnen
et al., 1990) suggests that the low deposition estimated for the Whiteface site resulted from a low
canopy surface area used for the calculations.
More relevant, perhaps, are the results of Miller et al. (1993a) calculations for the 1,350-m elevation
on Whiteface Mountain, which is listed as Whiteface Mountain (M) in Table 6-5. The MADPro
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
6-3
Harding ESE. Inc.
-------
MADPra Final Summary Report
water deposition rates are significantly higher than those of Miller et al, but Miller et al. were
simulating the entire year, including the winter season which they suggest has deposition rates of less
than half those of the warm season. They report a mean warm season deposition velocity
(45.3 cm/sec) that was actually higher than the MADPro result of 37.2 cm/sec. Miller et al. also
reported lower chemical concentrations in cloudwater (132 ueq/L of SOf versus 347 ueq/L of SO^ ).
This difference may be due to temporal variation, different sampling locations, or different data
treatment (e.g., data screening).
In other studies, the cloud deposition model-spruce (CDM-S) and the cloud deposition model-
deciduous (CDM-D) (Sigmon et al., 1989; Mueller et al., 1991) were used to predict deposition to
the forests on six eastern United States mountains (Table 6-6). The other large network reporting
data for eastern North America is the CHEF project. Deposition estimates from the CHEF project
(1985 through 1991) are not available because cloud sampling time extended over several hours and
no concurrent LWC measurements were made.
MCCP researchers have estimated that cloudwater deposition provides a substantial fraction of the
total chemical deposition to the forests that they studied in the eastern United States. Lindberg and
Johnson (1989) estimated that cloudwater contributes approximately 25 to 50 percent of total (rain +
cloud + dry) SOf, N, and H+ deposition at sites on Whiteface Mountain, New York, and in the Great
Smoky Mountains National Park, North Carolina. MCCP results indicate that cloud SOf, NOj, H+,
and NH} deposition exceeded wet deposition via precipitation for three sites located above 1,400 m.
Two sites located near 1,000-m elevation received cloudwater chemical inputs that were at least
50 percent of precipitation chemical deposition.
From a decade of observations (1986 through 1996) [using data from MCCP, MADPro CLOUD, and
Electric Power Research Institute-Integrated Forest Study (EPRI-IFS)], Miller and Friedland (1999)
estimated the atmospheric deposition at the Whiteface Mountain 1,050-m site for total nitrogen to
average 17.2 kg/ha/yr (37 percent deposited as NHJ and 63 percent as NO^) and sulfur deposition to
average 18.3 kg/ha/yr. Precipitation and cloudwater deposition contributed nearly equally to total
sulfate deposition at the lower (1,050 m) Whiteface Mountain site (Table 6-7).
In a separate study, Miller et al. (1993a) determined that cloudwater contributed 2, 39, and 73
percent of total annual SOf deposition at 600-, 1,025-, and 1,350-m elevation, respectively, on
Whiteface Mountain. Cloudwater contributed 1, 28, and 59 percent of total annual NOj deposition at
600-, 1,025-, and 1,350-m elevation, respectively, in that study. This shift in the fraction of
deposition contributed by cloudwater deposition at high elevations is confirmed by the results from
MADPro (Table 5-4). The MADPro results from the 1,350-m elevation at Whiteface Mountain show
that almost 91 percent and 84 percent of SOf and NO; deposition, respectively, are from cloudwater
interception. Again, the dry and wet deposition estimates from Whiteface Mountain are
underestimated, at the very least, by a probable factor of 2 due to the elevation difference in sampling
p/castnet2/wa02/fnlsum/fnlrpc_5.wpd
Harding ESE, Inc.
-------
MADPro Final Summary Repon
locations. Scaling the dry and wet deposition estimates to the 1,350-m cloud-modeling site would
reduce the percent composition of total deposition from cloud deposition to approximately 84 percent
for SOfand 71 percent for NOj. This provides a better agreement with the results of Miller et al.
(1993a).
The two southern sites also receive most of their deposition from clouds. Dry deposition has a larger
influence in NQ~3 deposition at the southern sites than at Whiteface Mountain. However, more than
65 percent of NO^ deposition at the southern sites is still received from clouds. The influence of dry
deposition may be even greater at the southern sites due to the limitations of the MLM model in
estimating deposition velocities and fluxes over complex and high-elevation terrain.
The variation in cloudwater interception with cloud frequency thus produces an elevational gradient
in total chemical deposition (Miller et al., 1993a). A summary by Lovett and Kinsman (1990)
suggests that eastern United States sites at elevations below 1,000 m receive less than 20 percent of
total SOf deposition via cloudwater, but sites above 1,500 m receive 45 to 80 percent of total SOf
deposition via cloudwater. These estimates are in agreement with independent summaries by
NAPAP (Hicks et al., 1991; Sisterson et al., 1991) and with the MADPro results presented here.
6.2.1 Comparison of Results with European Studies
European studies also suggest substantial chemical fluxes from cloud droplet deposition. In
Germany, Kroll and Winkler (1988) estimated that intercepted cloudwater contributed 25 to
150 percent of the water amount and 1 to 4 times the chemical deposition that occurs via rain for two
mountain sites at elevations of 840 and 1,440 m (cloud frequencies were 20 percent and 30 percent,
respectively). Harvey and McArthur (1989), using a gradient technique, estimated that fog droplet
deposition accounted for about 2 to 12 percent of total chemical deposition to moors in central
England. At another site in Great Britain with a cloud frequency of 23 percent, Fowler et al. (1990)
calculated that droplet deposition to moors would increase wet deposition of SOf, NO^, H+, and NHJ
by 12 percent, but if the site were forested, the increase might be 44 percent (based on differences in
aerodynamic roughness).
All investigations to date demonstrate the significant variability of the deposition rate with location
and time of year, which depends on several spatial and temporal factors, including:
• Seasonal and airmass (trajectories) dependence of pollutant ion-concentration in
cloud droplets;
• LWC variations with type of cloud, height above cloud base, and cloud droplet size
distribution (which in turn influences droplet interception);
• Wind velocity at any location and its variation within the forest canopy; and
• Type of vegetation and associated LAI at any location.
p/castnet2/wa02/fnlsum/fnlipt_5. wpd
6-5
Harding ESE, Inc.
-------
MADPro Final Summary Report
Consequently, cloud droplet capture by the forest canopy varies substantially from year to year in
response to changing meteorological conditions. These fluctuations impart significant variances to
total atmospheric deposition rates of all ions. Dry deposition is not an efficient delivery process at
high-elevation montane forests.
The observed deposition values for sulfur and nitrogen at high-elevation forest ecosystems represent
some of the highest sustained air pollution loadings reported for eastern United States forests.
The lack of a temporal trend in MADPro results for cloudwater ion deposition indicates that sensitive
mountain forest ecosystems may continue to experience high levels of sulfur and nitrogen input even
as lower elevation forests of the region may experience steady reductions in sulfur and nitrogen
deposition as a result of declining emissions.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
6-6
Harding ESE, Inc.
-------
MADPrn Final Summary Report
7.0 Conclusions and Recommendations
Since the onset of its operation in 1994, MADPro has produced a 6-year data set that is comparable
to data produced by past networks. The field SOPs, QA program, and laboratory analytical methods
for MADPro were designed, when possible, with this objective of comparability in mind.
Differences in equipment with respect to past networks consisted mainly of the use of an automated
cloudwater sampler and a continuous LWC measurement system (PVM-100). These changes were
implemented as improvements identified as necessary from past experience. Two of the three
permanent MADPro sites were identical to sites operated in past studies and all three sites collected
samples during the warm season, which also mirrored past networks. Over 5,300 valid hourly
samples were collected over the 6-year period of operation.
Therefore, MADPro has successfully achieved its two main objectives of development and
implementation of cloudwater measurement systems to be used in a network monitoring environment
and to update the cloudwater and deposition data collected in the Appalachian Mountains during
NAPAP. MADPro has produced the most substantial amount of cloudwater data (based on the
number of samples collected from 1994-1999) in the world. This high-quality data set enables
representative conclusions to be drawn.
7.1 Cloudwater Concentrations
Statistically significant decreasing temporal trends were exhibited at Whiteface Mountain for NHJ
and SOf for both the normalized and non-normalized concentrations. Whitetop Mountain exhibited a
statistically significant, but increasing, temporal trend for normalized SOf concentrations, even
though ambient SO2 levels have declined (Lavery et al, 2000). Since MADPro was not designed as a
trends/process study, no definitive conclusions are drawn as to the lack of trends for some or all of
the analytes at both Whitetop and Whiteface Mountains, or the SO*' trend in the opposite direction at
Whitetop Mountain. It can be speculated, however, that the lack of trends, weak trends, or trends in
the opposite direction, may be due to chemical non-linearity between emissions and cloudwater
concentrations. This non-linearity is probably due to oxidant limitation in clouds and/or year-to-year
variability in air mass trajectories, LWC, and local and regional meteorology (Miller and Friedland,
1999).
The CHngman's Dome site exhibits statistically significant temporal trends in concentrations of all
the major ions for both normalized and non-normalized results, but these trends are also in the
opposite direction (increasing) from what is expected given the SO, emission reductions. The trends
at this site may, in part, be due to the increase in regional summertime SOf concentrations as
documented by Mueller (2000). More research and analysis, especially that which involves back
trajectories, needs to be conducted to fully understand the pattern at Clingman's Dome.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
7-1
Harding ESE, Inc.
-------
MADPro Final Summary Report
For the major ions, Whiteface Mountain exhibited the lowest mean and median normalized values
and Clingman's Dome exhibited the highest mean and median concentrations. This is indicative of
the north-south gradient noted during the MCCP (Anderson et al., 1999; EPA, 1990). This gradient
is probably an effect of the different meteorological conditions experienced at northern versus
southern sites as well as a difference in the back trajectories of air masses reaching the sites.
Overall, 6-year mean concentrations at the three sites ranged from 289 to 456 ueq/L for SOf, 125 to
182 neq/L for NOj, 168 to 237 ueq/L for NHJ, and 225 to 359 ueq/L for H+.
7.2 Cloudwater Depositions
MADPro results have clearly demonstrated the significance of the additional deposition of SOf and
NOj compounds to high-elevation forests subjected to cloud exposure. These sites experience
additional loading on the order of 6 to 20 times greater for SOj and NOj compared to lower
elevation CASTNet sites. Approximately 80 to 90 percent of this extra loading is from cloud
deposition. The dry and wet deposition fraction estimates from the MADPro sites are relatively
comparable to dry and wet deposition amounts experienced at the CASTNet sites. The magnitude of
deposition and/or exposure at high elevation sites is substantial enough to cause ecological damage
(Eager and Adams, 1992).
A clear north-south pattern in total deposition amounts is not evident from the data presented here.
However, data completeness for deposition values for the southern sites must be improved before
further statements can be made. A review of the available monthly cloud and total deposition values
from 1994 through 1999 does show that, unlike cloudwater concentrations, Whiteface Mountain
experiences as much, if not greater, amounts of cloud and total deposition as the southern sites. This
increase in deposition rates at Whiteface Mountain with respect to cloudwater concentrations is due
in large part to the higher wind speeds and LWC experienced at this site.
Although MADPro was not designed for ecological studies, results from this program will aid
ecologists in damage assessment of high-elevation ecosystems through the provision of a data set of
known uncertainty.
7.3 Recommendations
1. To determine a trend in cloudwater concentrations and depositions, a longer
measurement period than 6 years is necessary, as well as application of more
sophisticated statistical procedures on the existing data set.
2. More back trajectory analyses are needed to elucidate any trends corresponding to
upwind source regions.
p/cascnet2/wa02/fnlsum/fnlrpt_5.wpd 7-2 Harding ESE, Inc.
-------
MADPro Final Summary Report
3. Cloud deposition model interconiparisons and evaluations should be conducted via
field studies of throughfall and eddy correlation.
4. Collection of spatial distribution of wind speed, cloud LWC, and cloud immersion
time information for expansion of cloud deposition modeling to a regional scale.
5. To further increase regional representativeness of the cloud deposition values,
canopy structure must be ascertained accurately and very specifically throughout the
eastern United States frequently impacted by clouds. Use of satellite sensors for this
purpose is recommended. This information will enable a more refined estimate of
deposition values than the homogenized deposition estimates provided in this report.
6. Collocated field testing of the Valente method and the PVM-100 is necessary to
determine whether MCCP LWC data need adjustment. If MCCP LWC data require
adjustment, then the MCCP deposition estimates need to be recalculated using the
adjusted LWC to allow better comparison with MADPro deposition estimates.
7. Cloudwater collection efforts must be collocated with ecological studies to further
elucidate the processes involved at these locations.
8. To establish an east-west gradient, additional sites in the Green or White Mountains
of New England are recommended. An additional site in West Virginia is also
recommended due to its proximity and location with respect to Ohio River pollutant
sources.
10.
Extend the program from seasonal to year-round for estimation of winter time
deposition values and comparison to other networks. The program should include
other impacted mountainous regions in the western United States (e.g., the Cascades
downwind of Seattle, the Sierra Nevadas downwind of the San Joaquin Valley, the
San Gabriels and/or San Bernadinos downwind of Los Angeles, etc.).
Continous SO2 measurements and PM2 5 sampling should be added to the suite of
measurements to better address questions on SOf trends or lack thereof.
p/castnet2/wa02/fnlsum/fnlrpt_5. wpd
7-3
Harding ESE, Inc.
-------
-------
MADPro Final Summary Report
References
Acker, K., Mdller, D., Wieprecht, W., Auel, R., and Kalass, D. 1999. The Mt. Brocken Air and
Cloud Water Chemistry Program and the QA/QC of Its Data. In: Workshop on
QualityAssurance of EMEP Measurements and Data Reporting. EMEP/CCC-Report 1/99,
pp. 161-165.
Anderson, J.B., Baumgardner, R.E., Mohnen, V.A., and Bowser, J.J. 1999. Cloud Chemistry in the
Eastern United States, as Sampled from Three High-Elevation Sites along the Appalachian
Mountains. Atmospheric Environment, 33:5105-5114.
Aneja, V.P. and Kim, D.S. 1993. Chemical Dynamics of Clouds at Mt. Mitchell, North Carolina..
Air & Waste, 43:1074-1083.
Anlauf, K.G., Wiebe, H.A., and Fellin, P. 1986. Characterization of Several Integrative Sampling
Methods for Nitric Acid, Sulphur Dioxide and Atmospheric Particles. JAPCA, 36:715-723.
Baumgardner, R.E., Kronmiller, K.G., Anderson, J.B., Bowser, J.J., and Edgerton, E.S. 1997.
Development of an Automated Cloudwater Collection System for Use in Atmospheric
Monitoring Networks. Atmospheric Environment, 31(13):2003-2010.
Best, A.C. 1951. Drop-Size Distribution in Cloud and Fog. Quarterly Journal of the Royal
Meteorological Society, 77:418-426.
Butler, T.J., Likens, G.E., and Sounder, B.J.B. (in press). Regional-Scale Impacts of Phase I of the
Clean Air Act Amendments: The Relation Between Emissions and Concentrations, Both Wet
and Dry. Atmospheric Environment.
Collett, Jr., J., Oberholzer, B., and Staehelin, J. 1993. Cloud Chemistry at Mt. Rigi, Switzerland:
Dependence on Drop Size and Relationship to Precipitation Chemistry. Atmos. Environ.,
27A:33-42.
Dasch, J.M. 1988. Hydrological and Chemical Inputs to Fir Trees from Rain and Clouds During a
1-month Study at Clingman's Peak, NC. Atmospheric Environment, 22:2255-2262.
Dollard, G.J. and Unsworth, M.H. 1983. Field Measurements of Turbulent Fluxes of Wind-Driven
Fog Drops to a Grass Surface. Atmospheric Environment, 17:775-780.
Dutkiewicz, V.A., Burkhard, E.G., and Husain, L. 1995. Availability of H2O2 for Oxidation of SO2
in Clouds in the Northeastern United States. Atmos. Environ., 29(22):3281-3292.
p/castnet2/wa02/fnlsum/fnirpt_5. wpd
REF-1
Harding ESE. Inc.
-------
MADPra Final Summary Report
Eager, C. and Adams, M.B., eds. 1992. Ecology and Decline of Red Spruce in the Eastern United
States. Springer Ecological Series, Vol. 96. pp.395-411. Springer. Berlin.
Falconer, R.E. and Falconer, P.D. 1980. Determination of Cloudwater Acidity at a Mountain
Observatory in the Adirondack Mountains of New York State. JCR, 85(C):7465-7470.
Finkelstein, P.L., Ellestad, T.G., Clarke, J.F., Meyers, T.P., Schwede, D.B., Herbert, E.G., and Neal,
J.A. 2000. Ozone and Sulfur Dioxide Dry Deposition to Forests: Observations and Model
Evaluation. Atmos. Environ., 105:D12:15,365-15,377.
Fowler, D., Morse, A.P., Gallagher, M.W., and Choularton, T.W. 1990. Measurements of
Cloudwater Deposition on Vegetation Using a Lysimeter and Flux Gradient Technique.
Tellus, 426:284-293.
Gerber, H., Arends, B.G., and Ackerman, A.S. 1993. New Microphysics Sensor for Aircraft Use.
Atmospheric Research, 00:1-18.
Gerber, H. 1984. Liquid Water Content of Fogs and Hazes from Visible Light Scattering. Journal.
of Climatology and Applied Meteorology, 23:1247-1252.
Gunz, D.W. and Hoffman, M.R. (1990). Atmospheric Chemistry of Peroxides: A Review. Atmos.
Environ., 24A: 1601-1633.
Harding ESE, Inc. (Harding ESE). 1999. Clean Air Status and Trends Network (CASTNet) Quality
Assurance Project Plan (Draft). Prepared for the U.S. Environmental Protection Agency
(EPA). Contract No. 68-D-98-112. Gainesville, FL.
Harding ESE, Inc. (Harding ESE). 1997. Clean Air Status and Trends Network (CASTNet)
Mountain Acid Deposition Program Data Report - 1996 Monitoring Season. Prepared for
the U.S. Environmental Protection Agency (EPA), OAQPS, Research Triangle Park, NC
27711. Contract No. 68-D2-0134, Work Assignment No. 2. Gainesville, FL.
Harvey, M.J. and McArthur, A.J. 1989. Pollution Transfer to Moor by Occult Deposition.
Atmospheric Environment, 23(5): 1073-82.
Heffler, J.L. 1983. Branching Atmospheric Trajectory (BAT) Model. NOAA Technical
Memorandum, ERL ARL-121. Air Resources Laboratory, Rockville, MD.
p/cascnet2/wa02/fnlsum/fnlrpt_5.wpd
REF-2
Harding ESE, Inc.
-------
MADPm Final Summary Report
Herckes, P., Wortbam, H., Wendling, R., and Mirabel, P.H. 1999. Cloudwater Studies at an High
Elevation Site in the Voges Mountains (France), (poster presentation). Sixth Scientific
Conference of the International Global Atmospheric Chemistry Organization, September 13-
17, 1999. Bologna, Italy.
Herterich, R. 1987. Eignung und Anwendung von Rotating Arm Collectoren zur Bestimmung von
Nebeleigenschaften Fliissigwasser-gehalt, ionische Spurenstoffe, Wasserstoffperoxid.
Diploma thesis. 85 p. Chair of Hydrology, University of Bayreuth. Bayreuth, Germany.
Hicks, B.B., Draxler, R. R., Albritton, D. L., Fehsenfeld, F. C., Davidson, C. I. 1991. Atmospheric
Processes Research and Process Model Development. State of Science and Technology
Report Number 2. National Acid Precipitation Assessment Program, Washington, DC.
Joslin, J.D., Mueller, S.F., and Wolfe, M.H. 1990. Test of Models of Cloudwater Deposition to
Forest Canopies Using Artificial and Living Forest Canopies and Living Collectors.
Atmospheric Environment, 24A:3007-3019.
Kalina, M.F., Zambo, E., and Puxbaum, H. 1998. Assessment of Wet, Dry and Occult Deposition of
Sulfur and Nitrogen at an Alpine Site. Environ. Sci. Pollut. Res., 1:53-58.
Kroll, G. and Winkler, P. 1988. Estimation of Wet Deposition via Fog. In: Environmental
Meteorology, pp. 227-36. K. Grefen and J. Lobel, eds. Kluwer. Dordrecht, Netherlands.
Lavery, T.F., Costakis, C., and Rogers, C.M. 2000. Trends in Atmospheric Deposition of Acid
Gases and Particles. Paper No. 1017. Annual Meeting of the Air and Waste Management
Association. June 2000.
Li, Z. and Aneja, V.P. 1992. Regional Analysis of Cloud Chemistry at High Elevations in the
Eastern United States. Atmospheric Environment, 26A:2001-2017.
Lindberg, S.E. 1992. Atmospheric Deposition and Canopy Interactions of Sulfur. In: Atmospheric
Deposition and Forest Nutrient Cycling, pp.74-89. S. E. Lindberg and D. W. Johnson, eds.
Springer-Verlag. New York.
Lindberg, S.E. and Johnson, J.W. 1989. 1988 Annual Report to the Integrated Forest Study. ORNL/
TM-11121. Report to EPRI (RP2621), Oak Ridge National Laboratory, Oak Ridge, TN.
p/castnet2/wa02/fnlsu m/fnlrpt_5. wpd
REF-3
Harding ESE, Inc.
-------
MADPro Final Summary Report
Lindberg, S.E., Silsbee, D., Schaefer, D.A., Owens, J.G., and Petty, W. 1988. A Comparison of
Atmospheric Exposure Conditions at High- and Low- Elevation Forests in the Southern
Appalachian Mountain Range. In: Acid Deposition at High Elevation Sites, pp. 321-44.
Unsworth, M.H. and Fowler, D., eds. NATO ASI Series C. Vol. 252. Kluwer Academic
Publishers, Boston.
Lovett, G.M. 2000. Modeling Cloudwater Deposition to the Sites of the CASTNet Cloud Network.
Prepared for Harding ESE, Inc. (Harding ESE). Gainesville, FL.
Lovett, G. M., and Kinsman, J. D. 1990. Atmospheric Pollutant Deposition to High-Elevation
Ecosystems. Atmospheric Environment, 24A:2767-2786.
Lovett, G.M. and Reiners,W.A. 1986. Canopy Structure and Cloudwater Deposition in Subalpine
Coniferous Forests. Tellus, 38B: 319-327.
Lovett, G.M. 1984. Rates and Mechanisms of Cloudwater Deposition to a Subalpine Balsam Fir
Forest. Atmospheric Environment. 18:361-371.
Lovett, G. M., Reiners, W.A., and Olson, R. K. 1982. Cloud Droplet Deposition in Subalpine
Balsam Fir Forests: Hydrological and Chemical Inputs. Science, 218:1303-1304.
McLaughlin, S.B., Anderson, C.P., Hanson, P.J., Tjoelker, M.G., and Roy, W.K. 1991. Increased
Dark Respiration and Calcium Deficiency of Red Spruce in Relation to Acidic Deposition at
High Elevation Southern Appalachian Mountain Sites. Canadian Journal of Forest
Research, 21:1234-1244.
McLaughlin, S.B., Anderson, C.P., Edwards, N.T., Roy, W.K., and Layton, P.A. 1990. Seasonal
Patterns of Photosynthesis and Respiration of Red Spruce Saplings from Two Elevations in
Declining Southern Appalachian Stands. Canadian Journal of Forest Research, 20:485-495.
Meyers, T.P., Finkelstein, P., Clarke, J., Ellestad, T.G., and Sims, P.F. (1998). A Multilayer Model
for Inferring Dry Deposition Using Standard Meteorological Measurements. J. Geophys.
Res., 103:22,645-22,661.
Miller, E.K. and Friedland, AJ. 1999. Local Climate Influences on Precipitation, Cloudwater, and
Dry Deposition to an Adirondack Subalpine Forest: Insights from Observations 1986-1996.
Journal of Environmental Quality, 28:270-277.
p/castnet2/wa02/fnisum/fnlrpt_5.wpd
REF-4
Harding ESE. Inc.
-------
MADPro Final Summary Report
Miller, E.K., Friedland, A.J., Arons, E.A., Mohnen, V.A., Battles, J.J., Panek, J.A., Kadlecek, J., and
Johnson, A.H. 1993a. Atmospheric Deposition Along an Elevational Gradient at Whiteface
Mountain, New York, USA. Atmospheric Environment, 27A:2121-2136.
Miller, E.K., Panek, J.A., Friedland, A.J., Kadlecek, J., and Mohnen, V.A. 1993b. Atmospheric
Deposition to a High-Elevation Forest at Whiteface Mountain, New York, USA. Tellus,
45B:209-227.
Mohnen, V.A., and Vong, R.J. 1993. A Climatology of Cloud Chemistry for the Eastern United
States Derived from the Mountain Cloud Chemistry Project. Environ. Rev., 1 -.251-212.
Mohnen, V.A., Aneja, V., Bailey, B., Cowling, E., Goltz, S.M., Healey, J., Kadlecek, J.A.,
Meagher, J., Mueller, S.M., and Sigmon, J.T. 1990. An Assessment of Atmospheric
Exposure and Deposition to High-Elevation Forests in the Eastern United States. Report
EPA/600/3-90/058 Edition. U.S. Environmental Protection Agency (EPA), Office of
Research and Development, Washington, DC.
Mohnen, V.A. and Kadlecek, J.A. 1989. Cloud Chemistry Research at Whiteface Mountain. Tellus,
41B:79-91.
Mohnen, V.A. 1988. Mountain Cloud Chemistry Project: Wet, Dry and Cloud-water Deposition.
EPA600/3-89-003, U.S. Environmental Protection Agency (EPA),OAQPS, Research
Triangle Park, NC.
MoHer, D., Acker, K., and Wieprecht, W. 1994. Cloud Chemistry at Mt. Brocken/Harz. In:
Proceedings of the 6th European Symposium on Physico-Chemical Behavior of Air
Pollutants, pp. 968-974. Rep. EUR 15609/2 EN. Brussels.
Moller, D., Acker, K. and Wieprecht, W. 1996. A Relationship Between Liquid Water Content and
Chemical Composition in Clouds. Atmos. Res., 41:321-335
Mueller, S.F. 2000. Recent Air Quality Trends in the Great Smoky Mountains, llth Joint
Conference on the Applications of Air Pollution Meteorology. Co-sponsored by American
Meteorological Society and Air, Waste, and Management Association.
Mueller, S.F. 1991. Estimating Cloud water Deposition to Subalpine Spruce-Fir Forests -1.
Modifications to an Existing Model. Atmospheric Environment, 25A:1093-1104.
Mueller, S.F., Joslin, J.D., and Wolfe, M.H. 1991. Estimating Cloudwater Deposition to Subalpine
Spruce-Fir Forests - II. Model Testing. Atmospheric Environment, 25 A: 1105-1122.
p/castnei2/wa02/fnlsum/fnlrpt_5. wpd
REF-5
Harding ESE. Inc.
-------
MADPro Final Summary Report
Mueller, S.F. and Weatherford, F.P. 1988. Chemical Deposition to a High Elevation Red Spruce
Forest. Water, Air and Soil Pollution, 38:345-363.
Pahl, S, and Winkler, P. 1995. Hdhenabhangigkeit der Spurenstoffdeposition durch Wolken auf
Walder. Abschlussbericht BMFT Nr. 07EU726A8.
Saxena, V.K. and Lin, N.H. 1990. Cloud Chemistry Measurements and Estimates of Acidic
Deposition on an Above Cloudbase Coniferous Forest. Atmospheric Environment, 24A:329-
352.
Saxena, V.K., Stogner, R.E., Hendler, A.H., DeFelice, T.P.,Yeh, R.J.Y., and Lin, N.H. 1989.
Monitoring the Chemical Climate of the Mt. Mitchell State Park for Evaluation of its Impact
on Forest Decline. Tellus, 413:92-109.
Schemenauer, R.S., Banic, C.M., and Urquizo, N. 1995. High Elevation Fog and Precipitation
Chemistry in Southern Quebec, Canada. Atmospheric Environment, 29: 2235-2252.
Schemenauer, R.S. and Winston, C. 1988. The 1986 Chemistry of High Elevation Fog (CHEF)
Project, pp. 1-16. 81st Annual Meeting of the Air Pollution Control Association. Dallas,
TX. 88-129.6.
Schemenauer, R.S. 1986. Acidic Deposition to Forests: the Chemistry of High Elevation Fog
(CHEF) Project. Atmosphere-Ocean, 24:303-328.
Scherbatskoy, T., and Klein, R.M. 1983. Response of Red Spruce and Birch Foliage to Leaching by
Acidic Mists. Journal of Environmental Quality, 12:189-195.
Schrimpff, E., Klemm, O., Eiden, R., Frevert, T., and Herrmann, R. 1984. Anwendung Eines
Grunow-nebelfangers Zur Bestimmung von Schadstoffgehalten in Nebelniederschlagen.
Staub-Reinh. Luff, 44:72-75.
Seinfeld, J.H. 1986. Atmospheric Chemistry and Physics of Air Pollution. Wiley, New York.
Shuttleworth, W.J. 1977. The Exchange of Wind-Driven Fog and Mist Between Vegetation and the
Atmosphere. Boundary-Layer Meteorology, 12:463-489.
Sigmon, J.T., Gilliam, F.S., and Partin, M.E. 1989. Precipitation and Throughfall Chemistry for
Montane Hardwood Ecosystem: Potential for Contributions from Cloud Water. Can. J. For.
Res., 19:1240-47.
p/castnet2/wa02/fnlsum/fnlipt_5. wpd
REF-6
Harding ESE, Inc.
-------
MADPro Final Summary Report
Sisterson, D.L., Bowersox, V.C., Meyers, T. P., Simpson, J.C., Mohnen, V. 1991. Deposition
Monitoring Methods and Results. State of Science and Technology Report Number 6.
National Acid Precipitation Assessment Program, Washington, DC.
Stogner, R.E. and Saxena, V.K. 1988. Acidic Cloud: Forest Canopy Interactions: Mt. Mitchell, NC.
Environ. Pollut., 53:456-58.
Thome, P.O., Lovett, G.M., and Reiners, W.A.. 1982. Experimental Determination of Droplet
Deposition on Canopy Components of Balsam Fir. J. Appl. Meteorol., 21:1413-1416.
Trautner, F. 1988. Entwicklung und Anwendung von Mefisystemen zur Untersuchung der
chemischen und physikalischen Eigenschaften von Nebelwasser und dessen Deposition auf
Fichten. Dissertation. 160pp. University of Bayreuth. Bayreuth, Germany.
U.S. Environmental Protection Agency (EPA). 2000. National Air Quality and Emission Trends
Report, 1998. EPA 454/R-00-003. March 2000.
U.S. Environmental Protection Agency (EPA). 1990. An Assessment of Atmospheric Exposure and
Deposition to High Elevation Forests in the Eastern United States. EPA/600/3-90/058.
September 1990.
U.S. Environmental Protection Agency (EPA). 1998. Clean Air Status and Trends Network
(CASTNet) Deposition Summary Report (1987-1995). EPA/600/R-98/027. OAQPS,
Research Triangle Park, NC 27711.
Valente, R.J., Mallant, R.K.A.M., McLaren, S.E., Schemenauer, R.S., and Stogner, R.E. 1989. Field
Intercomparison of Ground-based Cloud Physics Instruments at Whitetop Mountain,
Virginia. Journal of Atmospheric and Oceanic Technology, 6:396-406.
Verhoeven, W., Herrmann, R., Eiden, R., and Klemm, O. 1987. A Comparison of the Chemical
Composition of Fog and Rainwater Collected in the Fichtelgebirge, F.R. of Germany, and the
South Island of New Zealand. Theor. Appl. Clim., 38:210-221.
Vong, R.J. and Kowalski, A.S. 1995. Eddy Correlation Measurements of Size-Dependent Cloud
Droplet Turbulent Fluxes to Complex Terrain. Tellus, 476:331-352.
Vong, R.J., Sigmon, J.T., and Mueller, S.F. 1991. Cloudwater Deposition to Appalachian Forests.
Environmental Science and Technology, 25:1014-1021.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
REF-7
Harding ESE, Inc.
-------
MADPro Final Summary Report
Weathers. K.C., Lovett, G.M., Likens, G.E. 1995. Cloud Deposition to a Spruce Forest Edge.
Atmospheric Environment, 29:665-672.
Weathers, K.C., Likens, G.E.,Bormann, F.H., Bowden, W.B., Andersen, J.L., Cass, D.A.,
Galloway, J.N., Keene, W.C., Kimball, K.D., Huth, P., and Smiley, D. 1988. Cloudwater
Chemistry from Ten Sites in North America. Environmental Science and Technology,
22:1018-1026.
Weathers, K.C., Likens, G.E., Bormann, F.H., Bowden, W.B., Andersen, J.L., Cass, D.A.,
Galloway, J.N., Keene, W.C., Kimball, K.D., Huth, P., and Smiley, D. 1986. A Regional
Acidic Cloud/Fog Water Event in the Eastern United States. Nature, 319:657-658.
Wrzesinsky, T. and Klemm, O. 2000. Summertime Fog Chemistry at a Mountainous Site in Central
Europe. Atmos. Environ., 34:1487-1496. Abstract.
p/castnet2/wa02/fnlsum/fnlrpt_5.wpd
REF-8
Harding ESE, Inc.
-------
Tables
-------
-------
Figures
-------
-------
Appendix
-------
-------
Tables
-------
-------
MADPro Final Report
Table 2-1. Schedule of Routine QC Checks, Calibrations, and Audits Performed at the CASTNet Laboratory
(Harding ESE)
Observable
Total
Acidity
Ca2*. Mg2*,
Na*. K*
NH*
4
NO3, NO^,
Cl", SO*"
:ine/Coarse
Aerosol
Mass
Measurement
Device
Titration
Atomic
absorption
Automated wet
chemical
Suppressed ion
chromatography
Gravimetric
micro-balance)
Performance
Test
Frequency
One per batch
or daily
One per batch
or daily
One per batch
or daily
One per batch
or daily
One per batch
or daily
Performance
Standard
Blanks,
standards
Blanks,
standards,
replicates,
independent
reference
standards
Blanks,
standards.
replicates.
independent
reference
standards
Blanks,
standards,
replicates.
ndependent
reference
standards
Class S
weight
Calibration
Frequency
Daily
Daily
Daily
Daily
Daily
Calibration
Standard
NIST-traceable
pH4and
pH 7 potassium
hydrogen
phthalate (as
primary standard)
Reagent-grade
chemicals
Reagent-grade
chemicals
leagent-grade
chemicals
Class S weight
Audit
Frequency
Annual
Annual
Annual
Annual
Annual
Audit
Standard
Reference
standards
Reference
standards
Reference
standards
Reference
standards
Class S
weight
Primarv
EPA/NIST
reference
standards
EPA/NIST
reference
standards
EPA/NIST
reference
standards
EPA/NIST
reference
standards
NIST
reference
weight
Source: Harding ESE.
p/ecm/p/castnet2/wa02/fnlsum/!ables_fntsumrep.doc
-------
MADPm Final Report
Table 2-2. Precision and Accuracy Objectives for CASTNet Laboratory Data
Analyse
PH
Conductivity
Acidity
Ammonium (NH*)
Sodium (Na+)
Potassium (K+)
Magnesium (Mg2+)
Calcium (Ca2+)
Chloride (CO
Nitrite (NOJ)
Nitrate (NO,)
Sulfate (SO^")
Medium
W
w
W
W/F
W/F
W/F
W/F
W/F
W
W
W/F
W/F
Method
Electrometric
Electrometric
Titrimetric
Automated colorimetry
ICAP-AE
ICAP-AE
ICAP-AE
ICAP-AE
[on chromatography
ton chromatography
Ion chromatography
Eon chromatography
Acceptance Criteria
Precision
(RPD)
12
10
15
10
10
10
10
10
5
5
5
5
Accuracy
(%)
85-115
85-115
N/A
90-110
90 - 1 10
90 - 1 10
90-110
90-110
95 - 105
N/A
95 - 105
95 - 105
Nominal
Detection Limits
N/A
0.2 |j.ohms/cm
5 Jieq/L
0.02 mg-N/L
0.005 mg/L
0.005 mg/L
0.003 mg/L
0.003 mg/L
0.02 mg/L
0.004 mg/L
0.008 mg-N/L
0.04 mg/L
Note: For more information on analytical methods and associated precision and accuracy objectives, see CASTNet
Draft Quality Assurance Project Plan (QAPP) (Harding ESE, 1999).
W = wet or aqueous
F = filter
ICAP-AE = inductively coupled argon plasma-atomic emission
p/ecnVp/castnet2/wa02/fnlsum/tables_fntsurnrcp.doc
-------
MA DPro Final Report
Table 2-3. Data Quality Objectives for Continuous Measurements
Measurement
Parameter
Wind Speed
Wind Direction
Relative Humidity
(Rh)
Solar Radiation
Precipitation
Ambient Temperature
Delta Temperature
Surface Wetness
Ozone (O3)
Filter Pack Flow
Method
Anemometer
Wind Vane
Hygrometer
Pyranometer
Tipping Bucket Rain
Gauge
Platinum RTD
Platinum RTD
Conductivity Bridge
Ultraviolet (UV)
Absorbance
Mass Flow
Controller
Objectives*
Precision
±0.5 m/s
5%
±10% (of full scale)
±10% (of reading)
±10% (of reading)
±1.0°C
±0.5°C
Undefined
±10% (of reading)
±0.15Lpm
Accuracy
The greater of ±0.5 m/s for winds
<5 m/s or ±5% for winds >5 m/s
5°
±5%, Rh >85% ±20%, Rh >85%
10%
±0.05 inchf
±0.5°C
±0.5°C
Undefined
±10%
±5%
Note: °C = degrees Celsius
Lpm = liters per minute
m/s = meters per second
RTD = resistance-temperature device
* Precision criteria apply to collocated instruments, and accuracy criteria apply to calibration of instruments.
t For target value of 0.50 inch.
Source: CASTNet Draft Quality Assurance Project Plan, Harding ESE (1999).
p/ecm/p/castnet2/wa02/fnlsum/tables_fn!sumrep.doc
-------
MADPro Final Report
Table 2-4. Results From the QA Testing at ECN Facilities in the Netherlands
Run*
1
2
3
3a
4
4a
5
5a
6
7
8
8a
9
9a
10
lOa
lOb
11
12
13
13a
14
I4a
14b
14c
15
16a
16
Mean
ECN
Pall-Filters
(mg/m3)
883
882
140
94
393
412
345
352
360
687
105
97
88
689
675
437
441
438
687
862
128
97
388
383
388
384
647
114
ECN
PVM-100
(mg/m3)
808
832
163
110
392
412
361
369
360
686
107
94
85
695
681
427
431
429
689
857
116
88
406
401
406
402
680
120
RPD*
-8.49
-5.67
16.43
17.02
-0.25
0.00
4.64
4.83
0.00
-0.15
1.90
-3.09
-3.41
0.87
0.89
-2.29
-2.27
-2.05
0.29
-0.58
-9.38
-9.28
4.64
4.70
4.64
4.69
5.10
5.26
1.04
EPA-1
PVM-100
(mg/m3)
880
896
184
125
423
446
388
395
386
731
104
94
86
700
687
433
436
435
691
836
118
88
409
404
409 .
405
676
123
RPD
-0.34
1.59
31.43
32.98
7.63
8.25
12.46
12.22
7.22
6.40
-0.95
-3.09
-2.27
1.60
1.78
-0.92
-1.13
-0.68
0.58
-3.02
-7.81
-9.28
5.41
5.48
5.41
5.47
4.48
7.89
4.60
EPA-2
PVM-100
(mg/m3)
904
930
176
117
439
462
402
410
401
784
110
98
88
739
725
449
454
451
732
904
120
97
426
419
425
421
716
126
RPD
2.38
5.44
25.71
24.47
11.70
12.14
16.52
16.48
11.39
14.12
4.76
1.03
0.00
7.26
7.41
2.75
2.95
2.97
6.55
4.87
-6.25
0.00
9.79
9.40
9,54
9.64
10.66
10.53
8.36
EPA
Valente
(mg/m3)
1170
1110
75
112
510
472
477
932
72
823
517
491
841
1034
114
482
487
488
837
121
RPD
32.50
25.85
-46.43
19.15
23.79
34.09
32.50
35.66
-18.18
21.93
17.23
12.10
22.42
19.95
17.53
25.85
25.52
27.08
29.37
6.14
18.20
* Relative percent difference (RPD) was calculated using the Pall filter concentrations as references.
Source: ECN
p/ecntfp/castnet2/wa02/fnlsum/tab)es_fnlsumrep.doc
-------
MADPro Final Report
This page has been intentionally left blank.
p/castnet2/wa02/fti Isu m/figs-3. p6 5
Harding ESE. Inc.
-------
-------
MADPro Final Repon
Table 3-1. MADPro Cloud Frequency Summary (page 1 of 2)
Whiteface Mountain
May
June
July
August
September
October
November
% cloud frequency*
samples**
% completeness
% cloud frequency
Samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
1994
1995
1996
1597
1998
1999
Total
42.09
335
45.00
57.46
717
96.00
64.21
693
96.00
14.31
720
100;00
48.92
738
99.00
40.80
549
74.00
41.63
699
97.00
48.96
527
73.00
48.0!
579
78.00
45.24
714
96.00
50.43
704
98.00
33.88
546
76.00
26.32
475
64.00
42.86
686
92.00
57.89
596
83.00
40.68
708
98.00
37.56
639
86.00
35.67
600
81.00
45.15
660
92.00
20.11
378
53.00
33.24
737
99.00
38.01
734
99.00
33.09
556
77.00
34.46
41.93
43.34
48.73
31.51
384
52.00
Whitetop Mountain
May
June
July
August
September
October
November
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
1994
14.58
432
58.00
33.76
631
88.00
49.89
461
62.00
51.78
676
94.00
30.19
742
100.00
1995
28.05
656
88.00
42.43
634
88.00
20.95
592
80.00
33.22
578
78.00
36.76
710
99.00
30.29
733
99.00
1996
50.35
143
20.00
46.49
684
92.00
45.02
613
82.00
47.17
653
91.00
36.20
732
98.00
23.97
292
41.00
1997
43.47
352
47.00
49.13
692
96.00
29.66
725
97.00
33.17
627
84.00
26.55
708
98.00
28.63
716
96.00
65.41
266
37.00
1998
29.59
338
45.00
53.79
699
97.00
30.24
744
100.00
29.17
744
100.00
12.93
696
97.00
26.11
609
82.00
1999
15.45
479
64.00
36.43
700
97.00
46.06
469
93.00
14.67
634
85.00
25.03
719
100.00
Total
28.05
43.11
31.84
31.05
33.37
30.28
p/ecm/p/casmet2/wa02/fnlsum/lables_fnlsumrep.doc
-------
MADPro Final Report
Table 3-1. MADPro Cloud Frequency Summary (page 2 of 2)
Clingman's Dome
May
June
July
August
September
October
November
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
% cloud frequency
samples
% completeness
1994 1995
J996
1997
1998
1999
Total
81.78
82
11.00
29.47
285
38.00
49.44
710
95.00
32.41 30.37
395 349
55.00 48.00
40.27
663
89.00
46.64
298
40.00
23.64
330
44.00
61.63
172
24.00
34.34
661
89.00
41.49
617
93.00
33.18
639
89.00
35.52
563
76.00
48.58
422
59.00
55.42
720
97.00
71.43
7
1.00
43.93
387
54.00
30.32
696
94.00
41.38
667
93.00
44.75
733
99.00
24.93
742
100.00
27.65
622
86.00
41.38
44.84
38.62
30.42
35.37
59.70
67
9.00
* Cloud frequency is not used in subsequent analyses if the completeness criteria of greater than 70
percent is not met.
* * Number of samples
Source: Harding ESE.
p/ecm/pfca$toet2/wa02/fnlsum/tables_fnlsumrep.doc
-------
I
I
5
^
u.
o
&
8,
ON
ON
o\
r
00
3
1
«•
5
ON
P
<£
(A
S
(W
r>
s
U
<
2
3
1
u
&>
0
U
05
£
•o
_o
rj
Ui
o
ci
(A
O
'to
«
55
>,
w
3
CO
«?
rn
w
z
«
4)
e
Ci
so
C
U
e
S
o
IS
a
2
1
s
_c
5
c
s
1
§
£
S
ON
X
H
B 5
IE
< S
» -a
"S 3
O (0
^ =
n
IS
e
OS
-*
II e
11
&
5 &
•< "O
•3 S
C w
u C
C£ g
"5 E
o
£
x
A
3s
•w .S
is
8|
< '
m
0
U 0
QJ 5
5 1
« s
£
^O
tT3
o
s
0
s
o
s
0
oo
r-
o
o
c
o
o
PO
O
av
oo
in
m
O
0
-a-
r^>
O
S
0
PO
CM
U
^
vo
CM
oo
m
P
O
\D
CO
•t
Ov
vO
PO
V>
OO
CM
in
Tt"
o
o
oo
PO
r-
m
PO
r-
CM
S
O
r-
ov
PO
PO
CM
CL
8
Tf
•*
O
c~
01
V)
r~
PO
PO
in
CM
vo
ON
S
ON
r-
r-
o
00
(N
*n
§
S
ov
•*•
0
PO
(S
S
8
•*
r-
OV
ts
00
in
PO
VD
0
_
O)
•a
o
CJ
NO
ro
m
CM
r~
C
oo
oo
•*
PO
ON
PO
r-
o>
(N
00
00
•*s-
V~l
r-
m
TT
£
o
•*
{N
CN
r-
rr
s
g
s
oo
W!
o\
—
oc
&•
p~
r<^
*fi
Tf
vC
CM
f.
CVI
+
X
Tf
o
S
n
O
0
v£)
IT)
CM
O
cs
O
rJ
IT)
r^i
•f "T
I
*A
01
VO
>n
V)
CM
IO
£
m
>o
oo
CM
^
vO
ON
CM
S
o
oo
r~
Tl-
CM
CM
O
o
t—
0
p^
vO
O
m
CM
•r
u^
Tf
S
S
VC
s
o
pn
CM
ON
CM
ON
>n
oo
s
n
CM
+
"«
U
S
u-t
oo
CM
oo
v>
r~
£
vO
Ov
8
o
00
•v
r-
O
CM
ON
ON
<5
CM
OO
3
00
m
CM
oo
C
CM
f
Ov
^
vn
vO
m
CM
*
00
£
•*
oo
S
0
t-
CM
pn
r~
m
CM
m
CM
U!
r-
CM
r-
oo
•*
CM
^:
in
m
oo
01
en
•*
ON
P^l
00
O
m
m
m
r-
pn
oc
PO
ON
pn
r-
O4
O
m
S
\o
v>
0
£
po
Ov
CM
C
in
CM
U
c.
A
U
•M
i
x:
in
ev
Ps
x
(0
ve
"li" c
11
&
I «
•< -o
-g 2
C
O»
O
s
0
r-
CM
O
VO
0
CM
m
U
j
r-
01
VO
r-
vo
CM
pn
VO
O
r~
00
en
vO
pn
m
ol
n
0
m
VO
r-
m
CM
pn
vo
O
CM
Ov
pn
f>
in
o
o
1
o
m
Tf
r-
PO
Ov
m
\o
TT
vO
n
oo
CM
pn
in
VO
oo
in
CM
CM
«n
O
m
8
S
ON
—
o5
ON
in
O
ON
r-
pn
?s
m
•o
$
VD
ON
r-
en
CM
•*
m
O
r-
oo
m
oo
O!
r-
VI
CO
S
vO
PO
O
•*
O
r-
in
CM
Ov
OO
vn
01
oo
Ov
m
CM
oo
CM
S
o
in
PO
>n
Ov
S
R
0
%
VC
ro
PO
r-
O
R
PO
m
4
PO
1^-
r-
m
0
m
PO
m
pn
CM
Ol
s
CM
OO
m
PO.
PO
-*
CM
m
r-
o
•«•
r-
CM.
CM
g
8
CM
O
>n
a
•*
PO
r~
o
ON
in
PO
CM
OJ
PO
vO
CN
r*
«n
-»• •»
Z
•7,
CO
oc
ol
ro
PO
r~
Tf
PO
VO
§
•>3-
S
vO
PO
m
PO
PO
oo
>n
ON
r-
CM
X
NO
00
ON
ON
CM
C
tt
•V
1—
01
o
n
CM
o
PO
VO
PO
CM
PO
O
oo
m
VO
PO
00
VO
CM
0
oo
Ov
O
vC
r-
VO
vO
v*,
S
eo
S
m
i
ON
vO
0
OO
wo
PO
PO
PO
r-
00
CM
PO
PO
TT
m
00
r~
Ol
CO
01
o
ON
CM
ON
PO
OO
8
oo
m
in
oo
m
PO
01
PO
0
oo
m
PO
0
in
O
in
•f
es
Z
X
r-
Ol
0
r-
t—
Ol
ro
DO
Ol
PO
CO
oo
r-
m
t—
PO
ro
m
^**
vC
01
01
•t
Vl
r-
00
m
oo
0«
00
oc
0
0
r-
PO
•V
PO
CM
vC
in
•^
^
ON
CM
in
CM
vO
m
0
s
in
PO
r-
oo
PO
«o
PO
PO
CM
•^
ON
r-
o
m
ON
vO
-«-
NO
TT
CO
m
O
m
•*t
oo
5
0
0
c-
CM
•o
OV
-------
I
O
CN
a
03
CX
ON
ON
WJ
3
O\
£
o
U
u
.2
^
"g
b
s
03
41
0
Q
UQ
a
f,
_6
'c
§
«£
O
e
"S
.r
*
.s
3
§
S
S
£
*e
*••
0
"ii" =
u c
"S.
J «
< T)
u 1/>
u B
fi£ eg
_ *
2 £
^s
B
(B
Tf
"iF c
O E
£•
u >.
< T3
K «H
"53 <**
U W
1 c
Q£ CS
o
B
A
4 c
"a. *^
u >•
£j &
f 1
8 B
S6 2
B
CS
TT
0
S
o
0
o
es
O
Tf
r~
o
SO
^^
TT
o
sC
O
Os
S
o
o
SO
CS
0
TT
m
0
TT
m
U
in
oo
TT
SO
O
en
TT
O
en
r-
en
S
Tf
in
cs
oo
CN
in
O
^
en
oo
os
oo
m
oo
CN
SO
®
U
Vi
S
S
U
1
1
'1
*
JK
S
§
1
&
*
i
, g
fi
n _
i1
1 1
< -e
•o *
Z
0
0
o
2
o
oo
en
0
en
oo
CN
O
S
o
r-
0
O\
en
o
so
s
g
o
SO
0
t~-
o
m
en
en
U
T
oo
m
00
cs
so
o
so
en
oo
en
§
so
O
oo
CM
m
m
0
8s
en
O
O
SO
00
CN
S
0
oo
o
TT
Os
en
X
0.
B
0
oo
0
so
cs
s
en
r-
m
r-
r-
oc
O
oo
8
o
oo
00
o
TT
O
O
m
CN
en
m
SO
8
oo
OS
0
o\
—
s
oo
m
£
oo
en
oo
t-
en
•c
a
OS
oo
Tf
OO
«n
cs
en
eN
in
es
Os
OO
en
CN
en
oo
m
Os
oc
Tt-
OO
r-
oo
O
B
o
SO
en
en-
en
CM
O
ON
OO
^
O
CM
SO
en
cs
en
t-
OS
en
os
en
X
o
o
o
m
SO
O
en
SO
p:
in
S
oo
en
en
Tf
t--
r-
O
en
cs
SO
CS
s
0
es
O
en
rs
r-
0
so
OC
SO
OC
m
OC
es
OO
en
V
Z
£
Os
r-
cs
sC
en
S
s
TT
r-
r-
CN
en
TT
oo
en
O
O
in
t~-
en
CN
en
oo
8s
O
SO
en
_
en
en
O
iB
SO
a
oo
~
es
cs
oc
OS
en
oo
es
1
oo
en
V
CO
Tf
s
o
r-
vn
cs
i2
§
oo
m
en
oc
oo
en
r-
in
en
sO
oo
CN
CS
CS
OC
oc
m
r-
•n
o
8
0
o
Os
so
O
00
SO
CM
Tf
m
2
oo
m
Z
sO
oo
en
eN
0
O
m
§
oo
en
O
O
—
Os
es
O
Os
m
en
en
o
8
so
cs
O
£
in
Os
0
oo
en
'cf
Z
oo
cs
m
m
0
cs
1—
Os
IS
5
OO
en
OS
en
CN
O
so
m
en
t—
cs
O
P
OS
en
r—
m
o
00
en
r-
TT
SO
en
en
b
r--
en
f-
f~-
en
0
en
»n
n
o
CM
en
S
Tf
CM
OO
en
cs
>n
SO
m
0
SO
5
Os
CM
SO
CM
O
5
0
sO
in
O
o
TT
so
OO
CM
—
OO
en
U
-------
<*•
O
1
fTs
Os
ON
j-
00
•1
c\
" '
p
c:
bo
n
1
3
•o
o
U
"o
U
V3
«
tvt
U
a
1
(A
O
83
oo
£>
E
00
•
Z
es
H
o
O
cc
e
.1
6
e
's
0
a
3
a
*
e
1
§
£
IS
00
Os
a
8
IE
a
•S 2
C
e
oo
r-
0
s
0
— —
0
so
CO
O
CM
vn
OS
O
0
o
r—
0
oo
CN
O
m
sT-
CM
S
"
O
O
CM
O
Os
o
s
CO
|
CN
OC
v~i
CM
OO
CM
Os
CO
0
r-
co
oo
CN
v>
r-
CM
00
•<*•
0
IT)
CO
—
CN
O
00
m
SO
CM
r-
0
o
Tt
C—
oo
CO
a.
8
CO
Os
r-
O
t—
CN
OS
8
oo
O
CM
8
oc
O
Os
8
Os
SO
3
g;
oo
Os
—
CM
8
1
00
CO
•<*•
s
CO
R
oc
T3
I
SO
i
^
SO
CM
O
CM
OO
t—
so
00
1
CM
W)
r~
in
CM
sO
in
so
CO
r-
CM
CN
C—
CN
OO
—
g
o
oc
OS
CN
00
t-
00
3C
8
0
CO
Os
CO
OS
1
s
Os
t—
\b
E
r-
"3"
s
2
a
00
SO
r-
CM
r-
oo
CM
m
00
—
!2
g
CM
8
OO
00
CO
V
03
0
s
00
in
O
CM
oo
CO
SO
SO
m
CN
00
*"*
s
oo
•<*
oo
so
SO
r-
CM
r-
CM
OS
OS
-3-
CO
t
CM
S
VO
CM
in
r-
Os
CO
SO
oo
fo
2:
OS
CM
oo
i
0
00
CO
00
o
m
00
so
SO
CO
1
Os
t-
"""
r~
c
jj
Os
00
SO
1—
OO
Os
CN
—
oo
oo
oo
00
—
2
SO
OO
o"'
?
^
OS
CM
0
00
oo
0
00
0
so
o
o
oo
CN
0
m
co
SO
r-
B
vo
CM
O
t--.
0
oc
SO
O
vO
OO
Z
£
TT
o
00
CN
ON
SO
s
i
VO
0
Tf
so
£
0
Os
so
r-
CM
s.
0
oo
CO
£
O
Os
oo
^5
m
CM
t-
oo
5
^
m
r-
£
0
I—
0
fN
—
OO
CN
1
SO
•<*•
O
CN
Os
vo
—
so
t—
CN
CN
CN
CM
m
CM
O
CM
CN
OC
vO
t—
OO
^00
2
Os
m
1
5
0
m
m
CM
g;
S
SO
J^
0
CN
CN
CM
0
r~-
CM
CO
oo
vO
r-
CM
00
OO
o
CM
CN
O
ON
,_
£
^
r-
00
4-
a
CN
OC
J2
0
§
co
O
oo
CO
so
5
•>*
CO
0
CM
Os
'*
m
Os
CO
SO
r-
CN
^
CM
CO
O
oo
CN
r-
CN
OO
4-
^
in
vO
r—
vO
CN
O
s
so
m
CM
vO
m
0
00
o
CO
vo
«
VO
r-
CN
CO
SO
OO
CN
SO
m
O
s
in
*
Os
VO
oo
b
»
Ul
e
to
U
e
1
*H
•=
3
0
1
S
Z
OS
O\
n
ll
0 %
< -o
f 1
V &
? S
H
e
"8 s
«^
a
S a>
< tJ
^ 2
C n
0
i
t-
s
CO
CO
IT)
CN
^*
OO
m
3
r-
o
CM
CO
n
oo
oo
Os
CO
oc
r~
oc
vO
m
0
£
CM
CN
OO
-3-
0
CO
r-
U
1
a.
fi
h
^
|
5»
h
!
:aslne(2/v
i"
U
i
-------
."^
^J-
*o
^3-
&
a
^•*
Os
o
Os
.C
60
' 3
£
j=
4— '
^h
Os
0*
1
us
A\
Ss
35
o
£
D
2
*•>
TO
•8
1
o
U
«
_«
a,
c
C/D
£
«3
1
•o
0
3-2. Summary Statistics for Cl
V
3
C5
wi
0
1 c
5 .2
•a t3
max = maximum
mean = arithmetic average
min = minimum
n = sample size used in c
d dev = sample standard devi
if.
U
O
Z
i
(J
"S?
£
z.
>,
*_>
'>
*4^
U
a
13
O
U
"T3
C
ra
-
v;
t^
_W
4^
'£
3
1
««".
£
•*»
its are //eq/L except for LWC (
c
^
^M
<
1«
Bl
^ -S
11
o JJ
•« C3
^ i-
0 ,0
U <*H
S *o
S- ts
a +1
£ a
± *
t_^ ^
<— ' ^
>
+i ,0
« JJ
> —
? T3
y-x W
^ ,5
o 'S
• . «£ O
en 4) "D
s -0 ^
^11
b ,g as
pl
iei
c « -
.2 « J1
£ J d
J? e w
T5 H 5
C V "-J-
C » O
0 J -
'•C rr< C
« Sf «
w a j=
Sau
•£ S fe
c - «3
0 c Ji
"8 .= **
y ~ u
W £l
rt tn /D
* S3 $
•n ^^ f^i
^W j,» V3
8 S E
3^3
a * S
(lowing acceptance criteria wei
) If both cation and anion sum?
) If either of the cation or anioi
accepted.
<£ - ci
u
J=
H
o
1
c
8
c
o
o
O
c
a
-o
c
«
c
.2
CO
(J
U
jr
«ri
u-
0
D
W)
(3
V-i
flj
>
es
1
>,
•O
•a
T3
'>
"O
CO
e
o
'5
£
•*«»
C
jji
u
1
c
o
C3
•o
C
cd
C
0
RPD = The difference in cat
multiplied by 100
BJ
00
U
CO
1
o
y
3
O
CO
-------
MADPro Final Report
Table 3-3. Number of Records Accepted for Analysis
Site ID
Clingman's Dome
Whiteface Mountain
Whitetop Mountain
Year
1994
1995
1996
1997
1998
1999
1994
1995
1996
1997
1998
1999
1994
1995
1996
1997
1998
1999
Total Number of
Records
14
142
122
334
341
174
279
759
644
469
445
503
162
573
206
552
311
156
Number of Records
Accepted
9
136
105
324
269
174
235
525
597
448
387
478
141
550
194
501
276
143
Percent Accepted
64%
96%
86%
97%
79%
100%
84%
69%
93%
96%
87%
94%
87%
96%
94%
91%
89%
92%
Source: Harding ESE.
p/ecm/p/castnet2Ava02/fnlsum/tab les_fnlsumrep.doc
-------
MADPro Final Report
Table 3-4. Mountain Cloud Linear Regression Results
Concentrations
Normalized
Nonnormalized
Site ID
CLD303
WFM300
WTM302
CLD303
WFM300
WTM302
Two-sided p-values and directions of trend
H+
0.0001
0.0956
0.0650
0.0001
0.1046
0.5097
increase
increase
NH+4
0.0151
0,0002
0.6648
0.0328
0.0015
0.3613
increase
decrease
increase
decrease
NO;
0.0001
0.2159
0.4650
0.0001
0.5332
0.3594
increase
increase
so?
0.0002
0.0078
0.0305
0.0026
0.0029
0.4023
increase
decrease
increase
increase
decrease
Note: Bold italicized p-values indicate significant changes over time (alpha=0.05).
No diagnostics were performed to check the adequacy of the linear models.
Source: Harding ESE, 2000
p/ecm/p/castnet2/wa02/fnlsum/tabtes_fnlsurarep.doc
-------
MADPni Final Repnrt
Table 3-5. Summary Statistics of Major Ion Concentrations for June through
September 1994 through 1999
Clingman's Dome
Whiteface Mountain
Whitetop Mountain
mean
minimum
maximum
median
mean
minimum
maximum
median
mean
minimum
maximum
median
H+
358.92
1.51
2137.96
269.15
224.70
0.27
2754.23
104.71
297.63
4.37
2570.40
199.53
NH*
237.09
1.07
1650.00
193.21
167.69
0.56
1721.43 -
77.14
215.74
0.71
1721.43
133.57
so;
445.52
3.54
3686.91
337.50
288.77
1.35
3499.27
143.33
391.96
1.83
2395.83
262.45
NO;
182.31
1.64
1007.14
147.96
125.12
0.57
2251.54
59.86
174.70
2.50
1992.86
124.29
Source: Harding ESE.
p/ec m/p/castnet 2/wa02/f nl su m/iab les_fnlsumrep.doc
-------
-------
MADPm Final Report
Table 4-1. Monthly Deposition Estimates Produced with the CLOUD Model
Site
Clinsman's Dome
Whiteface Mountain
Whitetop Mountain
Year
1994
1995
1996
1997
1998
1994
1995
1996
1997
1998
1994
1995
1996
1997
1998
Month
10
8
—
7
8
9
10
7
10
8
9
7
8
9
6
7
8
9
6
8
9
6
7
8
9
6
9
10
6
7
8
9
10
7
8
9
10
6
7
8
9
10
6
7
8
9
10
Deposition (kg/ha)
H*
0.43
1.33
—
2.28
2.35
1.78
3.10
4.52
2.16
4.71
8.56
5.51
5.65
3.93
3.34
2.30
2.75
2.21
2.14
2.75
5.76
2.39
3.05
3.22
5.70
1.66
2.25
0.54
1.31
f.Ol
2.86
1.71
0.68
2.01
7.99
3.60
2.24
3.46
3.90
2.56
1.61
1.37
5.00
1.65
2.07
1.43
2.49
sof
1.30
3.11
—
4.71
4.72
3.70
6.57
7.86
3.93
12.10
—
11.10
8.11
5.41
7.50
6.14
6.76
4.45
5.46
5.31
12.40
4.11
5.38
6.29
10.02
3.16
3.52
1.26
2.78
2.44
5.73
3.40
1.37
4.37
15.91
6.11
5.18
7.30
9.67
5.46
3.74
2.83
10.56
3.47
3.82
2.44
0.88
NO;
0.52
1.12
—
1.55
1.89
1.02
2.76
3.01
2.22
3.38
7.84
4.10
2.78
3.12
2.13
2.38
2.36
2.16
2.90
1.72
4.71
1.77
1.75
2.11
4.08
1.10
1.80
0.62
1.04
1.02
2.56
1.55
0.85
1.55
5.17
2.95
2.34
3.11
2.84
2.06
1.30
1.37
4.03
1.07
1.22
0.90
0.37
NH*
0.82
1.30
—
2,01
2.36
1.58
3.66
5.92
2.35
6.47
11.33
6.36
4.24
3.42
4.63
3.57
3.77
2.48
3.57
2.09
5.86
2.11
2.81
3.71
5.77
1.59
2.86
0.83
1.48
1.27
2.64
1.68
0.79
1.74
5.15
3.97
3.20
4.02
4.35
2.70
1.98
1.64
4.35
1.64
1.79
1.01
0.27
H2O(cm)
6.43
9.83
—
5.54
8.74
10.43
7.02
10.76
9.10
22.23
50.83
23.47
21.07
19.85
27.17
21.80
17.77
21.70
14.34
17.87
23.78
17.44
17.89
15.79
20.90
5.59
10.27
7.09
7.16
2.07
8.10
5.64
6.55
10.28
10.21
17.24
11.65
13.77
17.09
7.79
6.83
8.47
16.43
3.63
4.40
1.68
5.63
p/ecm/p/castnet2/wa02/fn!sum/lables_fnlsumrep.doc
-------
MADPrn Final Report
Table 4-2. Seasonal Deposition Estimates Produced with the CLOUD Model*
Site
Clingman's Dome
Whiteface Mountain
Whitetop Mountain
Year
1997
1995
1996
1997
1998
1995
1996
1997
1998
IT
8.55
20.12
10.60
14.20
14.36
6.89
18.13
11.53
10.15
so2;
17.51
32.83
24.85
30.89
25.80
14.35
35.19
26.17
20.29
NO;
5.95
13.33
9.03
12.44
9.71
6.17
12.89
6.71
7.22
NH+4
7.93
18.69
14.45
15.36
14.40
7.07
14.48
13.05
8.79
Note: All measurements are in kg/ha
* Three of the four months were required to calculate seasonal deposition. Three-month depositions
were multiplied by 4/3 to obtain seasonal depositions. Season is defined from June through September.
p/ecm/p/castnet2/wa02/fnlsum/tables_fnlsumrep.doc
-------
j
I
_o
c
2
o
Q,
a
00
O
60
"E
1
a
rt
Q
o
u
UJ
*?
^f
-.B
||
a H
u <
U
w wi
8 *?
M A
Cd Z
sS
*A Q
O *
12 Q
2|
o g
Z ?
i S
A «s»
II A
1 §
O £^
i A
* §
*
u
* 1
•M
1
35
vo 5 r- Ti- oo
rn ^» cO r*^ 10
^~ *™™ *"~*
c c o o o
— O ON C O
J2 0 £ f. f-
3 2 ^ - ^
O ^ O ON f-
n
^n r**
ON ON
i i
F- F-
^ ^
u
o
u
I
s
OJ
o
2
-------
MADPro Final Report
Table 4-4. Combinations of Wind Speed and LWC Evaluated
LWC = 0.263 g/m3
LWC = 0.418 g/m3
LWC = 0.833 g/m3
WS = 3.618 m/s
V
•*
*
WS = 7.028 m/s
•/
^
v'
p/ecm/p/castnet2/wa02/fnlsum/tables_fn!sumrep.doc
-------
MADPm Final Report
Table 4-5. Alternate Forest Canopy Descriptions
Effect of Tree
Species
LAI = 10 mVm2
Height= 17m
Effect of Height
LA! = 10 m2/m2
Effect of Leaf
Area
Height = 10m
observed canopy
Broad Leaf =16% of LAI
Needle Leaf = 84% of LAI
specified*
Height = 10 m Height = 1 1 m
low
LAI = 6 LAI = 8
simple model
Needle Leaf =
Height = 12 m
typical**
LAI =10
canopy*
100% of LAI
^
modeled
Height = 14 m Height = 17 m
k
high
LAI =12
* A pure conifer canopy of 10-m height was specified.
** An observed LAI of 10 m2/m2 value at 1,225 m was selected as typical of the forests of interest.
p/ecm/p/castnet2/wa02/fnlsum/tabfes_fnlsumrep.doc
-------
MADPm Final Repiiri
Table 4-6. MCLOUD Seasonal Deposition Estimates
Site
Clingman's Dome
Whiteface Mountain
Whitetop Mountain
Year
1997
1999
1994
1995
1996
1997
1998
1999
1995
1997
1999
kg/ha
soj
9.4
21.3
27.3
17.8
16.6
28.0
20.4
12.4
13.1
18.7
6.7
NO;
3.2
7.3
9.5
7.1
6.1
10.4
7.4
4.8
6.1
6.6
2.4
NH*
4.3
7.9
14.6
8.8
8.6
14.2
10.8
5.7
6.8
8.8
2.7
p/ecm/p/castnet2/\va02/fn]sum/tables_fnlsumrep.doc
-------
MADProFinal Report
Table 4-7. Summary of CLOUD Model Sensitivity, Potential Bias, and Expected Differences with
MCLQUD Modeling Results
Model
Parameter
canopy type
canopy height
leaf area index nr/m2
wind speed & LWC
layer thickness
vertical distribution LAI
LWC
wind speed
SO, concentration
met & chemistry
vertical distribution LAI
drop size distribution
# drop size classes
wind speed height
pressure, temperature
Scenario
General Model Sensitivity
pure fir canopy vs. fir+birch canopy
10-m canopy height vs. typical observed 1 7-m height
LAI range of 6 to 12 nr/m2 @ 10-m height
6 m s"! vs. 12 m s'1 @LWC = 0.26 g m'3 WFdrop size distribution
6 m s'1 vs. 1 2 m s'1 @LWC = 0.83 g m'3 WF drop size distribution
6 m s"1 vs. 1 2 m s"1 @LWC = 0.26 g m'3 WT drop size distribution
6 m s"' vs. 12ms'1 @LWC = 0.83 g m'3 WT drop size distribution .
met conditions shift 1 150 - 1350-m a.s.l. w/o cloud frequency
met conditions shift 1 150 - 1350-m a.s.l. including cloud frequency
General Possibilities for Bias
1 -m layers vs. limit of progressively smaller layers
shift shape of curve + 1 meter
shift shape of curve - ! meter
effect of data screening procedure in some years
effect of data screening procedure in some years
effect of data screening procedure in some years
effect of completeness of record
Differences Between CLOUD and MCLOUD Modeling
CLOUD vs. MCLOUD
CLOUD vs. MCLOUD for Whiteface Mountain
CLOUD vs. MCLOUD for Whiteface Mountain
CLOUD vs. MCLOUD for CLD-97
CLOUD vs. MCLOUD for Whiteface Mountain
CLOUD vs. MCLOUD
Parameter
Shift
-41%
20%
-40%
-50%
-50%
-50%
-50%
-15%
-15%
10%
-10%
unknown
unknown
range of LWC &WS
WFM-97 average
20 to 500
3.2 vs. 1
Range of
Response
+1.5% to +2.5%
-2.5% to -3.5%
-2.0% to +4.0%
-9.8% to +1.8%
-48%
-34%
-49%
-39%
-31%
-65%
-8%
12%
-9%
+6.6% to -6.6%
-0.7% to -5.1%
+9.6% to +21%
possibly significant
at least ±10%
-24% to +6%
-3%
-10%
-20%
±<1%
p/ecm/p/castnet2/wa02/fnlsum/table*_fn!sumrcp.doc
-------
MADPro Final Report
Table 4-8. Comparison of Deposition Velocities and Water Depositions
Deposition Velocity (cm/sec)
Month CLOUD MCLOUD
Whiteface Mountain 1997
7 31.56 31.19
8 32.72 31.18
9 48.53 42.64
Clingman's Dome 1997
7 18.74 15.70
8 21.96 16.22
9 23.55 16.56
10 25.05 19.08
Whitetop Mountain 1997
5 32.60 22.20
6 28.10 20,00
7 30.80 21.50
8 27.20 19.80
9 28.10 . 20.30
10 33.20 23.40
H2O Deposition (cm)
CLOUD MCLOUD
Whiteface Mountain 1997
— * „*
22.23 16.15
50.8 34.41
Clingman's Dome 1997
5.54 3.81
8.74 6.16
10.43 6.7
7.02 5.07
Whitetop Mountain 1997
--* 6.25
13.77 9.96
17.09 12.1
7.79 5.46
6.83 4.88
8.47 5.99
* Not Available.
p/ecm/p/castnei2/wa02/fnlsum/tables_fn lsumrep.doc
-------
MADPro Final Report
Table 4-9. Best Estimates of Seasonal Depositions
Site
Clingman's Dome
Whiteface Mountain
Whitetop Mountain
Year
1997
1999*
1994*
1995
1996
1997
1998
1999*
1995
1996
1997
1998
1999*
H+
8.55
22.04
34.94
20.12
10.60
14.20
14.36
14.84
6.89
18.13
11.53
10.15
7.39
so2;
17.51
30.89
38.22
32.83
24.85
30.89
25.80
17.98
14.35
35.19
26.17
20.29
9.71
NO;
5.95
10.58
13.78
13.33
9.03
12.44
9.71
6.96
6.17
12.89
6.71
7.22
3.48
NH;
7.93
11.45
21.17
18.69
14.45
15.36
14.40
8.27
7.07
14.48
13.05
8.79
3.91
Note: AH measurements in kg/ha.
* All 1999 estimates and 1994 estimates for Whiteface Mountain were derived from MCLOUD model
results.
The MCLOUD results were multiplied by 1.45 to account for the differences discussed in
Section 4.3.
U.S. EPA Headquarters Library
Mail code 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
p/ecfn/p/cnsmet2/wa02/fnlsum/tab les_fnlsumrep.doc
-------
-------
MADPrn Final Report
Table 5-1. Summary of Cloud, Precipitation, Dry,
and Total
Deposition Estimates for Whiteface Mountain (page
1 of 2)
Total S Deoosition (ke/ha/month)
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995
Sep 1995
Jun 1996
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jul 1997
Aug 1997
Sep 1997
Jun 1998
Jul 199?
Aug 199S
Sep 199S
Clddep
SO4"as S
12.129
16.721
11.121
8.126
5.418
7.520
6.155
6.818
4.463
5.475
5.317
12.430
4.117
5.391
6.300
10.048
'Total Sulfur Pry Deposition =
drydep
S0j
0.302
0.083
0.098
0.090
0.100
0.077
0.050
0.125
0.078
0.255
0.273
0.172
0.233
0.083
0.209
SO] as S
(SOi* 0.5011)
0.151
0.042
0.049
0.045
0.050
0.039
0.025
0.063
0.039
0.128
0.137
0.086
0.117
0.041
0.105
so;
0.392
0.084
0.134
0.184
0.118
0.062
0.077
0.140
0.125
0.275
0.271
0.121
0.203
0.091
0.132
1-
SO4 asS
(S04'* 0.3338)
0.131
0.028
0.045
0.062
0.039
0.021
0.026
0.047
0.042
0.092
0.091
0.040
0.068
0.031
0.044
Total
DRYDEP1
-------
MADPro Final Report
Table 5-1. Summary of Cloud, Precipitation, Dry, and Total Deposition Estimates for Whiteface Mountain (page 2 of 2)
H" Deposition {kg/ha/month)
clddep
H*
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995
Sep 1995
Jun 1996
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jul 1997
Aug 1997
Sep 1997
Jun 1998
Jul 1998
Aug 1998
Sep 1998
0.474
0.863
0.555
0.570
0.396
0.337
0.232
0.277
0.222
0.216
0.277
0.581
0.241
0.308
0.325
0.575
drvdep
H* from HNO3
(HNOj* 0.0160)
0.0267
0.0146
0.025!
0.0220
0.0141
0.0129
0.0102
0.0192
0.0090
0.0392
0.0403
0.0194
0.0278
0.0161
0.0237
H*fromSOj
(SOj* 0.0157)
0.0047
0.0013
0.0015
0.0014
0.0016
0.0012
0.0008
0.0020
0.0012
0.0040
0.0043
0.0027
0.0037
0.0013
0.0033
Total DRYDEP1
(HNOj+SOs)
0.0314
0.0159
0.0266
0.0234
0.0157
0.0141
0.0110
0.0212
0.0102
0.0433
0.0446
0.0221
0.0314
0.0174
0.0270
wetdep Tot
H* H
n
0.078
0.060
0.042
a)
••
0.028 0.907
0.024
0.063 0.641
0.015 0.600
0.023 0.433
0.034 0.382
0.026 0.279
0.021 0.309
0.022
0.027 0.287
0.046
0.047 0.346
0.023
0.042
0.038 0.377
0.060 0.402
0.063 0.665
1 Total H* Dry Deposition = ((HNOj * 0.0160) + (SO2 * 0.0157)) as documented by Miller <1993a)
NH* as N Denosition (ke/ha/month)
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995
Sep 1995
Jun 1996
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jul 1997
Aug 1997
Sep 1997
Jun 1998
Jul 1998
Aug 1998
Sep 1998
clddep
NH*asN
6.474
11.338
6.365
4.241
3.422
4.630
3.570
3.776
2.480
3.563
2.092
5.866
2.109
2.816
3.708
5.773
drvdep
NH;
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
NH^asN
(NH; * 0.7765)
wetdep T
NH;
0.570
0.255
0.308
0.218
0.158
0.427
0.273
0.099
0.173
0.232
0.146
0.128
0.166
0.253
0.254
0.123
0.273
0.204
0.481
0.628
NH.asN
. 4 NH
(NH4* 0.7765)
0.443
0.198
0.239
0.169
0.123
0.331
0.212
0.077
0.134
0.180
0.113
0.100
0.129
0.197
0.197
0.096
0.212
0.159
0.373
0.488
otal
^asN
Source: Harding ESE.
p/ecnVp/castnct2/wa02/fn!sum/tab!es_fnlsumrcp.doc
-------
MADPro Final Report
Table 5-2. Summary of Cloud, Precipitation, Dry, and Total Deposition Estimates for Whitetop Mountain (page 1 of 2)
Total S Deposition fte/ha/month)
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995
Sep 1995
Jun 1996
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jul 1997
Aug 1997
Sep 1997
Jun 1998
Jul 1998
Aug 1998
Sep 1998
Jun 1999
Jul 1999
Aug 1999
Sep 1999
clddep
SO'asS
3.162
3.530
2.789
2.448
5.740
3.411
4.375
15.943
6.125
7.312
9.695
5.476
3.747
10.582
3.476
3.825
2.448
drydep
S02
0.346
0.212
0.302
0.4142
0.6365
0.587
0.261
0.445
0.591
0.568
SO, as S
(SO2* 0.5011)
0.173
0.106
0.151
0.208
0.319
0.294
0.131
0.223
0.296
0.284
so;
0.317
0.469
0.215
0.4564
0.3122
0.413
0.367
0.427
0.537
0.205
2.
SO4 asS
(SO2/* 0.3338)
0.106
0.157
0.072
0.152
0.104
0.138
0.123
0.143
0.179
0.069
Total
DRYDEP'
(SOj+SO^)
0.279
0.263
0.223
0.360
0.423
0.432
0.254
0.366
0.475
0.353
wetdep
so;-
0.682
0.725
0.702
0.685
0.9! 5
0.581
1.094
1.620
1.115
0.635
2.370
0.838
0.716
0.306
2_
SO^ as S
(SO;' * 0.3338)
0.22S
0.242
0.234
0.229
0.305
0.194
0.365
0.541
0.372
0.212
0.791
0.280
0.239
0.102
Total
S
4.883
16.511
6.542
6.208
4,382
2.982
'Total Sulfur Dry Deposition = <{SOj * 0.501 1) + (SO* * 0.3338))
Total N Deoosition kg/ha/month
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995
Sep !995
Jun 1996
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jul 1997
Aug 1997
Sep 1997
Jun 1998
Jul 1998
Aug 1998
Sep 1998
Jun 1999
Jul 1999
Aug 1999
Sep 1999
clddep
NOj as N
1.099
1.806
1.037
1.020
2.558
1.550
1.555
5.174
2.948
3.109
2.844
2.056
1.306
4.034
1.070
1.218
0.897
HNO3
3.596
3.483
2.708
3.600
3.694
3.188
2.996
3.926
3.721
2.808
HNOjasN
(HNOj* 0.2223)
0.799
0.774
0.602
0.800
0.821
0.709
0.666
0.873
0.827
0.624
1 Total Nitrogen Dry Deposition = ((HNO3 * 0.2223) + (NO3'
dryde
NO;
0.011
0.004
0.006
0.014
0.012
0.012
0.005
0.004
0.004
0.014
• 0.2260))
p
NO3" as N
NOJ* 0.2260
0.002
0.001
0.001
0.003
0.003
0.003
0.001
0.001
0.001
0.003
Total
DRYDEP'
(HNOj + NOJ)
0.802
0.775
0.604
0.803
0.824
0.711
0.667
0.873
0.828
0.627
wetdep
NO;
0.387
0.332
0.352
0.543
0.557
0.255
0.584
0.786
0.532
0.363
1.287
0.409
0.334
0,097
NQ," as N
(NO,'* 0.2260)
0.087
0.075
0.079
0.123
0.126
0.058
0.132
0.178
0.120
0.082
0.29!
0.092
0.075
0.022
Total
N
2.479
6.075
3.609
2.980
2.212
L631
p/ecm/p/castnet2/wa02/fn lsunVtables_fnlsumrep.doc
-------
MADPro Final Report
Table 5-2. Summary of Cloud, Precipitation, Dry, and Total Deposition Estimates for Whitetop Mountain (page 2 of 2)
H* Deposition fkg/ha/month)
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995
Sep 1995
Jun 1996
Jul 1996
Aug 1996
Sep 1996
Jan 1997
Jul 1997
Aug 1997
Sep 1997
Jun 1998
Jul 1998
Aug 1998
Sep 1998
Jun 1999
Jui 1999
Aug 1999
Sep 1999
clddep
H*
0.168
0.226
0.132
0.102
0.288
0.172
0.202
0.806
0.363
0.349
0.394
0.258
0.162
0.504
0.167
0.208
0.144
drydep
IT from HNO3
(HN03* 0.0160)
0.0575
0.0557
0.0433
0.0576
0.0591
0.0510
0,0479
0.0628
0.0595
0.0449
H* from SOj
(SOj* 0.0157)
0.0054
0.0033
0.0048
0.0065
0.0100
0.0092
0.0041
0.0070
0.0093
0.0089
Total DRYDEP
(HNOj+SOj)
0.0630
0.0590
0.0481
0.0641
0.0691
0.0602
0.0520
0.0698
0.0688
0.0538
wetdep
i
IT
0.014
0.015
0.014
0.017
0.021
0.015
0.022
0.032
0.022
0.014
0.046
0.018
0.015
0.005
Total
H*
0.282
0.886
0.426
0.344
0.245
0.209
1 Total H* Dry Deposition = «HNO3 * 0.0160) + (SO2 * 0.0157)) as documented by Miller (1993a)
NH as IV Deposition (kc/hfl/nioitth^
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995
Sep 1995
Jun 1996
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jul 1997
Aug 1997
Sep 1997
Jun 1998
Ju! 1998
Aug 1998
Sep 1998
Jun 1999
Jul 1999
Aug 1999
Sep 1999
clddep
NHlasN
4
1.591
2.858
1.48!
1.268
2.644
1.684
1.739
5.155
3.976
4.027
4.354
2.698
1.983
4.353
1.64!
1.790
1.015
drydep
NH*
4
0.064
0.077
0.039
0.092
0.069
0.108
0.097
0.075
0.116
0.057
NH^asN
(NB£ * 0.7765)
0.050
0.060
0.030
0.071
0.053
0.084
0.075
0.059
0.090
0.044
wetdep
NH*
4
0.126
0.286
0.082
0.000
0.000
0.039
0.140
0.194
0.131
0.084
0.393
0.101
0.099
0.049
NH'asN
(NH* * 0.7765)
0.098
0.222
0.064
0.000
0.000
0.030
0.109
0.151
0.102
0.066
0.305
0.079
0.077
0.038
TOTAL
NH*asN
1.789
5.215
4.036
2.871
2.102
1.136
Source: Harding ESE.
p/ecin/p/castnet2/wa02/fnisum/tables_fnlsumrep.doc
-------
MADPm Final Rtport
Table 5-3. Summary of Cloud, Precipitation, Dry, and Total Deposition Estimates for Clingman's Dome (page 1 of 2)
TotalS Deposition (kg/ha/month)
clddep
SO4 asS SO2 (SOj* 0.5011)
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995 3.117
Sep 1995
Jun 19%
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jul 1997 4.717
Aug 1997 4.725
Sep 1997 3.705
Jun 1998
Jul 1998 7.877 0.248 0.124
Aug 1998 0.306 0.153
Sep 1998
Jun 1999
Jul 1999 0.155 0.078
Aug 1999 0.302 O.J5I
Sep 1999 0.422 0.211
drydep
cr»z'«« Totai
SO* S,°'aSS DRYDEP' <
* (S04* 0.3338) (50,+ so,-)
3.994
2.388
3.659
1.127
4.600
2.528
2.855
1.767
2.149
1.763
4.988
2.756
2.365
4.04!
3.169
1.024
1.074
0.277 0.092 0.217 2.555
0.426 0.142 0.296 4.298
0.839
1.899
0.301 0.101 0.178 1.183
0.421 0.140 0.292
0.278 0.093 0.304
wetdep
SO| as S
(SO^' " 0.3338)
1.333
0.797
1.221
0.376
1.536
0.844
0.953
0.590
0.717
0.588
1.665
0.920
0.789
1.349
1.058
0.342
0.358
0.853
i.435
0.280
0.634
0.395
Total
S
8.946
' Total Sulfur Dry Deposition = ({SO2 * 05011) + (SO4" * 0.3338))
1 wetdep data taken from the NADP/NTN site at Mt Mitchell, NC
Total N Deoosition kfi/ha/month
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995
Sep 1995
Jun 1996
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jul 1997
Aug 1997
Sep 1997
Jun 1998
Jul 1998
Aug 1998
Sep 1998
Jun 1999
Jul 1999
Aug 1999
Sep 1999
clddep
NO,' as N
drvdep
HNOj
HNOj as N
(HNO3* 0.2223)
NO;
NQJasN
NO, * 0 .2260
Total
DRYDEP1
(HNOj + NOJ)
wetdep
NO;
NO3" as N
(NO,* 0.2260)
Total
N
1.118
1.555
1.889
1.022
3.007
2.555
2.324
2.367
2.930
2.898
0.568
0.517
0.526
0.651
0.644
0.007
0.007
0.002
0.003
0.010
0.002
0.002
0.001
0.001
0.002
0.570
0.518
0.527
0.652
0.646
1.905
1.321
1.531
0,609
2.731
1-.368
1.620
0.85!
1.374
0.927
2.122
1.590
1.495
1.882
1.201
0.621
0.874
1.193
1.715
0.346
1.091
0.552
0.431
0.298
0.346
0.138
0.617
0.309
0.366
0.192
0.310
0.209
0.480
0.359
0.338
0.425
0.271
0.140
0.197
0.270
0.387
0.078
0.246
0.125
3.846
1 Total Nitrogen Dry Deposition = ((HNOj * 0.2223) + (NQ," * 0.2260))
2 wetdep data taken from the NADP/NTN site at Mt. Mitchell, NC
p/ecm/p/castnet2/wa02/fn lsunVtables_fnisumrep.doc
-------
MADPro Final Report
Table 5-3. Summary of Cloud, Precipitation, Dry, and Total Deposition Estimates for Clingman's Dome (page 2 of 2)
H* Deposition (kg/ha/monthl
clddep
„. H* from HNOj
n (HNOj* 0.0160)
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995 0.135
Sep 1995
Jun 1996
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jnl 1997 0.230
Aug 1997 0.237
Sep 1997 0.179
Jun 199S
Jul 1998 0.456 0.0409
Aug 1998 0.0372
Sep .1998
Jun 1999
Jul 1999 0.0379
Aug 1999 0.0469
Sep 1999 0.0464
drydep wetdep
H* from SO: Total DRYDEP'
(SO2 * 0.0157) (HNOj +• SOj)
0.082
0.055
0.076
0.027
0.106
0.052
0.070
0.039
0.049
0.038
0.110
0.063
0.054
0.078
0.055
0.025
0.018
0.0039 0.0448 0.055
0.0048 0.0420 0.083
0.014
0.038
0.0024 0.0403 0.025
0.0047 0.0516
0.0066 0.0530
Total
IT
0.555
1 Total H* Dry Deposition = ((HNOj * 0.0160) + (SO2 * 0.0157)) as documented by Miller (1993a)
2 wetdep data taken from the NADP/NTN site at Mt. Mitchell, NC
clddep
NH*asN NH
Jun 1994
Jul 1994
Aug 1994
Sep 1994
Jun 1995
Jul 1995
Aug 1995 1.301
Sep 1995
Jun 19%
Jul 1996
Aug 1996
Sep 1996
Jun 1997
Jul 1997 2.014
Aug 1997 2.361
Sep 1997 1.584
Jun 1998
drydep wetdep Total
NrfasN
« NH.
' (NH4 * 0.776S) '
0.451
0.181
0.299
0.130
0.647
0.332
0.270
0.208
0.313
0.226
0.496
0.339
0.261
0.471
0.450
0.084
0.164
Jul 1998 5.920 0.0603 0.047 0.308
Aug 1998 0.0839 0.065 0.503
Sep 1998
Jun 1999
0.125
0.218
Jul 1999 0.0512 0.040 0.114
NH.asN
, 4 NH. as N
(NH, • 0.7765) '
0.350
0.140
0.232
0.101
0.502
0.257
0.210
0.16!
0.243
0.175
0.385
0.263
0.203
0.365
0.349
0.065
0.127
0.239 6.206
0.391
0.097
0.169
0.088
Aug 1999 0.0875 0.068
Sep 1999 0.0751 0.058
1 wetdep data taken from the NADP/NTN site at ML Mitchell, NC
Source: Harding ESE.
p/ecm/p/castnet2Ava02/fnlsum/iables_fnlsu rnrep.doc
-------
MADPro Final Repon
Table 5-4. Percent Composition of Total Deposition at the Three MADPro Sites
—
Whiteface Mountain Cloud Deposition
Wet Deposition
Dry Deposition
Whitetop Mountain Cloud Deposition
Wet Deposition
Dry Deposition
Clingman's Dome Cloud Deposition
Wet Deposition
Dry Deposition
Sulfur
90.95
7.58
1.47
89.27
3.96
6.78
81.14
14.54
4.32
Nitrogen
83.67
8.17
8.15
68.77
2.96
28.27
66.13
11.43
22.44
NH*
NA
NA
NA
95.37
1.80
2.84
90.42
7.97
1.91
H+
87.61
7.90
4.49
80.87
3.94
15.19
70.91
13.29
15.97
Note: Sulfur deposition includes SO: and/or S04".
Nitrogen deposition includes HNO3 and/or NOj.
NH* deposition is presented in terms of nitrogen (N).
Source: Harding ESE.
p/ecm/p/casinet2/wa02/fnlsum/tables_fn!sumrep.doc
-------
MADPm Final Reporl
Table 5-5. Dry, Wet, and Total Seasonal Depositions (June through September) for Whiteface Mountain and Two Nearby
CASTNet Sites for 1995 through 1998
Dry 1995
1996
1997
1998
Wet 1995
1996
1997
1998
Total 1995
1996
1997
1998
Sulfur
WFM300
0.349
0.321
0.765
0.541
1.945
1.489
2.328
3.588
35.180
26.770
34.060
29.990
WST109
0.136
0.116
0.162
0.119
2.762
2.822
2.996
2.944
2.898
2.938
3.158
3.063
CTH110
1.202
1.074
1.402
1.608
3.325
3.765
2.857
3.398
4.527
2.840
4.259
5.006
Nitrogen
WFM300
1.033
0.713
1.837
1.255
1.001
0.796
0.919
1.421
15.360
10.540
15.210
12.390
WST109
0.155
0.121
0.143
0.091
1.250
1.234
1.335
1.297
1.405
1.355
1.478
1.387
CTH110
1.235
0.971
1.086
0.930
1.594
1.615
1.249
1.490
2.828
2.586
2.335
2.420
H+
WFM300
0.080
0.057
0.147
0.101
0.125
0.103
0.143
0.203
2.230
1.230
1.720
1.750
WST109
0.013
0.031
0.013
0.009
0.196
0.188
0.196
0.189
0.942
0.840
0.903
0.932
CTH110
0.116
0.095
0.113
0.108
0.203
0.231
0.154
0.200
1.490
1.484
1.229
1.301
Notes: WST109 is Woodstock, NH.
CTH110 is Connecticut Hill, NY.
All measurements are in kg/ha.
Source: Harding ESE, 1999.
p/ecm/p/castnet2/wa02/fnlsum/tables_fnlsumrcp.doc
-------
MADPni Final Repnrl
Table 5-6. Dry, Wet, and Total Seasonal Depositions (June through September)
for Whitetop Mountain and Two Nearby CASTNet Sites for 1996
Dry Sulfur
Dry Nitrogen
Dry NH+4
DryH+
Wet Sulfur
Wet Nitrogen
Wet NH*
WetH*
Total Sulfur
Total Nitrogen
Total NH*
Total H+
WTM302
1.020
2.908
0.187
0.227
0.971
0.409
0.040
0.071
37.250
16.220
14.720
2.130
PNF126
0.980
0.942
0.201
0.085
1.772
0.655
0.567
0.124
2.752
1.597
0.768
0.209
VPI120
1.319
1.312
0.196
0.122
3.475
1.218
0.799
0.204
4.794
2.530
0.995
0.326
Note: PNF126 is Pisgnah National Forest, NC
VPI120 is Morton Station, VA
All measurements are in kg/ha.
Source: Harding ESE.
p/ecm/p/castnet2/wa02/fn]sum/iables_fnlsumrep.doc
_
-------
-------
MADPro Final Report
Table 6-1. Warm Season Average Ion Concentrations for the Six MCCP Sites 1986 through 1988
Site
Mt. Mitchell, NC
Whitetop, VA
Whiteface 1,NY*
Whiteface 2, NY**
Shenandoah, VA
MoosiJauke, NH
n
624
656
987
120
230
328
PH
3.4
3.76
3.77
3.59
3.77
3.58
NOj
174
144
73
92
94
132
cr
31
16
5
7
13
16
NH*
184
152
97
157
93
107
sol
489
321
205
352
176
257
H+
398
174
171
255
171
263
All measurements are in ueq/L
* 1,485-m elevation.
** 1,250-m elevation.
Table 6-2. Comparison (RPD) of MCCP versus MADPro Average Ion Concentrations
Site
Whiteface Mountain
Whitetop Mountain
Clingman's Dome
SO*
33.94
19.99
-9.31
NO;
28.31
19.27
4.66
NH;
26.71
34.67
25.22
H*
27.14
52.43
-10.37
Note: RPD = relative percent difference based on MADPro concentrations
p/ecrn/p/castnet2/wa02/fnlsum/tab)es_fnlsumrcp.doc
-------
MADPm Final Report
Table 6-3. Mean Chemical Composition Including Minima and Maxima of Cloudwater Collected at Mt. Brocken,
Germany**
Year
1992
1993
1994
n*
37
Min
Max
1054
Min
Max
847
Min
Max
sof
331
41
1194
265
4
4169
342
1
3922
NO;
387
21
1071
280
0.5
5946
408
0.5
3803
cr
127
6
579
68
0.3
2401
126
0.3
2065
TIC
845
613
876
NH*
391
9
1007
410
2
8083
545
2
6815
Na*
156
< 1
854
60
2
2444
124
2
2765
Ca2+
147
10
1311
54
3
2245
141
3
4859
Mg2+
40
1
212
26
5
693
40
5
679
H+
39
-------
1
MADPrit Final Report
Table 6-4. Statistical Characterization of the Chemical Composition of Fog Samples from European Investigations
Location and Investigator
Units
Waldstein arithmatic mean
(Wrzesinski & Klemm 2000)
median
mean derived from the logarithm
minimum
maximum
standard deviation
Easterly
mean derived from the logarithm
Westerly
mean derived from the logarithm
Confidence level
Wulfersreuth2 arithmatic mean
(Trautner 1988)
minimum
maximum
OchsenkopF arithmatic mean
(Hcnerich 1987)
minimum
maximum
Ochsenkopf minimum
(Verhocven et al. 1987)
Landeshupct* minimum
{Schrimpftctal. 1984)
maximum
Rigi Siaffcl
(Collet! ei. al.. 1993)
Number of samples
arithmatic mean
standard deviation
volume weighted average
minimum
maximum
Seeboden
Number of samples
arithmatic mean
standard deviation
volume weighted average
minimum
maximum
Cond.
nS cm-1
151
142
121
17
452
95
189
100
>99%
526
58
2500
255
49
900
175
69
1230
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
PH
4,3
4.3
4.3
3.3
5.7
0.5
4.0(c)
4.4(c)
>99%
3.1
2.4
4.9
3.5
2.9
4.3
3.7
2.5
4.2
38
4.63
1.05
5.05
6.88
2.95
20
5.28
0.96
5.22
6.76
3.78
H*
Na+
K*
^eq I/1
89
52
51
2
513
91
101
41
>99%
n.d.
n.d.
n.d.
314
45
1!90
224
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
65
30
39
<9
664
124
(a)
(a)
(a)
92
12
774
66
5
213
86
60
970
40
34
75
23
0
400
23
20
29
18
0
120
11.5
9.7
13.8
<6.7
68.5
10.5
(a)
(a)
(a)
80.2
9
246
28
4
80
37
n.d.
n.d.
40
8
7
6
0
26
22
15
18
16
0
76
NH;
Mg2<
Ca2*
cr
NO;
so;-
669
547
547
<21
2580
498
788
463
>99%
1370
87
5420
922
125
2820
395
n.d.
n.d.
40
660
930
460
46
4600
23
920
1680
1100
150
6600
19.5
10
14.8
<4.6
152
29.6
14
15.7
(b)
73.9
15.7
584
22
3
76
57
n.d.
n.d.
40
12
19
8
1
88
23
21
64
24
0
310
69
40
46
<11
493
84
74
40
>99.9%
375
51
3350
110
9
376
151
50
2530
40
46
78
32
2
330
23
88
240
98
6
1100
54
31
46
< 13
389
74
(a)
(a)
(a)
138
43
660
96
8
374
76
60
1510
[. 40
53
99
35
1
540
23
77
120
88
2
440
481
409
342
20
1740
370
596
268
>99%
708
127
5600
599
83
2610
376
80
3000
40
490
800
310
14
3700
23
440
880
520
19
3700
497
421
376
55
1800
360
690(c)
395(c)
>99%
1440
352
9040
669
83
2380
415
310
2890
40
300
390
213
24
1800
23
380
630
430
37
2500
250"03'54"N, 11"45'54"E. 650-680 nn a.s.l.
J50°01107"N. 11"4873"E. 1024 m a.s.l.
*50°22.5K ! I"38.5-E. 600-640 m a.s.l
(a) Data are not normally or lognormally distributed. No calculations were performed.
(b) no significant difference.
(c) arithmetic mean
Notes:
< = below the given determination limit
* = standard deviation, confidence level for Siudem's t-test for different means
n.d. = no data
p/eem/p/castnet2/wa02/fn lsunVtables_fn I s u mrep .doc
-------
MADPro Final Repon
Table 6-5. Comparison of MADPro Cloudwater Deposition Estimates to Previous Studies
Site
Moosilauke
Whiteface
Mt. Mitchell
Whitetop (B)
Moosilauke (L) *
Whiteface (M) *
CLD
WFML
WTM
Ref.
1
1
1
1
2
3
4
4
4
Water
1.7
2.3
4.9
11.5
7
12.8
8.1
19.1
8.3
H+
0.06
0.03
0.19
0.37
0.2
0.15
0.23
0.35
0.22
NH+4
0.44
0.39
I.I
3.3
1.4
1.6
3
4.9
2.8
sol
2.7
2.2
8.4
20.1
11.5
7.8
14.3
20.7
13.2
NO;
2
0.8
4
9.9
8.5
4.8
7.7
11.9
7.5
* Moosilauke (L) and Whiteface (M) arc annual averages; MCCP data and this study are warm-season only.
Note: Water deposition rates are in cm/month, ion deposition rates are in kg/ha/month.
Results from MADPro are in bold.
References: 1 = Mohnen el al. (1990) (MCCP project report).
2 = Lovett
-------
MADPra Final Report
Table 6-6. A Summary of Cloudwater Chemical Deposition via Droplet Interception for the Eastern United States
Author
(Reference)
Lovett et at.
(1982, 1983)
Mueller &
Weatherfbrd
(1988)
Lindberg et at.
(1988)
Sigmon et al.
(1989)
Saxena el al.
(1989)
Stogner &
Saxena (1988)
Dasch(1989)
Mohnen (1988)
Miohnen et al.
(1990)
Mohnen (1988)
Mohnen (1988)
Mohnen et al,
(1990)
Lindberg &
Johnson (1988)
Mohnen (1988)
Mohnen (1988)
Mohnen el al.
(1990)
Mohnen (1988)
Mohnen (1988)
Mohnen et al.
(1990)
Mohnen (1988)
Lindberg &
Johnson (1988)
Site
Moosilauke, NH
Whitetop, VA
Great Smokey Mt., NC
Shenandoah, VA
Mitchell, NC
Mitchell, NC
Mitchell, NC
Moosilauke, NH
Moosilauke, NH
Whiteface, NY
Whiteface, NY
Whiteface, NY
Whiteface, NY
Whitetop, VA
Whitetop, VA
Whitetop, VA
Mitchell NC
Mitchell, NC
Mitchell, NC
Shenandoah, VA
Great Smokey Mt., NC
Elevation
(meters)
1220
1686
1740
1014
2000
2000
1987
1000
1000
1483
1200
1250
1000
1686
1686
1686
2000
2000
2000
1014
1740
Year
1980-81
1986
1986
1987
1986
1986
1986
1987
1987-88
1987
1987
1986-88
1986-88
1987
1987
1986-88
1987
1987
1986-88
1987
1986-88
Cloud droplet deposition
(kg ha"' month'1)
H+
0.2
0.14
0.02
0.17
0.12
0.03
0.04
0.06
0.18
0.05
0.03
0.04
0.22
0.2
0.22
0.09
0.13
0.19
0.04
0.05
NH*
1.4
1
0.2
1.2
0.8
0.3
0.24
0.44
2.3
0.8
0.39
0.5
2.6
11.8
2
0.1
1.1
1.1
0.3
0.5
NO;
8.5
4.1
2.2
0.7
3.2
3.2
0.9
0.8
2
4.8
3
0.8
1
7.7
10.4
6.1
2.6
2.2
4
1.3
1
so:
11.5
7.2
7.2
0.9
9.8
7.3
2.2
1.7
2.7
11.2
7
2.2
2
13.8
13.1
12.1
5.6
8.9
8.4
1.3
4
H2O
(cm yr"')
68
-
13
35-77
-
18
18
20
127
-
28
13
-
-
90
-
-
59
-
37
Frequency
of cloud (%)
40
25
7
25.52
-
62
23
19
42
27
24
-
31
31
30
27
27
29
7
-
Source: Vong el al., 1991.
p/ecm/p/castnet2/wa02/fnlsum/tables_fnlsunuep.doc
-------
MADPro Final Report
Table 6-7. Comparison of the Proportion of Total Ion Deposition Delivered by the Dry, Cloudwater, and
Precipitation Deposition Estimates for Whiteface Mountain at an Elevation of 1,050 m
Ion
sof
N03
HH*
H*
Percent of total deposition
Precipitation
45
38
41
46
Cloudwater
46
42
58
36
Dry deposition
9
20
1
17
p/ecm/p/castnet2/wa02/fnlsum/tables_fn]sunirep.doc
-------
Figures
-------
-------
MADPro Final Report
WFM300
SLD301
HTJN301
WTM302
CLD303
Figure 2-1. Locations of Mountain Acid Deposition Sites
Source: Harding ESE.
US-
Whiteface Mountain
Slide Mountain
Hunter Mountain
Whitetop Mountain
Clingman's Dome
Libraiy
p/castnet2/wa02/fnlsum/figs-3.p65
Mail code 3201
'ennsy!van(aAve1,w
Washington DC 20460
Harding ESE, inc.
-------
MADPro Final Report
i-X*
•>, //
C~ r»-ftwi XammrjfltaHi ^ C- ,c--
^ // / c X14\x"
LEGEbJD
('2) Slate Route
e Ceo Feature
o Town, Small City
- Hill
. Couoty Boundary
Street, Road
Major Street/Road
Variable Trails
State Route
River
LandMass
| I Open Water
Scale 1:62,500 (at center)
i 1 Milge |
Wbite&ce Mtn., NY
Mag 13.00
ThuJiinOl 13:26:231995
Figure 2-2. Regional Map for Whiteface Mountain, NY
Source: Harding ESE.
p/castnet2/wa02/fnlsum/figs-3.p6S
Harding ESE, Inc.
-------
MADPro Final Report
\ I X
•, "X 7" .
"""», > ^
\ (.. 7v
LEGEND
„ Geo Feature
a Town, Small City
A Hill
.'•—'•• US Highway
- County Boundary
Street, Road
Major Street/Road
. US Highway
, , , Railroad
River
Intermittent River
i in Contours
Scale 1:31,250 (at center)
h
I nflO Meters
Wbitelop Mtn, VA
Mag 14.00
ThuJunOl 13:52:11 199S
Figure 2-3. Regional Map for Whitetop Mountain, VA
Source: Harding ESE.
p/castnet2/waQ2/fhlsum/figs-3.p65
Harding £S£. inc.
-------
MADPro Final Report
*,v,\ ''\ "/ *'\€
-------
MADPro Final Report
Cloud Collector
Spray Nozzle
Collector Funnel
3-Way Pinch
Solenoid
•Rinse Pump
Waste
Water
Drain
Pinch Solenoid
;BB3B»3»**5
Panicle Volume
Monitor
Rain Detector
ODESSA 3260
DATALOGGER
Serial Printer
LCD Display
Micro-Controller
Input-Output Opto Relays
Distributer Controller
Power Supply
Figure 2-5. MADPro Sampling System
Source: Harding ESE.
p/castnet2/wa02/fn)sum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
Packing and Shipping
. of Supplies for
Site Operators
01 Wash
Precipitation Buckets
Sample <10 g:
Discard
1
Sample 1
240gto<100g
i
r
Prioritize Analysis
Site Operators Ship
Sample to
laboratory
Set Up Field Groups
in CLASS
Sample Receipt:
Visual Inspection;
Assign Sample No.
Log Sample Into
CUSS Data System
Labels, Logsheets Given
to
Laboratory Technicians
See Precipitation
Sample Report form
and Rain Gauge Flowchart
(in Laboratory Operations
Manuan
Review
Laboratory Sample
Volume or Weight
J
Samples 2.100 g
Laboratory pH
and Conductivity
Filtration
of Sample
Dissolved
pH Conductivity
Store at fC
Until Analysis
Ion Chromalography
TRAACS®
NH;
ICAP-AE
Ca»*. Mg8*, Ma2*. K*
Acidity
Autotitrator
(precipitation samples only]
Laboratory pH and
Conductivity
Discard
Data Transferred
into CLASS:
Data Batches Created
QC Review,
MIST Recovery.
DrUt Checks,
Midranga Standards
QCFalls;
Meeds Explanation
or Rerun Samples
QC Passes;
Data Transmitted
toDMC
Data Finalized
forSubmtttal
to EPA
voiarcastnet_prop98ffigZ7.A[
Figure 2-6. Flowchart of Laboratory Operations for Cloudwater and Precipitation Sample Analyses
Source: Harding ESE.
p/castnet2/wa02/fnlsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
0
0.2
0.4
0.6
PVM1
Figure 2-7. PVM-100 Intel-comparison at Whitetop Mountain, VA, 1998
Source: Harding ESE.
p/castnet2/wa02/fnlsum/figs-3 .p65
Harding ESE, Inc.
-------
MADPro Final Report
June
August
September
Total
Figure 3-1. Monthly Cloud Frequency (1994 through 1999), Whiteface Mountain, NY
Source: Harding ESE.
July
August Septerrfcer October
Total
Figure 3-2, Monthly Cloud Frequency (1994 through 1999), Whitetop Mountain, VA
Source: Harding ESE.
H1994
01995
H 1 996
D1997
Hi 1998
ED 199 9
p/castnet2/wa02/fhIsum/figs-3. p65
Harding ESE, Inc.
-------
MADPro Final Report
60
50
40
30
0
sf
20
June
July
August
September
October
Total
Figure 3-3. Monthly Cloud Frequency (1994 through 1999), Clingman's Dome, TN
Source: Harding ESE.
0.1
0.0
August
September
Figure 3-4. Mean Liquid Water Content of Clouds (1994 through 1999), Whiteface Mountain, NY
Source: Harding ESE.
p/castnet2/wa02/fnlsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
June
July
August
September
October
Figure 3-5. Mean Liquid Water Content of Clouds (1994 through 1999), Whitetop
Mountain, VA
Source: Harding ESE.
0.40
0.30
0.10
0.05
July
August
September
October
Tocal
• 1994
El 1995
(U1996
D1997
H 1998
01999
Total
Figure 3-6. Mean Liquid Water Content of Clouds (1994 through!999), Clingman's Dome, TN
Source: Harding ESE.
p/eastnet2/wa02/fljlsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
0.1
1994
Figure 3-7. Mean Liquid Water Content of Clouds
Source: Harding ESE.
1999
• WFM300
---A--- WTW302
p/castnet2/wa02/ftilsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
Whiteface Mt. Whitetop Mt. Clingman's Dome
Figure 3-8. Mean pH of Cloudwater Samples at MADPro Sites (1994 through 1999)
Source: Harding ESE.
300
2.6 2.8
3.2 3.4 3.6 3.8
5.2 5.4 5.6 5.6
4.2 4.4 4.6 4.8
PH
Figure 3-9. Frequency Distribution for Cloudwater pH at Whiteface Mountain, NY
(1994 through 1999)
Source: Harding ESE.
p/castnet2/wa02/fhlsum/figs-3 .p6 5
6.2 6.4 6.6
Harding ESE, Inc.
-------
MADPro Final Report
26 2.8 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.$ 5 5.2 S.4 5.6 5.9 6 6.2
Figure 3-10. Frequency Distribution for Cloudwater pH at Whitetop Mountain, VA
(1994 through 1999)
Source: Harding ESE.
2.3
3.2 3.4 3.6 3.8
4.2 4.4 4.6 4.8
pH
52 5.4 5.6 5,8 6 6.2 6.4
Figure 3-11. Frequency Distribution for Cloudwater pH at Clingman's Dome, TN
(1994 through 1999)
Source: Harding ESE.
p/castnet2/wa02/fhlsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
I
o
Z
1200
1000
4 5
Arrival Sector
Figure 3-12. Normalized SO* Concentrations in Cloudwater at Whitetop Mountain, 1994
through 1996 — Calculated as LWC Weighted Means versus Arrival Sector
I
f?
o
600
500
400
300
200--€
100
4 5
Arrival Sector
Figure 3-13. Normalized SO^ Concentrations in Cloudwater at Whiteface Mountain, 1994 through 1996 —
Calculated as LWC Weighted Means versus Arrival Sector
p/castnet2Ava02/fnisum/figs-3. p6S
Harding ESE, Inc.
-------
MADPro Final Report
1995 Annual SO? Emissions
(thorium)
• 790.000 10 1.710.000
B 600,000 to 790,000
® 330,000 to 600,000
H 190,000 to 330,000
B 10,00010 150,000
Figure 3-14. Annual SO2 Emissions for 1995
p/castnet2/wa02/fhlsum/figs-3. p65
Harding ESE, Inc.
-------
MADPro Final Report
WFM300 Whiteface Mountain
WTM302 Whitetop Mountain
CLD303 Clingman's Dome
1994
1995
1996
Figure 3-15. Mean Normalized SO*" Concentrations (u.eq/L) Segregated by; Back Trajectory Arrival Sector
p/castnet2/wa02/fnlsum/figs-3 .p65
Harding ESE, Inc.
-------
MADPro Final Report
This page has been intentionally left blank.
p/castnet2/wa02/fn!sum/figs-3.p65
Harding ESE, Inc.
-------
-------
MADPro Final Report
7-
r
a
S/2C
Figi
Sour
7-
6
5-
4-
X
a.
3-
2-
1
0
6/9
Figu
Sourc
.
1 * f *
5 Jt
* **
i
* *
j
H
4 &
1
i r5r
i *
•t** ••
•
4*«
•
r*. 1 Lti
ji". 4? ^r •
*
•
•
*•
*
.1 .
• i :
4
* I
t*
?
185 6/12/95 7/2/93 7/22/95 6/11/95 8/31/95 9/20/95 10/10/95
ire 3-16. pH of Cloudwater Samples, Whiteface Mountain, NY (1995)
ce: Harding ESE.
_L| }i
'
'. ^
4
* 1
• -
.» t
w ,
8 *
i i
\
i
i
* ^
i
i
•*X AJ
•*^ 1"^
i
i i* ^
' # rti?
i
j^i:
»
IS7 S/29/97 7/19/97 8/8/97 8/28/97 9/17/97 10/7/97
re 3-17. pH of Cloudwater Samples, Whiteface Mountain, NY (1997)
e: Harding ESE.
p/castnet2/wa02/thlsum/figs-3.p65
Harding ESE. Inc.
-------
MADPro Final Report
6/8/99
6/28/99
7/18/99
BW99
KZ7/S9
Figure 3-18. pH of Cloudwater Sampies, Whiteface Mountain, NY (1999)
Source: Harding ESE.
•!
' •
: i i
* » :
i 'i
V i
: . 1
f 11 .» •
«/ f
i •
'I' *
t
•
• »
* «p :$
t • " t H
* V
1
9 i
5 S«^
* * *f t*
| t ** * * 5
Id o
H
•
«
t i
:^
i 2','
f 4>,
' 5"4
*• •
:5
•
*
f
•
*
: ?f
9/1689
10/6/99
p/castnet2/wa02/fhlsum/figs-3. p6 5
Harding ESE, Inc.
-------
MADPro Final Report
I »
±
i
4/24*5 5/14/95 6/3/95 6/23/95 7/13/95
S/22/95 9/11/9$ 10/1/95 ID/21/95 11/10/95
Figure 3-19. pH of Cloudwater Samples, Whitetop Mountain, VA (1995)
Source: Harding ESE.
6-
I
EL
1
0-
5/7
Fi
So
•
i
*
•*i
** * »
• i *& :
1 *
.' ytfijf
i
4
X ,t
1 -
J 'r
< *» i
* / l
<•
|
:
•
^
r* **
1 );
rit !.» i1 :
,! r^ *
1 # «'• t *«.
»•• t*
*
]
i
1
S7 5/27/97 6/16/97 7/6/97 7/26/97 8/15/97 9/4/97 9/24/97 10/14/97 U/3/97
gure 3-20. pH of Cloudwater Samples, Whitetop Mountain, VA (1997)
urce: Harding ESE.
p/castnet2/wa02/fhlsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
5 .
4 -
i
V
H;
? ii «J
$ «
|
r
ft *'
l*|l
' H
i
i
•
P
•
i
* ; ^
!
»*
A
»*
•
*
•
*
5/15/93 6/4/99 6/24/99 7/1 4/99 S/3/99 8/23/99 9/12/99 10/2/99
Figure 3-21. pH of Cloudwater Samples, Whitetop Mountain, VA (1999)
Source: Harding ESE.
p/castnet2/wa02/fhlsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Fatal Report
» 1
• *
*
• «*
«'*» «J
* :***
•t
' ** * i
* .
..
*:, * : i*
+
*
.' ' t
»
*
*
6/28/95 7/1W95 8/7/95 S/27/9S 9/16/95 10/6/95
Figure 3-22. pH of Cloudwater Samples, Clingman's Dome, TN (1995)
Source: Harding ESE.
7 •
I
3
2
1
0
5/1
Fi
So
•l
4*
1 » '•
«*'
4 .
* Tl
]
1 !,
lu i ^ *.
\ !| 1* 4*1
< .i
* * -
j i
!
I
«
!;
^ — ^ ^
J *L
5/97 5/30/97 6/19/97 7/9/97 7/29/97 8/18/97 9/7/97 9/27/97 10/17/97
gure 3-23. pH of Cloudwater Samples, Clingman's Dome, TN (1997)
urce: Harding ESE.
p/castnet2/wa02/fiilsum/figs-3 .p65
U.S. EPA Headquarters Library Harding ESE, inc.
Mai! code 3201
1200 Pennsylvania Avenue NW
Washington DC 20460
-------
MADPro Final Report
TT
"T^
« i
6/29/99
9/17/99
Figure 3-24. pH of Cloudwater Samples, Clingman's Dome, TN (1999)
Source: Harding ESE.
p/castnet2/wa02/fiilsufn/figs-3. p65
Harding ESE, Inc.
-------
MADPro final Report
500
450
01997
H1998
E31999
Whitetece Ml,
Whttatop Ml
Clingman's Dome
Figure 3-25. Mean Hydrogen Ion Concentrations of Cloudwater Samples at
MADPro Sites (1994 though 1999)
Source: Harding ESE.
500
450
—•—WFM300
.--A--- WTM302
--•--CLD303
50
1993
1994
1995
1996
1997
1998
1999
2000
Figure 3-26. Mean Hydrogen Ion Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Source: Harding ESE.
p/castnet2/wa02/fhlsum/figs-3.p65
Harding ESE. Inc.
-------
MADPm Final Report
500
400
300
200
100
April
-VT
—•—WFM300
---A--- W7M302
--•--CLD303
May
June
July
August September October November
Figure 3-27. Monthly Variation in H* Concentration in Cloudwater
(Means Across All Years, 1994 through 1999)
Source: Harding ESE.
p/castnet2/wa02/Msurn/figs-3.p65
Harding £5£, Inc.
-------
MADPro Final Report
600
300
200
100
Whiteface Mt
Whitelop M.
Clingman's Dome
Figure 3-28. Mean SO*" Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Source: Harding ESE.
H11994
E] 1995
m 1996
01997
H1998
H1999
p/castnet2/wa02/fn lsum/figs-3 .p65
Harding ESE. Inc.
-------
MADPro Final Report
soo
400
200
1993
1995
1996
1997
Figure 3-29. Mean SO*' Concentrations of Cloudwater Samples at MADPro
Sites (1994 through 1999)
Source: Harding ESE.
600
500 •
|
400
300
200
100
April
V* /
May June July August September October Noverrber
Figure 3-30. Monthly Variation in SO*' Concentrations in Cloudwater
(Means Across All Years, 1994 through 1999)
Source: Harding ESE.
—WFM300
.-W7M302
--CLD303
p/castnet2/wa02/fiilsum/figs-3. p65
Harding ESE, Inc.
-------
MADPro Final Report
Whiteface ML
WhitelOp Ml.
Clingman's Dome
Figure 3-31. Mean NOj Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Source: Harding ESE.
El 1994
O1995
01996
CM997
H1999
p/castnet2/wa02/fnlsum/figs-3.p6S
Harding ESE, Inc.
-------
MADPro Final Report
200
1993
1994
1996
1997
1998
2000
Figure 3-32. Mean NOj Concentrations of Cloudwater Samples at
MADPro Sites (1994 through 1999)
Source: Harding ESE.
2SO
200
150
100
April May June July August September October November
Figure 3-33. Monthly Variation in NO" Concentrations in Cloudwater
(Means Across All Years, 1994 through 1999)
Source: Harding ESE.
-•— WFM300
-A---WTM302
-•- - CLD303
p/castnet2/wa02/fiilsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
300
£50 -
I
Whitefece Ml
Whitelop M.
Clingman's Dome
Figure 3-34. Mean NH* Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Source: Harding ESE.
H1994
Q1995
01996
D1997
H1998
O1999
p/castnet2/wa02/fiilsum/figs-3 .p65
Harding ESE. Inc.
-------
MADPro Final Report
300
250
200
—•—WFM300
.--A---WTM302
--•--CLD303
1993
1994
199S
1996
1997
1998
1999
2000
Figure 3-35. Mean NH* Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Source: Harding ESE.
1994 1995 1996 1997 1998 1999 1994 1995 1996 1997 1998 1999 1994 1995 1996 1997 1998 1999
Whiteface Whitetop Clingman's Dome
Figure 3-36. Average Minor Ion Concentrations from 1994 to 1999
Source: Harding ESE.
p/castnet2/wa02/fiilsuni/figs-3 .p65
Harding ESE, Inc.
-------
MADPro Final Report
Whiteface M.
Whtetop ML Clingman's Corns
Figure 3-37. Normalized Mean Hydrogen Ion Concentrations of Cloudwater Samples
at MADPro Sites (1994 through 1999)
Source: Harding ESE.
Whiteface Ml
Whitetop ML
Clingman's Dome
• 1994
m 1995
m 1996
D 1997
• 1998
m 1999
Figure 3-38. Normalized Mean SOj- Concentrations of Cloudwater Samples at MADPro Sites
(1994 through 1999)
Source: Harding ESE.
p/castnet2/wa02/fh lsum/figs-3. p65
Harding ESE, Inc.
-------
MAD fro Final Report
Whiteface ML
Whitetap Ml
Clingman's Dome
Figure 3-39. Normalized Mean NO' Concentrations of Cloudwater Samples at
MADPro Sites (1994 through 1999)
Source: Harding ESE.
• 1994
m 1995
M 1996
D 1997
H 1998
H 1999
Whiteface W.
WtiitetopMt
Clingman's Dome
Figure 3-40. Normalized Mean NH* Concentrations of Cloudwater Samples at
MADPro Sites (1994 through 1999)
Source: Harding ESE.
p/castnet2/wa02/fiilsum/figs-3. p65
Harding ESE, Inc.
-------
MADPro Final Report
500
Figure 3-41. Concentrations for SLD/HUN3 01 (1995, 1997, 1998)
Source: Harding ESE.
p/castnet2/wa02/fhlsum/figs-3 .p65
Harding ESE, Inc.
-------
MADPro Final Report
600
|
500-
400-
300-
200-
100-
-Linear (H-)
Linear (NOj)
- - Linear (NH;)
—Linear (SO2,-)
0.1098x +370.28
R* = 0.0062
y = 0.1809x4- 246.43
R2 = 0.0258
y=*0.0441x
R2 = 0.0035
y = 0.0779X + 140.64
R*= 0.0206
994
1995
1993 1997
1998
1999
100 200 300 400 500 600 700
sample number with associated year
800
900
1000
Figure 3-42. CLD303 1994 through 1999 Regressions Using AH Data Points
Non-Normalized Concentrations
Source: Harding ESE.
180 •
160-
140-
1
| 120 -
I
3 100 •
1 80
60
40
20 -
0 -
Fig
Sou)
- -Linear (H-) - - - Unear (NH;> V = <"=* * 1 °7'17
— - Linear (NO") — — Unear (SOJ) R =°-0118
^ — ~~~~ " <-_ — **" y = 0.0724X + 73.591
^ «-•••"' Rz = 0.0309
^, -- • •"" y » 0.02X + 63.408
___—-••"' Rs = 0.0051
^- — ""
__.—•- — ""~" y = 0.0284X + 42.581
__,__. — - — •""" R2 = 0.0221
994
1995 1996 1997 1998 1999
III I I
0 100 200 300 400 500 600 700 800 900 1000
sample number with associated year
lire 3-43. CLD303 1994 through 1999 Regressions Using All Data Points -
Normalized Concentrations
rce: Harding ESE.
i
p/castnet2/wa02/fhlsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
400
350
300
~ 250
8 200
150-
100-
SO
• - - Linear (NH;)
•Linear (so;-)
-0.0318x-t-317.26
R2* 0.0037
y =-0.0185X +184.29
R2 = 0.0041
1994 1995
1996
1997
1998
1999
I
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800
sample number with associated year
Figure 3-44. WFM300 1994 through 1999 Regressions Using All Data Points -
Non-Normalized Concentrations
Source: Harding ESE.
250
200
150
100 •
so
- - - Linear (NH;> • Linear (SO;-)
y =-0.0168X1-167.75
R2 « 0.0033
= -0.0125x +100.99
R2 = 0.0063
1994 1995
.1996
,1997
1998
1999
0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800
sample number with associated year
Figure 3-45. WFM300 1994 through 1999 Regressions Using AH Data Points -
Normalized Concentrations
Source: Harding ESE.
p/castnet2/waQ2/fhlsum/figs-3.p65
Harding ESE. Inc.
-------
MADPro Final Report
200 -j
180-
160
140-
1 120-
| 100 •
§ 80-
60-
40-
20-
0-
•^Linear (so;-)
y = 0.012Sx + 95.461
Rz = 0.0027
" "
1994 1995 1996 1997 1993 1999
0 200 400 600 800 1000 1200 1400 1600 1800
sample number with associated year
Figure 3-46. WTM302 1994 through 1999 Regressions Using All Data Points
Normalized Concentrations
Source: Harding ESE.
p/castnet2/wa02/fhlsum/figs-3 ,p6S
Harding ESE, Inc.
-------
MADPro Final Report
200
150 -
100 -
50 -
0
0
0.2
0.4
LWC
0.6
Figure 4-1. Relationship Between Square of the Mean Droplet Diameter (D) and LWC for Clouds on
Whitetop Mountain
Note: Data are from Figure 2 of Joslin et al, 1990.
150
100
-250
o
1000
5000
2000 3000 4000
Total Ion Concentration (,ueq/L)
Figure 4-2. Relative Percent Difference Versus Total Anion Concentration for Each Sample
p/castnet2/wa02/fnlsum/figs-3.p65
Harding ESE. Inc.
-------
MADPro Final Report
60 T-
50 •
1 40
o
u
o
30 -
ID
i. 20
10 -i
o •i-
12
0 2 4 6 8 10
Canopy-top Wind Speed (m/sec)
Figure 4-3. Sensitivity of Deposition Velocity to Wind Speed
14
30 -
25 -
•S
a 20
u
-------
MADPro Final Report
slope o1 regression with 1-m
shift LAI profile +1m heigh!
shift LAI profie -1m heigh)
1.15
I 1'10
o
I 1.05
5 1.00
0.95
J 0.90
I
0.85
0
10
Model Layer Thickness (m)
Figure 4-5. Results of Numerical Experiments with Model Layer Thickness and Shifts in the Vertical
Distribution of Leaf Area.
Note: X-Axis is log scale. Y-axis represents the ratio of average model performance with a given
scenario versus average performance with 1-m layers and the CASTNet standard leaf area profile.
10 15
Wind Speed (ms-1)
20
25
Figure 4-6. Difference in Cloudwater Deposition Rate Between Model Runs Using the Whitetop Mountain
Cloud Droplet Size Distribution and the Whiteface Mountain Cloud Droplet Size Distribution
for a Range of LWC and Wind Speed
Note: Percent deviations are calculated with respect to the Whiteface Mountain distribution results.
p/castnet2/wa02/fhlsum/figs-3.p65
Harding ESE. Inc.
-------
MADPro Final Report
mmH2O:97cld-dep500.imp
0.9"
0.8 ~
0.7"
"
0.6 ~
0.5"
0.4"
0.3"
0.2 '
I
1
L
I
1
HS
"XK
0.1 -fS
0 0~"
3
1L,
S&Syyrf
•XXPPWI
iS&gffififtSa
;<';';'vl X B^^B
HHHH
-.
mm
*v
o
1
IQuantiles j
maximum
quartiie
median
quartiie
minimum
100.0% 0.80700
99.5% 0.80430
97.5% 0.53622
90.0% 0.38120
75.0% 0.26900
50.0% 0.15880
25.0% 0.10710
10.0% 0.06810
2.5% 0.04927
0.5% 0.01017
0.0% 0.00381
[Moments J
Mean
Std Dev
0.2029
0.1342
Std Error Mean 0.0082
Upper 95% Mean 0.2 1 90
Lower 95% Mean 0.1 869
N
271 .0000
Sum Weights 271 .0000
Sum
Variance
Skewness
Kurtosis
CV
54.9880
0.0180
1 .4532
2.7034
66.1362
[mmH20:97cld-dep20.csv j
iQuantiies 1
maximum 100.0%
99.5%
97.5%
90.0%
quartiie 75.0%
median 50.0%
quartiie 25.0%
10.0%
2.5%
0.5%
minimum 0.0%
[Moments j
Mean
Std Dev
Std Error Mean
Upper 95% Mean
Lower 95% Mean
N
Sum Weights
Sum
Variance
Skewness
Kurtosis
CV
0.73780
0.73445
0.51038
0.35372
0.24660
0.14150
0.09267 .
0.05942
0.03956
0.00759
0.00303
0.1830
0.1264
0.0077
0.1982
0.1679
271.0000
271.0000
49.6049
0.0160
1.4552
2.5110
69.0346
Figure 4-7. Frequency Distribution of Sample Water Deposition for Clingman's Dome 1997 Data Set
Calculated with Either 500 or 20 Droplet Size Classes Representing the Continuous Droplet
Size Distribution
p/castnet2/wa02/fhlsum/figs-3 .p65
Harding ESE, Inc.
-------
MADPro Final Report
PI-standard MCLOUD
mmH2O
.Quantiles
maximum 100.0%
99.5%
97.5%
90.0%
qua/tile
median
quaitile
1.8600
1.7742
1.3924
1.0935
75.0% 0.7736
50.0% 0.4547
25.0% 0.2407
10.0% 0.1395
2.5% 0.0726
0.5% 0.0211
0.0% 0.0187
Moments
Mean
SiaDev
Std Error Mean
Upper 95% Mean
lower 95% Mean
N
Sum Weighs
Sum
Variance
Slowness
KllltOSiS
cv
0.5447
0.3714
0.0231
0.5903
0.4992
258.0000
258.0000
140.5377
0.1379
0.9045
02106
68.1613
P2- Whitetop drop sizes
[lmmH20:Whitetop-drop-dist
P3- 3.2m ws ht
mmH2O:3.2m-ws-ht ]
2"
1"
L
i
PL,
Quantiles j
|
L :
maximum 100.0% 1.5510
99.5% 1.4725
37.5% 1.1299
90.0% 0.8703
quartile 75.0% 0.6063
median 50.0*
I, 0.3626
quartile 25.0% 0.1754
10.0*
1, 0.1058
2.5% 0.0509
0.5% 0.0153
minimum 0.0% 0.0140
Moments ]
Mean
StdDev
Std Error Mean
Upper 95% Mean
Lower 95% Mean
N
Sum Weights
Sum
Variance
Skewness
Kurtosis
CV
0.4260
0.3045
0.0190
0.4633
0.3886
258.0000
258.0000
109.9003
0.0927
0.9777
0.4188
71.4826
.Quantiles
maximum 100.0%
99.5%
97.5%
90.0%
quartile
median
1.9040
1.8232
1.4127
1.0903
0.7S33
quart ile
minimum
75.0%
50.0% 0.4359
25.0% 0.2361
10.0% 0.1341
2.5% 0.0712
0.5%
0.0%
0.0189
0.0178
.Moments
Mean
StdDev
Std Error Mean
Upper 95% Mean
Lower 95% Mean
N
Sum Weights
Sum
Variance
Skewness
Kurtosis
CV
0.5303
0.3737
0.0233
0.5761
0.4845
258.0000
258.0000
136.8074
0.1396
1.0033
0.4666
70.4663
P4-summit WS & LWC
ntntHSO-surnniit J
zH
1"
I
1
L
8
ffn
i
o
Quantiles ]
maximum
quaitile
mediar
1
quartile
minimum
.Moments
Mean
StdDev
100.0% 2.1530
99.5% 2.0589
97.5% 1.6373
90.0% 1.2918
75.0% 0.9242
50.0% 0.551S
25.0% 0.3009
10.0% 0.1736
2.5% 0.0942
0.5% 0.0264
0.0% 0.0232
D
0.653!
)
0.4328
Std Error Mean 0.0269
Upper 9S% Mean 0.7070
Lower 95% Mean 0.6009
N
256.000
Sum Weights 258.0000
Sum
Variance
Skewness
Kuttos
CV
s
168.7175
0.1873
0,8560
0.1027
66.1809
Figure 4-8. Comparison of Estimated Cloudwater Fluxes Using Four Potential Model Scenarios at White-
face Mountain, 1997
Note: PI - MCLOUD Model, P2 - using the Whitetop, VA droplet size distribution, P3 - using 3.2
meters as the height of the windspeed measurement above the canopy, P4 - applying (uncor-
rected) summit WS and LWC to the 1,350-m elevation site.
p/castnet2/wa02/fnlsum/figs-3 .p65
Harding ESE, Inc.
-------
MADPro Final Report
(LWC@1225m )
[Quantiles J
maximum 100.0%
99.5%
97.5%
90.0%
quartile 75.0%
median 50.0%
quartile 25.0%
10.0%
2.5%
0.5%
minimum 0.0%
Moments .
Mean
StdDev
Std Error Mean
Upper 95% Mean
Lower 95% Mean
N
Sum Weights
0.92X0
0.92300
0.91300
0.83300
0.60550
0.41800
0.26300
0.10400
0.03300
-0.0774
-0.087
0.4453
0.2512
0.0103
0.4656
0.4250
590.0000
590.0000
SWS@1225m
[Quantiles J
maximum 100.0%
99.5%
97.5%
90.0%
quartile 75.0%
median 50.0%
quartile 25.0%
10.0%
2.5%
0.5%
minimum 0.0%
Moments J
Mean
Std Dev
Std Error Mean
Upper 95% Mean
Lower 95% Mean
N
Sum Weights
17.678
17.068
15.098
12.278
9.466
7.028
5.058
3.618
1.578
-0.036
-0.542
7.4379
3.3411
0.1359
7.7049
7.1709
604.0000
604.0000
Figure 4-9. Frequency Distributions of LWC and Wind Speed Scaled to Values Representative of 1,225-m
Elevation on Whiteface Mountain
p/castnet2/wa02/fnlsum/figs-3, p65
Harding ESE, Inc.
-------
MADPro Final Report
Slope of Regression w/ STD 1 -m Layer Canopy
Jj KJ A o -» M W
»=•
' ' ' 1 ' '
a .
-~-=-~'s
— •— LWC=0.21 1
—• - LWC=0.42 1 ,
- •- - Lwc=o.as 1
~— - _
N
\
..._.-....«....._..
i
i
... j ...
' ' '
>S
N
\
\
>
,
am
2 4 6 8 10 12 14
Model Layer of Thickness (m)
Figure 4-10. Percent Deviation of Model Response to a Hypothetical Pure Balsam Fir Canopy from Model
Response to an Observed Average Canopy at 1,225-m Elevation on Whiteface Mountain over a
Range of Wind Speed and LWC Values.
Note: In both cases, the canopy height was 17m and LAI = 10 m2/m2.
0.50
0.0
-0.50
•1.0
-1.5
-2.0
-2.5
-3.0
-3.5
if
— •— LWC (0.833) SWS(12278)
• — LWC(0.418) SWS(12578)
- B- - LWC(0.263) SWS(12.27B)
we(.
— t- - LWC (0.263) SWS(7.028)
— jt- -Lwc(o.833) svvsiaeis)
— — LWC (0. 41 8) SVVS! 3. 6! fi)
~9~ LWC(0.263) SWS(a61 8)
8 10 12 14
Wind Speed (ms-1)
16
18
Figure 4-11. Percent Deviation of Model Response to Several Canopy Height Specifications from Model
Response to the Observed Height of 17 m at 1,225-m Elevation.
p/castnet2Ava02/fnlsum/figs-3. p6S
Harding ESE. Inc.
-------
MADPro Final Report
J
-8
10
/
- B- - LWC{0.2fl3) $WS(12.278}
• -*- • LWCftMIB) SWS(7028)
—H- - LWC(0.263) SWSJ7.028)
—*-- LWC{0.833) SWS(3.618)
•-UVC(0.41S}SWSf3.618)
-«•- LWC(0.263) SWS{3.616)
8 9 10 11 12 13
LAI (m'/m2)
Figure 4-12. Percent Deviation of Model Response to Variations in LAI with a
Constant Canopy Height of 10 m
.o
O
10
Wind Speed (ms-1)
Figure 4-13. Model Response (Hourly Water Flux) to Variation in Wind Speed for a Range of Cloud LWC
Note: Solid symbols with solid lines are model response using the Whiteface Mountain cloud droplet
size distribution (WF). Open symbols are model response using a droplet size distribution
representative of Whitetop Mountain (WT).
p/castnet2/wa02/fhlsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
300
Legend
800 meters and above
0 300 600 KMorretws
M>«s Code Equal-Area Projection
Figure 5-1. Areas in the Eastern United States with Elevations Above 800 Meters
p/castneC/wa02/fnlsum/figs-3.p65
Harding ESE, inc.
-------
MADPro Final Report
100%
80%
60%
40%
20%
ToWS Told do:
Whfteface Mt., NY
WhitetopMt.,VA
Clingman's Dome,
Figure 5-2. Percent Composition of Total Deposition Estimates at MADPro Sites
Source: Harding ESE.
@ Dry Deposition
D Wet Deposition
• Cloud Deposition
p/castnet2/waQ2/fnlsura/figs-3.p65
Harding ESE, Inc.
-------
Appendix
-------
-------
MADProFmal Report
1994
1999
Figure A-l. Mean Liquid Water Content of Clouds with Scaled 1998 WTM LWC
Source: Harding ESE.
p/castnet2/wa02/fnl sum/figs-3 .p65
Harding ESE, Inc.
-------
MADPro Final Report
Whiteface ML
WhitetopML
Figure A-2. Normalized Mean Hydrogen Ion Concentrations of Cloudwater Samples at
MADPro Sites (1994 through 1999) with scaled 1998 WTM LWC
Source: Harding ESE.
120
100
Whitelace ML
Whitetop ML
Clingman's Dome
Figure A-3. Normalized Mean NH* Concentrations of Cloudwater Samples at MADPro
Sites (1994 through 1999) with 1998 WTM LWC
Source: Harding ESE.
p/castnet2/wa02/fiilsum/figs-3.p65
Harding ESE, Inc.
-------
MADPro Final Report
Whiteface Mt
Whitetop Mt
Qingman's Dome
Figure A-4. Normalized Mean NO" Concentrations of Cloudwater Samples at MADPro
Sites (1994 through 1999) with 1998 WTM LWC
Source: Harding ESE.
200
180
1994
E3 1995
m 1996
D 1997
H 1998
m 1999
Whiteface Mt.
Whitetop Mt.
Clingman's Dome
Figure A-5. Normalized Mean SO*' Concentrations of Cloudwater Samples at MADPro
Sites (1994 through 1999) with 1998 WTM LWC
Source: Harding ESE.
p/castnet2/wa02/fhUum/fig$-3.p65
Harding ESE, inc.
-------
-------
|