"Great Lakes Total Phosphorus Models and Loads: A Fifteen Year Update"
Final Report, December, 2011
EPA Grant No. GL 00E58501
Submitted to
Jeffrey May
Project Officer
U.S. EPA Great Lakes National Program Office
and
U.S. EPA Region V
Chicago, IL
by
David M. Dolan
Principal Investigator
University of Wisconsin-Green Bay
Natural and Applied Sciences, ES-317
2420 Nicolet Dr.
Green Bay, Wl 54311
920-465-2986
doland@uweb.edu
and
Steven C. Chapra
Co-Principal Investigator
Tufts University
Civil & Environmental Engineering
Medford, MA 02155
617-627-3654
steven.chapra@tufts.edu

-------
Introduction and Background
This report summarizes the results of updated load estimates for total phosphorous (TP), total dissolved
phosphorus (TDP), nitrate, and chloride for Lakes Superior, Michigan, Huron and Ontario for the years
1994-2008. Modeling and data analysis methods were used to determine whether target loads for
phosphorus established by the Great Lakes Water Quality Agreement (GLWQA, Annex 3, 1978) were
being met. In addition, a mass-balance model was used to check the consistency of phosphorus loads
and to estimate interlake transport for each of the lakes (Michigan, Superior, Huron and Ontario.
The objectives for this project were to:
1.	Assess status of phosphorus (TP and TDP), nitrate, chloride, and hydrologic data needed for
Great Lakes load estimation for the period 1994-2008.
2.	Apply the experience gained from similar work on Lake Erie to the estimation of TP loads (and
associated uncertainty) to Lake Michigan for the fifteen-year period. These estimates also
include separate loading estimates for upper and lower Green Bay.
3.	Utilize estimation procedures developed for Lake Erie to calculate loads for TDP, chloride and
nitrate on the same basis as objective 2.
4.	Update the TP mass balance for Lake Michigan using components of the Lesht et al. (1991)
regional phosphorus model and Chapra's (2007) phosphorus budget model. Use the model to
assess the estimated phosphorus loads by comparison of the model-computed versus observed
in-lake TP concentrations.
5.	Collate data from the U.S. and Canada into common formats required for computation of loads
for Lakes Huron, Superior and Ontario. This task includes separate loading estimates for the
North Channel, Georgian Bay and Saginaw Bay.
Final Report, December, 2011
Page 2

-------
6.	Evaluate alternative load estimation techniques that may be able to utilize the wealth of historic
information to provide plausible estimates for current conditions. Use minimization of load
estimation uncertainty as a criterion for this evaluation.
7.	Make estimates for each component (tributary, point source and atmospheric) of total lake
loading including unmonitored areas on a watershed basis. Report the load estimation
uncertainty associated with each annual load to each lake.
8.	Update the TP mass balance for Lakes Huron, Superior, and Ontario using components of the
Lesht et al. (1991) regional phosphorus model and Chapra's (2007) phosphorus budget model.
Use the model to assess the estimated phosphorus loads by comparison of the model-computed
versus observed in-lake TP concentrations.
9.	Prepare tables and graphs of results showing subtotals on an annual-average basis for each
component of the regional model and source (point, nonpoint, etc.).
10.	Recommend the necessary level of monitoring for each lake to continue the assessment of
loading trends at a level of certainty necessary for use in models and load control programs.
Data Sources
All of the data used to estimate nutrients loads came from databases compiled by government agencies
in both the United States and Canada.
Tributary sources consisted of monitored and unmonitored tributary sub-basins; point sources included
direct and indirect industrial and municipal dischargers; and atmospheric sources refer to loadings from
the atmosphere that fall directly on the lake surface. Point source data was retrieved from the USEPA's
Permit Compliance System (PCS) and Canada's Municipal and Industrial Strategy for Abatement (MISA).
The USEPA PCS system has been replaced by a new system, Integrated Computer Information System
(ICIS), and some states began reporting to this new system beginning in 2006. The Wisconsin
Final Report, December, 2011
Page 3

-------
Department of Natural Resources (WDNR) and Michigan Department of Natural Resources and
Environment (DNRE) provided additional data on facilities permitted to discharge to surface waters in
the basins.
Tributary streamflow and nutrient concentration data were retrieved from the U.S. Geological Survey
(USGS) and Environment Canada (EC). The Minnesota Department of Natural Resources (MNDNR)
provided additional flow data when none was available from USGS.
Water quality concentration data for Lake Huron were retrieved from STORET (Legacy and Modern),
USGS, and the Provincial Water Quality Monitoring Network of Ontario.
Atmospheric deposition data was obtained from Environment Canada. Lake Michigan atmospheric
loads were estimated using the high spatial resolution 1994-1995 loads reported by Miller et al. (2000)
as part of the Lake Michigan Mass Balance (LMMB) study. No atmospheric sampling seems to have
occurred since 1995 in Lake Michigan, but monthly over-lake precipitation data from the National
Oceanic and Atmospheric Administration (NOAA) and chloride concentration data from the National
Atmospheric Deposition Program (NADP) (USGS, 2008) sampling stations were available for the study
period.
Results
Objective 1. Assess status of phosphorus (TP and TDP), nitrate, chloride, and hydrologic data needed for
Great Lakes load estimation for the period 1994 to 2008.
All data needed to determine Great Lakes load estimates for the period 1994-2008 were acquired.
Hydrologic data for the lakes were obtained, however some of the Canadian data took longer to acquire.
Water-quality concentration data for the 11 Lake Michigan tributaries that were sampled as part of
LMMB was successfully obtained and resulted in estimation of nutrient loads for Lake Michigan with
Final Report, December, 2011	Page 4

-------
roughly the same ratio (1.8) of monitored to unmonitored basin area as achieved in the LMMB.
Additional sources of water quality data have been identified and were used in the refinement of
loading to Green Bay.
Water-quality concentration data for Lake Huron were retrieved from STORET (Legacy and Modern),
USGS, and the Provincial Water Quality Monitoring Network of Ontario. Initially, less success was
achieved (compared to Lake Michigan) in obtaining consistent data from year to year to cover the major
tributary basins. However, the Ontario tributary data became available online for the years 2002 to the
present. Archived data covering the period 1994-2001 were requested from the Ontario Ministry of the
Environment and were received in December, 2009.
All available U.S. point source data have been de-archived and processed. Some Wisconsin data that
were not available in PCS, were received directly from the WDNR. Indiana data for 2007 and 2008 were
not available in PCS and data for those years were provided by EPA, Region V. Canadian point source
data were obtained in December, 2009 in summary form from the Ontario Ministry of the Environment.
Point source data for Lake Michigan and Huron (U.S) have been quality-assured, summarized, and
tabulated. See attached text files.
Objective 2. Apply the experience gained from similar work on Lake Erie to the estimation of TP loads
(and associated uncertainty) to Lake Michigan for the fifteen-year period. This task includes separate
loading estimates for upper and lower Green Bay.
The experience of Lake Erie has enabled estimation of TP loads for Lake Michigan including separate
estimates for upper and lower Green Bay. Lake Michigan was divided into nine segments, seven
segments in Green Bay (GB1-GB7) and two for Lake Michigan proper (Lake Michigan North and Lake
Michigan South). All problems with missing data were resolved. See attached Excel spreadsheets titled
LAKE MICHIGAN TP 1994-2008 and LAKE MICHIGAN TP 1994-2008 TRIB DETAIL TABLE. These
Final Report, December, 2011
Page 5

-------
spreadsheets are meant to provide the level of detail that was previously available in IJC reports, such as
the 1987 Appendix B (Rathke and McCrae, 1989). The 1994-2008 spreadsheet provides summary details
similar to Table 3.0-3 in Rathke and McCrae (1989). The summary loads have been broken down by
model segment and standard errors of estimated load have been included. The 1994-2008 TRIB DETAIL
TABLE is actually an Excel file composed of 15 worksheets; one for each year. Within each year, the
details of the load estimation for each tributary (organized by model segment) are included. The
monitored and unmonitored area estimates, direct and indirect point sources, and atmospheric
estimates by segment are all included.
Objective 3. Utilize estimation procedures developed for Lake Erie to calculate loads for TDP, chloride and
nitrate on the same basis as objective 2.
Load calculations have been completed for TDP, chloride, and nitrate. These estimates have been
reviewed and all problems with missing data have been resolved. See attached Excel spreadsheets titled
LAKE MICHIGAN TDP 1994-2008, LAKE MICHIGAN CHLORIDE 1994-2008, and LAKE MICHIGAN NITRATE
1994-2008. See Objective 2 for a description of these tables.
Objective 4. Update the TP mass balance for Lake Michigan using components of the Lesht et al. (1991)
regional phosphorus model and Chopra's (2007) phosphorus budget model. Use the model to assess the
estimated phosphorus loads by comparison of the model-computed versus observed in-lake TP
concentrations.
The updated mass balance with enhanced spatial resolution for Green Bay was the topic of Matthew
Maccoux's master's thesis, titled, "Total Phosphorus Loads (1994-2008) and a Mass Balance Budget
Model for Green Bay, Lake Michigan", and presented at the 54th Annual Conference on Great Lakes
Research (IAGLR 2011) in Duluth, Minnesota at a session organized by Dolan and Chapra (Co-PIs). The
total phosphorus model agreement was generally good, with the exception of a large difference in lower
Final Report, December, 2011
Page 6

-------
Green Bay during 1999-2004. This work helped in the development of a total phosphorus TMDL for
Green Bay and the Lower Fox River.
Objective 5. Collate data from U.S. and Canada into common formats required for computation of loads
for Lakes Huron, Superior and Ontario. This will include separate loading estimates for the North
Channel, Georgian Bay and Saginaw Bay.
An Access database was created to prepare both U.S. and Canadian tributary data for load
computations. The database has been structured to allow separation of North Channel loads. Load
calculations have been completed for TP, TDP, chloride, and nitrate for Lakes Huron, Superior and
Ontario. Lakes Superior and Ontario were treated as one completely mixed segment whereas Lake
Huron was divided into six segments, which include: North Channel, Georgian Bay, Saginaw Bay Inner,
Saginaw Bay Outer, Lake Huron North and Lake Huron South.
Objective 6. Evaluate alternative load estimation techniques that may be able to utilize the wealth of
historic information to provide plausible estimates for current conditions. Use minimization of load
estimation uncertainty as a criterion for this evaluation.
Since most of the missing data problems have been resolved for all the lakes (Michigan, Huron, Superior
and Ontario) this objective was not given priority. In general, traditional load estimation techniques
were considerate adequate given the availability of Great Lake data.
Objective 7. Make estimates for each component (tributary, point source and atmospheric) of total lake
loading including unmonitored areas on a watershed basis. Report the load estimation uncertainty
associated with each annual load to each lake.
The database referred to in Objective 5 has been structured to allow component estimates on a
watershed basis. Load calculations have been completed for TP, TDP, chloride, and nitrate for Lakes
Final Report, December, 2011
Page 7

-------
Huron, Superior and Ontario for each component including uncertainty estimates. These estimates have
been reviewed and all problems with missing data have been resolved.
See attached Excel spreadsheets titled: LAKE HURON TP 1994-2008, LAKE HURON TP 1994-2008 TRIB
DETAIL TABLE, LAKE HURON TDP 1994-2008, LAKE HURON TDP 1994-2008 TRIB DETAIL TABLE, LAKE
HURON CHLORIDE 1994-2008, LAKE HURON CHLORIDE 1994-2008 TRIB DETAIL TABLE, LAKE HURON
NITRATE 1994-2008, LAKE HURON NITRATE 1994-2008 TRIB DETAIL TABLE, LAKE SUPERIOR TP 1994-
2008, LAKE SUPERIOR TP 1994-2008 TRIB DETAIL TABLE, LAKE SUPERIOR TDP 1994-2008, LAKE
SUPERIOR TDP 1994-2008 TRIB DETAIL TABLE, LAKE SUPERIOR CHLORIDE 1994-2008, LAKE SUPERIOR
CHLORIDE 1994-2008 TRIB DETAIL TABLE, LAKE SUPERIOR NITRATE 1994-2008, LAKE SUPERIOR NITRATE
1994-2008 TRIB DETAIL TABLE, LAKE ONTARIO TP 1994-2008, LAKE ONTARIO TP 1994-2008 TRIB DETAIL
TABLE, LAKE ONTARIO TDP 1994-2008, LAKE ONTARIO TDP 1994-2008 TRIB DETAIL TABLE , LAKE
ONTARIO CHLORIDE 1994-2008, LAKE ONTARIO CHLORIDE 1994-2008 TRIB DETAIL TABLE, LAKE
ONTARIO NITRATE 1994-2008 and LAKE ONTARIO NITRATE 1994-2008 TRIB DETAIL TABLE.
Objective 8. Update the TP mass balance for Lakes Huron, Superior, and Ontario using components of the
Lesht et al. (1991) regional phosphorus model and Chopra's (2007) phosphorus budget model. Use the
model to assess the estimated phosphorus loads by comparison of the model-computed versus observed
in-lake TP concentrations.
Work on the TP mass balance for Lakes Huron, Superior and Ontario have been completed. The updated
mass balance with enhanced spatial resolution for Lake Huron was the topic of Heidi Schmitt Marquez's
master's thesis, titled, "Nutrient Loading in the Upper Great Lakes System: Chloride and Total
Phosphorus Estimates in Lakes Superior, Michigan, and Huron, 1994-2008", and presented at the 54th
Annual Conference on Great Lakes Research (IAGLR 2011) in Duluth, Minnesota at a session organized
by Dolan and Chapra (Co-PIs).
Final Report, December, 2011
Page 8

-------
Simulation results for Lakes Superior and Huron generally conform to observations and indicate that
these lakes have exhibited significant improvement following load reductions. However, data for Lake
Ontario suggests a much greater improvement than predicted by the model.
This work was coupled with ongoing load estimation for Lake Erie and allows for better resolution of the
upstream load to the Huron-Erie corridor.
Objective 9. Prepare tables and graphs of results showing subtotals on an annual-average basis for each
component of the regional model and source (point, nonpoint, etc.).
Graphs and tables have been completed. For each lake (Michigan, Superior, Huron and Ontario) tables
and graphs were prepared showing annual load estimates for TP, TDP, nitrate and chloride by loading
source. See attached Tables and Graphs.
Objective 10. Recommend the necessary level of monitoring for each lake to continue the assessment of
loading trends at a level of certainty necessary for use in models and load control programs.
Given the state of water quality monitoring in 2008, lakewide annual estimates of TP are feasible. High-
quality data are still available to allow reasonable estimates of total phosphorus to the Great Lakes.
Recent concerns about lakewide changes in trophic status as well as the recurrence of local and regional
water-quality problems should be ample justification for renewed monitoring efforts.
Project Outputs
1. Tables of annual TP loads with standard errors for each lake (Michigan, Huron, Ontario and Superior)
as well as for the key embayments (Green Bay, Georgian Bay, Saginaw Bay) for the fifteen-year period
1994-2008.
Final Report, December, 2011
Page 9

-------
Status: The final tables of annual TP loads for each lake (Michigan, Huron, Ontario and Superior) are
attached as a spreadsheet titled LAKE MICHIGAN TP 1994-2008, LAKE HURON TP 1994-2008, LAKE
ONTARIO TP 1994-2008, and LAKE SUPERIOR TP 1994-2008.
2.	Tables similar to those described for Output #1 for TDP, chloride, and nitrate. The loadings in the
tables described in Outputs #1 and #2 will be presented in two different ways: (1) The Summary of
Estimated Atmospheric, Industrial, Municipal, and Tributary Loading to each lake comparable to Table
3.0-1 in Appendix B (Rathke and McRae, 1989); and (2) The Summary of Estimated Atmospheric,
Industrial, Municipal, and Tributary Loading From Point and Nonpoint Sources to each lake comparable
to Table 3.0-8 in Appendix B (Rathke and McRae, 1989). The latter table includes indirect point sources.
These tables will also be consistent with those used by Chapra (1977) to display the components of
loading for mass-balance modeling and will be the basis for graphs and charts displaying this information
where appropriate. These tables will be in common formats for all lakes and will include Canadian data
that have been converted to the same basis (units and time periods) as U.S. data.
Status: Load estimates for TDP, chloride, and nitrate have been completed for all lakes (Michigan,
Huron, Superior and Ontario) and are included in this report. See attached tables.
3.	Calibrated TP and Chloride model analysis of lake response to loading trends using the latest available
data.
Status: Calibration is completed. The model was calibrated using chloride data due to its conservative
nature and relative ease and precision of measurement. Chloride does not settle out of the water
column, and thus can provide accurate information about the hydrodynamic behavior of a particular
body or segment of water. A chloride mass-balance model was used to estimate turbulent eddy
diffusion between model segments. Total phosphorus concentrations were then simulated based on
estimated historical and measured modern loading time series. Objectives 4 and 8 were then able to be
Final Report, December, 2011
Page 10

-------
addressed. Two masters-level students were hired and included various aspects of the calibration as the
topics of their theses. The theses identified below were defended in December, 2010 and submitted in
partial fulfillment of the requirements for a Master's of Science degree in Environmental Science and
Policy at the University of Wisconsin - Green Bay:
Total Phosphorus Loads (1994-2008) and a Mass Balance Budget Model for Green Bay, Lake
Michigan by Matthew J. Maccoux
Nutrient Loading in the Upper Great Lakes System: Chloride and Total Phosphorus Estimates in
Lakes Superior, Michigan, and Huron, 1994-2008 by Heidi S. Schmitt Marquez
4.	General software for evaluating the long-term trends of the contaminants addressed in this project
will be delivered along with model documentation.
Status: A general model software package, GL2K, has been developed and applied to simulate historical
trends of cxhloride and total phosphorus as a function of estimated loadings. The model is implemented
as an Excel Workbook with program code written as macros with VBA (Visual Basic for Applications).
This software along with supporting input files and a users manual (Chapra and Dolan 2011) are included
on the CD that accompanies this document.
5.	Monitoring recommendations for each lake in order to maintain a desired level of statistical certainty
in load estimation.
Status: Given the state of water quality monitoring in 2008, lakewide, annual estimates of TP are
feasible. High-quality data are still available to allow reasonable estimates of total phosphorus to the
Great Lakes. Recent concern about lakewide changes in trophic status as well as the recurrence of local
and regional water-quality problems should be ample justification for renewed monitoring efforts.
6.	Two interim reports and a final technical report.
Final Report, December, 2011
Page 11

-------
Status: The first interim report was delivered in April 2009 and the second in December 2009. This
report is the final technical report
7. One or more peer-reviewed publications.
Status: Four papers have been drafted and are ready for submission to peer-reviewed journals. These
draft publications include:
1.	Maccoux, M.J., D.M. Dolan, and S.C. Chapra. 2011. Chloride and total phosphorus
budgets for Green Bay, Lake Michigan.
2.	Chapra, S.C. and D.M. Dolan. 2011. Great Lakes total phosphorus revisited: 2. Mass-
balance modeling.
3.	Dolan, D.M. and S.C. Chapra. 2011. Great Lakes total phosphorus revisited: 1. Loading
analysis and update: 1994-2008.
4.	Schmitt-Marquez, H, D.M. Dolan, and S.C. Chapra. 2011. Interlake and total loading
estimates of total phosphorus and chloride in the upper Great Lakes system, 1994-2008.
2011
Several presentations on the preliminary and final results of this study have been made by co-
investigators at three Annual Conferences on Great Lakes Research (IAGLR). A presentation on the
preliminary results of this study was made by the co-investigators at the 52nd Annual Conference on
Great Lakes Research (IAGLR 2009) at the University of Toledo on May 20th. Also, a session was included
by the co-investigators at the 53rd IAGLR 2010 at the University of Toronto and at the 54th Annual
Conference on Great Lakes Research (IAGLR 2011) in Duluth, Minnesota that included much of this
work.
Final Report, December, 2011
Page 12

-------
In addition, a session titled "Eutrophication Management in the Great Lakes: Past, Present and Future" is
planned by the co-investigators at the 55rd Annual Conference on Great Lakes Research (IAGLR 2011) in
Cornwall, Ontario that will present much of this work.
Attachments
MIIN94_08loadsummary.txt: Load summary of Indiana dischargers to Lake Michigan basin
mimi94_08loadsumupdate.txt: Load summary of Michigan dischargers to Lake Michigan basin
miwi94loadsummaryfinal.txt: Load summary for Wisconsin dischargers to Lake Michigan basin
LAKE MICHIGAN TP 1994-2008.xlsx: Final TP Load Estimates for Lake Michigan
LAKE MICHIGAN TP 1994-2008 TRIB DETAILTABLE.xIsx
LAKE MICHIGAN TDP 1994-2008.xlsx: Final TDP Load Estimates for Lake Michigan
LAKE MICHIGAN TDP 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE MICHIGAN CHLORIDE 1994-2008.xlsx: Final Chloride Load Estimates for Lake Michigan
LAKE MICHIGAN CHLORIDE 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE MICHIGAN NITRATE 1994-2008.xlsx: Final Nitrate Load Estimates for Lake Michigan
LAKE MICHIGAN NITRATE 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE HURON TP 1994-2008.xlsx: Final TP Load Estimates for Lake Huron
LAKE HURON TP 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE HURON TDP 1994-2008.xlsx: Final TDP Load Estimates for Lake Huron
Final Report, December, 2011
Page 13

-------
LAKE HURON TDP 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE HURON CHLORIDE 1994-2008.xlsx: Final Chloride Load Estimates for Lake Huron
LAKE HURON CHLORIDE 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE HURON NITRATE 1994-2008.xlsx: Final Nitrate Load Estimates for Lake Huron
LAKE HURON NITRATE 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE SUPERIOR TP 1994-2008.xlsx: Final TP Load Estimates for Lake Superior
LAKE SUPERIOR TP 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE SUPERIOR TDP 1994-2008.xlsx: Final TDP Load Estimates for Lake Superior
LAKE SUPERIOR TDP 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE SUPERIOR CHLORIDE 1994-2008.xlsx: Final Chloride Load Estimates for Lake Superior
LAKE SUPERIOR CHLORIDE 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE SUPERIOR NITRATE 1994-2008.xlsx: Final Nitrate Load Estimates for Lake Superior
LAKE SUPERIOR NITRATE 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE ONTARIO TP 1994-2008.xlsx: Final TP Load Estimates for Lake Ontario
LAKE ONTARIO TP 1994-2008 TRIB DETAIL TABLE.xIsx:
LAKE ONTARIO TDP 1994-2008.xlsx: Final TDP Load Estimates for Lake Ontario
LAKE ONTARIO TDP 1994-2008 TRIB DETAIL TABLE.xIsx
LAKE ONTARIO CHLORIDE 1994-2008.xlsx: Final Chloride Load Estimates for Lake Ontario
Final Report, December, 2011
Page 14

-------
LAKE ONTARIO CHLORIDE 1994-2008 TRIB DETAILTABLE.xIsx
LAKE ONTARIO NITRATE 1994-2008.xlsx: Final Nitrate Load Estimates for Lake Ontario
LAKE ONTARIO NITRATE 1994-2008 TRIB DETAIL TABLE.xIsx
Final Report, December, 2011
Page 15

-------
References
Chapra, S,C. 1977. Total phosphorus model for the Great Lakes. J. Environ. Engin. Div., ASCE, 103:147-
161.
Chapra, S.C. 2007. Total Phosphorus Model for the Great Lakes Revisited: Success and Surprise. 50th
Conference on Great Lakes Research, International Association for Great Lakes Research,
Pennsylvania State University, State College, PA, June 1, 2007.
Chapra, S.C. and Dolan, D.M. 2011. GL2K: A Mass Balance Model for the Great Lakes, Users Manual for
Version 1.1. Civil and Environmental Engineering, Tufts University, Medford, MA.
Lesht, B.M., Fontaine III, T.D., and Dolan, D.M. 1991. Great Lakes total phosphorus model: post audit and
regionalized sensitivity analysis. J. Great Lakes Res. 17(1):3-17.
Miller, S.M., Sweet, C.W., DePinto, J.V., Hornbuckle, K.C. 2000. Atrazine and nutrients in precipitation:
results from the Lake Michigan Mass Balance Study. Environ. Sci. Technol., 34 (1): 55-61.
Rathke, D. E. and McCrae, G. 1989. Appendix B, Volume III, Report of the Great Lakes Water Quality
Board. International Joint Commission, Windsor, Ontario.
USGS. 2008. NADP/NTN: National Atmospheric Deposition Program/National Trend Network.
http://nadp.sws.uiuc.edu/, last accessed December 7, 2010.
Final Report, December, 2011
Page 16

-------