Sections 10-13
Preprint
July 'J3t,-' 1989
Future Effects of Long-Term Sulfur Deposition
on Surface Water Chemistry
in the Northeast and Southern Blue Ridge Province
(Results of the Direct/Delayed Response Project)
by
M. R. Church, K. W. Thornton, P. W. Shaffer, l^tb ^Stevens, Bfl^ftochelle,
G. R. Holdren, M. G. Johnson, J. J.'^ee,:B^ Turner, D. L^Gassell,:
D. A. Lamrners, W: G. Campbell, C.4 yffjfC^C. Brandt, Lffttege.\-
G. D. Bishop, D. C. Mortenson, S. MViPfeflSbn, D. D.
A Contribution to the
National Acid PrecipitatiQn7JlSiii^m0r>y^^ra^?-*-«--
U.S. Envirohm^iaJ:]p0
Office of Research and be
Environmental Research
CVJ
-------
NOTICE
The information in this document has been funded wholly (or in part) by the U.S. Environmental
Protection Agency. It has been subjected to the Agency's peer and administrative review, and it has
been approved for publication as an EPA document. Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.
-------
CONTENTS
SECTION PAGE
Notice ii
Tables xii
Figures xbc
Plates xxvii
Contributors xxix
Acknowledgments xxxi
1 EXECUTIVE SUMMARY 1-2
1.1 INTRODUCTION 1-2
1.1.1 Project Background 1-2
1.1.2 Primary Objectives 1-3
1.1.3 Study Regions 1-4
1.1.4 Time Frames of Concern 1-4
1.2 PROCESSES OF ACIDIFICATION 1-6
1.2.1 Sulfur Retention 1-6
1.2.2 Base Cation Supply 1-7
1.3 GENERAL APPROACH 1-7
1-3.1 Soil Survey 1-8
1,3.2 Other Regional Datasets 1-8
1.3.3 Scenarios of Atmospheric Deposition 1-10
1.3.4 Data Analysis 1-10
1.4 RESULTS 1-11
1.4.1 Retention of Atmospherically Deposited Sulfur 1-11
1.4.1.1 Current Retention 1-11
1.4.1.2 Projected Retention 1-12
1.4.2 Base Cation Supply 1-15
1.4.2.1 Current Control 1-15
1.4.2.2 Future Effects 1-15
1.4.3 Integrated Effects on Surface Water ANC 1-16
1.4.3.1 Northeast Lakes 1-16
1.4.3.2 Southern Blue Ridge Province 1-20
1.5 SUMMARY DISCUSSION 1-23
1.6 REFERENCES 1-24
2 INTRODUCTION TO THE DIRECT/DELAYED RESPONSE PROJECT 2-1
2.1 PROJECT BACKGROUND 2-1
2.2 PRIMARY OBJECTIVES 2-2
2.3 STUDY REGIONS 2-3
2.4 TIME FRAMES OF CONCERN 2-3
2.5 PROJECT PARTICIPANTS 2-6
2.6 REPORTING 2-6
3 PROCESSES OF ACIDIFICATION 3-1
3.1 INTRODUCTION 3-1
3.2 FOCUS OF THE DIRECT/DELAYED RESPONSE PROJECT 3-3
iii
-------
CONTENTS (continued) Page
3.3 SULFUR RETENTION PROCESSES 3-3
3.3.1 Introduction 3-3
3.3.2 Inputs 3-4
3.3.3 Controls on Sulfate Mobility within Forest/Soil Systems 3-5
3.3.3.1 Precipitation/Dissolution of Secondary Sulfate Minerals 3-7
3.3.3.2 Sulfate Reduction in Soils and Sediments 3-7
3.3.3.3 Plant Uptake 3-8
3.3.3.4 Retention as Soil Organic Sulfur 3-9
3.3.3.5 Sulfate Adsorption by Soils 3-10
3.3.4 Models of Sulfur Retention 3-14
3.3.5 Summary 3-16
3.4 BASE CATION SUPPLY PROCESSES 3-17
3.4.1 Introduction 3-17
3.4.2 Factors Affecting Base Cation Availability 3-20
3.4.2.1 Mineral Weathering 3-21
3.4.2.2 Cation Exchange Processes 3-25
3.4.3 Modelling Cation Supply Processes 3-28
3.4.3.1 Modelling Weathering 3-28
3.4.3.2 Modelling Cation Exchange Processes 3-29
4 PROJECT APPROACH 4-1
4.1 INTRODUCTION 4-1
4.2 SOIL SURVEY 4-3
4.2.1 Watershed Selection 4-3
4.2.2 Watershed Mapping 4-3
4.2.3 Sample Class Definition 4-3
4.2.4 Soil Sampling 4-4
4.2.5 Sample Analysis 4-4
4.2.6 Database Management 4-4
4.3 OTHER REGIONAL DATASETS 4-4
4.3.1 Atmospheric Deposition 4-5
4.3.2 Runoff Depth 4-5
4.4 DATA ANALYSIS 4-6
4.4.1 Level I Analyses 4-6
4.4.2 Level II Analyses 4-6
4.4.3 Level III Analyses 4-7
4.4.4 Integration of Results 4-8
4.4.5 Use pf a Geographic Information System 4-9
5 DATA SOURCES AND DESCRIPTIONS 5-1
5.1 INTRODUCTION 5-1
5.2 STUDY SITE SELECTION 5-1
5.2.1 Site Selection Procedures 5-1
5.2.2 Eastern Lake Survey Phase I Design 5-1
5.2.3 Pilot Stream Survey Design 5-2
5.2.4 DDRP Target Population 5-6
5.2.4.1 Northeast Lake Selection 5-6
5.2.4.2 Southern Blue Ridge Province Stream Selection 5-25
5.2.4.3 Final DDRP Target Populations 5-25
5.3 NSWS LAKE AND STREAM DATA 5-25
5.3.1 Lakes in the Northeast Region 5-25
5.3.1.1 Lake Hydrologic Type 5-25
5.3.1.2 Fall Index Sampling 5-30
5.3.1.3 Chemistry of DDRP Lakes 5-37
iv
-------
CONTENTS (continued) Page
5.3.2 Streams in the Southern Blue Ridge Province Region 5-37
5.3.2.1 Spring Baseflow Index Sampling 5-37
5.3.2.2 Chemistry of DDRP Stream Reaches 5-40
5.4 MAPPING PROCEDURES AND DATABASES 5-40
5.4.1 Northeast Mapping 5-42
5.4.1.1 Soils 5-43
5.4.1.2 Depth to Bedrock 5-49
5,4.1.3 Forest Cover Type 5-51
5.4.1.4 Bedrock Geology 5-51
5.4.1.5 Quality Assurance 5-52
5.4.1.6 Land Use/Wetlands 5-58
5.4.1.7 Geographic Information Systems Data Entry 5-73
5.4.2 Southern Blue Ridae Province Mapping 5-90
5.4.2.1 Soils 5-93
5.4.2.2 Depth to Bedrock 5-97
5.4.2.3 Forest Cover Type/Land use 5-98
5.4.2.4 Bedrock Geology 5-98
5.4.2.5 Drainage 5-98
5.4.2.6 Quality Assurance 5-100
5.4.2.7 Land Use/Wetlands 5-105
5.4.2.8 Geographic Information Systems Data Entry 5-106
5.5 SOIL SAMPLING PROCEDURES AND DATABASES 5-111
5.5.1 Development/Description of Sampling Classes 5-111
5.5.1.1 Rationale/Need for Sampling Classes 5-111
5.5.1.2 Approach Used for Sampling Class Development 5-112
5.5.1.3 Description of Sampling Classes 5-113
5.5.2 Selection of Sampling Sites 5-117
5.5.2.1 Routine Samples 5-117
5.5.2.2 Samples on Special Interest Watersheds 5-122
5.5.3 Soil Sampling 5-122
5.5.3.1 Soil Sampling Procedures 5-122
5.5.3.2 Quality Assurance/Quality Control of Sampling 5-123
5.5.4 Physical and Chemical Analyses 5-124
5.5.4.1 Preparation Laboratories 5-124
5.5.4.2 Analytical Laboratories 5-126
5.5.5 Database Management 5-140
5.5.5.1 Database Structure 5-140
5.5.5.2 Database Operations 5-143
5.5.6 Data Summary 5-148
5.5.6.1 Summary of Sampling Class Data 5-148
5.5.6.2 Cumulative Distribution Functions 5-150
5.6 DEPOSITION DATA 5-150
5.6.1 Time Horizons of Interest 5-161
5.6.1.1 Current Deposition 5-161
5.6.1.2 Future Deposition '. 5-161
5.6.2 Temporal Resolution 5-161
5.6.2.1 Level I Analyses 5-161
5.6.2.2 Level II Analyses 5-161
5.6.2.3 Level III Analyses 5-163
5.6.3 Data Acquisition/Generation 5-163
5.6.3.1 Level III Analyses - Typical Year Deposition Dataset 5-164
5.6.3.2 Level I and II Analyses - Long-Term Annual Average Deposition Dataset . . . 5-191
5.6.4 Deposition Datasets Used in DDRP Analyses 5-200
-------
CONTENTS (continued) Page
5.7 HYDROLOGIC DATA 5-200
5.7.1 Runoff 5-200
5.7.1.1 Data Sources 5-200
5.7.1.2 Runoff Interpolation Methods 5-203
5.7.1.3 Uncertainty Estimates 5-203
5.7.2 Derived Hydroloaic Parameters 5-204
5.7.2.1 TOPMODEL 5-204
5.7.2.2 Soil Contact (Darcy's Law) 5-209
5.7.2.3 Mapped Hydrologic Indices 5-211
6 REGIONAL POPULATION ESTIMATION 6-1
6.1 INTRODUCTION 6-1
6.2 PROCEDURE 6-1
6.2.1 Use of Variable Probability Samples 6-1
6.2.2 Estimation Procedures for Population Means 6-2
6.2.3 Estimators of Variance 6-4
6.2.4 Estimator of Cumulative Distribution Function 6-5
6.3 UNCERTAINTY ESTIMATES 6-6
6.4 APPLICABILITY 6-8
7 WATERSHED SULFUR RETENTION 7-1
7.1 INTRODUCTION 7-1
7.2 RETENTION IN LAKES AND WETLANDS 7-2
7.2.1 Introduction 7-2
7.2.2 Approach 7-4
7.2.3 Results 7-6
7.3 WATERSHED SULFUR RETENTION 7-9
7.3.1 Methods . 7-9
7.3.1.1 Input/Output Calculation 7-9
7.3.1.2 Data Sources 7-11
7.3.2 Uncertainty Estimates 7-11
7.3.2.1 Introduction 7-11
7.3.2.2 Individual Variable Uncertainties 7-12
7.3.2.3 Uncertainty Calculation - Monte Carlo Analysis 7-17
7.3.3 Internal Sources of Sulfur 7-19
7.3.3.1 Introduction/Approach 7-19
7.3.3.2 Bedrock Geology 7-20
7.3.3.3 Upper Limit Steady-State Sulfate Concentration 7-24
7.3.4 Results and Discussion 7-29
7.3.4.1 Northeast 7-31
7.3.4.2 Mid-Appalachians 7-41
7.3.4.3 Southern Blue Ridge Province 7-41
7.3.4.4 Conclusions 7-43
8 LEVEL I STATISTICAL ANALYSES 8-1
8.1 INTRODUCTION 8-1
8.1.1 Approach 8-2
8.1.2 Statistical Methods 8-7
8.2 RELATIONSHIPS BETWEEN ATMOSPHERIC DEPOSITION
AND SURFACE WATER CHEMISTRY 8-9
8.2.1 Introduction 8-9
8.2.2 Approach 8-9
8.2.3 Results and Discussion 8-9
8.2.3.1 Northeast 8-9
8.2.3.2 Southern Blue Ridge Province 8-11
8.2.3.3 Summary 8-11
vi
-------
CONTENTS (continued) Page
8.3 DERIVED HYDROLOGIC PARAMETERS 8-13
8.3.1 Soil Contact (Darcv's Law) 8-13
8.3.1.1 Introduction 8-13
8.3.1.2 Results and Discussion 8-18
8.3.2 Geomorphic/Hvdroloaic Parameters 8-21
8.3.2.1 Introduction 8-21
8.3.2.2 Results and Discussion 8-22
8.3.3 TOPMQDEL Parameters 8-37
8.3.3.1 Introduction 8-38
8.3.3.2 Results and Discussion 8-41
8.3.3.3 Summary 8-48
8.4 MAPPED BEDROCK GEOLOGY . . 8-48
8.4.1 DDRP Bedrock Sensitivity Scale 8-50
8.4.2 Results 8-51
8.4.2.1 Sulfate and Percent Retention 8-54
8.4.2.2 Sum of Base Cations, ANC, and pH 8-59
8.4.3 Summary 8-61
8.5 MAPPED LAND USE/VEGETATION 8-62
8.5.1 Introduction 8-62
8.5.2 Data Sources 8-63
8.5.3 Statistical Methods 8-63
8.5.4 Sulfate and Percent Sulfur Retention 8-64
8.5.4.1 Northeast 8-64
8.5.4.2 Southern Blue Ridge Province 8-73
8.5.4.3 Regional Comparisons 8-73
8.5.5 ANC. Ca plus Mo. and oH 8-75
8.5.5.1 Northeast 8-75
8.5.5.2 Southern Blue Ridge Province 8-76
8.5.5.3 Regional Comparisons 8-76
8.5.6 Summary and Conclusions 8-78
8.6 MAPPED SOILS 8-78
8.6.1 Introduction 8-78
8.6.2 Approach 8-79
8.6.3 Sulfate and Sulfur Retention 8-88
8.6.3.1 Northeast 8-88
8.6.3.2 Southern Blue Ridge Province 8-92
8.6.3.3 Regional Comparisons 8-97
8.6.4 ANC. Ca olus Md. and oH 8-102
8.6.4.1 Northeast 8-109
8.6.4.2 Southern Blue Ridge Province 8-111
8.6.4.3 Regional Comparisons 8-113
8.6.5 Summary and Conclusions 8-113
8.7 ANALYSES OF DEPTH TO BEDROCK 8-113
8.7.1 Introduction 8-113
8.7.2 Approach 8-113
8.7.3 Sulfate and Percent Sulfur Retention 8-115
8.7.3.1 Northeast 8-115
8.7.3.2 Southern Blue Ridge Province 8-119
8.7.3.3 Comparison of Regions 8-119
8.7.4 ANC. Ca plus Ma and oH 8-119
8.7.4.2 Southern Blue Ridge Province 8-122
8.7.4.3 Comparison of Regions 8-123
8.7.5 Summary and Conclusions 8-123
vii
-------
CONTENTS (continued) Page
8.8 INTEGRATED ANALYSIS OF ALL MAPPED VARIABLES 8-124
8.8.1 Introduction 8-124
8.8.2 Approach 8-124
8.8.3 Sulfate and sulfur retention 8-125
8.8.3.1 Northeast 8-125
8.8.3.2 Southern Blue Ridge Province 8-127
8.8.3.3 Regional Comparisons 8-130
8.8.4 ANC. Ca plus MQ. and oH 8-131
8.8.4.1 Northeast 8-131
8.8.4.2 Southern Blue Ridge Province 8-134
8.8.4.3 Regional Comparisons 8-137
8.8.5 Summary and Conclusions 8-137
8.9 SOIL PHYSICAL AND CHEMICAL CHARACTERISTICS 8-138
8.9.1 Introduction 8-138
8.9.2 Approach 8-138
8.9.2.1 Statistical Methods 8-140
8.9.3 Aggregation of Soil Data 8-143
8.9.3.1 Introduction 8-143
8.9.3.2 Aggregation of Soil Data 8-144
8.9.3.3 Assessment of the DDRP Aggregation Approach 8-145
8.9.3.4 Estimation of Watershed Effect 8-148
8.9.3.5 Evaluation of Watershed Effect 8-149
8.9.4 Regional Soil Characterization 8-155
8.9.5 Sulfate and Sulfur Retention 8-157
8.9.5.1 Northeast 8-157
8.9.5.2 Southern Blue Ridge Province 8-164
8.9.6 Ca Plus Ma fSOBQ. ANC. and pH 8-165
8.9.6.1 Northeast 8-169
8.9.6.2 Southern Blue Ridge Province 8-170
8.9.7 Evaluation of Alternative Aggregation Schemes 8-170
8.9.8 Summary and Conclusions 8-171
8.9.8.1 Alternative Aggregation Schemes 8-171
8.9.8.2 Sulfate and Sulfur Retention 8-174
8.9.8.3 Ca plus Mg (SQBC), ANC, and pH 8-175
8.9.9 Summary Conclusions 8-175
8.10 EVALUATION OF ASSOCIATIONS BETWEEN WATERSHED ATTRIBUTES
AND SURFACE WATER CHEMISTRY 8-176
8.10.1 Introduction 8-176
8.10.2 Approach 8-176
8.10.3 Regional Characterization of Watershed Attributes 8-177
8.10.3.1 Northeast Subregions 8-177
8.10.3.2 Northeast and Southern Blue Ridge Providence 8-182
8.10.4 Sulfate and Sulfur Retention 8-182
8.10.4.1 Northeast 8-192
8.10.4.2 Southern Blue Ridge Province 8-193
8.10.5 Ca plus Ma fSOBCl. ANC. and pH 8-193
8.10.5.1 Northeast 8-193
8.10.5.2 Southern Blue Ridge Province 8-197
8.10.6 Summary and Conclusions 8-197
8.10.6.1 Sulfate and Sulfur Retention 8-197
8.10.6.2 Ca plus Mg (SOBC), ANC, and pH 8-198
8.10.7 Summary Conclusions 8-198
viii
-------
CONTENTS (continued) Page
9 LEVEL II ANALYSES - SINGLE FACTOR RESPONSE TIME ESTIMATES 9-1
9.1 INTRODUCTION 9-1
9.2 EFFECTS OF SULFATE ADSORPTION ON WATERSHED SULFUR RESPONSE TIME . . 9-2
9.2.1 Introduction 9-2
9.2.2 Section Objectives 9-4
9.2.3 Approach 9-5
9.2.3.1 Model Description 9-5
9.2.3.2 Data Sources 9-6
9.2.3.3 Model Assumptions and Limitations 9-8
9.2.3.4 Adsorption Data 9-9
9.2.3.5 Evaluation of Aggregated Data and Model Outputs 9-14
9.2.3.6 Target Populations for Model Projections 9-17
9.2.4 Results 9-18
9.2.4.1 Comparison of Northeast and Southern Blue Ridge Province Isotherm
Variables 9-18
9.2.4.2 Model Results - Northeastern United States 9-20
9.2.4.3 Mode) Results - Southern Blue Ridge Province 9-35
9.2.4.4 Uncertainty Analyses and Alternative Aggregation Approaches 9-51
9.2.4.5 Summary of Results from the Southern Blue Ridge Province 9-59
9.2.5 Summary Comments on Level II Sulfate Analyses 9-62
9.2.6 Conclusions 9-64
9.3 EFFECT OF CATION EXCHANGE AND WEATHERING ON SYSTEM RESPONSE 9-66
9.3.1 Introduction 9-66
9.3.1.1 Level II Hypotheses 9-67
9.3.1.2 Approach 9-71
9.3.2 Descriptions of Models 9-75
9.3.2.1 Reuss Model 9-75
9.3.2.2 Bloom-Grigal Model 9-94
9.3.3 Model Forecasts 9-103
9.3.3.1 Reuss Model 9-105
9.3.3.2 Bloom-Grigal Model 9-154
9.3.4 Comparison of the Bloom-Grigal and Reuss Models 9-185
9.3.5 Summary and Conclusions 9-196
10 LEVEL 111 ANALYSES - DYNAMIC WATERSHED MODELLING 10-1
10.1 INTRODUCTION 10-1
10.2 DYNAMIC WATERSHED MODELS 10-3
10.2.1 Enhanced Trickle Down (ETD) Model 10-6
10.2.2 Integrated Lake-Watershed Acidification Study flLWASl Model 10-7
10.2.3 Model of Acidification of Groundwater in Catchments (MAGIC) 10-13
10.3 OPERATIONAL ASSUMPTIONS 10-14
10.4 WATERSHED PRIORITIZATION 10-14
10.4.1 Northeast 10-16
10.4.2 Southern Blue Ridae Province 10-18
10.4.3 Effects of Prioritization on Inclusion Probabilities 10-20
10.5 MODELLING DATASETS 10-20
10.5.1 Meteorological/Deposition Data 10-21
10.5.2 DPRP Runoff Estimation 10-22
10.5.2.1 Annual Runoff 10-22
10.5.2.2 Monthly Runoff 10-22
10.5.3 Morphometrv 10-24
10.5.4 Soils 10-25
10.5.5 Surface Water Chemistry 10-25
10.5.6 Other Data 10-25
10.5.7 Chloride Imbalance 10-25
10.6 GENERAL APPROACH 10-28
IX
-------
CONTENTS (continued) Page
10.7 MODEL CALIBRATION 10-30
10.7.1 Special Interest Watersheds 10-30
10.7.1.1 Northeast - 10-33
10.7.1.2 Southern Blue Ridge Province 10-33
10.7.2 General Calibration Approach 10-34
10.7.3 Calibration of the Enhanced Trickle Down Model 10-35
10.7.4 Calibration of the Integrated Lake-Watershed Acidification Model 10-38
10.7.5 Calibration of the Model of Acidification of Groundwater In Catchments 10-42
10.7.6 Calibration/Confirmation Results 10-44
10.8 MODEL SENSITIVITY ANALYSES 10-49
10.8.1 General Approach 10-50
10.8.2 Sensitivity Results 10-51
10.9 REGIONAL PROJECTIONS REFINEMENT 10-53
10.9.1 Enhanced Trickle Down 10-53
10.9.2 Integrated Lake-Watershed Acidification Study 10-54
10.9.3 Model of Acidification of Groundwater in Catchments 10-54
10.9.4 DDRP Watershed Calibrations 10-56
10.9.4.1 Integrated Lake-Watershed Acidification Study 10-56
10.9.4.2 MAGIC 10-59
10.9.4.3 Southern Blue Ridge Province 10-61
10.10 MODEL PROJECTIONS 10-66
10.10.1 General Approach 10-66
10.10.2 Forecast Uncertainty 10-67
10.10.2.1 Watershed Selection 10-69
10.10.2.2 Uncertainty Estimation Approaches 10-70
10.10.2.3 Relationship Among Approaches 10-74
10.10.2.4 Confidence Intervals 10-76
10.11 POPULATION ESTIMATION AND REGIONAL FORECASTS 10-76
10.11.1 Northeast Regional Projections . 10-77
10.11.1.1 Target Population Projections Using MAGIC 10-77
10.11.1.2 Target Population Projections Using MAGIC and ETD 10-91
10.11.1.3 Restricted Target Population Projections Using All Three Models .... 10-113
10.11.2 Southern Blue Ridae Province 10-141
10.11.2.1 Target Population Projections Using MAGIC 10-141
10.11.2.2 Restricted Target Population Projections Using ILWAS and MAGIC ... 10-155
10.11.3 Regional Comparisons 10-174
10.11.3.1 Northeastern Projections of Sulfate Steady State 10-174
10.11.3.2 Southern Blue Ridge Province Projections of Sulfate Steady State .... 10-178
10.11.3.3 ANC and Base Cation Dynamics - 10-178
10.12 DISCUSSION 10-195
10.12.1 Future Projections of Changes in Acid-Base Surface Water Chemistry 10-195
10.12.2 Rate of Future Change 10-197
10.12.2.1 Northeast 10-197
10.12.1.2. Southern Blue Ridge Province 10-202
10.12.3 Uncertainties and Implications for Future Changes in
Surface Water Acid-Base Chemistry 10-204
10.12.3.1 Deposition Inputs 10-205
10.12.3.2 Watershed Processes 10-207
10.13 CONCLUSIONS FROM LEVEL 111 ANALYSES . 10-210
11 SUMMARY OF RESULTS 11-1
1.1 RETENTION OF ATMOSPHERICALLY DEPOSITED SULFUR 11-1
11.1.1 Current Retention 11-1
11.1.2 Projected Retention 11-3
11.2 BASE CATION SUPPLY 11-6
11.2.1 Current Control 11-6
11.2.2 Future Effects 11-7
-------
CONTENTS (continued) Page
11.3 INTEGRATED EFFECTS ON SURFACE WATER ANC 11-8
11.3.1 Northeast Lakes 11-9
11.3.2 Southern Blue Ridae Province 11-17
11.4 SUMMARY DISCUSSION 11-26
12 REFERENCES 12-1
13 GLOSSARY 13-1
13.1 ABBREVIATIONS AND SYMBOLS 13-1
13.1.1 Abbreviations 13-3
13.1.2 Symbols 13-6
XI
-------
TABLES
TABLE PAGE
1-1 Lakes in the NE Projected to Have ANC Values <0 and <50 peq L"1 for Constant
and Decreased Sulfur Deposition 1-19
1-2 SBRP Stream Reaches Projected to Have ANC Values <0 and <50 ^eq L for
Constant and Increased Sulfur Deposition 1-22
3-1 Major Rock Forming Minerals and Their Relative Reactivities 3-22
5-1 Sampling Structure for Phase I, Region 1 (Northeast), Eastern Lake Survey 5-4
5-2 Sample Structure for the Direct/Delayed Response Project - Northeastern Sample 5-8
5-3 ANC Group, Lake Identification, ELS-I Phase I ANC, Weight and Inclusion
Probabilities for the Direct/Delayed Response Project Northeast Sample Watersheds ... 5-9
5-4 Lake Identification (ID) and Name, and State and Latitudinal/Longitudinal
Location of the Northeast Sample Watersheds, Sorted by Lake ID 5-13
5-5 Lake Identification (ID) and Name, Sorted by State - Northeast Sample Watersheds . . . 5-16
5-6 Stream Identification (ID), Weight, and Inclusion Probabilities for the Southern
Blue Ridge Province Direct/Delayed Response Project Sample Watersheds 5-26
5-7 Stream Identification (ID) and Name, and State and Latitudinal/Longitudinal
Location of the Southern Blue Ridge Province Sample Watersheds, Sorted by Stream ID 5-27
5-8 Stream Identification (ID) and Name, Sorted by State - Southern Blue Ridge
Province Sample Watersheds 5-28
5-9 DDRP Reclassification of Northeastern Lakes Classified as "Seepage" or "Closed"
by the NSWS 5-31
5-10 Depth-to-Bedrock Classes and Corresponding Level of Confidence 5-50
5-11 Interpretation Codes for Northeast Map Overlays - Land Use/Land Cover, Wetlands,
and Beaver Activity 5-59
5-12 Northeast Watersheds Studied for Independent Field Check of Land Use and
Wetland Photointerpretations 5-63
5-13 Chi-Square Test for General Land Use Categories 5-65
5-14 Comparison of Field Check (Matched) General Land Use Determinations with
Office Photointerpretations 5-66
5-15 Chi-Square Test for Detailed Wetland Categories 5-67
5-16 Comparison of Field Check (Matched) Detailed Wetland Determinations with
Office Photointerpretations 5-68
5-17 Comparison of Beaver Dam Number (#), Breached (B) and Unbreached
(U) Status, and Lodges (L), Identified via Field Check and Office Photointerpretation
Methods 5-70
5-18 Aggregated Land Use Data for Northeast Watersheds 5-72
5-19 Watershed No. 1E1062 Soil Mapping Units 5-87
5-20 Land Use Codes Used as Map Symbols 5-99
5-21 Percent Land Use Data for Southern Blue Ridge Province Watersheds 5-107
5-22 Laboratory Analysis of DDRP Soil Samples 5-125
5-23 Analytical Variables Measured in the DDRP Soil Survey 5-127
5-24 Data Quality Objectives for Detectability and Analytical Within-Batch Precision 5-131
5-25 Detection Limits for Contract Requirements, Instrument Readings, and
System-Wide Measurement in the Northeast 5-133
5-26 Detection Limits for the Contract Requirements, Instrument Readings, and
System-wide Measurement in the Southern Blue Ridge Province 5-134
5-27 Attainment of DQO's by the analytical laboratories as determined from blind
audit samples for the Northeast 5-136
5-28 Attainment of DQO's by the Analytical Laboratories as Determined from Blind
Audit Samples for the Southern Blue Ridge Province 5-138
5-29 Quality Assurance and Quality Control Checks Applied to Each Data Batch 5-146
5-30 Medians of Pedon-Aggregated Values of Soil Variables for the DDRP Regions
and Subregions 5-160
xii
-------
TABLES (continued) Page
5-31 Monthly Values of Leaf Area Index (LAI) Used to Apportion Annual Dry Deposition
to Monthly Values 5-176
5-32 Ratios of Coarse-to-Fine Particle Dry Deposition 5-180
5-33 Ratios of Dry Deposition to Wet Deposition for DDRP Study Sites for the
Typical Year (FY) Deposition Dataset 5-182
5-34 Deposition Datasets Used in DDRP Analyses 5-201
5-35 DDRP texture classes and saturated hydraulic conductivity (K) for the NE study systems. .5-206
5-36 SCS slope classifications 5-212
5-37 Mapped and calculated geomorphic parameters collected for the NE study sites 5-215
5-38 Mapped and calculated geomorphic parameters collected for the SBRP study sites. .... 5-219
7-1 Summary of Computed Sulfur Retention by In-take Reduction for Lake Systems in the
Eastern United States 7-5
7-2 Intensively Studied Sites Used in Surface Water Chemistry Uncertainty Analysis 7-13
7-3 Summary Statistics on Percent Differences Between Flow-weighted Average Annual
Sulfate Concentration and the Fail/Spring Flow-weighted Averages 7-18
7-4 Bedrock Geology Maps Used in the DDRP Internal Sources of Sulfur Bedrock
Geology Analyses 7-21
7-5 Potential for Sulfur Contribution by Geologic Type 7-23
7-6 Summary of Watersheds (by ELS and NSS Subregion) Dropped Due to Suspected
Internal Sources of Sulfur Identified by Steady-State Analysis 7-30
7-7 Percent Sulfur Retention - Summary Statistics by Region 7-33
7-8 Summary of Sulfur Retention Status and of Watershed Variables Contributing
to Sulfur Retention for 42 Watersheds in the Northeastern United States 7-39
8-1 Surface Water Chemistry and Percent Sulfur Retention Summary Statistics
for the Northeastern DDRP Sample of 145 Lake Watersheds 8-3
8-2 Surface Water Chemistry and Percent Sulfur Retention Summary Statistics
for the DDRP Sample of 35 SBRP Stream Watersheds 8-4
8-3 Summary Statistics for Wet and Dry Deposition on the DDRP Sample
of 145 Northeastern Lake Watersheds 8-5
8-4 Summary Statistics for Wet and Dry Deposition on the DDRP Sample of 35
SBRP Stream Watersheds 8-6
8-5 Results of Regressions Relating Surface Water Chemistry to Atmospheric
Deposition in the Northeast Region (n = 145) 8-10
8-6 Results of Regressions Relating Surface Water Chemistry to Atmospheric
Deposition in the Southern Blue Ridge Province (n = 32) 8-12
8-7 Estimated Population-Weighted Summary Statistics on the Darcy's Law Estimates
of Flow Rate and the Index of Flow Relative to Runoff 8-15
8-8 Estimated Population-Weighted Summary Statistics for Northeastern
Geomorphic/Hydrologic Parameters 8-23
8-9 Estimated Population-Weighted Summary Statistics for Southern Blue
Ridge Province Hydrologic/Geomorphic Parameters 8-24
8-10 Mapped and Calculated Geomorphic Parameters Collected
for the Northeastern Study Sites (Same as 5-37) 8-25
8-11 Mapped and Calculated Geomorphic Parameters Collected for the
SBRP Study Sites. 8-28
8-12 Stratification Based on Sulfur Deposition (Wet and Dry) 8-30
8-13 Results of Stepwise Regression Relating Surface Water Chemistry versus
Geomprphlc/Hydrologic Parameters for the Entire NE 8-31
8-14 Stepwise Regression Equations for Surface Water Chemistry and
Hydrologic/Geomorphic Parameters Based on Sulfur Deposition Stratification 8-33
8-15 Results of Stepwise Regression Relating Surface Water Chemistry
and Geomorphic/Hydrologic Parameters for the SBRP 8-34
8-16 Population-Weighted Summary Statistics for ln(a/KbTanB) for the Northeast 8-39
8-17 Population-Weighted Summary Statistics for ln(a/TanB) for the Southern Blue
Ridge Province 8-40
xiii
-------
TABLES (continued) Page
8-18 Spearman's Correlation Coefficients Between ln(a/KbTanB) and Surface Water Chemistry 8-42
8-19 Pearson's Correlation Coefficients Between fn(a/TanB) and NSS Pilot Chemistry 8-47
8-20 Tabulation of the Generic Bedrock Types Used to Classify the Mapped Units
Identified on State Map Legends 8-52
8-21 Tabulation of the Generic Bedrock Types Used to Classify the Mapped Units
Identified on State Map Legends 8-53
8-22 Regional and Subregional Statistics for the Bedrock Sensitivity Code Variables 8-55
8-23 Results of Regressions of Surface Water Chemistry on Bedrock Sensitivity
Code Statistics and Deposition Estimates for Northeast 8-56
8-24 Results for SBRP of Regressions of Surface Water Chemistry on Bedrock
Sensitivity Code Statistics and Deposition Estimates 8-58
8-25 Land Use and Other Environmental Variables Related to Surface Water
Chemistry of Northeastern Lakes 8-65
8-26 Factor Loadings for First 13 Principal Components after Varimax Rotation of
the Correlation Matrix of Land Use and other Environmental Variables for
Northeastern Lakes 8-66
8-27 Interpretation of the First 13 Principal Components After Varimax Rotation of the
Correlation Matrix of Land Use and Other Environmental Variables for Northeastern Lakes 8-68
8-28 Land Use and Other Environmental Variables Related to Surface Water Chemistry of
Southern Blue Ridge Province Streams 8-69
8-29 Composition of First 11 Principal Component Analysis (PCA) Factors After Varimax
Rotation of the Correlation Matrix of Land Use and Other Environmental Variables
Related to Surface Water Chemistry of Southern Blue Ridge Province Streams 8-70
8-30 Interpretation of the First 11 Principal Components after Varimax Rotation of
the Correlation Matrix of Land Use and Other Environmental Variables for Southern
Blue Ridge Province Streams 8-71
8-31 Results of Regressions Relating Surface Water Chemistry of Northeastern Lakes to
Land Use and Other Environmental Data 8-72
8-32 Results of Regressions Relating Sulfate and Percent Sulfur Retention of
Southern Blue Ridge Province Streams to Land Use Data 8-74
8-33 Results of Regressions Relating ANC, Ca plus Mg, and pH of Southern
Blue Ridge Province Streams to Land Use Data 8-77
8-34 Summary Statistics for Percent Area Distribution of the 38 Soil Sampling
Classes and the 4 Miscellaneous Land Areas on the DDRP Sample of 145 NE
Lake Watersheds 8-83
8-35 Summary Statistics for the Percent Area Distribution of the 38 Soil Sampling Classes
and the 4 Miscellaneous Land Areas in the GIS 40-ft Contour on the DDRP Sample of
145 NE Lake Watersheds 8-84
8-36 Summary Statistics for the Percent Area Distribution of the 38 Soil Sampling Classes
and the 4 Miscellaneous Land Areas in the Combined GIS Bufferson the DDRP
Sample of 145 NE Lake Watersheds 8-85
8-37 Summary Statistics for the Percent Area Distribution of the 12 Soil Sampling
Classes and the 2 Miscellaneous Land Areas on the DDRP Sample of 35 SBRP
Stream Watersheds 8-86
8-38 Summary Statistics for the Percent Area Distribution of the 12 Soil Sampling Classes
and the 2 Miscellaneous Land Areas in the 100-Meter Linear GIS Buffer on the
DDRP Sample of 35 SBRP Stream Watersheds 8-87
8-39 Lake Sulfate and Percent S Retention Regression Models Developed for NE Lakes
Using Deposition, Mapped Soils (as a Percentage of Watershed Area in Soil
Sampling Classes) and Miscellaneous Land Areas as Candidate Independent Variables . . 8-89
8-40 Regression Models of Sulfate In SBRP Streams, Developed Using Deposition,
Mapped Soils (as a Percentage of Watershed Area in Soil Sampling Classes) and
Miscellaneous Land Areas (as a Percentage of Watershed Area) as
Candidate Independent Variables 8-93
8-41 Regression Models of Percent Sulfur Retention In SBRP Stream Watersheds
Developed Using Deposition, Mapped Soils (as a Percentage of Watershed Area in Soil
Sampling Classes), and Miscellaneous Land Areas as Candidate Independent Variables . 8-96
xhv
-------
TABLES (continued) Page
8-42 Lake ANC and the Sum of Lake Calcium and Magnesium Regression
Models Developed for NE Lakes Using Deposition, Mapped Soils (as a
Percentage of Watershed Area in Soil Sampling Classes) and Miscellaneous
Land Areas as Candidate Independent Variables 8-99
8-43 Lake pH Regression Models Developed for NE Lakes Using Deposition,
Mapped Soils (as a Percentage of Watershed Area in Soil Sampling Classes) and
Miscellaneous Land Areas as Candidate Independent Variables 8-101
8-44 Regression Models of ANC in SBRP Stream Watersheds, Developed Using
Deposition, Mapped Soils (as a Percentage of Watershed Area in Soil Sampling
Classes) and Miscellaneous Land Areas as Candidate Independent Variables : ... 8-104
8-45 Regression Models of Calcium Plus Magnesium in SBRP Streams, Developed Using
Deposition, Mapped Soils (as a Percentage of Watershed Area In Soil Sampling
Classes) and Miscellaneous Land Areas as a Candidate Independent Variables 8-106
8-46 Regression Models of SOBC in SBRP Streams, Developed Using Deposition,
Mapped Soils (as a Percentage of Watershed Area in Soil Sampling Classes) and
Miscellaneous Land Areas as Candidate Independent Variables 8-107
8-47 Regression Models of Stream pH in SBRP Streams, Developed Using
Deposition, Mapped Soils (as a Percentage of Watershed Area in Soil Sampling
Classes) and Miscellaneous Land Areas as Candidate Independent Variables 8-110
8-48 Depth-to-Bedrock Classes for the Northeast and the Southern Blue Ridge Province .... 8-114
8-49 Regional and Subregional Statistics for the Depth-to-Bedrock Classes 8-116
8-50 Results for NE of Regressions of Surface Water Chemistry on Depth-to-Bedrock Classes
and Deposition Estimates 8-118
8-51 Results for SBRP of Regressions of Surface Water Chemistry on Depth-to-Bedrock
Classes and Deposition Estimates 8-120
8-52 Regression Models of Surface Water Sulfate and Sulfur Retention in the NE Lake
Watersheds Using Deposition, Derived Hydrologic Parameters, Bedrock Geology
Reaction Classes, Depth To Bedrock, Mapped Landuse/Vegetation, and Mapped
Soils as Candidate Regressor Variables 8-126
8-53 Regression Models of Surface Water Sulfate and Sulfur Retention in the SBRP
Stream Watersheds Using Deposition, Derived Hydrologic Parameters, Bedrock
Geology Reaction Classes, Depth To Bedrock, Mapped Landuse/Vegetation, and
Mapped Soils as Candidate Regressor Variables 8-128
8-54 Regression Models of Surface Water ANC, Ca plus Mg, and pH in the NE Lake
Watersheds Using Deposition, Derived Hydrologic Parameters, Bedrock Geology
Reaction Classes, Depth To Bedrock, Mapped Landuse/Vegetation, and Mapped
Soils as Candidate Regressor Variables 8-132
8-55 Regression Models of Surface Water ANC, Ca plus Mg, and pH in the SBRP
Stream Watersheds Using Deposition, Derived Hydrologic Parameters, Bedrock
Geology Reaction Classes, Depth To Bedrock, Mapped Landuse/Vegetation, and
Mapped Soils as Candidate Regressor Variables 8-135
8-56 Standard Deviations Within and Among Northeast Sampling Classes Estimated
from B Master Horizon Data 8-147
8-57 Means and Standard Deviations of Soil Characteristics by Aggregation Method
and Region 8-150
8-58 Population Means and Standard Errors for Selected Variables, by Subregion/
Region and Aggregation (Watershed Adjusted Data) 8-153
8-59 Non-parametric Correlations Between Lake Chemistry Variables and Selected
Soil Properties for the NE DDRP Watersheds 8-158
8-60 Non-parametric Correlations Between Stream Chemistry Variables and Selected
Soil Properties for the SBRP DDRP Watersheds 8-160
8-61 Results of Stepwise Multiple Regressions for DDRP Lake and Stream Sulfate
Concentrations (SO416) Versus Soil Physical and Chemical Properties 8-162
8-62 Results of Stepwise Multiple Regressions for DDRP Watershed Sulfur Retention
(SO4 NRET) Versus Soil Physical and Chemical Properties 8-163
xv
-------
TABLES (continued) Page
8-63 Results of Stepwise Multiple Regressions for DDRP Lake Calcium plus Magnesium
Concentrations (CAMQ16) and Stream Sum of Base Cation Concentrations (SOBC)
Versus Soil Physical and Chemical Properties 8-166
8-64 Results of Stepwise Multiple Regressions for ODRP Lake and Stream ANC
(ALKANEW and ALKA11) Versus Soil Physical and Chemical Properties 8-167
8-65 Results of Stepwise Multiple Regressions for DDRP Lake and Stream pH (PHEQ11)
Versus Soil Physical and Chemical Properties 8-168
8-66 Results of Stepwise Multiple Regressions for DDRP Lake and Stream ANC
(ALKANEW and ALKA11) Versus Unadjusted and Watershed Adjusted Soil
Physical and Chemical Properties 8-172
8-67 Results of Stepwise Multiple Regressions for DDRP Lake and Stream Sulfate
(SO416) Versus Unadjusted and Watershed Adjusted Soil Physical and Chemical
Properties 8-173
8-68 Population Means and Standard Errors for Selected Variables, by Subregion/
Region and Aggregation 8-178
8-69 Non-parametric Correlations Between Lake Chemistry Variables and Selected
Watershed Attributes for the NE DDRP Watersheds 8-183
8-70 Non-parametric Correlations Between Stream Chemistry Variables and Selected
Watershed Attributes for the SBRP DDRP Watersheds 8-187
8-71 Results of Stepwise Multiple Regressions for DDRP Lake and Stream Sulfate
Concentration (SO416) Versus Watershed Attributes 8-190
8-72 Results of Stepwise Multiple Regressions for DDRP Watershed Sulfur Retention
(SO4 NRET) Versus Watershed Attributes 8-191
8-73 Results of Stepwise Multiple Regressions for DDRP Lake Calcium Plus Magnesium
Concentrations (CAMG16) and Stream Sum of Base Cations (SOBC) Versus
Watershed Attributes 8-194
8-74 Results of Stepwise Multiple Regressions for DDRP Lake and Stream ANC
(ALKA11, ALKANEW) Versus Watershed Attributes 8-195
8-75 Results of Stepwise Multiple Regressions for DDRP Lake and Stream Air
Equilibrated pH (PHEQ11) Versus Watershed Attributes 8-196
9-1 Comparison of Summary Data for Sulfate Adsorption Isotherm Data for Soils in
the Northeastern United States and Southern Blue Ridge Province 9-19
9-2 Summary Statistics for Modelled Changes in Sulfate Concentration, Percent Sulfur
Retention, and Delta Sulfate for Northeast Watersheds Using Long-Term Average
Deposition Data 9-25
9-3 Summary Statistics for Modelled Changes in Sulfate Concentration, Percent Sulfur
Retention, and Delta Sulfate for Northeast Watersheds Using Typical Year Deposition
Data 9-26
9-4 Comparison of Measured and Modelled Base Year (1985) Sulfate Data for SBRP
Watersheds, Using Long-Term Average Deposition Data. 9-38
9-5 Comparison of Modelled Rates of Increase for [SO42"j in DDRP Watersheds in the
SBRP with Measured Rates of Increase in Watersheds in the Blue Ridge and
Adjoining Appalachians 9-41
9-6 Summary Statistics for Modelled Changes in Sulfate Concentration, Percent Sulfur
Retention, and Delta Sulfate for Watersheds in the Southern Blue Ridge Province,
Using Long-Term Average Deposition Data 9-45
9-7 Summary Statistics for Modelled Changes in Sulfate Concentration, Percent Sulfur
Retention, and Delta Sulfate for Watersheds in the Southern Blue Ridge Province 9-46
9-8 Summary Comparison of Watershed Sulfur Status and Model Forecasts
in the Northeastern United States and Southern Blue Ridge Province Using
Typical Year Deposition Data 9-63
9-9 List of the Chemical Species and Reactions Considered Within the Reuss Model
Framework 9-78
xvi
-------
TABLES (continued) Page
9-10 Effect of pCO2 on Changes Projected to Occur in Surface Water ANC Values at
50 and 100 Years Using the Reuss Model 9-90
9-11 List of Input Data for the Bloom-Grigal Soil Acidification Model 9-104
9-12 Summary Statistics for the Population Estimates of Current ANC Conditions for
Lakes in the NE Region for Five Different Deposition or Soils Aggregation Schemes .... 9-113
9-13 Descriptive Statistics of the Population Estimates for Changes in Lake Water
ANC for Systems in the NE 9-118
9-14 Summary Statistics Comparing the Projections Regarding Changes in Surface
Water ANC Values Obtained Using Different Soils Aggregation Schemes 9-122
9-15 Summary Statistics of the Differences Between the Population Estimates for
Future ANC Projections Made Using the Constant Level and Ramped Deposition
Scenarios 9-123
9-16 Summary Statistics for the Population Estimates of Current ANC Conditions for
Stream Reaches in the SBRP for Four Different Deposition Scenarios 9-126
9-17 Descriptive Statistics of the Population Estimates for Changes in Stream Reach
ANC Values for Systems in the SBRP 9-128
9-18 Summary Statistics of the Differences Between the Population Estimates for
Future ANC Projections Made Using the Constant Level and Ramped Deposition
Scenarios for Stream Reaches in the SBRP 9-133
9-19 Summary Statistics of the Projected Changes in Soil Base Saturations in the
NE Region, Obtained Using the Different Deposition Scenarios or Soil
Aggregation Schemes 9-138
9-20 Summary Statistics of the Projected Changes in Soil pH in the NE Region,
Obtained Using the Different Deposition Scenarios or Soil Aggregation Schemes. 9-139
9-21 Summary Statistics of the Projected Changes in Soil Base Saturations in the
SBRP, Obtained Using the Different Deposition Scenarios 9-147
9-22 Summary Statistics of the Projected Changes in Soil pH in the SBRP,
Obtained Using the Different Deposition Scenarios 9-148
9-23 Comparison of the Changes in Soil Base Saturation and Soil pH that Are
Projected to Occur in the NE and SBRP 9-152
9-24 Regionally Weighted Median Values of Initial Annual Deposition Inputs to the
Bloom-Grigal Model for the Northeastern Region and the Southern
Blue Ridge Province 9-156
9-25 Regionally Weighted Median Values of Annual Initial Soil Chemical Values Input
Into the Bloom-Grigal Model for the Northeastern Region and the Southern
Blue Ridge Province 9-159
9-26 Bloom-Grigal Model Regional Projections of the Change in Soil pH in the
Northeastern United States 9-163
9-27 Bloom-Grigal Model Regional Projections of the Change in Percent Base Saturation
in the Northeastern United States 9-165
9-28 Bloom-Grigal Model Regional Projections of the Change in Soil pH in the
Northeastern United States 9-170
9-29 Bloom-Grigal Model Regional Projections for the Change in Percent Base Saturation
in the Northeastern United States 9-172
9-30 Bloom-Grigal Model Regional Projections for the Change in Soil pH in the
Southern Blue Ridge Province 9-178
9-31 Bloom-Grigal Model Regional Projections for the Change in Percent Soil Base
Saturation In the Southern Blue Ridge Province 9-180
9-32 Summary of the Bloom-Grigal Projected Changes in Soil pH and Percent Base
Saturation in the NE and SBRP Under Constant LTA Deposition 9-183
9-33 Comparison of the Results from the Reuss and Bloom-Grigal Models with Regard to
the Magnitude of Changes in Soil pH and Base Saturation Projected in Soils
of the NE 9-187
9-34 Comparison of the Results from the Reuss and Bloom-Grigal Models with
Regard to the Magnitude of Changes in Soil pH and Base Saturation Projected
in Soils of the SBRP. Results Are Shown for 50 and 100 Years 9-193
xvii
-------
TABLES (continued) Page
10-1 Major Processes Incorporated in the Dynamic Model Codes 10-5
10-2 Meteorological Data Required by the Dynamics Model Codes 10-8
10-3 Chemical Constituents in Wet and Dry Deposition Considered
by the MAGIC, ETD, and ILWAS Codes 10-9
10-4 Chemical Constituents Included in Soil Solutions
and Surface Water for the MAGIC, ETD, and ILWAS Codes 10-10
10-5 Definitions of Acid Neutralizing Capacity (ANC) Used by the MAGIC, ETD,
and ILWAS Codes 10-11
10-6 Level III Operational Assumptions 10-15
10-7 Comparison of Calibration/Confirmation RMSE for Woods Lake Among
ETD, ILWAS, and MAGIC Models, with the Standard Error of the Observations 10-45
10-8 Comparison of Calibration/Confirmation RMSE for Panther
Lake Among ETD, ILWAS, and MAGIC Models, with the Standard Error
of the Observations 10-46
10-9 Comparison of Calibration RMSE for Gear Pond Among ETD,
ILWAS, and MAGIC Models, with the Standard Error of the Observations 10-47
10-10 Percent Change in RMSE for MAGIC and ETD for a Ten Percent Change in
Parameter Values. Parameters are Ranked in Descending Order of Sensitivity
from Left to Right 10-52
10-11 Watersheds, by Priority Class, for which Calibration Criteria
Were not Achieved 10-68
10-12 Deposition Variations Used in Input Uncertainty Analyses 10-72
10-13 Target Populations for Modelling Comparisons and Population Attributes 10-78
10-14 Descriptive Statistics of Projected ANC, Sulfate, pH, Calcium Plus Magnesium,
and Percent Sulfur Retention for NE Lakes in Priority Classes A -1 Using
MAGIC for Both Current and Decreased Deposition 10-81
10-15 Change in Median ANC and Sulfate Concentrations Over a 40-Year Period as
a Function of the Initial ELS-Phase I or NSS Pilot Survey ANC Groups 10-89
10-16 Descriptive Statistics of Projected ANC, Sulfate, and Percent Sulfur Retention
for NE Lakes in Priority Classes A - E Using MAGIC and ETD for Both Current
and Decreased Deposition 10-96
10-17 Descriptive Statistics for Projected ANC, Sulfate, Percent Sulfur Retention, and
Calcium Plus Magnesium for NE Lakes in Priority Classes A and B Using ETD,
ILWAS, and MAGIC for Both Current and Decreased Deposition 10-117
10-18 Descriptive Statistics of Projected ANC, Sulfate, and Percent Sulfur Retention,
and Calcium and Magnesium for SBRP Streams in Priority Classes A - E Using
MAGIC for Both Current and Increased Deposition 10-146
10-19 Descriptive Statistics of Projected ANC, Sulfate, Percent Sulfur Retention, and
Calcium Plus Magnesium for SBRP Streams in Priority Classes A and B Using
ILWAS and MAGIC for Both Current and Increased Deposition 10-159
10-20 Effects of Critical Assumptions on Projected Rates of Change 10-206
11-1 Weighted Median Projected Change in ANC at 50 Years for Northeastern DDRP
Lakes 11-11
11-2 Lakes in the NE Projected to Have ANC Values <0 and <50 ^eq L for
Constant and Decreased Sulfur Deposition 11-14
11-3 Weighted Median Projected Change in ANC at 50 Years for DDRP SBRP
Stream Reaches 11-19
11-4 SBRP Stream Reaches Projected to Have ANC Values <0 and <50 peq L"1 for
Constant and Increased Sulfur Deposition 11-23
xviii
-------
FIGURES
FIGURE PAGE
1-1 Steps of the Direct/Delayed Response Project (DDRP) approach 1-9
2-1 Activities of the Aquatic Effects Research Program within the National Acid
Precipitation Assessment Program 2-4
3-1 Diagram of sulfur cycle in forest ecosystems 3-22
3-2 Diagram of terrestrial base cation cyde 3-18
4-1 Steps of the Direct/Delayed Response Project (DDRP) approach 4-2
5-1 Northeastern subregions and ANC map classes, Eastern Lake Survey Phase I 5-3
5-2 Representation of the point frame sampling procedure for selecting NSS Stage I
reaches 5-5
5-3 DDRP site locations for Subregion 1A 5-19
5-4 DDRP site locations for Subregion 1B 5-20
5-5 DDRP site locations for Subregion 1C 5-21
5-6 DDRP site locations for Subregion 1D 5-22
5-7 DDRP site locations for Subregion 1E 5-23
5-8 DDRP stream reach study sites in the Southern Blue Ridge Province. 5-29
5-9 The pH-ANC relationship for (A) lakes of the ELS Phase I sampling in the
Northeast and (B) DDRP study lakes in the Northeast 5-38
5-10 The pH-ANC relationship for samples with ANC < 400 /*eq L"1 taken at the
downstream nodes of stream reaches sampled in the NSS 5-41
5-11 Location of Northeast field check sites and other DDRP watersheds. 5-62
5-12 Example of digitization log sheet : 5-81
5-13 Example of attribute entry log sheet 5-82
5-14 Definition of soil sampling classes for the DDRP Soil Survey in the Northeast 5-114
5-15 Definition of soil sampling classes for the DDRP Soil Survey in the Southern
Blue Ridge Province 5-116
5-16 Selection of watersheds for sampling 5-118
5-17 Selection of starting points for sampling .5-119
5-18 Field selection of a sampling point for sampling class on a watershed 5-120
5-19 Major steps and datasets from the DDRP database 5-141
5-20 Calculation percentage of regional or subregional area in each soil sampling 5-149
5-21 Relative areas of sampling classes in the Northeast subregions 5-151
5-22 Relative areas of sampling classes in the entire Northeast and Southern
Blue Region Province. 5-152
5-23 Aggregated soil variables for individual pedons in the Northeast. 5-153
5-24 Aggregated soil variables for individual pedons in the Southern Blue
Ridge Province 5-155
5-25 Calculation of cumulative distribution function for a soil variable in a
region or subregion 5-157
5-26 Cumulative distribution functions for pedon aggregated soil variables for the
Northeast and the Southern Blue Ridge Province 5-158
5-27 Sulfur deposition scenarios for the NE and SBRP for Level II and III Analyses 5-162
5-28 Example of average annual runoff map for 1951-80 5-202
5-29 Flow chart of Darcy's Law soil contact calculation as applied to the DDRP
study sites 5-213
xix
-------
FIGURES (continued) Page
7-1 Estimated percent sulfur retention by in-lake processes in drainage lakes in
ELS Region 1 (northeastern United States) 7-7
7-2 Percent sulfur retention for intensively studied sites in the United States and
Canada relative to the southern extent of the Wisconsinan glaciation 7-10
7-3 Model of flow-weighted average concentration calculations for Biscuit Brook 7-16
7-4 Flow chart for the determination of internal sources of sulfur using the
steady-state sulfate concentration 7-26
7-5. Scatter plot of the Monte Carlo calculated standard deviation versus the
calculated mean [SO42"]w 7-28
7-6. Comparison of percent sulfur retention calculated using (A) modtfied-LTA
deposition and (B) modified-LTA deposition adjusted with a 20 percent increase
in dry deposition 7-32
7-7. Population-weighted distribution of projected percent sulfur retention (upper and
lower bounds for 90 percent confidence interval): (A) Northeast; (B) Mid-Appalachians,
and (C) Southern Blue Ridge Province 7-34
7-8. Supplemental watersheds mapped for special evaluation of sulfur retention 7-36
7-9. Population-weighted distributions of projected percent sulfur retention, with upper
and lower bounds for 90 percent confidence intervals, for additional NSS subregions:
(A) Southern Appalachian Plateau, (B) Mid-Atlantic Coastal Plain, (C) Catskills/
Poconos, and (D) Piedmont 7-42
7-10 Combination regional population-weighted distributions of projected percent sulfur
retention, with upper and lower bounds for 90 percent confidence intervals, for
the Northeast, Mid-Appalachians, and Southern Blue Ridge Province 7-44
8-1 Distribution of estimated contact rate using Darcy's Law calculation 8-16
8-2 Distribution of index of contact (yr) using Darcy's Law calculatioa 8-17
8-3 Scatter plot of ANC versus contact rate calculated using Darcy's Law 8-19
8-4 Scatter plot of ANC versus index of soil contact calculated using Darcy's Law. 8-20
8-5 Scatter plot of ANC versus ln(a/KbTanB) 8-43
8-6 Scatter plot of Ca plus Mg versus ln(a/KbTanB) 8-44
8-7 Scatter plot of pH versus ln(a/KbTanB) 8-45
8-8 Data and regression model development flow diagrams 8-81
8-9 Model development procedure 8-141
8-10 Histograms of unadjusted and adjusted watershed means for selected SBRP
soils variables 8-151
8-11 The mean pH ± 2 standard errors for the SBRP watersheds estimated using
the common aggregation (bars) and the watershed effects adjusted aggregation
(lines) illustrate the lack of variation among the common aggregation values 8-152
9-1 Schematic diagram of extended Langmuir isotherm fitted to data points from
laboratory soB analysis 9-12
9-2 Comparison of measured lake (NE) or stream (SBRP) sulfate concentration with
computed soil solution concentration 9-16
9-3 Historic deposition inputs and modelled output for soils in a representative
watershed in the northeastern United States 9-22
9-4 Schematic of surface water response to changes in sulfur inputs 9-23
9-5 Comparison of measured, modelled and steady-state sulfate for Northeast lake
systems in 1984 9-27
9-6 Projected changes in percent sulfur retention and sulfate concentration for soils
in northeastern lake systems at 10, 20, 50 and 100 years 9-30
9-7 Box-and-whisker plots showing changes in sulfate concentration, percent sulfur
retention, and change in sulfate concentration for soils in northeastern lake
watersheds, using long-term average deposition data 9-31
9-8 Box-and-whisker plots showing changes in sulfate concentration, percent sulfur
retention, and change in sulfate concentration for soils in northeastern lake
watersheds, using TY deposition data 9-32
xx
-------
FIGURES (continued) Page
9-9 Projected time to steady-state concentration for sulfate in northeastern lakes (A)
at current deposition and (B) after end of decreasing input in ramp scenario 9-34
9-10 Historic deposition inputs and modelled output for soils in stream systems in
the Southern Blue Ridge Province 9-36
9-11 Comparison of measured, modelled, and steady-state sulfate for stream systems
in the Southern Blue Ridge Province in 1985 9-39
9-12 Comparison of forecasts based on two sulfur deposition datasets for soils in SBRP
watersheds ; 9-42
9-13 Projected changes in percent sulfur retention and in sulfate concentration for
stream systems in the Southern Blue Ridge Province at 0, 20, 50, 100 and 140 years. . . . 9-44
9-14 Box and whisker plots showing changes in sulfate concentration, percent sulfur
retention, and change in sulfate concentration for soils in watersheds of the
Southern Blue Ridge Province. 9-47
9-15 Box and whisker plots showing changes in sulfate concentration, percent sulfur
retention, and change in sulfate concentration for soils in watersheds of the
Southern Blue Ridge Province 9-48
9-16 Projected time to 95 percent of steady-state sulfur concentration of Southern Blue
Ridge Province stream systems 9-50
9-17 Comparison of model simulation results for DDRP Southern Blue Ridge watersheds 9-53
9-18 Projected base year sulfate concentration with upper and lower bounds for
90 percent confidence intervals for Southern Blue Ridge Province watersheds. 9-54
9-19 Projected time to sulfur steady state with upper and lower bounds for 90 percent
confidence intervals in Southern Blue Ridge Province watersheds 9-56
9-20 Effects of data aggregation on simulated watershed sulfur response for soils in
DDRP watersheds of the Southern Blue Ridge Province 9-57
9-21 Evaluation of alternate soil aggregation procedures for soils in SBRP watersheds 9-60
9-22 Schematic diagram of the principal process involved in the cycling of base
cations in surficial environments. 9-76
9-23 Plot of the log of the activity of AI3+ vs. soil solution pH for individual soil
samples collected for DDRP 9-83
9-24 Plot of the log of the selectivity coefficient for the calcium-aluminum exchange
reaction vs. the measured base saturation in A/E horizons in the NE 9-86
9-25 Histograms of the (unweighted for the population estimates) projected present-
day ANC values for lakes in the NE 9-87
9-26 Histograms of the (unweighted for the population estimates) projected, present-
day ANC values for lakes in the NE 9-89
9-27 Flow diagram for the one-box Bloom-Grigal soil simulation model 9-97
9-28 Cumulative distribution of projected, present-day ANC values for lakes in the
study population in the NE as projected using Reuss's cation exchange model 9-109
9-29 Scatter plot of the projected, present-day ANC values for lakes in the NE,
obtained using the Reuss model vs. observed (ELS) values 9-110
9-30 Scatter plot of the present-day lake ANC values projected using the Reuss
model in conjunction with the Watershed-Based Aggregation (WBA) soils data vs.
observed (ELS) ANC values ..9-115
9-31 Cumulative distribution of the projected surface water ANC values projected for
the study population of lakes in 50 years in the NE 9-116
9-33 Schematic illustration of the titration-like behavior displayed by soils in response
to constant loadings of acidic deposition 9-117
9-34 Cumulative distribution of projected present-day ANC values for stream reaches
in the study population in the SBRP, as projections using Reuss's cation
exchange model 9-125
9-35 Scatter plot of the projected present-day ANC values for stream reaches in the
SBRP, obtained using the Reuss model, vs. observed (NSS) values 9-127
9-36 Cumulative distribution of projected changes (at 50 years) in surface water ANC
obtained using the Reuss model for stream reaches in the SBRP 9-130
9-37 Cumulative distribution of projected changes (at 100 years) in surface water ANC
obtained using the Reuss model for stream reaches in the SBRP 9-131
xxi
-------
FIGURES (continued) Page
9-38 Comparison of measured vs. calculated soil pH values for the 580 aggregated
master horizons in the NE 9-136
9-39 Cumulative distribution of projected (a) base saturations and (b) soil pH values for
soils in NE 9-140
9-40 Cumulative distribution of projected (a) base saturations and (b) soil pH values for
soils in the NE 9-141
9-41 Plot of the measured (ELS) ANC values for lakes in the NE vs. the estimated,
watershed-level base saturations for mineral horizons in those watersheds 9-143
9-42 Plot of the changes in surface water ANC values at (a) 20, (b) 50, and (c) 100
years as projected by the Reuss model vs. the estimated, present-day, watershed-
level base saturations for mineral horizons in those watersheds 9-144
9-43 Plot of the projected changes in soil base saturations vs 9-145
9-44 Cumulative frequencies of changes in (a) soil base saturation and (b) soil pH
for the population of soils in the SBRP 9-149
9-45 Cumulative frequencies of changes in (a) soil base saturation and (b) soil pH for
the population of soils in the SBRP 9-150
9-46 Cumulative distributions of aggregate initial soil pH and percent base saturation
in the NE and SBRP, with and without organic horizons 9-160
9-47 Regional CDFs of the projected change in the pH of soils on NE lake watersheds
under constant and ramp down (30 percent i) deposition scenarios after 20, 50,
and 100 years of LTA, LTA-rbc, and LTA-zbc deposition 9-161
9-48 Regional CDFs of the projected change In the percent base saturation of soils
on NE lake watersheds under constant and ramp down (30 percent 4) deposition
scenarios after 20, 50, and 100 years of LTA, LTA-rbc, and LTA-zbc deposition 9-162
9*49 Regional CDFs of the projected change in the pH of soils on NE lake
watersheds under constant and ramp down (30% 4) deposition scenarios
after 20, 50, and 100 years of LTA, LTA-rbc, and LTA-zbc deposition 9-168
9-50 Regional CDFs of the projected change in the percent base saturation of soils on
NE lake watersheds under constant and ramp down (30% 4) deposition scenarios
after 20, 50, and 100 years of LTA, LTA-rbc, and LTA-zbc deposition 9-169
9-51 Regional CDFs of the projected change in the pH of soils on SBRP stream
watersheds under constant and ramp up (20% t) deposition scenarios after 20,
50, 100, and 200 years of LTA, LTA-rbc, and LTA-zbc deposition 9-176
9-52 Regional CDFs of the projected change in the percent base saturation of soils
on SBRP stream watersheds under constant and ramp up (20% f) deposition
scenarios after 20, 50, 100, and 200 years of LTA, LTA-rbc, and LTA-zbc deposition. . .. 9-177
9-53 Cumulative distributions of changes in soil base saturation for the population of
watersheds in the NE 9-188
9-54 Cumulative distributions of changes in soil pH for the population of watersheds
in the NE 9-190
9-55 Scatter diagrams of the projected changes in base saturation for individual
systems (not population weighted) in the NE obtained from the Reuss and
Bloom-Grigal models 9-191
9-56 Scatter diagrams of the projected changes in soil pH for individual systems (not
population weighted) in the NE obtained from the Reuss and Bloom-Grigal models 9-192
9-57 Cumulative distributions of changes in soil base saturation for the population of
watersheds in the SBRP 9-194
9-58 Distributions of changes in soil pH for the population of watersheds in the SBRP 9-195
10-1 Modelling priority decision tree: Northeast 10-17
10-2 Modelling priority decision tree: Southern Blue Ridge Province 10-19
10-3 Decision tree used to identify watersheds with net chloride export and procedures
for determining chloride imbalance 10-26
10-4 Approach used in performing long-term projections of future changes in surface
water chemistry 10-29
10-5 Schematic of modelling approach for making long-term projections 10-31
XXII
-------
FIGURES (continued) Page
10-6 Representation of horizontal segmentation of Woods Lake, NY, watershed for
MAGIC and ETD 10-36
10-7 Representation of vertical layers of Woods Lake Basin for ETD 10-37
10-8 Representation of horizontal segmentation of Woods Lake Basin for ILWAS 10-39
10-9 Representation of vertical layers of Woods Lake Basin for ILWAS 10-40
10-10 Representation of vertical layers of Woods Lake, NY, watershed for MAGIC 10-43
10-11 Comparison of population histograms for simulated versus observed (Eastern
Lake Survey Phase I 1984 values) ANC for ILWAS and MAGIC 10-57
10-12 Comparison of population histograms for simulated versus observed (Eastern Lake
Survey - Phase I 1984 values) sulfate concentrations for ILWAS and MAGIC,
Priority Classes A and B 10-58
10-13 Comparison of population histograms for simulated versus observed (Eastern Lake
Survey Phase I 1984 values) ANC and sulfate concentrations for MAGIC, Priority
Classes A - E 10-60
10-14 Comparison of population histograms for simulated versus observed (Eastern
Lake Survey Phasee I 1984 values ) ANC and sulfate concentrations for MAGIC,
Priority Classes A -1 10-62
10-15 Comparison of population histograms for simulated versus observed (NSS Pilot
Survey values) ANC, Priority Classes A and B using ILWAS and MAGIC 10-63
10-16 Comparison of population histograms for simulated versus observed (NSS Pilot
Survey values) sulfate concentrations, Priority Classes A and B using ILWAS and
MAGIC. • 10-64
10-17 Comparison of population histograms for simulated versus observed (NSS Pilot
Survey values) ANC and sulfate concentrations, Priority Classes A - E using MAGIC ... 10-65
10-18 Comparison of projection standard errors as a function of ANC (top figure)
and sulfate (bottom figure) concentrations for the NE uncertainty analysis
watersheds using ETD and MAGIC 10-75
10-19 Projections of ANC and sulfate concentrations for NE lakes, Priority Classes
A -1, using MAGIC for 20, 50, and 100 years, under current deposition and a
30 percent decrease in deposition 10-79
10-20 pH projections for NE lakes, Priority Classes A -1, using MAGIC for 20, 50,
and 100 years, under current deposition and a 30 percent decrease in deposition 10-84
10-21 Box and whisker plots of ANC distributions at 10-year intervals for NE Priority
Classes A -1 using MAGIC 10-85
10-22 Box and whisker plots of sulfate distributions at 10-year intervals for NE Priority
Classes A -1 using MAGIC 10-86
10-23 Box and whisker plots of pH distributions at 10-year intervals for NE Priority
Classes A -1 using MAGIC 10-87
10-24 Comparison of population histograms for ANC under current levels of
deposition and a 30 percent decrease in deposition for NE lakes, Priority Classes
A -1, using MAGIC 10-90
10-25 Comparison of population histograms for sulfate concentrations at current levels
of deposition and a 30 percent decrease for NE lakes, Priority Classes A -1, using MAGIC. 10-92
10-26 Comparison of MAGIC and ETD projections of ANC for NE lakes, Priority
Classes A - E, under current and decreased deposition 10-93
10-27 Comparison of MAGIC and ETD projections of sulfate concentrations for NE lakes,
Priority Classes A - E, under current and decreased deposition 10-94
10-28 Comparison of MAGIC and ETD projections of pH for NE lakes, Priority Classes
A - E, under current and decreased deposition 10-95
10-29 Comparisons of projected change in ANC under current and decreased
deposition for NE Priority Classes A - E, using ETD and MAGIC 10-99
10-30 Comparisons of projected change in sulfate concentrations under current and
decreased deposition for NE Priority Classes A - E, using ETD and MAGIC 10-100
10-31 Comparisons of projected change in pH under current and decreased deposition
for NE Priority Classes A - E, using ETD and MAGIC 10-101
10-32 Box and whisker plots of ANC distributions projected using ETD in 10-year
intervals for NE lakes, Priority Classes A - E 10-103
xxiii
-------
FIGURES (continued) Page
10-33 Box and whisker plots of sulfate distributions projected using ETD in 10-year
intervals for NE lakes, Priority Classes A - E 10-104
10-34 Box and whisker ptots of pH projected using ETD in 10-year intervals for NE lakes,
Priority Classes A - E 10-105
10-35 Box and whisker plots of ANC distributions in 10-year intervals using MAGIC for
NE lakes, Priority Classes A - E 10-106
10-36 Box and whisker plots of sulfate distributions in 10-year intervals using MAGIC
for NE lakes, Priority Classes A - E 10-107
10-37 Box and whisker plots of pH In 10-year intervals using MAGIC for NE lakes,
Priority Classes A - E 10-108
10-38 ETD ANC distributions at year 10 and year 50 for NE lakes, Priority Classes A -
E, under current and decreased deposition 10-109
10-39 MAGIC ANC distribution at year 10 and year 50 for NE lakes, Priority Classes
A - E, under current and decreased deposition 10-110
10-40 ETD sulfate distributions at year 10 and year 50 for NE lakes, Priority Classes
A - E, under current and decreased deposition 10-111
10-41 MAGIC sulfate distributions at year 10 and year 50 for NE lakes, Priority Classes
A - E, under current and decreased deposition 10-112
10-42 Comparison of ANC projections using ETD, ILWAS, and MAGIC for NE lakes,
Priority Classes A and B, under current and decreased deposition 10-114
10-43 Comparison of sulfate projections using ETD, ILWAS, and MAGIC for NE lakes,
Priority Classes A and B, under current and decreased deposition 10-115
10-44 Comparison of pH projections using ETD, ILWAS, and MAGIC for NE lakes,
Priority Classes A and B, under current and decreased deposition 10-116
10-45 Comparison of ANC projections under current and decreased deposition for NE
lakes, Priority Classes A and B, at year 20 and year 50 using ETD, ILWAS, and MAGIC. 10-122
10-46 Comparison of sulfate projections under current and decreased deposition for NE
lakes, Priority Classes A and B, at year 20 and year 50 using ETD, ILWAS, and MAGIC. 10-123
10-47 Comparison of pH projections under current and decreased deposition for NE
lakes, Priority Classes A and B, at year 20 and year 50 using ETD, ILWAS, and MAGIC. 10-124
10-48 Box and whisker plots of ANC distributions in 10-year intervals projected using
ETD for NE lakes, Priority Classes A and B 10-126
10-49 Box and whisker plots of ANC distributions in 10-year intervals projected using
ILWAS for NE lakes, Priority Classes A and B 10-127
10-50 Box and whisker plots of ANC distributions in 10-year intervals projected using
MAGIC for NE lakes, Priority Classes A and B 10-128
10-51 Box and whisker plots of sulfate distributions in 10-year intervals projected using
ETD for NE lakes, Priority Classes A and B 10-129
10-52 Box and whisker plots of sulfate distributions in 10-year intervals projected using
ILWAS for NE lakes, Priority Classes A and B 10-130
10-53 Box and whisker plots of sulfate distributions in 10-year intervals projected using
MAGIC for NE lakes, Priority Classes A and B 10-131
10-54 Box and whisker plots of pH distributions in 10-year intervals projected using
ETD for NE lakes, Priority Classes A and B 10-132
10-55 Box and whisker plots of pH distributions in 10-year intervals projected using
ILWAS for NE lakes, Priority Classes A and B 10-133
10-56 Box and whisker plots of pH distributions in 10-year intervals projected using
MAGIC for NE lakes, Priority Classes A and B 10-134
10-57 ETD ANC population distributions at year 10 and year 50 for current and
decreased deposition 10-135
10-58 ILWAS ANC population distributions at year 10 and year 50 for current and
decreased deposition 10-136
10-59 MAGIC ANC population distributions at year 10 and year 50 for current and
decreased deposition 10-137
10-60 ETD sulfate population distributions at year 10 and year 50 for current and
decreased deposition 10-138
xxiv
-------
FIGURES (continued) Page
10-61 ILWAS sulfate population distributions at year 10 and year 50 for current and
decreased deposition .- 10-139
10-62 MAGIC sulfate population distributions at year 10 and year 50 for current and
decreased deposition 10-140
10-63 MAGIC ANC and sulfate projections for SBRP streams, Priority Classes A - E,
at year 20, year 50, year 100, and year 200 under current and increased deposition. ... 10-142
10-64 MAGIC pH projections for SBRP streams, Priority Classes A - E, at year 20, year
50, year 100, and year 200 under current and increased deposition 10-143
10-65 Box and whisker plots of ANC distributions in 10-year intervals projected using
MAGIC for SBRP streams, Priority Classes A - E, for current and increased deposition. . 10-149
10-66 Box and whisker plots of sulfate distributions in 10-year intervals projected
using MAGIC for SBRP streams, Priority Classes A - E, for current and
increased deposition 10-150
10-67 Box and whisker plots of pH distributions in 10-year intervals projected Using
MAGIC for SBRP streams, Priority Classes A - E, for current and increased
deposition 10-151
10-68 MAGIC ANC population distributions at year 10 and year 50 for current and
increased deposition, SBRP streams, Priority Classes A - E 10-153
10-69 MAGIC sulfate population distributions at year 10 and year 50 for current and
increased deposition, SBRP streams, Priority Classes A - E 10-154
10-70 Comparison of ILWAS and MAGIC projections for ANC at years 0, 20, and 50 for
SBRP streams, Priority Classes A and B, under current and increased deposition 10-156
10-71 Comparison of ILWAS and MAGIC projections for sulfate concentration at years
0, 20, and 50 for SBRP streams, Priority Classes A and B, under current and
increased deposition 10-157
10-72 Comparison of ILWAS and MAGIC projections for pH at years 0, 20, and 50 for
SBRP streams, Priority Classes A and B, under current and increased deposition 10-158
10-73 Box and whisker plots for ANC distributions in 10-year intervals projected
using ILWAS for SBRP streams, Priority Classes A and B, for current and
increased deposition 10-164
10-74 Box and whisker plots for ANC distributions in 10-year intervals projected using
MAGIC for SBRP streams, Priority Classes A and B, for current and
increased deposition 10-165
10-75 Box and whisker plots for sulfate distributions in 10-year intervals projected
using ILWAS for SBRP streams, Priority Classes A and B, for current and
increased deposition 10-166
10-76 Box and whisker plots for sulfate distributions in 10-year intervals projected
using MAGIC for SBRP streams, Priority Classes A and B, for current and
increased deposition 10-167
10-77 Box and whisker plots for pH distributions in 10-year intervals projected using
ILWAS for SBRP streams, Priority Classes A and B, for current and
increased deposition 10-168
10-78 Box and whisker plots for pH distributions in 10-year intervals projected
using MAGIC for SBRP streams, Priority Classes A and B, for current and
increased deposition 10-169
10-79 ILWAS ANC population distributions at year 10 and year 50 for current and
increased deposition, SBRP Priority Class A and B streams 10-170
10-80 MAGIC ANC population distributions at year 10 and year 50 for current and
increased deposition, SBRP Priority Class A and B streams 10-171
10-81 ILWAS sulfate population distributions at year 10 and year 50 for current and
increased deposition, SBRP Priority Class A and B streams 10-172
10-82 MAGIC sulfate population distributions at year 10 and year 50 for current and
increased deposition, SBRP Priority Class A and B streams 10-173
10-83 Comparison of projected sulfate versus sulfate steady-state concentrations using
ETD, iLWAS, and MAGIC for NE lakes 10-175
XXV
-------
FIGURES (continued) Page
10-84 Comparison of projected sulfate concentrations under decreased deposition
with the current sulfate steady-state concentrations using ETD, ILWAS, and
MAGIC for NE lakes 10-176
10-85 Comparison of projected sulfate concentrations between models for NE lakes
after 50 years under current and decreased deposition 10-177
10-86 Comparison of projected sulfate versus sulfate steady-state concentrations for
SBRP streams using ILWAS and MAGIC under both current and increased deposition. . . 10-179
10-87 Comparison of projected ANC between models in NE lakes after 50 years under
current and decreased deposition 10-180
10-88 Projected changes in ANC as a function of changes in sulfate for NE lakes
using ETD, ILWAS, and MAGIC for current and decreased deposition 10-181
10-89 Comparison of pH - ANC relationship for each of the models 10-183
10-90 Comparison of projected pH values between models for NE lakes after 50 years
under current and decreased deposition 10-184
10-91 Comparison of projected changes in calcium and magnesium versus changes in
sulfate using ILWAS and MAGIC for NE lakes 10-185
10-92 Change in median ANC, calcium and magnesium, and sulfate concentrations
projected for NE lakes using MAGIC under current and decreased deposition 10-186
10-93 Comparison of the change in pH after 50 years as a function of the initial calibrated
pH for MAGIC, ETD and ILWAS on northeastern lakes 10-188
10-94 Comparisons of projected ANC and sulfate concentrations and pH between
ILWAS and MAGIC after 50 years for SBRP streams 10-189
10-95 Comparison of projected AANC and Asulfate relationships in SBRP Priority
Class A and B streams using ILWAS and MAGIC 10-190
10-96 Comparison of projected AANC and Asulfate relationships and A(calcium and
magnesium) and Asulfate relationships for SBRP Priority Class A - E streams
using MAGIC . 10-191
10-97 Comparison of projected A(calcium and magnesium) and Asulfate relationships
for SBRP Priority Class A and B streams using ILWAS and MAGIC 10-193
10-98 Change in median ANC, calcium and magnesium, and sulfate concentrations
projected for SBRP streams under current and increased deposition using MAGIC 10-194
10-99 Comparison of the change in pH after 200 years as a function of the initial
calibrated pH for MAGIC on SBRP streams, Priority Classes A - E 10-196
10-100 Comparison of projected MAGIC change in pH versus derived pH after 50
years for NE lakes 10-201
XXVI
-------
PLATES
PLATE PAGE
1-1 Direct/Delayed Response Project study regions and sites 1-5
1-2 Sulfur retention and wet sulfate deposition for National Surface Water Survey and
National Stream Survey regions In the eastern United States. 1-13
1-3 Changes In sulfur retention in the Southern Blue Ridge Province as projected by
MAGIC for constant sulfur deposition 1-14
1-4 Changes in median ANC of northeastern lakes at 50 years as projected by MAGIC . . 1-18
1-5 Changes In median ANC of Southern Blue Ridge Province stream reaches at 50
years as projected by MAGIC 1-21
2-1 Direct/Delayed Response Project study regions and sites 2-5
5-1. ANC of DDRP lakes by ANC group 5-24
5-2 Final DDRP classification of lake hydrologic type - Subregion 1A 5-32
5-3 Final DDRP classification of lake hydroiogic type - Subregion 1B 5-33
5-4 Final DDRP classification of lake hydrologic type - Subregion 1C 5-34
5-5 Rnal DDRP classification of lake hydrologic type - Subregion 1D 5-35
5-6 Final DDRP classification of lake hydrologic type - Subregion 1E 5-36
5-7 Example of watershed soil map (including pedon site location) 5-75
5-8 Example of watershed vegetation map 5-76
5-9 Example of depth-to-bedrock map 5-77
5-10 Example of watershed land use map 5-78
5-11 Example of watershed geology map 5-79
5-12 Example of 40-ft contour delineations on a 15' topographic map 5-89
5-13 Example of combination buffer: (A) stream and 30-m linear buffer for streams,
(B) wetlands and 30-m linear buffer for wetlands. (C) elevational buffer for lake,
and (D) combination of all preceding buffers 5-91
5-14 ADS and NCDC sites linked with DDRP study sites for NE Subregion 1A 5-166
5-15 ADS and NCDC sites linked with DDRP study sites for NE Subregion 1B 5-167
5-16 ADS and NCDC sites linked with DDRP study sites for NE Subregion 1C 5-168
5-17 ADS and NCDC sites linked with DDRP study sites for NE Subregion 1D 5-169
5-18 ADS and NCDC sites linked with DDRP study sites for NE Subregion 1E 5-170
5-19 ADS and NCDC sites linked with DDRP study sites for the SBRP 5-171
5-20 DDRP study sites relative to distance from Atlantic Coast (<10 km, 10-50 km, >50 km). 5-178
5-21 Pattern of typical year sulfate deposition for the DDRP NE study sites 5-184
5-22 Pattern of typical year sulfate deposition for the DDRP study sites in Subregion 1A. ... 5-185
5-23 Pattern of typical year sulfate deposition for the DDRP study sites in Subregion 1B. ... 5-186
5-24 Pattern of typical year sulfate deposition for the DDRP study sites in Subregion 1C. ... 5-187
5-25 Pattern of typical year sulfate deposition for the DDRP study sites in Subregion 1D. ... 5-188
5-26 Pattern of typical year sulfate deposition for the DDRP study sites in Subregion 1E. ... 5-189
5-27 Pattern of typical year sulfate deposition for the DDRP SBRP study sites 5-190
5-28 Pattern of LTA sulfate deposition for the DDRP NE study sites. 5-193
5-29 Pattern of LTA sulfate deposition for the DDRP study sites in Subregion 1A 5-194
5-30 Pattern of LTA sulfate deposition for the DDRP study sites in Subregion 1B 5-195
5-31 Pattern of LTA sulfate deposition for the DDRP study sites in Subregion 1C 5-196
5-32 Pattern of LTA sulfate deposition for the DDRP study sites in Subregion 1D 5-197
5-33 Pattern of LTA sulfate deposition for the DDRP study sites in Subregion 1E 5-198
5-34 Pattern of LTA sulfate deposition for the DDRP SBRP study sites 5-199
xxvii
-------
PLATES (continued) Page
7-1 Median percent sulfur retention by NSWS Subregion 7-35
7-2 Regional percent sulfur retention by major land resource area (MLRA) based on
target populations (ELS and NSS sites) 7-45
11-1 Sulfur retention and wet sulfate deposition for National Surface Water Survey subregions
In the eastern United States 11-2
11-2 Changes in sulfur retention in the Southern Blue Ridge Province as
projected by MAGIC for constant sulfur deposition 11-5
11-3 Changes in median ANC of northeastern lakes at 50 years as projected by MAGIC . . . 11-10
11-4 ANCs of northeastern lakes versus time, as projected by MAGIC for
constant suifur deposition 11-12
11-5 ANCs of northeastern lakes versus time, as projected by MAGIC for
decreased sulfur deposition 11-13
11-6 Changes in median pH of northeastern lakes at 50 years as projected by MAGIC ... 11-16
11-7 Changes In median ANC of Southern Blue Ridge Province stream reaches at 50
years as projected by MAGIC 11-18
11-8 ANCs of Southern Blue Ridge Province stream reaches versus time, as projected
by MAGIC for constant sulfur deposition 11-21
11-9 ANCs of Southern Blue Ridge Province stream reaches versus time, as projected
by MAGIC for increased sulfur deposition 11-22
11-10 Changes In pH of SBRP stream reaches as projected by MAGIC 11-24
11-11 Granges in pH of SBRP stream reaches as projected by ILWAS 11-25
xxviii
-------
PRIMARY CONTRIBUTORS TO THE DORP REPORT
The Direct/Delayed Response Project and this Review Draft Report represent the efforts of many
scientists, technical and support staff. The primary contributors to this report are noted here.
Section 1: Executive Summary
M. R. Church, U.S. Environmental Protection Agency
Section 2: Introduction
M. R. Church, U.S. Environmental Protection Agency
Section 3: Processes of Acidification
P. W. Shaffer, NSI Technology Services Corp.
G. R. Holdren, NSI Technology Services Corp.
M. R. Church, U.S. Environmental Protection Agency
Section 4: Project Approach
M. R. Church. U.S. Environmental Protection Agency
Section 5: Data Sources and Descriptions1
L J. Blume, U.S. Environmental Protection Agency
G. E. Byers, Lockheed Engineering and Sciences Co.
W. G. Campbell, NSI Technology Services Corp.
M. R. Church, U.S. Environmental Protection Agency
D. A. Lammers, U.S.D.A. Forest Service
J. J. Lee, U.S. Environmental Protection Agency
L H. Liegel, U.S.D.A. Forest Service
D. C. Mortenson, NSI Technology Services Corp.
C. J. Palmer, NSI Technology Services Corp.
M. L Papp, Lockheed Engineering and Sciences Co.
B. P. Rochelle, NSI Technology Services Corp.
D. D. Schmoyer, Martin Marietta Energy Systems, inc.
K. W. Thornton, FTN & Associates, Ltd.
R. S. Turner, Oak Ridge National Laboratory
R. D. Van Remortel, Lockheed Engineering and Sciences Co.
Section 6: Regionalization of Analytical Results
D. L Stevens, Eastern Oregon State University
K. W. Thornton, FTN & Associates, Ltd.
Section 7: Watershed Sulfur Retention
B. P. Rochelle, NSI Technology Services Corp.
M. R. Church, U.S. Environmental Protection Agency
P. W. Shaffer, NSI Technology Services Corp.
G. R. Holdren, NSI Technology Services Corp.
Section 8: Level I Statistical Analyses
M. G. Johnson, NSI Technology Services Corp.
R. S. Turner, Oak Ridge National Laboratory
D. L Cassell, NSI Technology Services Corp.
D. L Stevens, Eastern Oregon State University.
M. B. Adams, Automated Systems Group, \nc.
C. C. Brandt, Oak Ridge National Laboratory
W. G. Campbell, NSI Technology Services Corp.
M. R. Church, U.S. Environmental Protection Agency
G. R. Holdren, NSI Technology Services Corp.
L H. Liegel, U.S.D.A. Forest Service
xxix
-------
Section 8: Level I Statistical Analyses (continued):
B. P. Rochelle, NSI Technology Services Corp.
P. F. Ryan, University of Tennessee
D. D. Schmoyer, Martin Marietta Energy Systems, Inc.
P. W. Shaffer, NSI Technology Services Corp.
D. A. Wolf, Martin Marietta Energy Systems, Inc.
Section 9: Level II Single-Factor Time Estimates1
G. R. Holdren, NSI Technology Services Corp.
M. G. Johnson, NSt Technology Services Corp.
C. I. Lift. Utah State University
P. W. Shaffer, NSI Technology Services Corp.
Section 10: Level III Dynamic Watershed Models
K. W. Thornton, FTN & Associates, Ltd.
D. L Stevens, Eastern Oregon State University
M. R. Church, U.S. Environmental Protection Agency
C. I. Lift, Utah State University
Extramural Cooperators Providing Modelling Expertise and Support:
C. C. Brandt, Oak Ridge National Laboratory
B. J. Cosby, University of Virginia
S. A. Gherini, Tetra-Tech, Inc.
G. M. Hornberger, University of Virginia
M. Lang, Tetra-Tech, Inc.
S. Lee, University of Iowa
R. K. Munson, Tetra-Tech, Inc.
R. M. Newton, Smith College
N. P. Nikolaidis, University of Connecticut
P. F. Ryan, University of Tennessee
J. L Schnoor, University of Iowa
R. S. Turner, Oak Ridge National Laboratory
D. M. Wolock, U.S. Geological Survey
Section 11: Integration and Summary
M. R. Church, U.S. Environmental Protection Agency
P. W. Shaffer, NSI Technology Services Corp.
Contributors to this section listed alphabetically
Beginning on this line, remaining contributors listed alphabetically
XXX
-------
ACKNOWLEDGMENTS
The performance of this portion of the Direct/Delayed Response Project (DDRP) and the
preparation of this report have required the efforts of hundreds of scientists and support personnel. We
acknowledge here a few of those persons who made particularly outstanding contributions. To all the
others who helped us, but who are not named here, we also extend our sincere thanks.
William Ruckieshaus led the way in calling for the initiation of the DDRP and Lee Thomas showed
a continued and very patient interest in seeing that it was completed properly. We thank them for their
foresight and leadership.
Courtney Rlordan and Gary Foley of the EPA Office of Research and Development (ORD) provided
much encouragement and support for the Project throughout its development and implementation. We
thank them for their appreciation of the technical complexity of the task.
Rick Linthurst, the first Director of the Aquatic Effects Research Program (AERP), played an
absolutely critical role in the development and nurturing of the Project during its early years. We greatly
appreciate his early and continuing commitment to the DDRP. Dan McKenzie, as Director of the AERP,
provided important continuing support for the Project and we thank him for his efforts in helping guide
this phase of the Project to its conclusion.
Tom Murphy, Laboratory Director for EPA's Environmental Research Laboratory-Corvallis (ERL-C),
and Ray Wilhour, Bob Lackey and Spence Peterson, Branch Chiefs for ERL-C, have all supported the
Project and its staff from the first to the last. We thank them for their support.
Dwain Winters and Brian McLean from the Office of Air and Radiation at EPA-Headquarters
provided insight and suggestions for analyses of particular relevance to questions of Agency policy.
We thank them for their interest and assistance.
xxxi
-------
Dixon Landers, Technical Director of the National Surface Water Survey, Jay Messer, Technical
Director of the Pilot Stream Survey, and Phil Kaufmann, Technical Director of the National Stream Survey
and their staffs all provided valuable help in interpreting and correctly using their surface water chemistry
data. We thank especially Tim Sullivan, Joe Eilers, Jim Blick, Mark DeHaan, Alan Herlihy and Mark Mitch.
Jim Omemik (EPA), Andy Kinney (NSI) and Andy Herstrom (NSI) provided many interesting hours
of instruction and discussion on the topics of physical geography and the proper use and application of
Geographic Information Systems. Our efforts in these technical areas have certainly profited from their
valuable advice and counsel.
Bill Fallen (ORD), Chuck Frank (EPA) and his staff, Linda Looney (EPA), and Cindy Burgeson (NSI
Technology Services Corp.) all have provided much administrative assistance to help keep the Project
moving in the right direction and at the pace required. We thank them all for their efforts and assistance.
Many landowners and state and government agencies allowed us to map and sample soils on
their properties. We thank them for permission to do so.
The cooperation of the U.S. Department of Agriculture (USDA) Soil Conservation Service (SCS)
was essential to the completion of the DDRP Soil Survey. People in the SCS state offices who were
responsible for mapping of DDRP watersheds and obtaining the soil descriptions and samples included
Ed Sautter, Roy Shook (Connecticut and Rhode Island); Gene Grice, Steve Hundley (Massachusetts); Dick
Babcock, Bob Joslin, Kenny LaFlamme (Maine); Sid Pilgrim, Henry Mount (New Hampshire); Fred Gilbert,
Keith Wheeler, Will Hanna (New York); Garland LJpscomb, George Martin (Pennsylvania); Dave Van
Houten (Vermont); Talbert Gerald, Bob Wilkes (Georgia); Horace Smith, Andy Goodwin (North Carolina);
Darwin Newton, David Lewis (Tennessee); Niles McLoda (Virginia). In addition, more than 100 soil
scientists were involved in mapping and sampling.
Regional consistency and comparability was greatly assisted by. the efforts of people at the SCS
National Technical Centers, especially Oliver Rice, Ted Miller (Northeast) and Larry Ratliff (South). The
xxxii
-------
continuing support of DDRP activities by Milt Meyer, Ken Hinkiey, and Dick Arnold of the SCS National
Office was extremely helpful.
John Warner, former SCS Assistant State Soil Scientist for New York was the Regional
Correlator/Coordinator of the Soil Survey for both the Northeast and Mid-Appalachian Regions. Hubert
Byrd, former State Soil Scientist for North Carolina, served as RCC for the SBRP Soil Survey.
Elissa Levine and Harvey Luce (University of Connecticut), Bill Waltman and Ray Bryant (Cornell
University), Cheryl Spencer and Ivan Fernandez (University of Maine), Steve Bodine and Peter Veneman
(University of Massachusetts), Bill Smith and Lee Norfleet (Clemson University), and Dave Utzke and
Marilew Bartling (University of Tennessee) supervised the operation of the soils preparation laboratories
for the DDRP Soil Survey.
A large and dedicated staff at EPA's Environmental Monitoring and Systems Laboratory-Las Vegas
(EMSL-LV) played an absolutely crucial role in support of the DDRP Soil Survey. Gareth Pearson and
Bob Schonbrod provided supervisory guidance for the DDRP Soil Survey activities at EMSL-LV. Lou
Blume (EPA) served as Technical Monitor for the program and was responsible for delivery of verified
field, soil preparation laboratory, and analytical databases. Lou Blume was responsible for contracting
and management of soil preparation laboratories and analytical laboratories and for the delivery of
operations reports, quality assurance reports, methods manuals and field sampling manuals for the Soil
Survey. Mike Papp of Lockheed Engineering and Sciences Corporation (LESC) was responsible for
delivery of verified field, soil preparation and analytical databases for the Soil Survey. Rick Van Remortel
(LESC) assisted in the verification of the SBRP analytical database and in the preparation of laboratory
operations and quality assurance reports. Bill Cole (LESC) was the Task Lead for the verification of the
analytical database for the NE and assisted in the preparation of the methods manual and quality
assurance report for the NE Soil Survey. Gerry Byers (LESC) assisted in the preparation of methods
manuals and quality assurance reports for the NE and SBRP. Marilew Bartling (LESC) served as the Task
Lead for the verification of Soil Survey data for the SBRP, served as a manager of a soil preparation
laboratory for the SBRP Soil Survey and contributed to the operations and quality assurance reports for
xxxiii
-------
the SBRP. Rod Slagle (LESC) served as the DDRP soils database manager at EMSL-LV. Steve Simon
and Dan Hillman (LESC) assisted in methods development and project implementation early in the
Project. Craig Palmer of the Environmental Research Center of the University of Nevada-Las Vegas
provided Invaluable technical assistance on quality assurance of soils analytical data.
Deborah Coffey (NSI) played a critical role in ensuring the quality of the watershed and soils data
gathered for the Project. She either had a major responsibility for, or assisted in, the development of
data quality objectives, field sampling manuals, laboratory methods manuals, field operations reports, field
quality assurance reports and numerous other facets of the Soil Survey. We thank her for her unswerving
attention to detail. Jeff Kern (NSI) has also assisted in helping to assure the quality of field and
laboratory data.
Other scientists who made major contributions to the design of the soil survey activities included
Stein Buol (North Carolina State University), John Ferwerda (University of Maine-Orono), Maurice
Mausbach (Soil Conservation Service), Ben Hajek (Auburn University), John Reuss (Colorado State
University), Mark David (University of Illinois), and Fred Kaisaki (Soil Conservation Service).
Phil Arberg (EPA) and Dave Williams (LESC) of EMSL-LV were responsible for acquisition and
Interpretation of aerial photography of the DDRP watersheds.
Numerous extramural cooperators assisted in this Project. Jack Cosby, George Homberger, Pat
Ryan and David Wolock (University of Virginia), Jerry Schnoor, Tom Lee, Nikofaos Nikolaldis, Konstantine
Georgakakos and Harihar Rajaram (University of Iowa), Steve Gherini, Ron Munson and Margaret Lang
(Tetra-Tech, Inc.), Carl Chen and Louis Gomez (Systech, Inc.) all assisted in watershed modelling
analyses. Bob Newton of Smith College assisted in gathering supplementary watershed data for use in
calibrating the models to the Special Interest lake/watersheds in the Adirondacks. John Reuss and Mark
Walthall of Colorado State University and Tom Voice of Michigan State University performed investigations
of processes of base cation supply and sulfate adsorption, respectively, that assisted us in interpreting
our Soil Survey data and in modelling soil responses. Warren Gebert, Bill Krug, David Graczyk and Greg
xxxiv
-------
Allord of the U.S. Geological Survey (Madison, Wisconsin) supplied runoff data and maps that were
crucial to the Project. Wayne Swank and Jack Waide of the USDA Forest Service cooperated with the
Project in allowing us to use data gathered by the Coweeta Hydrologic Laboratory. Jack Waide also
provided many insights into the workings of watersheds in the Southern Blue Ridge and in the application
of watershed simulation models. Tony Olsen, Sally Wampler and Jeanne Simpson of Battelle Pacific
Northwest Laboratories provided a great deal of information on estimates of wet deposition to sites of
interest in the Eastern United States. Tony Olsen also assisted in editing text describing analyses of the
wet deposition data. Robin Dennis and Terry Clark of the EPA's Atmospheric and Exposure Assessment
Laboratory-Research Triangle Park and Steve Seilkop of Analytical Services, Incorporated, provided key
information on estimates of atmospheric dry deposition. Steve Undberg of Oak Ridge National
Laboratory and Bruce Hicks and Tilden Myers of the National Oceanographic and Atmospheric
Administration provided considerable assistance in the form of discussions and preliminary data on rates
of atmospheric dry deposition. We thank all of these cooperators for their assistance.
No project of the magnitude of the DDRP can be successfully completed without the assistance
of peer reviewers. The DDRP benefitted immensely from peer review comments all the way from its
inception to the completion of this report.
The following scientists served as reviewers of the initial Review Draft Report: David Grigal of the
University of Minnesota, Peter Chester, R. Skeffington and D. Brown of the Central Electricity Generating
Board (London), Jerry Elwood of Oak Ridge National Laboratory, John Melack of the University of
California - Santa Barbara, Phil Kaufmann of Utah State University, Bruce Hicks of the National
Oceanographic and Atmospheric Administration, Gary Stensland of the Illinois State Water Survey, Jack
Waide of the USDA Forest Service, David Lam of the National Water Research Institute (Burlington,
Ontario), Nils Christophersen of the Institute of Hydrology (Wallingford Oxon, Great Britain), Bill McFee
of Purdue University, Steve Norton of the University of Maine, Scott Overton of Oregon State University,
Ken Reckhow of Duke University, Dale Johnson of the Desert Research Institute (Reno, Nevada), and
Gray Henderson of the University of Missouri. We thank these scientists for their efforts in reviewing a
long and complex document. We especially thank Dave Grigal (Chairman), Jerry Elwood, John Melack
xxxv
-------
and Phil Kaufmann who served on the Overview Committee of reviewers. This report benefitted greatly
from the comments and constructive criticisms of all of these reviewers.
f
Numerous other scientists also served as reviewers over the years of individual aspects of the
Project or of the Project as a whole. We thank them also for helping us to improve the quality of the
work that we performed.
Dave Marmorek, Mike Jones, Tim Webb and Dave Barnard of ESSA, Ltd. provided much valuable
assistance in the planning of various phases of the DDRP. Their assistance in this planning was
invaluable.
John Berglund of InstaGraphics, Inc. prepared many of the figures that appear in this report. We
thank him for the fine job that he did.
A majority of the word processing throughout the DDRP and, especially, for this report was done
by Carol Roberts of NSI. We thank Carol for her many, many hours of diligent work and for her
forbearance in helping us in our attempts to get everything "exactly right". Significant word processing
support was also provided by Laurie Ippolfti (NSI), Amy Vickland (USDA Forest Service), Lana McDonald,
Rose Mary Hall and Deborah Pettiford of Oak Ridge National Laboratory, and Eva Bushman and Suzanne
Labbe of Action Business Services.
Penelope Kellar and Perry Suk of Kilkelly Environmental Associates performed truly amazing tasks
in editing both the Review Draft and Final Draft of this report. The job could not have been completed
on time without their efforts. Ann Hairston (NSI), Amy Vickland (USDA Forest Service), Susan Christie
(NSI) and Linda Allison (ORNL) also provided important editorial assistance.
The DDRP Technical Director sincerely thanks all of the Project staff and extramural cooperators
for their unquenchable enthusiasm and dedication to seeing that this very tough job was done correctly.
Good work gang...thank you.
xxxvi
-------
SECTION 10
LEVEL III ANALYSES - DYNAMIC WATERSHED MODELLING
10.1 INTRODUCTION
Previous sections have discussed (1) the principal theories and basic processes of acidification
(Section 3); (2) the relationship among atmospheric deposition, watershed attributes, and surface water
chemistry (Section 8); and (3) future changes that might occur in watershed sulfate adsorption and base
cation exchange (Section 9) for up to the next 200 years. This section discusses the Level ill Analyses -
the application of dynamic watershed models in projecting future changes in surface water chemistry.
Three terms are used to describe simulations of future change:
• Predict - to estimate some current or future condition within specified confidence limits on
the basis of analytical procedures and historical or current observations.
• Forecast - to estimate the probability of some future event or condition as a result of rational
study and analysis of available data.
• Project - to estimate future possibilities based on rational study and current conditions or
trends.
The distinction among these three terms and definitions is the intended use of the model output. Level
III analyses are defined, and intended to be used, as projections.
Predictions typically are made to compare different scenarios, controls, or management options.
Predictions can be performed within specified confidence limits because of previous model evaluations,
testing, applications, and comparisons with measured data for a variety of system types. Model
predictions of various surface water attributes are legally required for many proposed management
strategies that range, for example, from examining potential alterations of hydrologic regimes due to land
use modifications to estimating mixing zones for effluent discharges to estimating phytoplankton response
to nutrient reduction. Predictions generally are performed for short time periods (e.g., single events, parts
10-1
-------
of a season, or a few years) and focus on before-after comparisons such as water quality before and
after wasteload reductions or plankton biomass before and after nutrient reductions.
Forecasts convey some estimate of the likelihood or probability that various conditions or events
will occur in the future. Daily weather forecasting, with associated probabilities of showers,
thunderstorms, etc., is a classic example of forecasting. This represents a short-term forecast. Weather
forecasts also are made for annual or decadal time frames. Rood forecasts can be short term (daily or
weekly), but also are made for long-term events such as the probability of 1QO-, 1000-, and 1,000,000-
year events (NRC, 1988).
Projections, in contrast, are not accompanied by estimates of the probability that any of the
conditions or events might occur in the future. Projections can be used as a basis for relative
comparisons among various emission or deposition scenarios. While the probability that a scenario will
occur cannot be estimated, projections do provide a relative basis for comparing costs and beneficial or
deleterious effects associated with different control or management strategies. This information generally
is relevant to policymakers and decisionmakers for evaluating different control strategies. The models
in the Level III analyses are being used in projecting, not in forecasting, the effects of alternative acidic
deposition scenarios on future changes in surface water acid-base chemistry.
In Level III Analyses integrated, process-oriented watershed models are used to project long-term
changes (i.e., up to 100 years) in surface water chemistry as a function of current and alternative levels
of sulfur deposition. The watershed models integrate our current understanding of how various processes
and mechanisms interact and respond to acidic deposition. These mechanisms include soil-water
interactions (including soil-water contact time), sulfur retention, base cation exchange and replacement
of base cations through mineral weathering, and other watershed processes (e.g., vegetative uptake,
in-lake processes, organic interactions). However, the present study does not establish the adequacy of
the formulations that implement these processes, the mode of spatial aggregation of data, and the
calibration approaches used for long-term acidification projections.
10-3
-------
The three watershed models that have been applied are the Enhanced Trickle Down (ETD),
Integrated Lake-Watershed Acidification Study (ILWAS), and Model of Acidification of Qroundwater in
Catchments (MAGIC). The DDRP is an applied project and has used existing techniques and models for
these analyses. The use and application of these models to achieve the objectives of the DDRP was
approved by peer reviewers in accordance with the Agency's standard competitive funding process and
requirement for external review of environmental data collection programs (Section 4.4.3).
This section presents
• dynamic watershed models used in the Level III Analyses,
• operational assumptions of these analyses,
• watershed prioritization procedures,
• preparation of the modelling datasets (specifically identifying any differences required for Level
III Analyses compared to Level I and II Analyses),
• general modelling approach,
• model calibration and confirmation,
• model sensitivity analyses,
* regional projection refinements,
• model projection and uncertainty procedures,
• regional population estimates and uncertainties,
• regional comparisons and uncertainties, and
• discussion and conclusions.
10.2 DYNAMIC WATERSHED MODELS
Processes that influence the acid-base chemistry of surface water, and that were considered by the
NAS Panel (NAS, 1984), were described In Section 3. Although these processes can be individually
identified, discussed, and represented empirically, they are not independent and do not occur in isolation
from other processes. The observed lake or stream response to acidic deposition represents the
integrated response of many watershed and lake/stream processes controlling surface water chemistry.
10-3
-------
To project the future response of a lake or stream to acidic deposition, therefore, requires dynamic
watershed models that incorporate and integrate the important processes controlling the acid-base
chemistry of surface water.
Both dynamic and steady-state models can be used to project changes in surface water chemistry
as a function of changes in acidic deposition. A dynamic watershed model, however, simulates the time
trends of various lake, stream, and watershed constituents, such as ANC, pH, sulfate, calcium, soil base
saturation, and sulfate adsorption. A steady-state model can project conditions at only one time in the
future, the time at which steady state is achieved (i.e., ultimate constituent concentration or value), and
does not provide any indication of the changes that occurred between the initial conditions and steady-
state condition or concentration. It is the computation of concentrations and processes as a function of
time that distinguishes dynamic models from steady-state models.
Three dynamic watershed models were used to.project surface water chemistry for the next 50 to
100 years both at current and alternative levels of acidic deposition in the Northeast (NE):
* Enhanced Trickle Down (ETD) (Nikolaidis et al., 1988; Nikolaidis et al., 1989)
• Integrated Lake-Watershed Acidification Study (ILWAS) (Chen et al., 1983; Gherini et al., 1985)
• Model of Acidification for Groundwater in Catchments (MAGIC) (Cosby et al., 1985a,b,c)
Two of these three watershed models also are being used to project changes in surface water chemistry
in the Southern Blue Ridge Province (SBRP) • MAGIC and ILWAS.
Although each model incorporates the processes considered important in controlling the acid-base
chemistry of surface water, process resolution and detail vary significantly among the models. Some of
the processes included in the three models and their spatial/temporal resolution are compared in Table
10-1. The use of multiple models is important because:
10-4
-------
Table 10-1. Major Processes Incorporated in the Dynamic Model Codes (Parentheses Indicate
Umited Treatment of Process, and Dashes Indicate Processes not Included in a Code) (Jenne
et al., 1989)
MAGIC/TOPMODEL ETD/PEN ILWAS
Atmospheric Processes
- Dry deposition XXX
- Wet deposition XXX
Hvdroioaical Processes
• Snow sublimation - X X
- Evapotranspiration XXX
- Interception storage (X)a - X
-Snowmelt XXX
- Overiand flow XXX
- Soil freezing - X X
- Macropore flow X - -
- Unsaturated subsurface flow XXX
- Saturated subsurface flow XXX
- Stream flow X - X
- Lake stratification - - X
- Lake ice formation . - . - X
Geochemical Processes
- Carbonic acid chemistry X X X
- Aluminum chemistry X - X
- Organic acid chemistry X - X
-Weathering XXX
- Anion retention XXX
- Cation exchange XXX
Bioaeochemical Processes
- SO/' reduction in lake (X)b X X
- Nitrification in soil (X)b - X
- Nutrient uptake (X)b - X
- Canopy interactions (X)a - X
- Litter decay (X)a - X
- Root respiration (X)a - X
a Cosby, B.J. (written review comments, 1988) considers that canopy interactions and root decay and respiration are
implicitly included in the MAGIC code by use of a dry deposition factor and by designation of CO2 partial pressure in
soils and surface waters.
Sulfate reduction, nitrification, and uptake of ions can be simulated with the MAGIC code by specifying uptake rates of
SO* and NH4* for various hydroiogic compartments.
10-5
-------
• the level of detail by which each process or mechanism is represented varies between
models, reflecting the relative importance of each process in the systems for which the model
was first developed;
• Identification of similar key watershed parameters and processes in each model and their
relationship to measured watershed characteristics provides greater confidence in conclusions
about which factors influence the acid-base chemistry of surface water; and
• long-term data sets for model validation/verification do not exist, so model accuracy and
precision for long-term projections is unknown; however, similar projections of watershed
responses among the models provides greater confidence In the conclusions.
10.2.1 Enhanced Trickle Down (ETD1 Model
The ETD is a lumped parameter model based on the concept of ANC mass balance. Various
chemical processes in the ETD model, such as mineral weathering, sulfate adsorption and desorption,
and cation exchange, are incorporated as either consuming or producing ANC (Schnoor and Stumm,
1985). Rate expressions are used to describe mineral weathering, cation exchange, and sulfate reduction
reactions. Equilibrium expressions are used to describe carbonic acid chemistry and sulfate adsorption
and desorption. ETD explicitly incorporates mineral weathering rate reactions and sulfate reduction but
does not explicitly incorporate chemical reactions involving aqueous aluminum, nitrate, or organic acid
chemistry, although the ETD code does implicitly consider contributions to total acidity from these
constituents. Mass balance calculations are considered for ANC (equivalent to the modified Gran ANC),
sulfate, and chloride, with chloride considered to be a conservative constituent.
The Trickle Down model, a precursor to ETD, was formulated to perform assessments of the effects
of acidic deposition on a number of seepage lakes in the Upper Midwest. The objective of the modelling
effort was to provide a model with sufficient detail to calculate alkalinity concentrations in surface water,
soils, and ground water, but sufficiently simple to apply using a microcomputer with one master variable
(alkalinity) for acidic deposition assessments (Schnoor et al., 1984, I986a). Enhanced Trickle Down was
modified to include sulfate adsorption and desorption and improved hydrologic flowpaths (Nikolaidis,
1987). The hydrologic submodel simulates snowmelt, interflow, overland flow, groundwater flow, frost-
driven processes, seepage, and evapotranspiration (Nikolaidis et al., 1989). ETD is spatially partitioned
into three vertical components within the watershed: soil, unsaturated zone, and ground water. The
10-6
-------
watershed contributes to a lake compartment. The lake and terrestrial compartments are considered
areaily homogeneous. The temporal resolution of the ETD output generally is daily.
The meteorological and deposition input requirements for ETD are illustrated in Tables 10-2 and 10-
3. The chemical constituents projected in soil solution and surface water are listed In Table 10-4.
Because of the importance and pivotal rde that ANC has in the projections of surface water acidification
and chemical improvement, the components of the ANC calculations are shown for each of the three
models in Table 10-5. The ANC calculation for ETD corresponds with the ANC calculation for the
modified Gran titration method. These input requirements and output constituents are contrasted with
those included in ILWAS and MAGIC.
The ETD model was originally applied to Lakes Clara and Vandercook in northcentral Wisconsin
(Lin and Schnoor, 1986), as a joint effort by the U.S. EPA, U.S. Geological Survey, the Wisconsin
Department of Natural Resources, and the University of Iowa. ETD reproduced the seasonal and annual
changes in water chemistry for the short periods of record on these two lakes. ETD has since been
applied to several Adirondack lakes including Woods Lake, Panther Lake, and Clear Pond (see Appendix
A), other lakes in the Upper Midwest, and several lakes in the Sierra Nevada mountains of California
(Nikolaidis et al., 1988, 1989; Lee et al., in press).
10.2.2 Integrated Lake-Watershed Acidification Study (ILWAS1 Model
The ILWAS model is a process-oriented model that uses both equilibrium and rate-limited
expressions to describe mass balances for 18 chemical constituents (Table 10-4). The ILWAS algorithms
represent the effects of biogeochemical processes on surface water chemistry (see Table 10-1). Mass
balances for the major cations and anions and the effects of aqueous aluminum and organic acids on
surface water chemistry are incorporated in the model. The ILWAS model was formulated to simulate
the seasonal and long-term changes in water chemistry caused by acidic deposition. As a result, ILWAS
has a strong assessment capability (Huckabee et al., 1989). The ILWAS model contains three modules:
10-7
-------
Table 10-2. Meteorological Data Required by the Dynamics Model Codes (from Jenne et
al., 1989)
Meteorological
Data
MAGIC/TOPMODEL ETD/PEN
ILWAS
Interval for data
measurement
Precipitation
Relative humidity
or dewpoint
Monthly3
yearly
m
Dally
mm
a TOPMODEL runs with a daily time step.
Average values per month required.
Daily
cm
Min. air temperature
Max. air temperature
Ave. air temperature
Mean daylight hours
Cloud cover (fraction)
Atmospheric Pressure
Wind Speed
oC °c
°C °C
oC
% %
(unitiess) (unitless)b
mbars
km day"1 . m sec"1
10-8
-------
Table 10-3. Chemical Constituents in Wet and Dry Deposition Considered
by the MAGIC, ETD, and ILWAS Codes (from Jenne et al., 1989)
Constituent
SOx(g)
NO (g)
H*
Al (total)
car
Mg
rV
Na*
NH+4
so42-
X
NO,"
cr3
F
PO
ANC
TOC°
TIC3
H4Si04
Units
Interval
MAGIC
Wet Drya
OQb
(X)b
X X
X X
X X
X X
X X
X X
X
X X
X X
X X
Meq L*1 /interval
monthly or
yearly ave.
ETD ILWAS
Wet Dry Wet
XX X
X
X
X
X
X
X
X
X
XX X
X
X X
X X
X
X
meq m meq m mg L
-daily- volume
Dry
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
mg m"3
weighted
monthly average
The MAGIC code requires that dry deposition be expressed by means of a
dry deposition factor.
Cosby, B.J. (written communication, 1988) considers that SO(g) and N0x(g)
are implicitly included by means of the dry deposition factor.
Total organic carbon
Total inorganic carbon
10-9
-------
Table 10-4. Chemical Constituents Included in Soil Solutions
and Surface Water for the MAGIC, ETO, and ILWAS Codes (from
Jenne et al., 1989)
Chemical Constituent
ANC
Ca2+
Mg2+
K+
Na+
NH4+
H+
AI3+
AKOH)/" (n=1 to 4)
AKF),,3"" (n=1 to 6)
AKSO^r,3"" (n=1 to 2)
AI-R(a)
S042'
N03"
cr
F
H2P04-
H4Si04(aq)
C02(g)
C02(aq)
H2C03(aq)
HC03-
C032'
HR'°, R-
H2R"0, HR"', R"2'^
HnvifO t_l D»<"(b)
«tt . rlyiri
3 ' 2
LJQ*»*2- Qt>i3*
nr» i n
MAGIC ETD
X X
X
X
X
X
X
X X
X
X
X
X
X
X X
X
X X
X
X X
X X
X X
X X
X X
X
ILWA
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
AI-Refers to the various organic complexes of aluminum.
b R', R", and R1" refer to monoprotic, diprotio, and
triprotic organic acids, respectively.
10-10
-------
Table 10-5. Definitions of Acid Neutralizing Capacity (ANC) Used by the MAGIC, ETD,
and ILWAS Codes (Brackets indicate concentration in molar or molal units, and R',
R", and R'" represent mono-, di-, and triprotic organic acids, respectively.) ANC
Simulated by All Three Models is Equivalent to the Modified Gran ANC (from Jenne et
al., 1989).
Code
Units
ANC Definition
MAGIC
ETDa
ILWAS
L'
meq m
MeqL"
-1
ANC - [HCOgT + 2[CO321 + [OHT + [HR'T
+ 2[R"2T + [AI(OH)4T - [H+] - 3[AI3+
- 2[AI(OH)2+] - [AIOH2+]
ANC - [HCOgl + 2[CO32'J + [OH'] - [H*] + [R'l
ANC = [HC03*] + 2[CO32T + [OH'] + fH2R"T
+ 2[HR'"2"] + 3[R"13"] + [R1']
*] + 2[AI(OH)2+] + 3[AI(OH)3°
'2+
4[AI(OH)41 + 3[AIR'"0] + [AIR'+]
2[AI(r')2+] + 3[AI(R')3°] - [H+]
The ETD code operates on the principle of ANC mass balance.
10-11
-------
(1) a canopy module to simulate forest canopy interactions with both wet and dry deposition, (2) a
hydrology and watershed soil module to route precipitation through the soil horizons and simulate soil-
water physicochemical processes and biotic transformations, and (3) a lake module to simulate aquatic
biochemical reactions (Chen et al., 1983; Gherini et al., 1985).
The vertical resolution in the ILWAS code includes the canopy, a snow component, stream
segments, a lake component, and up to 10 soil layers for each subcatchment in the watershed (Chen
et al., 1983). The ILWAS model can simulate up to 20 subcatchments and associated stream segments.
For most DORP watersheds, only one or two subcatchments were used. To calculate the distribution of
water between flowpaths, the ILWAS model uses various forms of the continuity equation, Darcy's law
for flow in unsaturated and saturated permeable media, and Manning's equation, Muskingum routing, and
stage-flow relations for surface waters (Chen et al., 1982, 1983). The vertical layers within each
subcatchment are assumed to be areally homogeneous. The lake is vertically one-dimensional with up
to 80 vertical layers including snow and ice layers. For DORP application, the layer thickness was set
at 1 m so most lakes had between 3 and 7 layers. The temporal resolution of ILWAS output is generally
daily, but more frequent output can be obtained (with added computational effort and increased input
data).
The meteorological and deposition input requirements are shown in Tables 10-2 and 10-3. The
output variables in the soil solution and water chemistry are listed in Table 10-4. The components of the
ANC calculation are shown in Table 10-5. The ANC simulated by ILWAS is equivalent to the modified
Gran ANC.
The ILWAS model was developed to further the understanding of how atmospheric and terrestrial
acid-base processes interact to produce observed surface water chemistry. The model was developed
as part of the Electric Power Research Institute's (EPRI) Integrated Lake-Watershed Study of three
Adirondack lakes, Woods, Sagamore, and Panther (Chen et al., 1983; Gherini et al., 1985; Goldstein et
al., 1984). The model reproduced the seasonal and annual changes in water chemistry in these three
10-12
-------
lakes for the 5 years of record (see Appendix A). The model has subsequently been applied to 25
watersheds in Wisconsin, Minnesota, North Carolina, Tennessee, Utah, and California (Chen et al., 1988;
Davis et al., 1986; Gilbert et al., 1988; Greb et al., 1987; Munson et al., 1987). Regional assessments
have been conducted as part of the EPRI- funded Regional Integrated Lake Watershed Acidification Study
(RILWAS) and through other independent applications (Gherini et al., 1989).
10.2.3 Model of Acidification of Groundwater in Catchments (MAGIC)
MAGIC Is a lumped-parameter model of intermediate complexity, originally developed to project the
long-term effects (i.e., decades to centuries) of acidic deposition on surface water chemistry. One of the
model's principal assumptions is that a minimum number of critical processes in a watershed Influence
the long-term response to acidic deposition. The model simulates soil solution chemistry and surface
water chemistry to project the monthly or annual average concentrations of the water chemistry
constituents listed in Table 10-4. Hydrologic flow of water through soil layers to the receiving system is
simulated using a separate hydrologic model, TOPMODEL (Hornberger et al., 1985). TOPMODEL is a
topography-based, variable contributing area, catchment model adapted from the version of Beven and
Kirkby (1979). The model considers overland flow, macropore flow, drainage from the upper zone to the
lower zone and to the stream, and baseflow from the lower zone. Row routing through the watershed
is provided from TOPMODEL to MAGIC, a model in which both equilibrium and rate-controlled
expressions are used to represent geochemical processes. Mass balances for the major cations and
anions and the effects of aqueous aluminum and organic acid species on ANC are incorporated in the
model. The ANC simulated by MAGIC is equivalent to the modified Gran ANC. These processes are
listed in Table 10-1.
MAGIC represents the watershed with two soil-layer compartments. These soil layers can be
arranged vertically or horizontally to represent the vertical or horizontal movement, respectively, of water
through the soil. A vertical configuration was used in the DDRP, and the soil compartments were
10-13
-------
assumed to be areally homogeneous. Annual output is the typical temporal resolution of the model, but
monthly output also can be obtained.
The meteorological and depositional input requirements for MAGIC are shown in Tables 10-2 and
10-3. The output soil solution and water chemistry constituents are shown in Table 10-4. The
components included in the ANC calculation are shown in Table 10-5.
MAGIC was originally formulated to be parsimonious in selecting processes for inclusion and was
intended to be used as a heuristic tool for understanding the influences of the selected processes on
surface water acidification. The spatial/temporal formulations in the model reflect the intended use for
assessment and multiscenario evaluations. It was originally developed on two southeastern streams but
has subsequently been applied to many watersheds in the Southeast, lakes In the Adirondacks, and
watersheds in England, Scotland, Norway, Finland, and Sweden (Cosby et al., 1985a,b, I986a,b,c, 1987;
Lepisto et al., 1988; Musgrove et al., 1987; Neal et at., 1986; Whitehead et at., in press). MAGIC
reproduced the annual changes in water chemistry for these systems for the short period of available
record. It also has been used for a regional assessment of Norwegian lakes using the Norwegian lake
resurvey data (Cosby et al., 1987; Homberger et al., I987a,b).
>
10.3 OPERATIONAL ASSUMPTIONS
There are several operational assumptions associated both with DDRP and the individual models
(Table 10-6). These assumptions underlie the DDRP Level 111 Analyses in toto. Each of the models has
specific assumptions in addition to those made for the DDRP. These specific assumptions, summarized
by Jenne et al. (1989), are described in more detail by the authors and developers of the models (Chen
et al., 1983; Cosby et al., 1985a,b,c; Gherini et al., 1985; Nikolaidis, 1987; Nikolaidis et al., 1988, 1989).
10.4 WATERSHED PRIORITIZATION
The general approach for selecting the DDRP watersheds was described in Section 5.2. This
section presents the approach for prioritizing watersheds for Level III Analyses in the NE and S8RP.
10-14
-------
Table 10-6. Level III Operational Assumptions
1. Index sample water chemistry from the NSWS provides an index of chronically acidic systems and
systems with low ANC that are susceptible to acidic deposition.
2. Index soil data from the DDRP Soil Survey adequately characterize watershed attributes influencing
surface water chemistry.
3. Projections of future acidification consider primarily chronic acidification. Episodic acidification is
considered in the EPA Episodic Response Project.
4. Surface water acidification is a sulfur-driven process. Sulfur is assumed to be the primary acidifying
agent in acidic deposition. Eastern deciduous forests generally are nitrogen-limited (Ukens et al.,
1977; Swank and Crossley, 1988) so there Is low export of nitrate. In addition, annual nitrate
deposition exceeds annual ammonium deposition in the eastern United States (Kulp, 1987) and
nitrate has a slight alkalizing effect in the watershed (Lee and Schnoor, 1988).
5. The watershed processes controlling the effects of sulfur deposition on surface waters are sulfate
adsorption and desorption and base cation depletion and resupply through mineral weathering and
exchange.
6. The effects of organic acids on acid-base chemistry are constant through time and independent
of sulfate.
7. These major processes are known well enough to be incorporated in the projection models used
in the DDRP.
8. Current watershed attributes and conditions (e.g., climate, land use, basin characteristics) will remain
relatively constant over the next 50 years.
9. Long-term projections using models are plausible and are the only feasible approach for evaluating
the regional, long-term effects of sulfur deposition scenarios on surface water chemistry.
10. Typical" year projections are not intended to represent future forecasts of water chemistry but
rather to provide a common basis for comparisons among deposition scenarios to assess potential
changes in surface water chemistry.
11. Acidification is reversible and the processes in the models are adequate to describe both chemical
acidification and chemical improvement.
12. Physical and chemical processes are adequately considered in the Level III models.
13. Uncertainty calculations provide estimates of relative error for long-term comparisons among models
and deposition scenarios but are not absolute error estimates.
10-15
-------
10.4.1 Northeast
Developing a priority order for performing the watershed calibrations and forecasts permitted early
comparisons among model outcomes, identified data problems of general concern to all three models
as the problems developed, and permitted joint resolution of these problems by all modelling groups.
The priority ordering ensured that problems encountered with regard to the watershed classes of highest
interest or greatest concern could be addressed early in the projection period, so that if additional
projections were precluded due to time or manpower constraints, projections for the highest priority
systems would be completed by all three modelling groups.
A decision tree was developed for the watersheds in the NE (Figure 10*1). The decision tree was
based on several criteria including previous calibration and projections, internal sources of sulfur,
hydrologic type, sulfur retention, chloride status, and ANC [based on values from the Eastern Lake
Survey Phase I (ELS-I), Linthurst et al., I986a)]. These criteria were used to rank the watersheds in
descending order of priority with the highest priority given to Class A watersheds and the lowest priority
to Class I watersheds. The number of lakes in each priority class {A -1) is shown on the right-hand side
of the priority class box.
Class A watersheds are those that previously had been investigated as part of an internal EPA
evaluation for the Administrator. Two of the models previously had been calibrated on these watersheds,
so minimal problems were anticipated in recalibration with aggregated soils data and site-specific
deposition data Watersheds with unequivocal internal sources of sulfate confound the effects of sulfur
deposition on surface waters, so these systems were ranked lowest priority. Systems with no inlets or
outlets, i.e., seepage lakes (see Section 5.3), also confound interpretation of sulfur deposition effects on
surface water chemistry because of internal alkalinity generation; these systems also were ranked as a
lower priority class for projections. Although drainage lakes with long residence times (i.e., > 1 yr) also
can have significant internal sulfate reduction, the estimated median hydraulic residence time for
northeastern lakes was 0.20 yr so internal alkalinity generation for the DORP lakes was not considered
10-16
-------
BSSSSBSSSSS^^
ODELLING PR1O
DECISION TREE: NE
Special interest
watersheds
Internal source
of sulfur
Priority
Class
Drainage lake
or reservoir
Positive sulfur
retention
CI class
AorB
Priority
Class
Priority
Class
Priority
Class
Priority
Class
Figure 10-1. Modelling priority decision tree: Northeast.
10-17
-------
to be a confounding factor. Watersheds that appear to be currently retaining sulfur were judged to be
higher priority than watersheds that appear to be at or near sulfur steady state, because of the potential
for acidification as sulfate (acting as a mobile anion) depletes soil base cations. Many northeastern
watersheds are influenced by the application of road salt (calcium chloride, magnesium chloride, sodium
chloride). The DDRP watersheds were screened to identify those systems for which the output chloride
was greater than the input from atmospheric sources. Those watersheds with significant net chloride
export were given a lower priority. Finally, those systems with initial ANC < 100 jieq L"1 were designated
higher priority than watersheds with ANC >. 100 fieq L"1.
Ail three modelling groups followed this priority order when making projections. The objective was
to complete analyses on at least the first 60 watersheds, which included those with ANC <100 /*eq L*
1, those that were the least disturbed with respect to road salt additions, those near sulfur steady state,
and those currently retaining sulfur within the watershed (i.e., Classes A - C).
10.4.2 Southern Blue Ridae Province
A decision tree also was developed for the SBRP watersheds (Figure 10-2) using criteria similar to
those for the NE with two exceptions: watersheds previously were screened for internal sources of
sulfate, and none of the dynamic models was calibrated previously on SBRP watersheds. Therefore, the
first criterion for prioritization was whether the watersheds are currently retaining sulfur, followed by
whether chloride concentrations are less than 50 /xeq L*1 (indicating little impact or disturbance by road
salting practices). The chloride criterion was the same as that used for northeastern lakes. Those
systems with ANC < 100 /ieq L"1 [based on values from the National Stream Survey (NSS) Pilot Survey
(Kaufmann et al., 1988)] also were given higher priority than those with ANC >_ 100 /Lteq L'1. The number
of streams in each priority class is shown on the right-hand side of the priority class box (i.e., A - E).
Of the 35 total watersheds for the SBRP, 25 were placed in the first two priority classes, which also
resulted in a restricted target population. This priority order was followed by the ILWAS and MAGIC
10-18
-------
!C^«<^^«SK«is^^!«!S&.w*^^\«1»'>%*vPv*"iVST ^'^X^iyX "«"SS.
-------
modelling groups in performing projections for watersheds in at least the first two priority classes. ETD
was not applied to streams, so SBRP watersheds were not simulated using ETD.
10.4.3 Effects of Prioritization on Inclusion Probabilities
Watersheds were ranked in priority order to minimize comparability problems among models in the
event that not every group could complete simulations on all 145 watersheds in the NE or 35 watersheds
in the SBRP. This prioritization scheme does not affect the inclusion probability of any watershed.
Inclusion probabilities are based on the statistical design and the manner by which the sample watersheds
were selected from the population of watersheds in the region (see Section 5.2.6). Simulating only
selected classes of watersheds, however, does affect the target population about which inferences can
be drawn. For example, if no watersheds with initial ANC >_ 100 Lieq L"1 are included in the projection,
no inferences or conclusions can be drawn about the future response of this class of systems to acidic
deposition scenarios. Even though the original target population had ANC concentrations ranging from -
87 to 400 peq L'1 the new target population about which inferences can be reached refers only to that
portion of the original population with ANC concentrations ranging from -53 to 100 jueq L"1 . The DORP
target population for the NE that corresponds to these Class A - C watersheds represents 1,851
watersheds compared with the full northeastern DDRP target population of 3,667 watersheds. The first
two priority groups in the SBRP also represent a restricted target population of 1,051 watersheds
compared with the full SBRP DDRP target population of 1,531 watersheds.
10.5 MODELLING DATASETS
A major objective of the Level III Analyses was to ensure that all three modelling groups were given
the same datasets, developed using identical procedures so that differences among model projections
reflected differences in model process formulations and not differences in input, data. Different process
formulations among the models requires that different averaging or aggregation procedures be used to
prepare model input and parameter data. The source data provided to each modelling group (e.g.,
10-20
-------
meteorology, deposition, morphometry, soils, and water chemistry) on which these procedures operated,
however, were identical for each model to minimize problems of comparability among model projections.
10.5.1 Meteorological/Deposition Data
The meteorological and deposition data for both the NE and SBRP were discussed in Section 5.6,
with the exception of daily meteorological data, which were specific to the Level III Analyses.
Meteorological data for daily temperature, dew point, pressure, wind speed, cloud cover, and solar
radiation also were required as model simulation input for ETD and ILWAS. These data are not measured
at as many locations as daily precipitation Is. Typical meteorological year (TMY) data have been
produced for 238 locations across the United States. These locations are usually in major cities. Ten
different TMY sites were selected and matched to each DORP site, based on geographic location and
elevation. Temperature and dew point were adjusted to match 30-year normal temperatures for the period
1951-1980. TMY temperature data also were adjusted to closely match long-term monthly average
temperatures at the TMY site. Hence, the monthly and daily temporal pattern for each TMY site was
representative of the long-term norm. Because temperature is elevation dependent, the TMY data were
adjusted to match the annual 30-year normal at a nearby site with an elevation comparable to the
National Climatic Data Center (NCDC) site assigned to a DDRP lake. The adjustment was additive based
on the difference between the annual average TMY temperature and the annual 30-year normal
temperature. A similar adjustment was applied to dew point.
Watershed specific "typical year" meteorological and deposition data were provided to the modelling
groups for each watershed in the NE and SBRP. These typical year data were repeated year after year
for 50 years in performing the watershed projection under current deposition levels. The altered
deposition scenarios for the NE and SBRP followed a temporal sequence of current deposition levels for
the first 10 years of the projection, altered, sulfur deposition for the next 15 years to the desired
percentage change relative to current deposition (30 percent decrease in the NE, 20 increase in the
SBRP), and then constant sulfur deposition at this altered level for the final 25 years.
10-21
-------
10.5.2 DDRP Runoff Estimation
Runoff is an important variable for the models used in Level III Analyses. The DORP study sites
are not gaged, so measured estimates of runoff were unavailable. A combination of techniques was used,
therefore, to obtain estimates of annual and monthly runoff for the northeastern and SBRP watersheds.
10.5.2.1 Annual Runoff
Annual runoff was estimated for each of the 145 northeastern and 35 SBRP watersheds, as
discussed in Section 5.7. Long-term average annual runoff estimates were based on 1951-1980 records.
The annual runoff was partitioned into average monthly fractions for use in calibrating the hydrologic
submodels.
10.5.2.2 Monthly Runoff
10.5.2.2.1 Northeast -
Monthly runoff estimates were calculated for the NE based on USGS long-term monthly averages.
USGS calculated a 30-year average monthly runoff proportion for a 12-month period (October -
September) using 1951-1980 runoff data for stations that had complete records for the 30-year period and
did not have diversions or regulations (D. Graczyk, personal communication). The final database
contained runoff data for 134 USGS gaging stations.
The USGS sites were linked then to the 145 DDRP study sites and 3 intensive study sites. For the
NE, a "nearest neighbor" linkage was used with physiographic considerations included when appropriate
(R. Nusz, personal communication). Using the Geographic Information System (GIS), a map was
prepared that depicted locations of USGS gaging stations and DDRP study sites. A USGS station was
linked to each study site based on station proximity, in areas like the White Mountains of New
Hampshire, physiographic considerations (e.g., elevation and topography) also were included in the linking
process. Physiographic data were obtained from Krug et al. (in press). In the Adirondack Subregion
(Subregion 1A in the ELS-I), USGS station density was extremely sparse relative to the number of ELS
10-22
-------
sites. In this area, a Thiessen polygon weighting system was used to link the few USGS stations to the
ELS sites. In many cases, more than one ODRP site was linked to a single USGS station.
Monthly runoff for each study site was calculated by applying the 12 monthly proportions for each
USGS station to the linked ODRP study sites. The average annual runoff value at each study site,
interpolated from the map of Krug et al. (in press), was multiplied by each of the 12 monthly proportions
to obtain 12 monthly runoff values (in inches) for each site.
10.5.2.2.2 Southern Blue Ridge Province -
Monthly runoff for the SBRP watersheds was estimated using a USGS database prepared similarly
to the one for the NE. The resulting database contains 30-year average monthly proportions (October -
September) for 41 USGS stations in the SBRP.
The USGS monthly proportion data were linked to the interpolated annual runoff at each site to
calculate a long-term monthly runoff estimate for each of the 12 months for the water year. The USGS
stations had limited spatial coverage of this region and did not overlap consistently with the DDRP study
sites. The GIS was used to link the USGS sites and DDRP study sites based on topographic features
and similar site characteristics. An average monthly proportion for each of the 12 months was calculated
for the USGS sites within the major land resource area (MLRA) to obtain a single file of 12 monthly
proportions. For the SBRP, all but one of the watersheds were located in a single MLRA.
Monthly runoff for each DDRP study site was calculated by applying the single file of 12 monthly
proportions for each MLRA to the study sites located in the respective MLRAs. The average annual runoff
value at each study site, interpolated from the map of Krug et al. (in press), was multiplied by each of
the 12 monthly proportions to obtain 12 monthly runoff values (inches) for each site.
10-23
-------
10.5.3 Morphometry
Basin, lake, and stream morphometry and characteristics were discussed in Section 5.4. These
data, obtained from the DDRP Soil Surveys and the NSWS for the NE and SBRP, were provided to each
modelling group for use in model calibration for each watershed in the NE and SBRP.
Lake volume and stage-discharge empirical relationships for the northeastern watersheds were
formulated using data obtained from the ILWAS, RILWAS, ME, and VT lakes for which bathymetric
information was available. These lakes were partitioned by surface area and relationships established
between lake volume and lake area. Lake volume relationships were Improved if the lakes were
partitioned by surface area (i.e., surface area < 100 acres and surface area > 100 acres). These
relationships are
Volume (acre-ft) « 4.486[Lake Area (acres)]1l382
for lake areas < 100 acres, r2 = 0.87
Volume (acre-ft) - 5.670[Lake Area (acres)]1'227
for lake areas > 100 acres, r2 = 0.96
Empirical relationships also were established between lake stage and discharge based on data
available for ILWAS and RILWAS lakes. A relationship between discharge (Q), height of the lake spillway
(h) and lake depth was established for different classes of lakes based on their watershed areas. These
relationships, which were used in the ILWAS model; are
Q (cfs) = 2.694h (ft)3'538
for lakes with watershed areas < 350 ha, r2 = 0.98
Q (cfs) » 0.897h (ft)5'279
for lakes with watershed areas 350 - 600 ha, r2 = 0.98
Q (cfs) = 3.160h (ft)3'70
for lakes with watershed areas 601 - 3000 ha, r2 = 0.96
10-24
-------
Lake volume for the ETD and MAGIC simulations was assumed to be constant (i.e., inflow volume =
outflow volume + evaporation volume + seepage volume) so a stage-discharge relationship was not
required.
10.5.4 Soils
The soils data, discussed in Section 5.5, were aggregated (Sections 9.2.3.2 and 9.3.1.2) and
provided to each modelling group on a model-specific basis. The soils data used by each modelling
group were Identical, but the aggregation procedures used by each group were model specific. The
ILWAS modelling group used unaggregated data.
10.5.5 Surface Water Chemistry
As described in Section 5.3, surface water chemistry data were obtained from the NSWS and were
described in detail by Kanciruk et a!. (1986a) and Messer et ai. (1986a). Both 1984 ELS-I and 1986 ELS-
II data were provided for the northeastern watersheds, and NSS Pilot stream data were provided for the
SBRP watersheds for all sampling periods in 1985.
10.5.6 Other Data
Watershed data such as bedrock geology, land use, vegetative cover, estimated depth to bedrock,
and other data (discussed in Sections 5.4 and 5.5.6) were provided to each of the modelling groups for
calibration of the individual watersheds. Because of different model formulations, the data were used
differently during model calibration but the information provided to each modelling group was identical.
10.5.7 Chloride Imbalance
Preliminary mass balances for chloride indicated that chloride export exceeded deposition input for
a significant number of northeastern watersheds. Road salt, watershed disturbances, and other factors
might account for these additional chloride Inputs. A decision tree approach (Figure 10-3) was used to
identify watersheds with net chloride export and to correct this imbalance. First, a chloride concentration
below which sites could be considered "unaffected" by any unusual sources was investigated. A
10-25
-------
Cl STATUS
DECISION TREE
[Cl]>50
ueqL'1
Seasalt
Influence
only
Roadsalt
Influence
only
•
24
Na:CI
<0.8
NO
YES
ci
I balanced
by other
Roadsalt
&
seasalt
Influence
37
Na:CI
<0.8
NO
?;
IYES
ci
\ balanced
fby seasalt
Cl
i balanced
by other
Cl
balanced
by seasalt
Figure 10-3. Decision tree used to identify watersheds with net chloride export and procedures for
determining chloride imbalance.
10-26
-------
concentration of 50 Ateq L"1 was chosen following an examination of the data. This concentration was
at the upper end of the range in concentrations found in "undisturbed" (see below) lakes. Some
"disturbed* lakes had concentrations below 50 neq L*1, but just because a lake was classified as
disturbed by our criteria does not mean that it actually was unusually or adversely affected. Rather,
disturbed simply implies that the site has the potential for being affected because of the level of human
activity associated with its watershed. "Unaffected" systems were designated as Class A (Figure 10-3),
to which 80 sites were assigned.
Disturbance was based on a number of factors including location of roads in the watershed (e.g.,
proximity to lakes, streams, etc.), as well as the occurrence of mines, waste disposal sites, urban
industrial sites, or residential areas. If a lake had chloride > 50 /zeq L*1 but was undisturbed, then its
distance from the coast was examined. A distance of 50 km from the coast was selected as a cutoff
based on (1) plots of chloride vs. distance from the coast for undisturbed sites, (2) deposition data (A.
Olsen, personal communication), and (3) sea salt effects relative to distance (R. Dennis, personal
communication). Four undisturbed sites had chloride > 50 /zeq L"1 but were within 50 km of the coast.
These sites were classified as Class B lakes with sea salt influences only. The chloride inputs at these
sites were treated as sea salt inputs distributed uniformly across the watersheds. The anion inputs were
balanced by cations consistent with sea salt composition. The occurrence of undisturbed sites with
chloride > 50 /ieq L'1 at distances greater than 50 km from the coast was examined, and no such sites
were found in the DDRP dataset, resulting in Class C.
Disturbed sites then were examined relative to distance from the coast. Disturbed sites exceeding
distances of 50 km from the coast are probably influenced only by road salt practices. Sites close to
the coast might show both road salt and sea salt effects. Thirty-seven sites fell into the latter category
and 24 into the former. For those sites apparently affected by road salt only, point source inputs should
be assumed. For the other sites, the Inputs might be both "broadcast" and point source, but no method
for discrimination between these possibilities was available. Therefore, these sites were treated as having
10-27
-------
point source inputs of salts directly to the lake rather than input sources spread evenly across the
watershed.
The last factor to be examined was the composition of the salts. The molar ratio of sodium
chloride in seawater is 0.864 (Harvey, 1969). Given the measurement uncertainty, a ratio less than 0.8
was used to screen lakes. Only three of the remaining sites fell into this category, with 0.72 the lowest
ratio observed. These low ratios might be due to uncertainty and, because of the difficulty in developing
a procedure for deciding which ions to use to balance the chloride, these sites also were balanced by
base cations of sea salt composition "added" directly to the lake.
10.6 GENERAL APPROACH
The following general approach was used In performing the long-term projections of future change
in surface water chemistry by each of the modelling groups:
* Model calibration
• Sensitivity analyses
• Regional projection refinement
• Future projections
• Uncertainty analyses
* Regional population estimates
This approach, illustrated in Figure 10-4, is fully consistent with the recommendations made by the
Environmental Engineering Committee of the EPA Science Advisory Board (EPA-SAB) on the use of
mathematical models by EPA for regulatory assessment and decision-making (EPA-SAB, 1988).
All three models were calibrated to three watersheds in the NE - Woods Lake, Panther Lake, and
Clear Pond. In the SBRP, MAGIC was calibrated to White Oak Run, VA, and ILWAS was calibrated to
Coweeta watershed 36. These watersheds are discussed in the next section. All three models performed
10-28
-------
DYNAMIC MODELLING METHODOLOGY
Model Calibration
- Intensively studied
watersheds
- NE & SBRP
i
Sensitivity Analysis
- ETD
- ILWAS
- MAGIC
i
r
Refine Calibration/
Projection Approach
for DDRP Watersheds
- ETD
- MAGIC
\
r
Mode! Forecasts
Northeast SBRP
- Projection - Projection
- Uncertainty - Uncertainty
Estimate Estimate
^
r
Regional Population
Estimates
- Northeast
- SBRP
Figure 10-4. Approach used in performing long-term projections of future changes in surface water
chemistry.
10-29
-------
sensitivity analyses to determine those parameters and inputs to which the models were most sensitive.
Particular attention was given to these parameters and inputs during model calibration in preparation for
the long-term projections. Sensitivity analyses are discussed in Section 10.9. Intensive site calibration
and sensitivity analyses were conducted to document model behavior, to demonstrate that these models
can predict short-term watershed responses, and to identify areas for improvement in calibration and
projection procedures for the regional sets of watersheds. These refinements are discussed in Section
10.10.
The general approach for long-term projections followed by each of the modelling groups is
illustrated schematically in Figure 10-5. The models were calibrated to each of the DORP watersheds,
or some subset, in the Northeast and SBRP. Long-term projections (i.e., for 50 years) were performed
on the individual watersheds and the results presented as population estimates for the NE or SBRP. The
population estimates were generated as indicated in Section 6. Results from individual watersheds were
of interest only with respect to their representation of the target population. Uncertainty analyses,
described in Section 10.11, were incorporated in the confidence intervals about the regional population
estimates. The following sections discuss each of these general topics in greater detail.
10.7 MODEL CALIBRATION
10.7.1 Special Interest Watersheds
Three intensively studied watersheds were selected for model calibration in both the NE and SBRP.
Selecting multiple intensively studied watersheds for calibration was important for the following reasons:
• Watershed characteristics and parameter values vary from watershed to watershed, and thus
a range of values can be simulated. For example, watersheds can be selected with varying
combinations of sulfate adsorption, percent base saturation, depth of till and other watershed
and lake attributes.
• The relationship among model parameters and measured parameters can be examined
because extensive information on watershed processes, watershed characteristics, and system
responses is available and an intensive time-series record exists.
• The short-term behavior of the models on a variety of systems can be evaluated.
10-30
-------
Schematic Modelling Approach
Measured
Data
Data Preparation
Aggregation
Function
Population Mean
Population Median
Population Variance
Uncertainty
Population Estimation
Lumped Watershed
Chatacteristics,
Soil Data, & Chemistry
I
Time Dependent
Data
Model
Output
Model
Model Projections
Scaling
Function
o
a
2.
s
£
o
Model
Parameters
Figure 10-5. Schematic of modelling approach for making long-term projections.
10-31
-------
• Comparable results among the models simulating a varying combination of watershed
processes and responses provides greater confidence in using them for long-term projections.
The datasets for the six watersheds were each subdivided into a calibration dataset and
confirmation dataset. The calibration dataset was provided to each modelling group for use in calibrating
the model to the respective watershed. The confirmation dataset was retained by Oak Ridge National
Laboratory until calibration was complete. The confirmation dataset consisted only of the model inputs
and not the lake or stream water chemistry record. The modelling groups applied the calibrated models
to the confirmation datasets and then compared the predicted output to the observed water chemistry
record. For comparisons among models, calibration and confirmation root mean square errors (RMSE)
were computed for the following output variables:
• Instantaneous flow (m3 s"1)
• Cumulative flow (m yr"1)
• Chloride fceq L'1)
• Sulfate G*eq L'1)
• Gran alkalinity (peq L'1)
• Calcium (^eq L"1)
• Magnesium (^eq L"1)
• Sodium (/jeq L'1)
• Potassium fceq L"1)
• Total aluminum (^g L"1)
• pH
10.7.1.1 Northeast
The three northeastern intensively studied watersheds are Woods Lake, NY, Panther Lake, NY, and
Clear Pond, NY. Woods and Panther Lakes were EPRIILWAS research sites (Chen et al., 1983; Goldstein
10-32
-------
et al., 1984; Gherini et al., 1985). Clear Pond was an EPRI RILWAS site. All basin and lake morphometry,
soil chemistry, mineralogy, and hydrology data were obtained from EPRI (Valentin! and Gherini, 1987; R.
Goldstein, personal communication). Water chemistry data were collected approximately weekly during
the study periods. All three watersheds also were sampled during the DORP Soil Survey, and these data
were provided to the modelling groups. The calibration and confirmation periods for these three sites
were
Site Calibration Confirmation
Woods Lake 9/78 - 8/80 9/80 - 8/81
Panther Lake 8/78 - 8/80 9/80 - 8/81
Clear Pond 7/82 - 7/84
10.7.1.2 Southern Blue Ridge Province
The three intensively monitored stream watershed sites in the SBRP are Coweeta watershed 34, NC,
Coweeta watershed 36, NC, and White Oak Run, VA. Watershed and stream morphometry, soil
chemistry, water chemistry, and historical site information were obtained for the Coweeta sites from the
USDA Forest Service's Coweeta Hydrological Laboratory (W. Swank and J. Waide, personal
communication) and for White Oak Run from B. Cosby and G. Homberger (personal communication).
Water chemistry samples were collected approximately weekly during the study period. These sites also
were sampled during the DDRP Soil Survey and these data provided to the modelling groups. The
calibration and confirmation periods for these three sites were
Calibration Confirmation
WS 34 6/82 - 5/86 6/73 - 5/82
WS 36 6/73 - 5/79 6/79 - 5/86
WOR 1/80 - 12/82 1/83 - 12/84
10-33
-------
The period of record at the Coweeta sites permitted partitioning the datasets for the purposes of
both projecting and hindcasting. Because of time constraints, the Coweeta watersheds were not
simulated. The ILWAS and MAGIC models will be calibrated on the Coweeta watersheds and the results
presented as part of the DORP Mid-Appalachian report in mid-1990. MAGIC was calibrated for White Oak
Run using data collected for period January 1980 to December 1984. The MAGIC model was developed
using data from White Oak Run and Deep Run, VA (Cosby et al., 1985a).
10.7.2 General Calibration Approach .
The general approach for model calibration was, first, to calibrate the hydrologic submodel or
companion model to the discharge records; next, calibrate the chemical submodel or model to a
conservative constituent such as chloride; and finally to calibrate the watershed model to the suite of lake
or chemical concentrations simulated by the model. Calibration was an Interactive process. The
hydrologic submodel can route flow through various soil horizons and still predict the observed stream
hydrograph or lake discharge. Calibration to a conservative constituent provides confidence that mass
balance Is maintained in the models and also provides confidence in, and constraints for, the hydrologic
calibration. If evapotranspiration, overland flow, subsurface flow, or other components of the hydrologic
budget were not properly calibrated, a flow balance might be achieved, but it is unlikely that the model
outputs would match the observed conservative constituent concentrations. Calibrating the model to
additional noncortservative chemical constituents, such as anions (other than chloride) and cations, further
constrains the hydrologic flowpaths through various soil horizons, because the physical and chemical
attributes of the soils restrict the range of parameters for each compartment. If the observed stream or
lake concentrations could not be predicted within these ranges, the hydrologic calibration was revised
to provide additional flow through different soil horizons or along different flowpaths to achieve the
observed water chemistry concentrations. Calibration of the models to predict the observed
concentrations of multiple constituents provided relatively restrictive constraints on calibration parameters.
Variables that could be calculated from measured soil or lake attributes were incorporated directly Into
the model without modification. These variables included constituents such as soil cation exchange
capacity, exchangeable fractions of base cations, base saturation, porosity, and lake hydraulic residence
10-34
-------
time. The sampling and measurement errors in these variables were included in the uncertainty analyses.
The mass-weighted mean or median values of the aggregated variables were used in the calibration
process.
10.7.3 Calibration of the Enhanced Trickle Down Model
The ETD model represented the watershed horizontally as a homogeneous catchment with no
subcatchments (Figure 10-6) and vertically with a snow compartment, soils, unsaturated zone, and
saturated or groundwater compartment (Figure 10-7). Watershed data for ETD were lumped or aggregated
to provide average or weighted average values for each of the soil layers. In the NE, the top soil
compartment represented the mass-weighted average conditions of the O, A, and upper B horizons, the
unsaturated zone was represented by mass-weighted averages of the lower B and upper C horizons, and
the groundwater compartment was represented by mass-weighted averages for the lower C horizons.
The processes represented in the ETD model are shown in Table 10-1. ETD calibrations were
achieved by decoupling the hydrologic, chloride, sulfate, and ANC-weathering submodels. The hydrologic
calibrations were conducted first. Next the chloride submodel was calibrated. Chloride was assumed
to be conservative and was used to evaluate the hydrologic calibration. Calibration was an iterative
process because of the coupling between hydrology and chemistry. The sulfate submodel was calibrated
next and the final calibration involved the ANC-weathering submodel. The calibration of each submodel
was achieved by using a standardized optimization package, IDESIGN (Arora et al., 1985) coupled with
ETD, and a trial-and-error procedure. The range of parameters for calibration was input to IDESIGN.
A complete two-year simulation was performed at each iteration and a cost or penalty function evaluated.
(DESIGN performs minimization of the cost function using the Fletcher-Reeves algorithm (gradient
method). A priori bounds on physical quantities and parameters were included as constraints in the
optimization process. In all gradient methods when the objective function is of the least-squares type,
it is assumed that the residuals are homoscedastic, independent, and sufficiently small to assume
normality. For the DDRP simulations, the residuals were not homoscedastic, which resulted in the
10-35
-------
Woods Lake
1000 0
I I I I . t
1000 2000 3000 Feet
i t i
0.5
1 Kilometer
Approximate mean
declination 1979
Figure 10-6. Representation of horizontal segmentation of Woods Lake, NY, watershed for MAGIC
and ETD.
10-36
-------
Prototype
Ocm
Model
Layerl
r
Upper
till
(C)
Layer 2
r
r
Lower
till /
(C) ^
Layers
Rgure 10-7. Representation of vertical layers of Woods Lake Basin for ETD.
10-37
-------
(DESIGN optimizations being biased toward the extreme values of the residuals. The (DESIGN does,
however, bring the parameters within reasonable range of their optimal values. Following the (DESIGN
simulations, therefore, a trial-and-error method was employed to achieve optimal calibration. There were
three main guidelines used for trial and error calibration:
(1) Obtain closure of cumulative flow or mass during the entire calibration period.
(2) Capture the seasonal variability of the state variables.
(3) Capture the peaks and valleys of the daily flows and concentrations.
This calibration process was followed for each of the submodels. The parameters for the previous
submodel calibration were fixed during calibration of the succeeding submodel. In some instances, this
was an iterative process. Calibration of the sulfate and ANC submodels might indicate that the flowpaths
through the watershed would have to be changed to match observed water chemistry and maintain
parameter values within a reasonable range for the soils on that watershed. The hydrology submodel,
therefore, would be recalibrated and the process repeated. Calibration was an iterative process of
constraining parameters to achieve an optimal calibration. Additional information on the ETD model
calibration has been presented by Nikolaidis et al. (1987).
10.7.4 Calibration of the Integrated Lake-Watershed Acidification Model
In the ILWAS model the watershed was partitioned into a series of subcatchments to represent the
horizontal variation in the watershed (Figure 10-8) and a series of vertical layers to represent various soil
horizons (Figure 10-9). Basin data are used quantitatively to characterize the system to be simulated and
delineate the appropriate number of subcatchments.
The ILWAS model requires specification of over 200 parameters, coefficients, and Initial conditions
for model calibration to represent the processes listed in Table 10-1. These values can be classified into
three groups: constants, measured values, and calibration parameters. Constant values include
thermodynamic constants or other factors that do not vary from watershed to watershed. Measured
10-38
-------
Woods Lake
1000
1000
i
2000
0.5
3000 Feet
i
\ Kilometer
Approximate mean
declination 1979
Figure 10-8. Representation of horizontal segmentation of Woods Lake Basin for 1LWAS.
10-39
-------
Prototype
%VxVxVxV>'N->'%V%V%VN'x
fSftfSffffffttStt*
Upper
till
(C)
Lower
till
(C)
Ocm
Model
15cm
25cm
75cm
Layer 1
Layer 2
Layers
/I
Layer 4
Layers
Figure 10-9. Representation of vertical layers of Woods Lake Basin for ILWAS.
10-40
-------
values included watershed area, base saturation, lake volume, and other attributes that were either directly
measured or calculated from measured data at a specific site but were not varied during model
calibration. The third set of values was calibration parameters such as mineral weathering rates, hydraulic
conductivity, nitrification rates, and other parameters that are not well known and were modified during
calibration to match the observed watershed and lake constituent concentrations. The general rules for
calibration were: calibrate the system's hydrologic behavior before calibrating the chemical behavior;
calibrate in the same order as water flows through the basin; and calibrate on an annual basis first, then
seasonally, and finally to the instantaneous (daily) behavior.
The hydrologic calibration typically involves first matching the annual cumulative lake/stream
discharge by adjusting the basin evapotranspiration coefficient. Seasonal flow variations are matched by
varying the seasonal evapotranspiration coefficient. Flow through the watershed, both laterally and
vertically, is adjusted by varying the hydraulic conductivity to match the instantaneous discharge.
Chemical calibration involves varying canopy, snowpack, and soil parameters to match the observed
surface and groundwater chemical concentrations.
Although there was a significant number of parameters, and, therefore, significant degrees of
freedom in selecting parameter values, only certain combinations of parameter values resulted in predicted
constituent concentrations that matched observed concentrations. The interactions among parameters
and parameter combinations placed limitations on the number of feasible parameter combinations. For
example, increasing the in-lake nitrification rate coefficient will lead to decreased ammonium, increased
nitrate, decreased ANC, and decreased pH values. These feedbacks provide a robust set of constraints
for calibration.
The calibration exercise involved identifying the set of parameters that minimized the differences
between the set of predicted versus observed constituent concentrations. Additional information on the
ILWAS model calibration has been presented by Munson et al. (1987).
10-41
-------
10-7.5 Calibration of the Model of Acidification of Groundwater In Catchments
MAGIC represents the horizontal dimension of the watershed as a homogeneous unit with no
subcatchments (Figure 10-6) and the vertical dimension as two soil layers (Figure 10-10). Watershed data
for MAGIC were lumped or aggregated to provide average or weighted average values for each of the
soil layers. The top soil compartment represented the mass-weighted average conditions of the A and
B Master horizons in both the NE and SBRP. The lower soil compartment represented the mass-weighted
average conditions in the C Master horizon in both regions.
Projecting long-term effects of acidic deposition on surface water chemistry using MAGIC involves
coupling MAGIC with TOPMODEL (Cosby et al., I985a,b,c). Both models were calibrated using an
optimization procedure that selected parameters so that the difference between the observed and
predicted measurements was minimized. The calibration exercise was a three-step process. The first
step was to specify the model inputs such as precipitation, deposition (both wet and dry), an estimate
of historical inputs for the long-term chemical model, and fixed parameters or parameters whose values
correspond directly to (or can be computed directly from) field measurements, e.g., topographic variables
such as slope, aspect, area. This approach, in effect, assigns all of the uncertainty associated with
sampling, aggregation, and intrinsic variability to the "adjustable" parameters. The adjustable parameters
are those that are calibrated or scaled to match observed field measurements.
The second step was the selection of optimal values for the adjustable parameters. These
adjustable parameters were selected using optimization. The method of Rosenbrock (i960) was used.
Optimal values were determined by minimizing a loss function defined by the sum of squared errors
between simulated and observed values of system state variables. Different loss functions were used for
the hydrologic and chemical models. The hydrdogic model used daily stream flow volumes while the
chemical model used weekly lake outflow chemistry or observed soil chemistry.
The final step was to assess the structural adequacy of the model in reproducing the observed
behavior of the criterion variables and parameter identifiability or the uniqueness of the set of optimized
10-42
-------
Ocm
Model
Lower
till
(C)
Layer 1
Layer 2
Figure 10-10. Representation of vertical layers of Woods Lake, NY, watershed for MAGIC.
10-43
-------
parameters. Structural adequacy was assessed by examining the mean error in simulated values of
observed state variables for those variables used in the calibration procedure as well as for an additional
state variable which was not used during calibration. Parameter identrfiability was assessed using
approximate estimation error variances for the optimized parameters (Bard, 1974). Additional information
on the MAGIC calibration process has been presented by Cosby et al. (1989).
10.7.6 Calibration/Confirmation Results
Calibration/confirmation results from the three models currently are available only for the intensively
studied watersheds in the NE - Woods Lake, Panther Lake, and Clear Pond. ILWAS and MAGIC have
been used previously, however, to predict water chemistry for southeastern streams. The ILWAS model
was used on Coweeta streams as part of RILWAS (Gherinl et al., 1989), and MAGIC was developed for
southeastern streams (Cosby et al., 1985a,b,c).
The calibration/confirmation results, expressed as RMSE for each model applied to each lake, are
shown in Tables 10-7 to 10-9. ILWAS was calibrated on the complete datasets for Woods and Panther
Lakes and Clear Pond as pan of the ILWAS/RILWAS studies (Chen et al., 1983; Gherinl et al., 1989)
immediately before the initiation of DDRP, so only calibration RMSE values are provided for ILWAS. The
RMSE was calculated because it is similar to a standard deviation and has the same units as the original
measurements. Therefore, RMSEs can be compared among models and with the standard deviations of
the observed data (Tables 10-7 to 10-9). For an unbiased model (i.e., a model in which the mean of the
observations is equal to the simulated mean), the model RMSE should be equal to the standard deviation
of the observations.
Comparing the RMSEs among models and among lakes indicates model results are similar for all
three lakes (Tables 10-7 to 10-9). Instantaneous discharge, chloride, sulfate, and ANC (the four variables
predicted by all three models) were within 0.01 - 0.15 m3 s"1 , 2 - 5 Meq L"1 , 9 - 18 p.eq L"1 , and 18 -
82 jueq L"1, respectively, of the observed values for these constituents for all three models. The model
RMSEs were similar to the standard errors of most of the constituents in each lake, indicating unbiased
10-44
-------
Table 10-7. Comparison of Calibration/Confirmation RMSE for Woods Lake Among ETD, ILWAS,
and MAGIC Models, with the Standard Error of the Observations
Calibration
Constituent3
Inst. Discharge
Chloride
Sulfate
Alkalinity
Calcium
Magnesium
Sodium
Potassium
Tot. Aluminum
Hydrogen
ETD
0.05
5.5
17.5
27.9
ILWAS
0.09
3.3
17.3
31.4
6.6
1.8
4.4
1.9
7.9
8.1
MAGIC
0.05
1.9
11.4
17.9
16.6
6.7
6.5
3.2
3.5
1.3
Observed
SE
.
3.1
16.4
18.6
55.8
56.9
5.82
1.4
16.5
—
ETD
0.07
1.9
10.5
14.7
Confirmation
ILWAS MAGIC
0.07
3.8
16.9
16.4
6.9
2.9
3.2
1.7
6.2
6.9
All units in peq I'1 except instantaneous discharge (m3 s"1} and total aluminum IJQ I'1).
ILWAS was calibrated prior to the DORP using all the data so the dataset could not be split for confirmation.
10-45
-------
Table 10-8. Comparison of Calibration/Confirmation RMSE for Panther
Lake Among ETD, ILWAS, and MAGIC Models, with the Standard Error
of the Observations
Constituent3
Calibration Observed
ETD ILWAS MAGIC SE
Confirmation
ETD ILWAS MAGIC
Inst. Discharge
Chloride
Suifate
Alkalinity
Calcium
Magnesium
Sodium
Potassium
Tot. Aluminum
Hydrogen
0.03
5.1
11.3
82.6
0.01
4.5
17.6
47.1
36.7
8.8
8.1
2.0
3.2
1.9
0.03
5.6
16.0
87.4
40.4
9.4
9.9
1.7
4.6
3.7
.
4.3
14.0
71.0
150.3
154.8
8.7
1.7
11.1
—
0.04
2.4
11.7
70.0
0.05
2.1
15.0
57.7
40.0
8.4
9.9
2.8
2.1
1.4
j*AII units in jieq L except instantaneous discharge (m3 s ) and total aluminum
ILWAS was calibrated prior to the OORP using all the data so the dataset could not be split for confirmation.
10-46
-------
Table 10-9. Comparison of Calibration RMSE for Clear Pond Among ETD,
ILWAS, and MAGIC Models, with the Standard Error of the Observations
Constituent*
Inst. Discharge
Chloride
Sulfate
Alkalinity
Calcium
Magnesium
Sodium
Potassium
Tot Aluminum
Hydrogen
ETD
0.16
2.4
8.9
18.6
23.6
5.0
6.3
1.0
1.1
1.0
Calibration
ILWAS
0.03
1.4
10.6
17.9
21.1
4.7
5.0
0.7
1.2
0.2
MAGIC
0.15
4.7
9.5
18.6
21.2
4.7
Observed
SE
1.6
9.7
18.8
All units in f/eq L'1 except Instantaneous discharge (m3 s"1) and total aluminum pg L~l)
10-47
-------
estimates of mean constituent concentration in the lakes. The similarity between the observed standard
error and model RMSE indicated the seasonal and annual changes predicted by the models were
consistent with the seasonal and annual constituent dynamics observed in the three lakes. The RMSEs
also were similar among models. All three models predicted similar seasonal and annual changes in flow
and constituent concentrations for all three lakes indicating unbiased estimates of mean constituent
concentration in the lakes. Although the RMSE for ANC appears large in Panther Lake, the standard
deviation of observed ANC values for Panther Lake for this same period was 71 peq L"1 . The RMSEs
during the confirmation period were equivalent to, or smaller than RMSEs calculated during the calibration
period, for both the ETD and MAGIC simulations on Woods and Panther Lakes.
The RMSEs were equivalent for other constituents predicted by ILWAS and MAGIC (Tables 10-7
to 10-9). Smaller RMSEs than observed standard errors for calcium and magnesium in Woods and
Panther Lakes imply MAGIC and ILWAS did not predict as large a deviation from the mean for these
constituents as was measured in the lake outflow. Many of these large deviations occurred during
snowmelt (Chen et a)., 1983; Gherini et al., 1985). The volume averaging by the models conserves mass
but will result in lower predicted constituent concentrations if this snowmelt moves as a thin lens under
the ice (Gherini et al., 1985). In the models, the higher/lower concentrations in this lens will be mixed
with the rest of the volume in the lake or that layer to compute the constituent concentrations.
Although these models were (1) developed for different systems in different regions of the United
States, (2) calibrated independently using model-specific procedures, and (3) run using different
computational time steps (daily versus monthly), the RMSEs for all constituents are similar. Woods Lake,
a chronically acidic lake (ANC = -10 /zeq L'1 ), Clear Pond, with an average annual ANC of
approximately 100 ju,eq L*1, and Panther Lake, with an average annual ANC of approximately 150 Meq
L*1, span the range of DORP systems of interest in the NE, and the results indicate all three models can
predict acid-base water chemistry with precision similar to that observed in the measured data for short-
10-48
-------
term periods of record. These results do not, however, necessarily ensure calibration for long-term
projections. Long-term calibrations can be achieved only with long-term data (Simons and Lam, 1980).
Comparisons of the time sen'es of predicted versus observed values for each of the model
applications to Woods Lake, Panther Lake, and Clear Pond are described and shown graphically in
Appendix A.1. The calibration and confirmation exercise indicated the three models produced comparable
results for three watersheds with a range of watershed characteristics, from deep to shallow till depth,
and lake chemistry, from ANC concentrations of -40 to over 200 /ieq L*\ Although, there is variability
for individual daily values, the models reproduce the flow-weighted average annual constituent values.
Average annual estimates, and the change in these estimates, represent the focus of the DDRP.
This calibration/confirmation exercise and calculation of RMSEs is consistent with the
recommendations of the Environmental Engineering Committee of the Science Advisory Board for model
confirmation with field data (EPA-SAB, 1988). The next step recommended for using environmental
models was to conduct sensitivity analyses (EPA-SAB, 1988).
10.8 MODEL SENSITIVITY ANALYSES
Sensitivity analysis is a formalized procedure to identify the impact of changes in various model
components on model output. Sensitivity analysis is an integral part of simulation experiments and model
applications. Models represent aggregations and simplifications of watershed and soil processes,
including physical, chemical, and biological processes. Parsimony is introduced to represent these
multiple processes by a single (or few) aggregated process(es) and a transfer coefficient or parameter.
Sensitivity analysis is an approach used to determine if model output or system response is sensitive or
responsive to small changes in these transfer coefficients. Those parameters for which the model output
was sensitive received greater attention during model calibration for long-term projections.
Sensitivity analyses were performed on the three intensively studied watersheds in the NE.
Examining the sensitivity of model output over a short simulation period (e.g., 3-5 years) provides useful
10-49
-------
information about model behavior; however, these short-term analyses might not reveal the full
dependence of the simulated system response on these parameter values. Certain parameters, for
example, might have little effect on short-term model behavior but might be critical for long-term
projections. This caveat must be considered in evaluating the model sensitivity analyses and the long-
term projections.
10.8.1 General Approach
Sensitivity analyses were performed using the Woods and Panther Lakes and Clear Pond datasets
following the model calibration and confirmation exercises for the MAGIC model and those for ETD on
Woods and Panther Lakes. The ETD sensitivity analyses were conducted as part of a Ph.D. Thesis
(Nikolaidis, 1987). The Clear Pond dataset was not available in time to be included in this Thesis.
Sensitivity analyses were performed for ILWAS prior to the initiation of the DORP. These qualitative
analyses for ILWAS are included In Appendix A.1. The classical approach of Tomovic (1963) was used
in performing sensitivity analyses on each of the models. Each coefficient or parameter was individually
varied by ±10 percent with all other coefficients retaining their original, calibrated values. The relative
change in model output RMSE for different variables or model components was noted to determine their
sensitivity to this parameter change. If the increase in the RMSE was large, the model was considered
sensitive to this parameter. RMSEs permitted a quantitative estimate of the increase in variance
associated with each parameter.
The optimization procedures used with ETD and MAGIC also indicated relative parameter sensitivity.
The response surface around an optimum parameter value was evaluated to determine if the surface was
relatively flat or steep. A relatively flat response surface indicated several parameter values could be
selected without influencing the optimum system response. A steep surface, however, indicated that small
changes in the parameter value would affect the optimum system response or that the system response
would be sensitive to that parameter.
10-50
-------
All three modelling groups selected coefficients or parameters for processes expected to control
or strongly influence both hydrologic and water chemistry output variables, including hydrologic routing
parameters, sulfate adsorption, ion exchange, and weathering rate parameters. Selection of these
parameters was based on previous analyses and published results by each modelling group (Cosby et
al., 1985a,b,c; Gherini et al., 1985; Nikolaidis et al., 1988; Lee et al., in press; Georgakakos et al., In
press). The specific parameters evaluated are listed in Table 10-10. This evaluation provided an estimate
of the variability introduced by the parameter in the system response, and, therefore, an indication of the
range over which the parameter could vary without significantly altering the model output.
10.8.2 Sensitivity Results
The parameters selected for sensitivity analyses are ranked in priority order from most sensitive to
least sensitive in Table 10-10. The effects of a ± 10 percent change in these parameters on the RMSE
for predicted lake ANC concentrations also are listed with these parameters.
Parameters related to weathering and hydrologic transport processes, in general, were sensitive in
each of the three models. Parameters related to sulfate adsorption and bulk soil processes such as
cation exchange capacity were not particularly sensitive in any of the models. Sulfate adsorption would
not be expected to be an important process in the NE if lakes are near sulfate steady state. The models
all appeared to be robust to small changes in bulk soil properties, which also can be measured directly
in the field.
The specific parameters that were sensitive for each model differ because of different process
formulations among the models. For example, a 10 percent change in weathering rates for MAGIC
resulted in a 2 to 5 percent change in the RMSE for predicted average annual ANC concentration. A 10
percent change in weathering parameters in ETD resulted in a 4 to 9 percent change in the RMSE for
predicted average annual ANC. A 10 percent change in ILWAS weathering parameters resulted in a
minimal change in the RMSE for predicted average annual ANC because the change was compensated
for by ion exchange in these short-term simulations (see Appendix A).
10-51
-------
Table 10-10. Percent Change in RMSE for MAGIC and ETD for a Ten Percent Change in Parameter
Values. Parameters are Ranked in Descending Order of Sensitivity from Left to Right
MAGIC
Factor Weathering Capacity S04 Adsorp.
Parameter"1 Weath. + Weath.- Depth* Depth- EMax+ EMax-
Woods Lake
Alkalinity -2.1 2.1 -1.0 2.1 -1.0 2.1
Hydrogen ion 0.5 -0.6 0.0 -0.6 0.0 -0.6
Panther I aka
Alkalinity 0.0 2.6 0.0 0.2 0.0 0.2
Hydrogen ion 0.0 0.0 0.0 0.0 0.0 0.0
Gear Pond
Alkalinity 4.3 4.9 0.1 0.2 -0.2 0.1
Hydrogen ion 0.0 0.0 0.0 0.0 0.0 0.0
ETD
Factor Weathering Ion Exchange Snowmelt Rate
Parameter4" KH5+ KH5 - RE+ RE- KAPPA+ KAPPA-
Woods Lake
Alkalinity -7.0 9.4 -3.1 3.7 4.7 -3.8
Panther Lake
Alkalinity 4.6 3.9 6.2 -2.6 -3.8 3.1
Hydro). Ion Exchange
PMAC+ PMAC- Select Selec-
1.1 -1.0 0.0 1.1
-0.6 0.0 0.0 0.0
0.0 -0.1 0.0 0.1
0.0 0.0 0.0 0.0
-0.1 -0.1 0.0 0.0
0.0 0.0 0.0 0.0
Lat/Vert. Hydraul.Cond.
KLAT3+ KLAT3- KPERC3 +KPERC-
3.1 2.2 -2.5 8.9
-1.0 -0.9 3.5 -2.2
"MAGIC Parameters
Weath = Weathering rate for base cations (meq m yr )
Depth a Estimated average depth to bedrock of the watersheds (m)
EMax - Maximum sulfate adsorption capacity (meq kg* }
PMAC « Unsaturated zone channeling parameter
Selec « Specific base cation (e.g.. Ca) to aluminum selectivity coefficient
b ETD Parameters
KH5 = Hydrolysis rate constraint for water body (eq m d)
RE = Ion exchange reaction rate coefficient (nr/eq"1 d"1)
KAPPA = Snow melt rate pn d'1 "C'1)
KLAT3 = Lateral flow recession constraint for the soil compartment (1 d"1)
KPERC3 = Vertical hvdraulic conductivity for soil (md"1)
10-52
-------
Weathering rate parameters (which are calibration parameters) were not, and generally are not,
measured in the field. These weathering rate parameters, however, are not completely unconstrained.
Weathering rates are constrained, in part, by cation-anion balances and ratios in surface waters and by
ranges observed In the literature for watersheds with similar geology, mineralogy, and soil and water
chemistry.
Hydrologic parameters also are constrained during calibration. Maintaining mass balance for
conservative constituents constrains evapotranspiration and runoff processes. Calibration of the sulfate
adsorption and ion exchange submodels constrains lateral and vertical hydraulic conductivity parameters
and, therefore, flowpaths through the watershed.
While there are similarities in several sensitive parameters among models, process formulations are
different among the three models. This is evident by the different parameters to which the predicted
output is sensitive among models. There is not a 1:1 mapping between parameters and processes for
any of the models. Mass balance, electroneutrality, and other requirements, however, constrain the
parameter values for all models including sensitive parameters. Similarities in calibration/confirmation
RMSEs indicate parameter values for all the processes can be constrained by watershed and lake
attributes to achieve calibration within the range of observed values.
10.9 REGIONAL PROJECTIONS REFINEMENT
The intensively studied watershed calibration/confirmation and sensitivity analysis exercises were
conducted to evaluate model behavior and the calibration procedures. These exercises resulted in
improvements and refinements in the calibration procedures used in the long-term DDRP projections.
These refinements are discussed below.
10.9.1 Enhanced Trickle Down
Calibration of the northeastern DDRP watersheds using ETD followed a similar procedure as that
used for the three intensively studied watersheds. The optimization program, (DESIGN, was used initially
10-53
-------
to optimize parameters followed by trial-and-error procedures. For the DDRP watersheds, the hydrologic
optimization focused on the evaporation rate for the lake and the lateral and vertical hydraulic
conductivities for the soil compartments. Sensitivity analyses indicated the model output was sensitive
to these three hydrologic parameters. The watershed/soil physical and chemical attributes and lake water
chemistry were used to estimate values for the other hydrologic parameters for each DDRP watershed,
which were fixed during calibration. All of the chemistry parameters were optimized and calibrated for
each watershed. ETD used the aggregated soil data discussed in Section 8.8.3 to determine the
watershed hydrologic and biogeochemical parameter values (e.g., cation exchange capacity, sulfate
adsorption, base saturation) used in calibration.
Initial conditions for each watershed simulation were set at the recalculated ELS-I ANC values and
the 1984 ELS • I value for other appropriate chemistry variables. Subsequent comparisons of calibrated
versus observed values in 1984 for ETD, therefore, will be identical.
10.9.2 Integrated Lake-Watershed Acidification Study
The ILWAS calibration procedure for the DDRP watersheds was similar to the calibration procedure
used for the intensively studied watersheds. The primary difference was a reduced number of
subcatchments for each of the DDRP watersheds. In general, only one or two subcatchments were used
to represent the DDRP watersheds. Individual soil pedon data, instead of aggregated soils date, were
used to calibrate the soil parameters, similar to the procedure used in calibrating these parameters on
the intensively studied watersheds.
10.9.3 Model of Acidification of Groundwater in Catchments
The MAGIC calibration sequence was similar to that used for the intensively studied watersheds but
the procedure was refined and automated, where possible, for the DDRP watersheds. First, TOPMODEL
was calibrated using daily rainfall and monthly runoff to derive flow routing parameters for the two-layer
structure of MAGIC. Next, MAGIC was calibrated using annual time steps to simulate average
volume-weighted lake chemical concentrations for comparison wrth the 1984 index lake chemistry values.
10-54
-------
No calibration was attempted for chloride because it was assumed to be a conservative Ion (except for
those northeastern watersheds in Priority Classes F -1 where the chloride balance was completed by sea
salt correction). Sulfate was not calibrated in the NE. The aggregated sulfate adsorption parameters
computed during aggregation of the soils data were used directly in MAGIC for the northeastern
watersheds. Sulfate adsorption parameters, however, were calibrated in the SBRP. The aggregated half-
saturation constant for sulfate adsorption was scaled by a constant factor for all catchments in the SBRP.
In the NE, the Baker et al. (!986b) model and coefficients for in-lake sulfate reduction were used. This
model computes sulfate reduction, in part, based on theoretical hydraulic residence times in the laka
No sulfate reduction was used in SBRP stream projections.
Finally, base cation concentrations were calibrated using an optimization procedure based on the
Rosenbrock (1960) algorithm. The base cation calibration involved fitting the results of long-term model
simulations to currently observed water and soil base cation data (i.e., target variables). The target
variables were both soil exchangeable fractions (for both soil compartments) and lake Index
concentrations of calcium, magnesium, sodium, and potassium. The target variables comprised a vector
of measured values, all of which must be reproduced by the model for a successful calibration. The use
of multiple, simultaneous targets in an optimization procedure provided robust constraints on model
calibration (Cosby et al., I986a).
Those priysicochemical soil and surface water attributes measured in the field in the DDRP Soil
Surveys were considered "fixed" parameters in the model, and the measurements were used directly in
the models during the calibration procedure. The maximum sulfate adsorption capacity and sulfate half-
saturation coefficient, determined for individual horizons and aggregated to the watershed, were used
directly in the model and were not calibrated. Base cation weathering rates and base cation exchange
selectivity coefficients for the soils were not directly measured and were used as "adjustable* or optimized
parameters in the calibration process. The calibrations were performed on simulations run from 1844 to
1984 for the NE and 1845 to 1985 in the SBRP. The historical deposition sequence over this period was
estimated by scaling the present-day deposition provided in the DDRP database to a reconstruction of
10-55
-------
sulfur emissions for the NE or Southeast (OTA, 1984). This scaling procedure has been described by
Cosby et at. (1985b). After each simulation, the 1984 and 1985 simulated versus observed values were
compared; the adjustable parameters were modified as necessary to improve the relationship between
simulated and observed values; the simulation was re-run and the procedure repeated until no further
improvements in these relationships were achieved.
10.9.4 DDRP Watershed Calibrations
The three models can be compared for the target population of lakes with ANC < 100 Meq L*1,
which corresponds to 495 northeastern lakes. The comparisons for ANC and sulfate for the three models
are shown as population histograms in Figures 10-11 and 10-12, respectively. Estimated aluminum
concentrations were added to estimated MAGIC alkalinity concentrations so the MAGIC ANC projections
are consistent with the ILWAS ANC estimates. The ILWAS and MAGIC models are calibrated on base
cations and acid anions so ANC is a computed, not a calibrated value (e.g., ANC = sum base cations -
sum acid anions). The ETD histograms are not discussed here because ETD assumed the 1984 ELS-I
lake index chemistry value was the calibrated value for initiating the long-term forecasts. Comparison of
the histograms for the calibrated 1984 ETD value versus the ELS-I value, therefore, are nearly identical.
The discrepancies between the ELS-I distributions and the ETD values represent lakes that were not
simulated by ETD in Classes A - E.
10.9.4.1 Integrated Lake-Watershed Acidification Study
ILWAS was not calibrated on the very acidic lakes (i.e., ANC < -30 jneq L*1) but generally matched
the ANC of moderately acidic lakes (ANC ~ -15 jieq L'1) (Figure 10-11). In general, the calibrated ANC
for the low ANC lakes (0 < ANC < 75 jieq L'1) was greater than the observed as evidenced by the
larger number of calibrated lakes with higher ANC than observed in the DDRP Priority Class A-8 target
population (Figure 10-11).
10-56
-------
2001
150
50
Northeast Lakes
Priority Class A - B
Model = Magic
Deposition a Constant
-40
10
35
60
2001
150-
100'
so-
85
Northeast Lakes
Priority Class A - B
Model = ILWAS
Deposition = Constant
110
135
160
-40
-15
10
35
60
110
135
160
Figure 10-11. Comparison of population histograms for simulated versus observed (Eastern Lake
Survey Phase I 1984 values) ANC for ILWAS and MAGIC. ETD used the ELS-I values as initial
model conditions, so the simulated values are nearly Identical to the observed values.
10-57
-------
200
M50-
1100
so-
Northeast Lakes
Priority Class A - B
Model = Magic
Deposition = Constant
D PHASE 1
B MAGIC Year 0
30 40 50 60 70 80 90 100110120130140150160170180190200210220230240250260270
200i
§150
5
««•»
2100
50
Northeast Lakes
Priority Class A - B
Model = ILWAS
Deposition = Constant
a PHASE 1
9 ILWAS Year 0
30 40 50 60 70 80 90 100110120130140150160170180190200210220230240250260270
Figure 10-12. Comparison of population histograms for simulated versus observed (Eastern Lake
Survey - Phase I 1984 values) suifate concentrations for ILWAS and MAGIC, Priority Classes A and
B. ETD used the ELS-I values as initial conditions, so the simulated values are nearly Identical
to the observed values.
10-58
-------
Calibrated sulfate concentrations generally were underestimated for lakes with observed sulfate
concentrations less than 75 neq L"1 and overestimated for lakes with observed sulfate concentrations
between 75 and 125 jteq L*1 (Figure 10-12). Calibrated sulfate values were overestimated for lakes with
suifate concentrations greater than 125 jueq L"1.
10.9.4.2 MAGIC
\
10.9.4.2.1 Priority Classes A and B -
MAGIC was not calibrated for the very acidic lakes (i.e., ANC < -30 /ieq L"1) but generally matched
the observed ANC for the moderately acidic lakes (ANC —15 /ieq L*1) (Figure 10-11). Calibrated ANC
concentrations, in general, were consistently higher than observed ELS-I ANC concentrations as indicated
by the underestimated number of lakes with lower ANC and overestimates of the number of lakes with
higher ANC (Figure 10-11). .
The low sulfate lakes (e.g., SO42' < 50 Meq L*1) were not represented in the MAGIC calibration
(Figure 10-12). The calibrated and observed sulfate concentrations were comparable for lakes with sulfate
concentrations between 75 and 115 /ieq L*1 (Figure 10-12). The calibrated sulfate concentrations typically
exceeded observed sulfate concentrations for lakes with observed sulfate concentrations greater
150 Meq L"1.
10.9.4.2.2 Priority Classes A - E -
The calibrated ANC concentrations generally were higher than observed ANC concentrations for
the low ANC lakes (i.e., < 100 jueq L'1) but were lower than observed for moderate ANC lakes (i.e., 120-
175 jueq L'1) (Figure 10-13). The calibrated ANC concentrations for higher ANC lakes (i.e., > 175
L'1) were similar to or greater than observed ANC concentrations.
10-59
-------
400
J300
S2°0
3
z
100
Northeast Lakes
Priority Class A - E
Model = Magic
Deposition = Constant
-40 -15 10 35 60 85 110 135 160 185 210 235 260 285 310 335 360 385 410
ANCOieqL-1)
400]
300
2 2001
£
5 100
30 40 50 60 70 80 90100110120130140150160170180190200210220230240250
Rgure 10-13. Comparison of population histograms for simulated versus observed (Eastern Lake
Survey Phase I 1984 values) ANC and sulfate concentrations for MAGIC, Priority Classes A - E.
10-60
-------
The calibrated sulfate concentrations were higher than observed for the low sulfate lakes (e.g., < 75
L"1) and generally lower than observed sulfate concentrations for the moderate sulfate systems (i.e.,
75 < SO4*2 < 135 jueq L*1). The calibrated sulfate concentrations generally exceeded observed sulfate
concentrations for the high sulfate lakes (Figure 10-13).
10.9.4.2.3 Priority Classes A -1 -
The calibrated ANC concentrations generally were higher than observed for the low ANC lakes (i.e.,
<100 Meq L*1) but similar for the higher ANC lakes (Figure 10-14). Calibrated sulfate concentrations
exhibited a varied pattern compared with the observed pattern that, in general, was slightly lower for low
sulfate lakes and slightly higher for high sulfate lakes (Figure 10-14).
10.9.4.3 Southern Blue Ridge Province
10.9.4.3.1 Priority Classes A and B -
Calibrated versus observed ANC and sulfate concentrations are shown for the ILWAS and MAGIC
models in Figure 10-15. The MAGIC calibrations overestimated the number of low ANC and high ANC
streams and underestimated the number of moderate (i.e., 75 to 125 jueq L"1) ANC streams (Figure 10-
15). ILWAS-calibrated ANC was similar to observed ANC for the low ANC streams (e.g., < 75 /*eq L*
1) but overestimated the number of moderate ANC systems and underestimated the number of high ANC
streams (Figure 10-15).
MAGIC underestimated the number of low sulfate streams and overestimated the number of higher
sulfate streams (Figure 10-16). ILWAS also underestimated the number of low sulfate streams and
overestimated the number of higher sulfate streams (Figure 10-16).
10.9.4.3.2 Priority Classes A - E -
Calibrated versus observed ANC and sulfate concentrations for the MAGIC model in Priority Classes
A - E are shown in Figure 10-17. MAGIC overestimated the number of low ANC streams but, in general.
10-61
-------
Northeast Lakes
Priority Class A -1
Model = Magic
Deposition = Constant
D PHASEI
MAGIC Year 0
•40 -15 10 35 60 85 110 135 160 185 210 235 260 285 310 335 360 385 410
ANCOieqL'1)
30 40 50 60 70 80 90100110120130140150160170180190200210220240260280300
Figure 10-14. Comparison of population histograms for simulated versus observed (Eastern Lake
Survey Phase I 1984 values ) ANC and sulfate concentrations for MAGIC, Priority Classes A - 1.
10-62
-------
SBRP Stream Reaches
Priority Class A -B
Model = Magic
Deposition = Constant
200
jj 150
k.
«••
W
o 1001
5
1
I 50
-40 -15
35
60 85 110
ANCftieqL-')
135 160 185 210
200
10
E150
a
£
CO
0.100
z 50
SBRP Stream Reaches
Priority Class A -B
Model = ILWAS
Deposition = Constant
-40 -15 10 35 60 85 110 135 160 185 210
ANCftiflqL'1)
Figure 10-15. Comparison of population histograms for simulated versus observed (NSS Pilot
Survey values) ANC, Priority Classes A and B using ILWAS and MAGIC.
10-63
-------
400
(0
£300
3
55
"5200
z 100
SBRP Stream Reaches
Priority Class A -B
Model = Magic
Deposition = Constant
10 20 30 40 50 60 70 80 90 100 110 120 130 140
400
«
E 300
|
w
o 200
E
Z 100
SBRP Stream Reaches
Priority Class A -B
Model = ILWAS
Deposition = Constant
10 20 30 40 50 60 70 80 90 100 110 120 130 140
Figure 10-16. Comparison of population histograms for simulated versus observed (NSS Pilot
Survey values) sulfate concentrations, Priority Classes A and B using ILWAS and MAGIC.
10-64
-------
SBRP Stream Reaches
Priority Class A -E
Model = Magic
Deposition = Constant
400
E300
0
£
-------
the distribution of calibrated ANC was similar to the observed ANC distribution. The calibrated sulfate
distribution for MAGIC underestimated the number of low sulfate streams (e.g., < 40 jueq L"1) and slightly
overestimated the number of higher sulfate streams (Figure 10-17).
10.10 MODEL PROJECTIONS
10.10.1 General Approach
The general approach for performing long-term projections of the effects of sulfate deposition on
surface water chemistry over the next 50 years in the NE and SBRP was illustrated schematically in
Figure 10-4. The two simulated deposition scenarios were illustrated previously in Figure 5-27. In the
first scenario in the NE, the current deposition rate at each individual site was maintained over the 50-
year interval. The models then projected changes in lake water chemistry over that 50 years. In the
second deposition scenario, current deposition rates at each site were held constant for the first 10 years,
decreased by a total of 30 percent over the next 15 years, and then held constant at this reduced
deposition rate for the next 25 years. This phasing corresponded with one possible scenario of how
deposition rates might change if emissions controls were implemented in the NE in year 1 of the
simulation (OPPE, personal communication). Deposition rates have been declining in the NE over the
past 15 years and would decline further if emissions controls were promulgated. Current deposition rates
at each individual watershed in the SBRP also were simulated, but the alternative deposition scenario was
an increase in deposition. For the alternative scenario, deposition rates were held constant for the first
10 years, increased by a total of 20 percent over the next 15 years, and then held constant at this
increased rate for the next 25 years (Figure 5-27). Deposition rates are expected to increase in the
Southeast (OPPE, personal communication).
Each model was calibrated on individual watersheds using the data sources indicated in Section
10.5. The projected change In surface water chemistry in the individual lake or stream was simulated for
the next 50 years using the typical year meteorology and deposition data, discussed in Section 10.5.
Output for each model represents flow-weighted annual average constituent concentration. The projected
ANC is defined similarly and is consistent among all three models. Not all watersheds in the NE and
10-66
-------
Not all watersheds in the NE and SBRP were simulated by all three modelling groups. For the
MAGIC model, there were some watersheds for which the optimization and calibration criteria were not
satisfied. Long-term projections for these watersheds, therefore, were not performed (Table 10-11). In
the NE, optimizations criteria generally could not be achieved either because of chloride imbalances or
because cation inputs exceeded outputs. Time and funding constraints restricted the number of ETD
simulations to northeastern watersheds in Priority Classes A - E (Figure 10-1). Similar constraints
occurred with the ILWAS model; in the NE, only watersheds in Priority Classes A and B were simulated,
while only watersheds in Priority Class A were simulated in the SBRP. The ILWAS simulations are
ongoing and the anticipated number of watersheds simulated in the NE and SBRP should be alt of those
in Priority Class A - C and Class A - B, respectively, by June 30, 1989. Optimization and calibration
criteria for some northeastern watersheds also were not satisfied for the ETD and ILWAS models. These
watersheds are listed in Table 10-11. The individual watersheds simulated by each model in each region
are presented in Appendix A.2 and are listed by NSWS lake or stream identification number, name (if
available), state, latitude and longitude, and initial NSWS ANC. Comparisons among models are made
only for similar target populations, and the target population is clearly identified for each comparison.
The results discussed below have been obtained by weighting the individual watershed estimates by the
appropriate inclusion probability. Weighting by the appropriate inclusion probability is critical for any
analyses performed on these data. The individual watershed estimates are of interest only as they relate
to the distribution of the population attributes.
10.10.2 Forecast Uncertainty
An integral part of all the analyses performed in the DDRP is the estimate of error associated with
the analyses or projections. Each modeling group conducted error analyses on Its respective model,
which were incorporated in the confidence intervals about the population estimates.
10-67
-------
Table 10-11. Watersheds, by Priority Class, for which Calibration Criteria
Were not Achieved
Priority
Region Class
Northeast A
B
C
D
E
F
Q
H
SBRP A
B
C
D
E
ETD
0
0
0
1E2-069
0
NA
NA
NA
NA
NA
NA
NA
NA
Watershed
Model
ILWAS
0
0
NA
NA
NA
NA
NA
NA
2A07821
2A08906
2A08802
2A08803
NA
NA
NA
NA
ID
MAGIC
1D2-027
1C1-068
183-056
1E1-106
1D3-002
1A2-004
0
1A2-058
1B1-043
1D3-003
1 02-094
101-067
103-029
1C2-054
102-049
101-031
1C3-055
1A3-028
102-036
101-068
2A07811
2A07816
2A08803
2A07803
0
0
10-68
-------
10.10.2.1 Watershed Selection
The computational time required to conduct uncertainty analyses on all simulated DDRP watersheds
would have been prohibitive. Therefore, six northeastern watersheds were selected from Priority Classes
A - D. The watersheds were sorted by the following criteria:
* previously simulated for the EPA Internal Staff Paper,
• no internal sulfur sources
• drainage lakes versus seepage lakes (drainage lakes preferred)
• percent sulfur retention (positive sulfur retention preferred)
• watershed disturbance as indicated by chloride balance (undisturbed watersheds preferred)
• ANC class (watershed/lake systems with ANC < 100 /*eq L"1 preferred)
This sorting emphasized (1) those systems considered likely to show a response, (2) those for
which early modelling output might be available, and (3) those that provided a representative cross-
section of potential watershed responses. The six watersheds randomly selected for uncertainty analyses
were
(1) one watershed (1A3-048) from Class A - previously simulated
(2) two watersheds (1A2-045, 1E1-111) from Class B - low ANC, positive sulfur retention
(3) two watersheds (1A1-003, 1C2-035) from Class C • low ANC, negative sulfur retention
(4) one watershed (1D3-025) from Class D • high ANC, positive sulfur retention
Characteristics of these six watersheds follow:
Watershed ID Priority Class
ANC
% S Ret. Soil Type WA LA WA1A
1A3-048
1A2-045
1E1-111
1A1-003
1C2-035
1D3-025
A
B
B
C
C
D
14.6
13.2
11.0
-9.9
73.6
149.3
-99.0
4.0
12.2
-36.7
-15.6
14.6
Spodosols
Spodosols
Mixed
Mixed
Spodosols
228
168
80
96
215
57
54
26
24
13
11
8
4.7
6.4
3.4
7.5
19.0
7.4
10-69
-------
These watersheds include a lake with negative ANC, systems with low ANC and relatively high ANC (i.e.,
74 and 150 jieq L'1), watersheds in four of the five subregions, a distribution of watersheds across the
deposition gradient, and watersheds selected from clusters representing the majority of watersheds with
ANC <100 Meq L*1. Uncertainty analyses conducted on these watersheds were assumed to be
representative of the other watersheds within these priority classes.
In the SBRP, five watersheds were randomly selected for uncertainty analyses. The watersheds
were sorted in priority order, as illustrated in Figure 10-2. Two watersheds were selected from Priority
Class A: watersheds with ANC < 100 /ieq L"1 and chloride < 50 /ieq L"1 , and with positive sulfur
retention. Two watersheds were selected from Priority Class B: watersheds with ANC >100 jieq L"1 but
<200 jueq L"1 , chloride < 50 /teq L"1, and with positive sulfur retention. One watershed was selected
from Priority Class D: this watershed had ANC < 200 /ieq L'1 but chloride > 50 peq L'1 , indicating
possible watershed disturbance. This watershed also had positive sulfur retention. Characteristics of
these five watersheds follow:
Watershed ID Priority Class ANC % S Ret. Soil Type WA
2A07828 A 37.0 81.1 Acid crys., high org. 19.1
2A08802 A 71.0 83.6 Acid crys., low org. 5.7
2A08810 B 114.4 77.9 Acid crys., low org. 4.9
2A08811 B 95.3 49.0 Acid crys.,/meta sedmt., 3.3
low org.
2A07830 D 1630 58.2 Acid crys., low org. 14.0
Uncertainty analyses conducted on these watersheds were assumed to be representative of other
watersheds or similar watersheds within these classes.
10.10.2.2 Uncertainty Estimation Approaches
Three different approaches were used with the individual models to estimate the error associated
with the projections. The three approaches were first order second moment analyses (ETD), first order
error analyses (ILWAS), and "fuzzy" optimization and multiple simulations (MAGIC). These approaches
10-70
-------
reflect differences in mode! formulations, which required different uncertainty estimation approaches.
Uncertainty estimates were computed for both input and parameter error. However, parameter error and
input error were computed separately for each of the models.
Variance estimates were derived from the aggregated soils data and from the deposition monitoring
sites. Variance algorithms were used with the soils aggregation procedures to obtain variance estimates
for the physical and chemical variables aggregated to the watershed level. These variance estimates were
used to scale the parametric variance associated with these different variables. For example, hydraulic
conductivities were a function of soil porosity, sulfate half-saturation coefficients were a function of soil
sulfate adsorption, and ion exchange selectivity coefficients were a function of soil cation exchange
capacity and percent base saturation. Estimating the effect of parametric uncertainty on model output
involved propagating this range of watershed parameter values through the model and observing the
range in model output constituent concentrations or values.
Wet deposition uncertainty estimates were computed by calculating the variance for individual
chemical species over the period of record at each Acid Deposition System (ADS) station used in the
NE or SBRP, and adding the variance associated with (1) kriging of precipitation to the individual NE or
SBRP sites and (2) extrapolating from the nearest ADS site to the simulated watershed. This latter
variance component was obtained using resampling or jackknifing procedures following random deletion
on an ADS site. The wet deposition relative standard deviations are listed, by species, in Table 10-12.
Dry deposition estimates for these species were assumed to be +. 50 percent of the estimated annual
dry deposition values at each individual site. Deposition uncertainty estimates were evaluated by varying
the deposition (both wet and dry) consistently up or consistently down for all chemical species.
Meteorological variability was not specifically investigated. The operational assumption is that typical year
projections provide a common basis for comparisons among deposition scenarios to assess potential
changes in surface water chemistry. Deposition uncertainty is an important part of these comparisons.
Meteorological variance is implicitly incorporated in the deposition uncertainty but there is no intent to
investigate interannual or interdecadal variance in meteorology.
10-71
-------
Table 10-12. Deposition Variations
Used in Input Uncertainty Analyses
Decosition
Chemical
Species
Sulfate
Nitrate
Chloride
Ammonium
Sodium
Potassium
Calcium
Magnesium
Hydrogen
Wet
%RSD
17.8
17.2
46.9
38.1
57.6
70.9
35.5
29.6
Used to complete
Dry
% RSD
50
50
50
50
50
50
50
50
charge balance
10-72
-------
10.10.2.2.1 Enhanced Trickle Down -
Input, initial condition, and parameter errors were evaluated for the ETD model using first order
second moment analyses. First order second moment analyses involve replacing the nonlinear model
with a first order Taylor series approximation for error covariance propagation. First order second
moment analyses includes not only the simultaneous effects of all state variables, inputs, and parameters
on each state variable, but also the propagation of uncertainties in inputs, parameters, and state variables
(Lee et al., in press). Initial condition and parametric error were evaluated for the six northeastern
watersheds listed above. Input (i.e., deposition) uncertainty was evaluated only for Kalers Pond (1E2-
063) (Georgakakos et al., in press).
10.10.2.2.2 Integrated Lake-Watershed Acidification Study -
First order error analysis was used to estimate uncertainty in the ILWAS model. First order error
analysis is similar to sensitivity analysis. Parameters or input variables were subjected to small
perturbations and the change in selected output variables relative to the perturbation provided an estimate
of the first derivatives. The first derivative estimates were used as weights to propagate the parametric
or Input variance to an output variance. This procedure is similar to the first order second moment
analysis but the covariance matrix is not estimated with the first order error analysis.
10.10.2.2.3 Model of Acidification of Groundwater in Catchments -
Uncertainty estimates for the MAGIC model were obtained using a fuzzy optimization procedure.
The fuzzy optimization procedure consisted of multiple calibrations of each catchment using perturbations
of the values of the fixed parameters to reflect the sampling and measurement error In these parameters.
These error estimates were obtained from the aggregated soils data. Each of the multiple calibrations
began with (1) a random selection of perturbed values of fixed parameters, and (2) a random selection
of the starting values of the adjustable parameters. The adjustable parameters then were optimized using
the Rosenbrock algorithm to achieve a minimum error fit to the target variables. Using the fuzzy
optimization on multiple calibrations (i.e., an average of 7 calibrations for each DORP catchment with a
minimum of 3 and a maximum of 10 calibrations/catchment), uncertainty bands of maximum and
10-73
-------
minimum values were computed for each output variable for each year. These uncertainty bands
encompass the range of variable values that were simulated given the specified uncertainty in fixed
parameter values and measured target variables. The difference between maximum and minimum
simulated values defines an uncertainty width about the simulated value arising from parametric
uncertainty for each output variable for each DDRP catchment. The values of the uncertainty widths for
each variable in the calibration year were regressed against the simulated value of the variable in the
calibration year across all the DDRP catchments to derive a percentage uncertainty value for each value,
representative of the region.
10.10.2.3 Relationship Among Approaches
Each of the approaches used for uncertainty analyses is appropriate for the model being used for
DDRP Level III projections. MAGIC runs on a microcomputer, for example, and was coded for the fuzzy
optimization analyses because computational time was not a consideration for MAGIC simulations. This
approach was developed for the DDRP to incorporate both input and parameter uncertainty. The ETD
model is intermediate In complexity and requires greater computational effort. The first order second
moment analysis provides an estimate of simulation uncertainty and permits partitioning this uncertainty
into input and parameter components (Lee et al., in press). The ILWAS model has the greatest number
and most complex set of formulations. The ILWAS model, therefore, is not compatible with optimization
or first order second moment analyses but is compatible with first order error analyses, which provide
an estimate of simulation uncertainty.
The relationship between the uncertainty estimates using the two different procedures for MAGIC
and ETD is shown for ANC and sulfate in Figure 10-18. For both models, the magnitude of the standard
deviation was a function of the ANC or sulfate concentration as indicated by the slope of regression line.
A multiplicative error term, therefore, was used in the uncertainty analyses. The slope of the line for both
models was nearly identical with the differences occurring in the offset. This offset resulted in greater
uncertainty estimates for MAGIC than ETD. The MAGIC simulations, however, incorporated both input
and parameter error while the ETD simulations (with the exception of Kaler's Pond) included only
10-74
-------
55
o
Q -
I ^
1151
CO -
..a—
•**
Northeast Lakes
Model Uncertainty
ANC
KaleraPond
O
..o-
-20 0 20 40 60
ANC (ueq L'1) at 50 Years
80
o-— ETD
o— MAGIC
45
O
-230
Q -j
"2 -•
CO
Northeast Lakes
Model Uncertainty
SO4
D KalersPond
55 70 85 100 115
SO42"(ueq L1) at 50 Years
130
o ETD
a— MAGIC
Figure NM8. Comparison of projection standard errors as a function of ANC (top figure) and
sulfate (bottom figure) concentrations for the NE uncertainty analysis watersheds using ETD and
MAGIC.
10-75
-------
parameter error. Kaler's Pond included both input and parameter error and is comparable to the MAGIC
standard deviation estimates. The offset between the two models represents the input error. Because
of the similarity between MAGIC and ETD for Kaler's Pond, the MAGIC standard deviations were used
to compute confidence intervals for both ETD and MAGIC.
The estimates of parametric uncertainty using the ILWAS model also are consistent with the regional
estimates of both MAGIC and ETD (i.e., Kaler's Pond). Input uncertainty estimates were computed by
all three models using the individual uncertainty estimation procedures discussed above. The input
uncertainty estimates computed for all three models were on the same order as the parametric
uncertainty. Parametric variance estimates, therefore, were doubled to obtain estimates of projection
uncertainty for the ILWAS model. The procedures described in Section 6 were used to integrate this
projection error with sampling error and compute confidence intervals for the population estimates
presented in the next section.
10.10.2.4 Confidence Intervals
Upper and lower bounds for a 90 percent confidence interval were computed for ANC, pH, and
sulfate projections from all three models and for calcium and magnesium for MAGIC and ILWAS
(Appendix A.3). The confidence intervals were computed using the variance estimator discussed in
Section 6. This variance estimator includes estimates of sampling and measurement error, parameter and
input error, and regional estimation error. Time constraints prevented the inclusion of confidence intervals
for the ILWAS SBRP projections. These figures will be incorporated in the Mid-Appalachian Report in mid-
1990.
10.11 POPULATION ESTIMATION AND REGIONAL FORECASTS
Population estimation procedures were discussed in Section 6. The uncertainty estimation
procedures were discussed in the previous section, Section 10.10. Comparisons among model
projections are presented only for comparable target populations. The number of watersheds simulated
with each model differed; therefore, it is critical that comparisons be made only among or between
10-76
-------
models that simulated the same watersheds, i.e., those that represent the same target population. The
population estimates discussed below represent three different target populations. The target population
discussed In each of the following sections is dearly defined. These definitions are essential for the
proper interpretation of the results. The models, target populations, and population attributes for both
the NE and SBRP are shown in Table 10-13.
10.11.1 Northeast Regional Projections
10.11.1.1 Target Population Projections Using MAGIC
An estimated 3227 lakes in the target population in the NE were simulated using MAGIC, compared
to the DDRP total target population of 3667. The smaller target population reflects an exclusion of lakes
for which MAGIC was unable to satisfy the calibration criteria. The simulated target population represents
Priority Classes A - I, which includes both disturbed and undisturbed watersheds based on chloride
concentrations (See Section 10.5.7), watersheds that had both positive and negative sulfur retention, and
watersheds that had initial ELS-I ANC concentrations ranging from -53 to 392 fteq L"1. The MAGIC
simulations for this target population extend to 100-year projections. Projections using other models were
restricted to 50 years, because time and computational requirements prohibited longer simulations. The
100-year time frame provides additional insight into the cumulative effects of sulfur deposition on changes
In surface water chemistry.
10.11.1.1.1 Deposition scenarios -
Projected changes in ANC and sulfate concentrations that might occur over a 100-year period,
assuming current and decreased deposition, are shown in Figure 10-19. The confidence limits about the
individual simulations are included in Appendix A.3. Confidence intervals are not included on the figure
In order to Increase the contrast among model projections or between deposition scenarios.
The projected changes in median ANC concentrations over a 100-year period assuming either
current and decreased deposition were small (Figure 10-19). The median ANC concentration projected
after 50 years for constant deposition at current levels was 124 /jeq L*1 and for a 30 percent deposition
10-77
-------
Table 10-13. Target Populations for Modelling Comparisons and Population Attributes
Region
Northeast
Total DDRP Target
SBRP
Priority
Class
A and B
A-E
A-l
Population (NE)
A and B
A-E
Models
ETD.ILWAS
MAGIC
ETD.MAGIC
MAGIC
ILWAS,
MAGIC
MAGIC
Target
Population
502
1813
3227
3667
567
1323
Population
Attributes
ANC<100/ieqL"1(lnt
Staff Paper)
ANC<400, Undisturbed
Watersheds
ANC<400, Represent-
ative of entire NE
population
ANC< 100, Undisturbed
Watersheds
Qi-iB-iBinruXBatnti JUT. j-rf .rmtivM.
nepreseiwuMe or enure
SBRP population
Total DDRP Target Population (SBRP)
1531
10-78
-------
tOr
a.
8 0.6
a.
s
00
-KM
NE Lakes
Model - MAGIC
Priority Class - A - I
Year - 20
SknuMkm Ywv 0
—-- Constant Deposition
-••—-» Rtmp Dvpoiltton
0 100 200 300 400
ANC ftiaq I/O
tOr
NE Lakes
Model - MAGIC
Priority Class - A - I
Year - 20
300
Year - SO
tOr
om
-
Simutatlon YMT 0
---- Cantttnl Dtposltlen
0 100 200 100
ANC (|ieq LI)
400
to
on
04
U
OJJ
Year • SO
100
ISO,*!
— Sknotatton Year 0
---- Constant
SfflQ
L"«)
900
Year - 100
to,
o ^
0.0
-wo
'Simulation Y*v 0
Constant Deposition
Rtnp Deposition
0 100
ANC
200 100 400
Figure 10-19. Projections of ANC and sulfete concentrations for NE lakes, Priority Classes A - I,
using MAGIC for 20, 50, and 100 years, under current deposition and a 30 percent decrease in
deposition.
10-79
-------
decrease was 135 /ieq I"1, representing a difference of 11 /ieq L"1 (Table 10-14). The change projected
in median sulfate concentration after 50 years for current deposition was 99 /*eq L"1 and for decreased
deposition was 71 /teq L*1, representing a -28 jueq L"1 difference. The changes projected in median ANC
concentration over a 100-year period between current and decreased deposition were 121 versus 134 jieq
L*1, respectively, or a difference of 13 M@q L'1. The projected changes in median sulfate concentration
after 100 years for current deposition and a 30 percent deposition decrease were 99 versus 70 Meq L*
1, respectively, or a difference of -29 Meq L"1. A small decline in ANC concentrations, (less than 1
L'1) was indicated between year 50 and year 100 for both current and decreased deposition. Projected
calcium and magnesium concentrations also showed a small but continual decline over the 100-year
period under both current deposition and a 30 percent deposition decrease (Table 10-14). Sulfate
concentrations declined during the initial 50-year period, asymptotically approaching steady state, and
were projected to remain essentially constant from year 50 to year 100 under current deposition and to
decrease slightly over this same period for decreased deposition. The confidence limits about the
projected CDFs represented a projection error of about ± 36 /ieq L*1 in ANC and ± 32 /xeq L"1 in sulfate
concentrations. Both the changes projected for 50 and 100 years, assuming different deposition
scenarios, were within the uncertainty bounds of the projections.
The projections of the sulfate concentrations indicated the watersheds would be near sulfate steady-
state after 50 years under current deposition with the median projected watershed sulfur retention of
about 5 percent in both year 50 and year 100 (Table 10-14). The interquartile range varied from about
1 to 11 percent in both year 50 and 100, indicating the majority of the watersheds were projected to be
near sulfate steady state. The median projected sulfur retention under decreased deposition for these
watersheds was nearly zero. From year 50 to year 100, the projected median sulfur retention changed
from 0 to a slightly positive sulfur retention (-5 percent) in the watersheds. The interquartile ranges for
the 50- and 100-year periods were -7 to +8 percent and -1 to +9 percent, respectively.
10-80
-------
Table 10-14. Descriptive Statistics of Projected ANC, Sulfate, pH, Calcium Plus
Magnesium, and Percent Sulfur Retention for NE Lakes In Priority Classes A - I
Using MAGIC for Both Current and Decreased Deposition
Year
Maaic All.
YrO
Yr20
YrSO
Yr 100
Maaic All.
YrO
Yr20
YrSO
Yr 100
Maaic AM.
YrO
Yr20
YrSO
Yr 100
Maaic All.
YrO
Yr20
YrSO
Yr 100
Maaic All.
YrO
Yr20
YrSO
Yr 100
Mean
ANC
151.43
151.16
149.49
145.55
SO 2'
1116.87
108.42
106.46
106.29
pH
6.01
6.02
5.99
5.90
Ca + Mq
228.92
220.88
217.17
213.11
Std.
Dev.
113.96
113.91
113.88
112.84
47.51
43.55
43.09
43.42
0.60
0.60
0.62
0.66
122.10
120.57
120.13
119.87
Min.
Current
-21.25
-21.10
-21.62
-21.93
50.09
47.47
46.24
45.34
4.47
4.49
4.48
4.47
41.09
39.78
38.39
36.54
P_25
Deposition
70.18
70.24
67.27
61.76
70.78
70.06
67.82
66.36
6.73
6.73
6.71
6.67
128.49
123.75
121.29
118.41
Median
126.18
125.88
123.57
121.39
110.81
101.28
99.09
98.73
6.97
6.98
6.98
6.96
197.27
190.01
186.25
182.18
P_75
222.99
223.68
223.78
215.92
151.58
146.74
142.08
140.47
7.22
7.22
7.22
7.21
295.65
288.32
283.65
279.54
Max.
416.47
416.66
414.05
408.96
245.60
221.44
214.91
213.83
7.49
7.49
7.49
7.48
559.60
544.06
540.20
531.30
% S Retention
-3.59
3.85
5.72
5.97
10.06
8.57
8.14
8.43
-24.58
-19.88
-17.86
-18.98
-11.05
-1.47
1.07
1.75
-3.47
3.37
4.96
5.33
3.55
10.40
11.13
10.92
19.34
25.35
26.49
26.65
continued
10-81
-------
Table 10-14. (Continued)
Year
Mean
Std.
Dev.
Min.
P_25
Median
P_75
Max.
30% Decrease in Deposition
Maaic All.
YrO
Yr20
YrSO
Yr 100
Maaic All.
YrO
Yr20
YrSO
Yr 100
Maaic All.
YrO
Yr20
YrSO
Yr 100
Maaic All.
YrO
Yr20
YrSO
Yr 100
Maaic All.
YrO
Yr20
YrSO
Yr 100
ANC
151.43
156.21
160.06
158.41
so,2-
116.87
99.22
79.77
76.86
.fiH
6.01
6.13
6.28
6.26
Ca^+ Ma
228.92
217.43
202.97
198.25
113.96
114.70
114.90
113.71
47.51
40.02
32.60
31.89
0.60
0.56
0.52
0.53
122.10
120.13
117.77
116.66
-21.25
-20.34
-18.20
-18.97
50.09
44.01
34.87
33.78
4.47
4.52
4.58
4.58
41.09
39.42
35.37
32.03
70.18
73.66
75.84
73.66
70.78
63.29
51.52
47.88
6.73
6.74
6.77
6.75
128.49
120.27
112.92
109.72
126.18
130.61
135.18
134.49
110.81
93.44
71.30
69.77
6.97
6.99
7.01
7.01
197.27
183.87
172.87
169.94
222.99
226.79
230.68
230.88
151.58
131.68
107.14
102.58
7.22
7.23
7.24
7.24
•
295.65
285.77
267.49
267.89
416.47
424.86
430.94
428.36
245.60
202.13
157.31
153.46
7.49
7.50
7.50
7.50
559.60
539.01
529.47
510.94
% S Retention
-3.59
-9.94
-0.98
2.97
10.06
11.37
12.64
11.00
-24.58
-43.83
-44.44
-35.93
-11.05
-16.41
-6.69
-1.27
-3.47
-9.80
0.54
4.15
3.55
•0.96
7.94
9.06
19.34
16.53
23.90
25.82
10-82
-------
Projected changes in pH that might occur over a 100-year period, assuming current and decreased
deposition are shown in Figure 10-20 and listed in Table 10-14. The projected changes in pH over a 100
year period under constant and decreased deposition were small (Figure 10-20). The projected
differences between the median pH under constant and decrease deposition after 100 years were less
than 0.05 pH units (Table 10-14). The projected differences in pH for the lower quartile after 100 years
under the two deposition regimes were less than 0.1 pH units, varying from 6.67 under current deposition
to 6.75 under decreased deposition (Table 10-14).
Projections of the number of lakes currently not acidic that might become acidic in the next 50
years for current deposition were 87 (3 percent) lakes and 50 (2 percent) for a 30 percent deposition
decrease based on an estimated target population of 3227 lakes. Projections of the number of lakes
currently not acidic that might become acidic in the next 100 years under current deposition and a 30
percent deposition decrease were 100 (3 percent) and 50 (2 percent), respectively. The number of
currently acidic lakes (i.e., 162 lakes in the target population with ANC <0 /ieq L*1) that might chemically
improve (increase in ANC) under current deposition levels and a 30 percent deposition reduction after
50 years were projected as 64 (39 percent) and 125 (77 percent), respectively. The number of currently
acidic lakes that might chemically improve after 100 years at current and decreased deposition levels was
projected as 52 (32 percent) and 113 (70 percent), respectively. The percentages for chemical
improvement are based on the number of currently acidic lakes estimated in the target population (i.e.,
162 lakes).
10.11.1.1.2 Rate of change of ANC, sulfate, and pH over 100 years -
The projected changes in ANC, sulfate concentrations, and pH over the next 100 years are
displayed as box and whisker plots (Figures 10-21 through 10-23). Box and whisker plots illustrate how
both the target population constituent median and interquartile range vary through time.
10-83
-------
NE Lakes
Model * MAGIC
Priority Class » A - I
Year - 20
tOr
0.6
« 0.4
— SknutaUen YMT 0
Con«tam Dopoittfon
— Ramp D»po«ttton
4J> 42 SA IS 64 M 7.0 73
PH
Year - 50
to
So.
8
e
•*
0.4
Skmrt»Uon YMT 0
Conttmt Mpatlttan
Mmp Mporitlm
4J) 4A S.O
7J»
Year • 100
u>r
StowttUoa YMT 0
Bwnp
4J 44 M 10 M U 7.0 7*
PH
Figure 10-20. pH projections for NE lakes, Priority Classes A - I, using MAGIC for 20, 50, and
100 years, under current deposition and a 30 percent decrease in deposition.
10-84
-------
3rd OuartB* *
(1.5 x hwquuflte Rang*)"
1KC
1«t{_
<1 S X Mvquwtf* HMIB*)"
"Net to noted mMM wkw
Constant
250-
200-
100-
50-
o-
-inn—
o
e
w
flW
•••
o
f
^*
••
^
•riH
•
Ml
S
8E
Ml
W
^M
•P
IWHI
§
-
^
Figure 10-21. Box and whisker plots of ANC distributions at 10-year intervals for NE Priority
Classes A - I using MAGIC.
10-85
-------
3ri Quarts* +
MQwrtl*
(UxMv
•ftatkitttMd
Constant
8^%
»
£
s
PC
8 i
250-
200
f' —
8 100-
50-
^•A
•
^
^ ^
•^
^i^
•1
M
^
— — -
HI
HI
•1M
i— — «
^^ ^^
^
^
^
^•V
^
^
"
HI
^
--
Ramped
8' S
8
250—1
200
iJ 1SO—
*- _,
g 100.-S
50-
Figure 10-22. Box and whisker plots of sulfate distributions at 10-year intervals for NE Priority
Classes A -1 using MAGIC.
10-86
-------
3rd Quart* +
(1.5 x Inuitquaitto Fl»ng«)-
Mwn
Mwiiin
tMOuutfli
1MQu*itl*
(1 J5 x InOfquwtto Rang*)
Constant
7-
16-
5-
o
oc
oc
oc
8
>-
CC
CC
I
CC
7-
5-
o
CC
Ramped
^3 ^^
« »
CC
S
CC
I
CC
Figure 10-23. Box and whisker plots of pH distributions at 10-year intervals for NE Priority
Classes A -1 using MAGIC.
10-87
-------
The median ANC concentrations projected for current deposition changed from an initial calibrated
concentration of 126 /ieq L*1 to 124 /ieq L"1 after 50 years and to 121 /ieq L"1 after 100 years (Table
10-14). For a 30 percent deposition decrease, the median ANC was projected to change from 126 /ieq
L"1 at year 0 to 135 /ieq L"1 at year 50, remaining essentially unchanged over the next 50 years. The
median calcium plus magnesium concentration decreased linearly over the entire 100-year simulation
period for current deposition with a projected decrease from 197 to 182 /ieq L*1 (about 0.15 /ieq L*1
yr1). Median calcium plus magnesium concentrations also declined from 197 to 170 /ieq L"1 over the
100-year period for decreased deposition and from 197 /ieq L"1 in year 0 to 173 /ieq L"1 in year 50
(approximately 0.5 /ieq L'1 yr'1). The projected rate of change for the next 50 years decreased (less
than 0.1 /ieq L'1 y'1) but retained a negative slope.
The projected change in median sulfate concentration for current deposition was an asymptotic
decrease toward sulfate steady state (or to a small positive retention due to In-lake retention) with
concentrations near steady state after the first 10 years. For the scenario of decreased deposition, the
projected change in the median sulfate concentration was -40 /ieq L"1 and was essentially complete by
year 50. The mean projected change in the median sulfate concentration over the subsequent 50 years
was slightly less than -2 /ieq L*1.
The median pH values changed from 6.97 to 6.96 under current deposition and from 6.97 to 7.01
under decreased deposition, a change of less than 0.05 units (Table 10-14). The variance in pH
remained relatively constant through time (Figure 10-23).
Neither the change in ANC nor change in sulfate concentration was a function of the initial ELS-
I ANC, for either current deposition or for a 30 percent decrease in deposition (Table 10-15). Shifts in
the population distribution of median ANC and sulfate concentrations over the 40-year period from year
10 to year 50 indicate a relatively uniform change among ANC and sulfate intervals (Figures 10-24 and
10-88
-------
Table 10-15. Change in Median ANC and Sulfate Concentrations Over a 40-Year Period as
Function of the Initial ELS-Phase I or NSS Pilot Survey ANC Groups
ANC (uea L'1l
Sulfate fuea
NE
Priority Class AB
ETD, cons.
ETD, ramp
ILWAS, cons.
ILWAS, ramp
MAGIC, cons.
MAGIC, ramp
Priority Class A-E
ETD, cons.
ETD, ramp
MAGIC, cons.
MAGIC, ramp
Priority Class A-l
MAGIC, cons.
MAGIC, ramp
SBRP
Priority Class AB
ILWAS, cons.
ILWAS, ramp
MAGIC, cons.
MAGIC, ramp
Priority Class A-E
MAGIC, cons.
MAGIC, ramp
< 0
-0.41
7.17
-2.79
7.75
-0.58
5.34
-0.48
3.68
-1.97
4.99
-1.97
4.99
.
-
.
-
.
-
0 -25
4.84
12.37
-3.33
10.39
0.36
13.31
0.94
8.82
-2.98
6.21
-2.06
10.34
-
-
-
-
.
-
25 - 100
2.86
13.59
-5.65
5.37
-2.90
6.33
2.41
13.80
-0.52
6.60
-1.34
9.75
-15.26
-15.60
-13.96
-20.74
-13.96
-20.74
100 - 400 < 0
-
-
.
-
.
-
-3.93
10.61
0.50
7.20
-2.83
14.67
-5.79
-6.96
-14.46
-24.22
-24.04
-33.80
2.5
-43.65
4.9
-51.9
-5.29
-48.41
-4.72
-36.7
-0.5
-30.43
-0.5
-30.4
-
•
.
'
-
-
0-25 25 - 100 100 - 400
-1.5
-33.9
-21.49
-5.36
-36.72
-1.09
-31.42
-5.0
-20.57
-1.8
-33.4
.
-
-
-
-
-
-9.0
-24.02
-19.7
-5.53
-26.93
3.9
-24.6
-0.6
-26.91
-3.3
-29.9
36.53
52.88
26.58
43.95
26.58
43.95
-
-
-
.
-
-10.03
-28.22
-4.3
-20.47
-6.7
-34.2
24.92
33.46
30.93
47.4
30.93
47.4
10-89
-------
Northeast Lakes
Priority Class A -1
Model = Magic
Deposition = Constant
*r f
1
3
0>
.a
700-
600-
500-
400-
300-
200-
100-
-40-15 10 35 60 85 110135160185210235260285310335360385410
ANC&ieqL-1)
D MAGIC Year 10
B MAGIC Year 50
Northeast Lakes
Priority Class A • I
Model = Magic
Deposition = Ramped 30% Decrease
-------
10-25). The 40-year interval was selected as the period for comparison because the ramp change in
deposition did not occur until year 10; the first 10 years of all projections represent, therefore, current
deposition levels. The change in ANC was smaller for acidic lakes for decreased deposition but similar
among the other three ANC groups (Table 10-15).
10.11.1.2 Target Population Projections Using MAGIC and ETD
An estimated target population of 1920 lakes was simulated using both ETD and MAGIC. These
lakes represent Priority Classes A - E (Figure 10-1), which have ANC concentrations ranging from
-53 to 392 Meq L*1. These priority classes contains watersheds that have both positive and negative
sulfur retention but, based on chloride concentrations (Section 10.5.7), are relatively undisturbed.
10.11.1.2.1 Deposition scenarios -
ETD and MAGIC projected similar changes in ANC, sulfate, and pH over the 50-year period for both
current deposition and a 30 percent deposition decrease (Figures 10-26 through 10-28). Confidence
intervals for each of the projections are included in Appendix A.3. For current deposition levels, the
median ANC concentrations projected after 50 years using ETD and MAGIC were 74 and 110 /ieq L"1,
respectively (Table 10-16). Under a 30 percent deposition decrease, the median ANC concentrations
projected using ETD and MAGIC after 50 years were 85 and 119 Meq L"1, respectively. The differences
between the model projections result primarily from the initially calibrated ANC concentrations. The
median calibrated ANC concentration for MAGIC was 116 /ieq L'1, while the median (ELS-I) ANC
concentration assumed as the initial model condition for ETD was 77 Meq L'1. The difference between
the ETD initial and 50-year projected ANCs was 4 /ueq L*1 , similar to the 6 peg L"1 difference observed
for MAGIC (Table 10-16). Similar differences between the initial and 50-year projections occurred for
sulfate and pH. These relatively minor discrepancies reflect differences in the calibration procedures for
both models (See Section 10.9) and are within the uncertainty bounds on the projections.
10-91
-------
Northeast Lakes
Priority Class A -1
. Modei = Magic
Deposition = Constant
700
600
®
* 400
o
I 300
Z- 200
100
0
ESSS
1
In Pn n
30 40 50 60 70 80 90100110120130140150160170180190200210220230
[SOf](ueqL")
B MAGIC Year 50
Northeast Lakes
Priority Class A -1
Modei = Magic
Deposition = Ramped 30% Decrease
700
600
I500
3 400-
"o
£ 300-
200-
100-
-I
g
i
n
n .n
30 40 50 60 70 80 90100110120130140150160170180190200210220230
D MAGIC Year 10
B Year 50 Ramped
Figure 10-25. Comparison of population histograms for sulfate concentrations at current levels of
deposition and a 30 percent decrease for NE lakes, Priority Classes A -1, using MAGIC.
10-92
-------
to
I
£
>
o.e
0.6
0.4
NE Lakes
Priority Class - A • E
Deposition « Constant
Year - 0
-»0 0 WO 200 MO
ANC (|ieq Lt)
NE Lakes
Priority Class - A - E
Deposition - Ramp 30% Decrease
Year - 0
to
I"
I
2 0.6
O4
1X0
-100 0 100 200 400 400
ANC (jieq L-i)
to
OB
Deposition • Constant
Year - 20
0 100 200 300 400
ANC (jieq Li)
Deposition • Ramp 30% Decrease
Year - 20
to
OL«
^tf
(L2
OJ>
•100 - 0 100 200 SOO 400
ANC (fieq LI)
to
0 04
S
I 0.4
Deposition - Constant
Year - 50
•WO 0 100 200 300 4QO
ANC dieq Li)
Deposition - Ramp 30% Decrease
Year - 50
0 100 200 800
ANC (|ieq L-i)
Figure 10-26. Comparison of MAGIC and ETD projections of ANC for NE lakes, Priority Classes
A - E, under current and decreased deposition.
10-93
-------
to
O OJ
i
a.
S 0.6
a
5 0.4
0.2
0.0
NE Lakes
Priority Class - A - E
Deposition - Constant
Year - 0
100 200
ISO,*! (|ieq L-i)
aoo
NE Lakes
Priority Class - A - E
Deposition • Ramp 30% Decrease
Year - 0
tOr
O OS
i
!«*
I-
02-
a wo 200
[SO.*-} (jieq L-i)
300
to
o
Deposition • Constant
Year - 20
100 200
(jieq L-«)
too
Deposition • Ramp 30% Decrease
Year - 20
to
"
0.6
o
OJO
100
[S0t*-]
200
300
Deposition - Constant
Year - SO
Deposition - Ramp 30% Decrease
Year - SO
Wr
M
04
ISO4*1 (jieq
100
ISO,*]
200
SOD
Figure 10-27. Comparison of MAGIC and ETD projections of sulfate concentrations for NE lakes,
Priority Classes A - E, under current and decreased deposition.
10-94
-------
NE Lakes
Priority Class « A • E
Deposition - Constant
Year •> 0
NE Lakes
Priority Class • A • E
Deposition = Ramp 30% Decrease
Year - 0
to
(U
I
0.4
°fc
<5 SJ>
8.0
PH
7J> 7.5 SJJ
Deposition » Constant
Year - 20
4JB 45
Deposition •> Ramp 30% Decrease
Year - 20
W>
Deposition » Constant
Year - 50
S-0 SJ «J»
PH
7.0 7J
Deposition • Ramp 30% Decrease
Year • 50
to
O2
4.0
e.0
PH
Figure 10-28. Comparison of MAGIC and ETD projections of pH for NE lakes, Priority Classes A
E, under current and decreased deposition.
10-95
-------
Table 10-16. Descriptive Statistics of Projected ANC, Sulfate, and Percent Sulfur
Retention for NE Lakes in Priority Classes A - E Using MAGIC and ETD for Both
Current and Decreased Deposition
Model
Mean
Std.
Oev.
Min.
P 25 Median P 75
Max.
Current Deposition
MAGIC vs. ETD. ANC
Model Year 0
ETD 106.63 109.52 -53.00 19.50 76.90 190.90 391.60
MAGIC 134.47 115.64 -21.25 44.71 115.77 179.23 409.99
Model Year 20
ETD 106.19 108.50 -51.51 16.47 71.21 191.54 383.18
MAGIC 134.26 115.66 -21.10 44.08 113.69 178.56 408.27
Model Year 50
ETD 104.63 106.97 -51.60 15.03 73.82 177.25 383.63
MAGIC 132.67 115.92 -21.62 42.45 110.50 177.80 407.34
MAGIC vs. ETD. SO,'2
Model Year 0
ETD 105.52 34.23 33.80 79.40 104.40 124.70 199.00
MAGIC 106.59 43.00 59.66 67.33 100.13 125.33 245.60
Model Year 20
ETD 104.08 38.01 52.55 72.93 101.39 121.32 221.95
MAGIC 98.44 39.35 54.58 62.51 91.62 110.74 221.44
Model Year 50
ETD 102.85 40.28 54.25 67.77 100.47 118.20 215.67
MAGIC 96.50 38.49 52.76 60.99 91.27 109.48 214.91
MAGIC vs. ETD. oH
Model Year 0
ETD 5.61 0.81 4.27 6.24 6.82 7.22 7.53
MAGIC 5.78 0.72 4.47 6.53 6.94 7.12 7.48
Model Year 20
ETD 5.64 0.80 4.29 6.17 6.79 7.22 7.52
MAGIC 5.80 0.71 4.49 6.53 6.93 7.12 7.48
Model Year 50
ETD 5.62 0.79 4.29 6.13 6.80 7.18 7.52
MAGIC 5.77 0.73 4.48 6.51 6.92 7.12 7.48
MAGIC vs. ETD. % S Retention
Model Year 0
ETD -4.65 32.15 -97.21 -18.78 -7.07 17.97 68.55
MAGIC -1.42 8.03 -15.54 -7.99 -1.28 4.27 19.07
Model Year 20
ETD 2.37 8.96 -23.68 -4.25 2.19 6.79 30.42
MAGIC 6.25 6.56 -4.93 1.64 4.73 10.82 23.73
Model Year 50
ETD 4.36 7.83 -17.20 -0.85 2.99 8.99 27.32
6.06
-1.16
3.52
6.37
12.59
23.74
continued
10-96
-------
Table 10-16. (Continued)
Model Mean
Std.
Dev. Min.
P_25 Median
P_75
Max.
30% Decrease in Deposition
MAGIC vs. FTP. ANC
Model Year 0
ETD 106.63
MAGIC 134.47
Model Year 20
ETD 109.29
MAGIC 138.92
Model Year 50
ETD 112.26
MAGIC 142.77
MAGIC vs. ETD. SO*
Model Year 0
ETD 105.52
MAGIC 106.59
Model Year 20
ETD 93.64
MAGIC 89.42
Model Year 50
ETD 74.15
MAGIC 70.55
MAGIC vs. ETD. oH
Model Year 0
ETD 5.61
MAGIC 5.78
Model Year 20
ETD 5.72
MAGIC 5.91
Model Year 50
ETD 5.82
MAGIC 6.08
109.52
115.64
108.45
116.20
107.29
116.35
34.23
43.00
33.61
36.13
29.12
28.55
0.81
0.72
0.76
0.67
0.71
0.61
-53.00
-21.25
-46.81
-20.34
-39.64
-18.20
33.80
59.66
44.10
49.43
38.82
38.27
4.27
4.47
4.33
4.52
4.40
4.58
19.50
44.71
17.95
48.62
22.48
51.18
79.40
67.33
62.65
55.37
50.95
44.20
6.24
6.53
6.20
6.57
6.30
6.59
76.90
115.77
76.92
118.24
84.79
119.41
104.40
100.13
89.76
82.98
70.44
64.16
6.82
6.94
6.82
6.95
6.86
6.95
190.90
179.23
203.85
192.93
197.91
204.04
124.70
125.33
107.91
102.65
86.13
81.24
7.22
7.12
7.24
7.16
7.23
7.18
391.60
409.99
389.03
414.52
399.03
417.16
199.00
245.60
185.65
202.13
161.51
157.31
7.53
7.48
7.52
7.48
7.53
7.49
MAGIC vs ETD. % S Retention
Model Year 0
ETD -4.65
MAGIC -1.42
Model Year 20
ETD -10.09
MAGIC -6.32
Model Year 50
ETD 1.31
MAGIC 4.04
32.15
8.03
11.50
8.93
11.17
8.71
-97.21
-15.54
-49.91
-21.87
^9.87
-18.42
-18.78
-7.99
-16.65
-13.29
-4.29
-1.22
-7.07
-1.28
-9.45
-6.13
2.05
4.08
17.97
4.27
-3.09
-0.08
7.16
10.31
68.55
19.07
27.01
16.25
24.01
23.69
10-97
-------
Results of within-model comparisons of the effects of alternative deposition scenarios on surface
water chemistry are shown in Figures 10-29 through 10-31. Changes in median ANC projected for both
deposition scenarios after 50 years using either model were small and were consistent with the MAGIC
projections discussed in the previous section. For example, differences in the median ANC projected
after 50 years using ETD versus MAGIC, between current deposition and a 30 percent deposition
decrease were 74 /neq L"1 versus 85 /ueq L"1 (+11 peq L"1 ) and 110 /ieq L"1 versus 119 /*eq L*1 (+9
/ieq L*1 ), respectively. The differences in the median sulfate concentrations projected after 50 years
using ETD and MAGIC between current deposition and a 30 percent deposition decrease were 100 versus
70 jieq L"1 (-30 jxeq L."1) and 91 versus 64 /ieq L"1 (-27 neq L"1 ), respectively.
The differences in the median pH at the end of 50 years under current and decreased deposition
for MAGIC were 6.92 versus 6.95 (+0.30) and for ETD were 6.80 versus 6.86 (+0.06), respectively.
The sulfate concentrations projected using both models indicated the watersheds were near sulfate
steady state after 50 years. The median sulfur retention for the watersheds projected using both models
ranged from 3 to 6 percent for current deposition and from 2 to 4 percent for decreased deposition
(Table 10-16). Although there was a range in this distribution of sulfur retention, the upper and lower
quartile values for current deposition ranged from -3 to 9 percent for ETD and 4 to 13 percent for MAGIC;
under decreased deposition, percent sulfur retention ranged from -4 to 7 for ETD and -1 to 10 for MAGIC
after 50 years, indicating most of the watersheds were near sulfur steady state (Table 10-16).
Projection of the number of lakes not currently acidic that might become acidic in the next 50 years
for current deposition using ETD and MAGIC were 49 (3 percent) and 87 (5 percent), respectively. The
number of lakes currently not acidic that might become acidic for a 30 percent deposition decrease using
ETD and MAGIC were 37 (2 percent) and 50 (3 percent), respectively. The number of currently acidic
lakes that might chemically improve under current deposition after 50 years was projected by ETD and
MAGIC to be 52 (23 percent) and 64 (39 percent), respectively. Under a 30 percent deposition reduction,
10-98
-------
NE Lakes
Model = MAGIC
Priority Class = A - E
Year - 20
o
•&
o
Q.
O
1
O
1-Or
0.8
o.e
0.0'—
-100
Simulation Year p
---- Constant Deposition
Ramp Deposition
1.0r
O 0.8
5
a.
e 0.6
o
£ 0.4
3
O °-2
NE Lakes
Model = MAGIC
Priority Class = A - E
Year = 50
0 100 200
ANC
~
CO
3
0.4
-100
0 100
ANC
200 300
L'i)
400
0.0"—
•100
Simulation Year 0
Constant Deposition
Ramp Deposition
o too 200
ANC (|ieq L
soo
400
Figure 10-29. Comparisons of projected change in ANC under current and decreased deposition
for NE Priority Classes A - E, using ETD and MAGIC.
10-99
-------
1.0
0 0.8
0.8
a
*3 0.4
s
3
O °-2
0.0
NE Lakes
Model = MAGIC
Priority Class « A - E
Year * 20
Simulation Year 0
Constant Deposition
Ramp Deposition
100
200
300
to
o 0.6
o
Q.
0.6
0
*! 0.4
JS
o o-2
0.0
NE Lakes
Model * MAGIC
Priority Class = A - E
Year = 50
Simulation Year 0
Constant Deposition
Ramp Deposition
100 200
ISO,*!
-------
1.0r
<5 0.8
O
0'6
»= 0.4
3
O <
0.0
NE Lakes
Model = MAGIC
Priority Class - A - E
Year = 20
Year 0
---- Constant
Ramp
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
PH
1.0
O 0.6
O
Q.
£ 0.6
I"
3
O 0-2
0.0
NE Lakes
Model = MAGIC
Priority Class = A - E
Year = 50
Year 0
---- Constant
Ramp
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
PH
1.0r
O
Q.
S 0.6
*= 0.4
o
0.0
Model = ETO
Year = 20
Year 0
Constant
Ramp
4.0 4.5 5.0 S.5 6.0 6.5 7.0 7.5 8.0
pH
to
0.8
0.
I-
0
0-4
0.0
Model = ETD
Year - 50
Year 0
Constant
Ramp
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
pH
Figure 10-31. Comparisons of projected change In pH under current and decreased deposition for
NE Priority Classes A - E, using ETD and MAGIC.
10-101
-------
the projections of chemical improvement were estimated to be 103 lakes (46 percent) for ETD and 125
lakes (77 percent) for MAGIC.
10.11.1.2.2. Rate of change of ANC, suifate, and pH over 50 years -
The changes in ANC and suifate concentrations and in pH projected over the next 50 years using
both the ETD and MAGIC models are shown in Figures 10-32 through 10-37. The change In median ANC
projected using ETD and MAGIC under current deposition over the next 50 years was a total decrease
of -3.1 and -5.3 jueq L"1, respectively (Table 10-16). The change in median suifate concentrations using
ETD and MAGIC was -3.9 and -8.9 /ieq L*1, respectively, for current deposition. The change in median
pH using ETD and MAGIC under current deposition was -0.02 units for each model. With a 30 percent
deposition reduction, the ANC increase projected using ETD and MAGIC was from 77 to 85 (+8) and 116
to 119 (+4) /*eq L"1, respectively, similar to the change projected for the larger target population in the
previous section. The decrease projected in suifate concentrations with a 30 percent deposition
decrease using ETD and MAGIC was from 104 to 70 (-34) and 164 (-36) /ieq L*1, respectively, over 50
years. These values are roughly equivalent to the measurement or projection error determined for suifate.
The pH increase projected under decreased deposition using either model was less than +0.05 units
over 50 years. The variance in ETD projections, although larger than MAGIC projections, also remained
relatively constant through time (Figures 10-34 and Figure 10-37).
The changes in ANC or suifate concentration were not functions of the initial ELS-i concentrations
using either MAGIC or ETD for either deposition scenario (Table 10-15). Histograms of projected change
in the population distribution of median ANC and suifate from year 10 to year 50 indicate a relatively
uniform change among ANC and suifate intervals using both ETD and MAGIC (Figures 10-38 through 10-
41). A slightly greater change In ANC was projected for non-acidic lakes with decreased deposition.
10-102
-------
3nIQuwlte+
(1.5 x kiterqiMili* Ftange)~
3rd Quart*
KtQuvta*
IKQiMftto
(1.S « htmniwH* Hong*)-
Mwttn
Constant
a a
400-
350-
300-
250-
•T" 200-
flSO-
} 100-
50-
0-
-50-
3
•*m
•
M
M y
^
H
K
^ __
- ^
^^ >^
_ —
** J
^H ^^
^
^
••
*~ — .
^ ^
^V ^^
™
H
^
dM
- *
^w ^HB
^
•
— . —
PV
Ramped
400-
350-
300-
250—
150-
1OO—
50-
0-
-5O-
2 8 $
$ §E SE j
^
^
•
••
!••
*"* -~
•
0+
••M
••
•*•*•
•HIM-
^
•HI
•
•••h ^^
$ s
E §
^B ^^
•
^^ ^^
? s
L J
•H
*••
__ —
5
e
Figure 10-32. Box and whisker plots of ANC distributions projected using ETD in 10-year intervals
for NE lakes, Priority Classes A - E.
10-103
-------
2SO-,
20
-------
3rd Ouartlb +
(1.5 x Irrtorquartta Ring*)"
3rd Quart to
IstQiurtl*
(1.5 * JntMquanto Ringt)~
"NtttocxoMdwtnnravafeM
Constant
288
Ramped
o
SE
oc
tr
ir
S
8-
7-
B-
5-
A—
>
^
iv
M
IB
^
^
M
>
^
1
»
«
>
1^
V*^ ^^
V
N^
>
M
^
^^ ^^
^
^^
^
5
^
— __
)•
^
i>
^
p— __
•
^H
QC
s
S5
Figure 10-34. Box and whisker plots of pH projected using ETD in 10-year intervals for NE lakes,
Priority Classes A - E.
10-105
-------
(15 it Morqwtife flmga)~
ardduartto
1«t Quart*
(1.51 kiMiqutftil* Range)"
"Mat ID ciOMd «xtrem» value
Constant
500-
400-
300-
lj
JL200-
100-
0—
o S
SE £
WM
^V ^^
•V B^H
-J LH
g
>-
•bri
^
0
§E
h^
•
0
§E
VH
^
iwtf
^
^«
s
§£
••
^
^
^
HIVM
Ramped
SOO-i o S
400-
_* 300~
"Li
r
A 200-
1 '
100-
o-
£ £
^— «B
— _
••J !•—
•••* v*V
4*k
•HUH-
8
§E
^
8
5
«w
^
,
M
'
S
?
^
^
8
5
^^
^
Figure 10-35. Box and whisker plots of ANC distributions in 10-year intervals using MAGIC for NE
lakes, Priority Classes A - E.
10-106
-------
250-1
200-
I
A
160—
so-
SriOuartfle*
(1i x (nMnyartB* Rang*
*dQu«rfte
IBOuartfl.
Constant
2 a § s s
> > £ £ >-
Ramped
250-
200-
^1150-
SO-
85
Figure 10-36. Box and whisker plots of sullate distributions in 10-year intervals using MAGIC for
NE lakes, Priority Classes A - E.
10-107
-------
3rddutrtl« +
(1.5 x Inttiquwtl* R«ng»)
Mwtan
IMOuutl*
IttQuaftl*
(1.5 x Iffluquwtt* Rang*)"
'"I |
7-
6-
5-
o
>
Constant
88?
DC DC
> >•
Ramped
8-1
7-
5-
$ $
S 3
Figure 10-37. Box ami whisker plots of pH in 10-year Intervals using MAGIC for NE lakes, Priority
Classes A - E.
10-108
-------
Northeast Lakes
Priority Class A - E
Model = ETD
Deposition = Constant
500-
400
8
« 300
200
100
r
nl
-40-15 10 35 60 85110135160185210235260285310335360385410
Northeast Lakes
Priority Class A - E
Model a ETD
Deposition = Ramped 30% Decrease
500
400-
w
*
to 3001
$
E 200-
100-
-I
npri n Fl
Li II inl i il
— i I - - ]- --- - ( ------- i T- ---- r T — i t i — i
-40-15 10 35 60 85 110135160185210235260285310335360385410
D ETD Year 10
B Year 50 Ramped
Figure 10-38. ETD ANC distributions at year 10 and year 50 for NE lakes, Priority Classes A - E,
under current and decreased deposition.
10-109
-------
Northeast Lakes
Priority Class A - E
Model = Magic
Deposition = Constant
I
500-
400
300
o
fe
E 200
100
ri
I
-40 -15 10 35 60 85 110 135 160 185 210 235 260 285 310 335 360 385 410
O MAGIC Year 10
H MAGIC Year 50
Northeast Lakes
Priority Class A - E
Model = Magic
Deposition = Ramped 30% Decrease
5001
400
«
^ 300
o
E
E
200-
100-
n
-40-15 10 35 60 85110135160185210235260285310335360385410
ANCUieqL-1)
O MAGIC Year 10
B Year 50 Ramped
Figure 10-39. MAGIC ANC distribution at year 10 and year 50 tor NE lakes, Priority Classes
A - E, under current and decreased deposition.
10-110
-------
Northeast Lakes
Priority Class A - E
Model s ETD
Deposition = Constant
500'
400'
S
« 300-
CD
e 200-
3
100-
•
*•
^
1 1 1 1
G
A
\
i
\
fy
I
%
r
^
^
%
|
|
|
tf
I^H
|
^
-.
«L
$
I
\
fy
1
n
^
I
|
\
y
\
k
\
f
1 n (~S
aj-| |
1 ^ ^
y 1 ^rHZ3 ? 1 ^ ^ ^
^ 1_ H. 'Sl-r*' f? (~~1 1 « f** P* f"~P91
30 40 50 60 70 80 90100110120130140150160170180190200210220230
rc/^2-1/ i n O ETD Year 10
[SCMOieqL") B ETDY^SO
v>
500
400
300
E 200
3
100-
Northeast Lakes
Priority Class A - E
Model = ETD
Deposition = Ramped 30% Decrease
^
i
30 40 50 60 70 80 90100110120130140150160170180190200210220230
D ETOYaarlO
g YearSORamped
Figure 10-40. ETD suffete distributions at year 10 and year SO for NE lakes, Priority Classes
A - E, under current and decreased deposition.
10-111
-------
Northeast Lakes
Priority Class A-E
Model = Magic
Deposition = Constant
I
500
400-
300-
200-
100
|
I
I
%
J
%
-rqa.
!0l
-pa.
30 40 50 60 70 80 90100110120130140150160170180190200210220230
O MAGIC Year 10
B MAGIC Year 50
Northeast Lakes
Priority Class A-E
Model = Magic
Deposition = Ramped 30% Decrease
ouu
400
M
l^
2 300
"o
£
1 200
3
z
100-
I
a
y
n
^
i
^
.
i
I
\
^
g
|
i
1
|
A
\
v^
•
p
8
^
O
j
5la
r
i
I
g
i
—
i
-
?3
i
111 j ji [LI n FL
30 40 50 60 70 80 90 100110120130140150160170180190200210220230
ISO 2~\ (uea L 'M ^ MAGIC Year 10
J W-BM g Year 50 Ramped
Figure 10-41. MAGIC sulfate distributions at year 10 and year 50 for NE lakes, Priority Classes
A-E, under current and decreased deposition.
10-112
-------
10.11.1.3 Restricted Target Population Projections Using All Three Models
There were an estimated 495 lakes in the target population simulated using all three Level III
models. This target population represents Priority Classes A and B (Figure 10-1). Lakes in this target
population had initial ELS-I ANC < 100 /ieq L'1, ranging from -43 to 86 (ieq L'\ The watersheds were
undisturbed, based on chloride concentrations (See Section 10.5.7), and in general had positive sulfur
retention.
10.11.1,3.1 Deposition scenarios -
All three models simulated comparable changes in ANC, sulfate, and pH over the 50-year period
assuming current deposition or a 30 percent deposition decrease (Figures 10-42 through 10-44).
Confidence intervals computed for each of the projections are included in Appendix A.3. Projections for
all three models were comparable at the lower ANC concentrations (i.e., ANC < 25 /ieq L'1) but deviated
at higher ANC concentrations. Projections from MAGIC deviated the most at the higher ANC
concentrations but were still within the uncertainty bounds about the projections (Appendix A.3).
Projected ANC values were similar, however, among all three models for lakes with ANC in the lower
quartile of the population (Table 10-17). Lower quartile values of ANC projected using ETD, ILWAS, and
MAGIC with current deposition after 50 years were 2.6, 5.9, and 11.9 /ieq L*1, respectively. Lower quartile
values for sulfate projected using the ETD, ILWAS, and MAGIC models after 50 years of current deposition
were 70.4, 80.6, and 69.4 /xeq L"1, respectively. Assuming a 30 percent deposition decrease, lower
quartile values of ANC projected using ETD, ILWAS, and MAGIC after 50 years were 10.5, 19.6, and 25.1
/teq L*1, respectively: Sulfate concentrations projected under similar conditions using ETD, ILWAS, and
MAGIC were 51.7, 66.5, and 53.2 /ieq L*1, respectively. These values were all within the uncertainty
bounds for the projections (Appendix A.3).
The projected pH values were similar at the higher pHs but deviated at low pH with the greatest
difference between ILWAS and the other two models. The projected median pH after 50 years with ETD,
10-113
-------
10
0.6
0.4
-MO
NE Lakes
Priority Class - A & B
Deposition * Constant
Year - 0
— Ptmtt 1
— — MAGIC
--- ETO
NE Lakes
Priority Class - A & B
Deposition = Ramp 30% Decrease
Year - 0
VOr
0.6
o no wo
ANC <|»q l/
300 400
(US
-MO
RlM* 1
MAGIC
--- ETO
0 WO MO 100
ANC (|ieq LI)
400
to
oe
0.4
Deposition * Constant
Year - 20
•wo
Deposition - Ramp 30% Decrease
Year - 20
tOr
— — MAGIC
ETD
(.WAS
o
0.2
0 MO 200 300
ANC (jieq L-i)
400
•5T
— — MAQIC
ETO
0 MO 300 MO
ANC (Jieq Li)
to
§
O4
•wo
Deposition * Constant
Year - 50
Deposition - Ramp 30% Decrease
Year - 50
— PtWM 1
MAQIC
--- ETC
R.WAS
0 MO
ANC
200
MO 400
PIMM 1
MAQIC
ETD
LWA3
-MO 0 MO 200 300
ANC (neq L-I)
400
Rgure 10-42. Comparison of ANC projections using ETD, ILWAS, and MAGIC for NE lakes, Priority
Classes A and B, under current and decreased deposition.
10-114
-------
ME Lakes
Priority Class - A & B
Deposition » Constant
Year - 0
—— PIMM 1
MAGIC
- - - E7O
NE Lakes
Priority Class - A A B
Deposition « Ramp 30% Decrease
Year - 0
I0r
04
to
iu
M
0.4
OJ»
Deposition • Constant
Year - 20
^— Phu* i.
— — MAGIC
ETD
LWA8
tSO.*l
wo
Deposition - Ramp 30% Decrease
Year - 20
Ur
0.6
PMW 1,
MAGIC
610
"*"**"*"• W.WAS
g MO 200
ISO.*] <|ieq L-')
soo
Deposition • Constant
Year - 50
* ...-•"
»*•••
Deposition • Ramp 30% Decrease
Year - 50
tOr
900
—• — MAGIC
6TD
E.WA3
o c5>mP5ri80n of 8ulfa*e projections using ETD, JLWAS, and MAGIC for NE lakes,
Priority Classes A and B, under current and decreased deposition.
10-115
-------
NE Lakes
Priority Class - A 4 B
Deposition • Constant
Year - 0
tOr
4X U U «J> U 7.0 74
NE Lakes
Priority Class » A & B
Deposition - Ramp 30% Decrease
Year - 0
Deposition * Constant
Year - 20
Deposition - Ramp 30% Decrease
Year - 20
Deposition - Constant
Year • SO
tOr
Deposition • Ramp 30% Decrease
Year •> 50
Ur
Rgure 1«M4. Comparison of pH projections using ETD, ILWAS, and MAGIC for NE lakes, Priority
Classes A and B, under current and decreased deposition.
10-116
-------
Table 10-17. Descriptive Statistics for Projected ANC, Sulfate, Percent
Sulfur Retention, and Calcium Plus Magnesium for NE Lakes in Priority Classes
A and B Using ETD, ILWAS, and MAGIC for Both Current and Decreased Deposition
Model Mean
All Models. ANC
Model Year 0
ETD 31.91
ILWAS 44.15
MAGIC 56.99
Model Year 20
ETD 29.91
ILWAS 43.01
MAGIC 56.86
Model Year 50
ETD 29,94
ILWAS 39.46
MAGIC 55.58
All Models. SO,2"
Model Year 0 '
ETD 90.11
ILWAS 118.05
MAGIC 113.96
Model Year 20
ETD 110.27
ILWAS 118.44
MAGIC 105.01
Model Year 50
ETD 110.43
ILWAS 118.49
MAGIC 102.65
All Models. nH
Mode) Year 0
ETD 5.55
ILWAS 5.07
MAGIC 5.39
Mode! Year 20
ETD 5.50
ILWAS 5.04
MAGIC 5.41
Model Year 50
ETD 5.48
ILWAS 5.01
MAGIC 5.40
Std.
Dev.
32.60
52.09
51.58
33.02
53.06
51.11
33.60
53.21
50.91
32.20
52.36
45.53
43.66
52.91
41.63
45.43
52.70
40.76
0.64
0.95
0.79
0.66
0.98
0.78
0.66
0.99
0.79
Min.
Current
-43.10
-66.86
-21.25
-48.68
-70.47
-21.10
-51.60
-73.69
-21.62
33.80
42.19
50.09
52.55
43.84
47.47
54.25
43.61
46.24
4.36
4.15
4.47
4.31
4.13
4.49
4.29
4.11
4.48
P_25
Deposition
6.30
7.77
14.76
1.32
7.94
13.72
2.61
5.90
11.91
67.20
82.60
77.63
70.24
80.75
71.08
70.39
80.59
69.36
5.83
4.93
5.83
5.55
4.86
5.95
5.63
4.91
5.94
Median
25.60
35.98
66.31
30.40
33.19
66.64
30.86
30.29
64.92
81.10
101.90
106.45
103.24
102.50
96.47
106.50
103.60
95.16
6.36
6.02
6.40
6.42
6.24
6.40
6.43
6.09
6.40
P_75
58.80
86.48
84.09
60.63
86.48
83.53
62.56
82.00
81.70
112.90
135.90
126.20
146.63
137.20
126.36
153.81
137.80
127.03
6.71
6.75
6.79
6.72
6.78
6.79
6.73
6.74
6.79
Max.
89.90
158.60
174.54
105.35
160.50
172.99
106.01
161.30
170.86
185.30
266.30
245.60
221.95
267.10
221.44
215.67
264.30
214.91
6.89
7.26
6.97
6.96
7.27
6.97
6.96
7.27
6.97
continued
10-117
-------
Table 10-17. (Continued)
Model Mean
Std.
Dev.
Min.
P_25
Median
P_75
Max.
All Models. % S Retention
Model Year 0
ETD 20.15
ILWAS 1.39
MAGIC 1.13
Model Year 20
ETD 6.42
ILWAS 1.19
MAGIC 8.95
Model Year 50
ETD 7.09
ILWAS 1.16
MAGIC 11.09
ILWAS vs. MAGIC. Ca
Model Year 0
ILWAS 122.07
MAGIC 131.47
Model Year 20
ILWAS 121.77
MAGIC 124.33
Model Year 50
ILWAS 119.42
MAGIC 120.93
Delta Ca+Mg
ILWAS -3.17
MAGIC -5.84
21.15
19.51
9.14
10.84
19.31
7.01
8.69
19.25
6.30
+ Mq
40.03
51.49
42.15
51.97
41.93
52.58
4.79
2.74
-17.04
-32.60
-13.89
-15.87
-30.18
-4.93
-7.44
-29.92
-1.16
45.24
41.09
43.03
39.78
41.42
38.39
-10.80
-12.43
6.07
-15.37
-9.09
1.82
-14.06
3.37
-0.44
-14.21
6.37
102.04
98.75
102.82
90.32
102.18
85.75
-6.60
-8.60
26.82
-0.02
2.67
4.21
2.78
10.40
6.98
2.95
10.85
118.55
121.60
115.58
114.27
110.69
110.89
-3.25
-5.44
32.73
12.34
7.09
12.20
12.34
14.21
11.51
11.34
16.99
141.38
144.68
143.07
138.92
142.69
135.37
0.30
-3.88
68.55
63.24
19.07
30.42
63.53
20.51
27.32
62.91
21.29
203.69
281.03
219.50
280.01
210.50
279.54
8.40
-1.74
30% Decrease in Deposition
All Models. ANC
Model Year 0
ETD 31.91
ILWAS 44.15
MAGIC 56.99
Model Year 20
ETD 33.48
ILWAS 50.25
MAGIC 60.73
Model Year 50
ETD 37.72
ILWAS 55.86
MAGIC 64.08
32.60
52.09
51.58
33.43
51.62
51.20
33.85
50.47
50.37
-43.10
-66.86
-21.25
-44.27
-51.07
-20.34
-39.64
-39.65
-18.20
6.30
7.77
14.76
3.50
12.11
16.56
10.54
19.62
25.11
25.60
35.98
66.31
35.81
36.40
68.71
31.54
48.86
70.62
58.80
86.48
84.09
62.69
91.43
86.67
67.56
93.02
88.05
89.90
158.60
174.54
108.21
161.10
177.51
111.24
162.50
178.67
continued
10-118
-------
Table 10-17. (Continued)
Model Mean
All Models. SO"
Model Year 0 '
ETD 90.11
ILWAS 118.05
MAGIC 113.96
Model Year 20
ETD 98.70
ILWAS 107.32
MAGIC 96.70
Model Year 50
ETD 80.11
ILWAS 90.68.
MAGIC 77.01
All models. pH
Model Year 0
ETD 5.55
ILWAS 5.07
MAGIC 5.39
Model Year 20
ETD 5.59
ILWAS 5.18
MAGIC 5.52
Model Year 50
ETD 5.67
ILWAS 5.37
MAGIC 5.67
Std.
Dev.
32.20
52.36
45.53
38.83
47.24
38.33
32.05
37.91
31.15
0.64
0.95
0.79
0.63
0.93
0.73
0.58
0.86
0.66
Min.
33.80
42.19
50.09
44.10
37.11
44.78
38.82
31.30
38.25
'
4.36
4.15
4.47
4.35
4.25
4.52
4.40
4.36
4.58
P_25
67.20
82.60
77.63
61.87
75.95
66.06
51.68
66.53
53.20
5.83
4.93
5.83
5.68
5.11
6.11
6.00
5.57
6.17
Median
81.10
101.90
106.45
95.01
95.60
86.88
76.17
82.51
67.55
6.36
6.02
6.40
6.49
6.41
6.48
6.44
6.53
6.54
P_75
112.X
135.90
126.20
135.24
123.70
120.54
109.91
113.40
104.10
6.71
6.75
6.79
6.73
6.84
6.81
6.76
6.91
6.82
Max.
185.30
266.30
245.60
185.65
237.70
202.13
161.51
189.90
157.31
6.89
7.26
6.97
6.97
7.30
6.98
6.98
7.32
6.99
All Models. % S Retention
Model Year 0
ETD 20.15
ILWAS 1.39
MAGIC 1.13
Model Year 20
ETD -4.75
ILWAS -12.15
MAGIC -4.74
Model Year 50
ETD 3.06
ILWAS -8.71
MAGIC 4.90
21.15
19.51
9.14
14.01
22.43
8.89
12.55
21.32
8.95
-17.04
-32.60
-13.89
-36.70
-51.38
-24.17
-38.32
-40.16
-11.47
6.07
-15.37
-9.09
-14.32
-26.47
-11.69
-2.83
-25.14
-1.09
26.82
-0.02
2.67
-6.18
-9.58
-3.94
3.89
-13.60
6.18
32.73
12.34
7.09
4.69
0.71
1.48
8.97
2.18
11.02
68.55
63.24
19.07
27.01
59.58
13.73
24.01
59.57
20.37
continued
10-119
-------
Table 10-17. (Continued)
Model
Mean
ILWAS vs. MAGIC. Ca
Model Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
0
122.07
131.47
20
119.56
121.41
50
111.94
108.38
Std.
Dev.
+ Mq
40.03
51.49
41.74
51.84
41.17
52.80
Min.
45.24
41.09
40.91
39.42
35.21
35.37
P_25
102.04
98.75
100.12
84.83
89.13
74.12
Median
118.55
121.60
114.91
112.30
105.97
97.78
P_75
141.38
144.68
139.80
137.59
126.65
125.27
Max.
203.69
281.03
212.80
278.55
200.50
274.23
Delta Ca+Mg
ILWAS
MAGIC
-10:64
-18.38
7.41
9.37
-32.02
-53.70
-13.48
-23.42
-11.01
-15.44
-3.74
-12.88
-0.09
-5.09
10-120
-------
ILWAS, and MAGIC under current and decreased deposition was 6.4, 6.1, 6.4 and 6.4, 6.5, 6.5,
respectively. The projected lower quartile pH values after 50 years with ETD, ILWAS, and MAGIC under
current and decreased deposition were 5.6, 4.9, 5.9 and 6.0, 5.6, 6.2, respectively. At the lower pHs, the
projected ILWAS pH was from 0.5 to 1.0 pH unit less than projected by the other two models.
Changes in surface water chemistry under different deposition scenarios were compared within
models (Figures 10-45 through 10-47). Differences in median ANC concentrations between current
deposition and a 30 percent deposition decrease projected using ETD, ILWAS, and MAGIC after 50 years
were 30.9 versus 31.5 (+0.6), 30.3 versus 48.9 (+18.6), and 64.9 versus 70.6 (+5.7) jieq L'1, respectively.
Differences in median sulfate concentrations between current deposition and a 30 percent deposition
decrease projected using ETD, ILWAS, and MAGIC after 50 years were 106.5 versus 76.2 (-30.3), 103.6
versus 82.5 (-21.1), and 95.2 versus 67.6 (-27.6) Meq L"1, respectively. Differences in median pH between
current and decreased deposition projected using ETD, ILWAS, and MAGIC after 50 years were 6.43
versus 6.44 (+0.01), 6.09 versus 6.53 (+0.04), and 6.40 versus 6.54 (+0.14), respectively.
All three models indicated northeastern watersheds were near sulfate steady state or near zero
percent net sulfur retention after 50 years for scenarios of either current or decreased deposition (Table
10-17).
Projections of the number of lakes currently not acidic that might become acidic in the next 50
years under current deposition and a 30 percent decrease in deposition using ETD, ILWAS, and MAGIC
were 25 (5 percent), 74 (17 percent), 75 (17 percent) and 25 (5 percent), 25 (5 percent), and 50 (11
percent), respectively. Projections of the number of currently acidic lakes that might chemically improve
under current deposition and a 30 percent deposition decrease using ETD, ILWAS, and MAGIC were 27
(36 percent), 25 (32 percent), 0 (0 percent), and 52 (68 percent), 25 (32 percent), 13 (16 percent),
respectively.
10-121
-------
NE Lakes
Model - MAGIC
Priority Class - A & B
Y«ar - 20
to
OB
s
i"
i
04)
•100
—— Staiutatton Ywr 0
--— Constant Opposition
Rama 0«po«man
0 WO 200 300
ANC )
Model * ILWAS
Year - 20
— Simulation YMT 0
---- Constant Deposition
RMip MpetWan
0 100 200 300
ANC (|ieq LI)
Model - ILWAS
Year - SO
iMOon YMT 0
Ramp Opposition
0 100 200 WO 400
ANC
-------
to
S 0.6
0.
« 0.4
NE Lakes
Model » MAGIC
Priority Class - A & B
Year - 20
Simulation Year 0
ConsUnt Deposition
.Ramp Deposition
100 300
ISO,*] (peq L-i)
300
to
1
ae
i
NE Lakes
Model - MAGIC
Priority Class - A & B
Year - 50
,-~...._
Constant Deposition
Ramp Deposition
100 200
ISO,*"] (|ieq L-i)
300
to
OLB
QjO
Model - ETD
Year - 20
SknUaUan Y«w 0
— -- Constant D*po*ltion
—— Ramp Opposition
100
IS04*I
200
Li)
too
10r
0.6
SimuMOon YMT 0
—— Constant Deposition
— Ramp Deposition
ISO««-1
200
L-i)
300
to
OJ6
Model - ILWAS
Year - 20
100
ISO,*]
200
300
to
02-
OJD
Model - ILWAS
Year « SO
". — 1
Constant Deposition
Ramp Deposition
MO 200
ISO,*! (|ieq L-i)
300
Figure 10-46. Comparison of sulfate projections under current and decreased deposition for NE
lakes, Priority Classes A and B, at year 20 and year 50 using ETD, ILWAS, and MAGIC.
10-123
-------
NE Lakes
Model . MAGIC
Priority Class - A & B
Year • 20
to
O 0.8
a.
a
£
0.4
Simulation Y«w 0
Content Deposition
•"•«--• Ramp Dvpotttton
44) *S 5.0 U «J» U 7.0 74 8.0
PH
Q.
0.6
0.4
0.2
1X0
NE Lakes
Model » MAGIC
Priority Class - A 4 8
Year - 20
—— Stmutatlon Y«at 0
Constant Dapotltton
Ramp (MpoiMon
10
4.0
7JO
tOr
«
I"
£0,4
Model « ETD
Year - 20
Shnutatton Y««r 0
"-- ConsUnt D«oatttfen
— R«m» Mpotltten
5.0
tOr
Model = ETD
Year - 50
pH
Simulation Year 0
*"• Constant Dapovldon
Ramp
pH •
to
I
0.0
Model - ILWAS
Year - 20
4J> 4A S.O
«J) «.S 7jO
PH
to
I"
S 0.8
0.4
£
o «
OjO
Model . ILWAS
Year - 50
4JS 5J» « »,0 «.S 7.0
8.0
Figure 10-47. Comparison of pH projections under current and decreased deposition for NE lakes,
Priority Classes A and B, at year 20 and year SO using ETD, ILWAS, and MAGIC.
10-124
-------
10.11.1.3.2 Rate of change of ANC, sulfate, and pH over 50 years -
The changes in ANC, sulfate, and pH projected over the next 50 years using the three models are
shown in box and whisker plots (Figures 10-48 through 10-56). The relative change in the median ANC
projected using all three models and assuming current deposition levels was less than 0.1 jueq L* yr*
while the rate of change of median ANC for a 30 percent deposition decrease was about 0.3 /ieq L*1 yr*
1 for ILWAS and MAGIC and remained less than 0.1 peg L*1 yr*1 for ETD. These rates, while three
times greater for ILWAS and MAGIC, are still small and indicate little change in ANC over the 50-year
period under either deposition scenario.
The rates of change projected for median sulfate concentrations under current deposition ranged
from -0.2 /ieq L*1 yr"1 for MAGIC to less than 0.1 Aieq L*1 yr"1 for ILWAS to 0.4 yeq L"1 yr"1 for ETD.
Assuming a 30 percent deposition decrease, these rates of change in median sulfate concentrations
changed sign and magnitude, varying from -0.1 /ieq L"1 yr"1 for ETD to -0.4 fieq L"1 yr "1 for ILWAS and -
0.8 /ieq L'1 yr*1 for MAGIC.
The change in median pH projected over 50 years under current deposition ranged from 0.0 for
MAGIC to +0.07 for both ETD and MAGIC. The change In median pH projected over 50 years under
decreased deposition ranged from 0.1 for ETD to 0.15 for MAGIC and 0.5 for ILWAS. The variance in
pH was greatest for ILWAS and varied overtime (Figure 10-56).
There also was no indication that the rates of change in ANC or sulfate concentrations were
functions of the initial ELS-I ANC concentration for either deposition scenario (Table 10-17). Histograms
of projected change in median ANC and sulfate concentrations over 40 years using all three models
indicate a relatively uniform change among lakes regardless of their initial ANC concentrations (Figures
10-57 through 10-62).
10-125
-------
<1.5 x Marquwtfle tone*)"
(1.5 « Mnquufl* Ftono«r
Constant
150-
100-
Ii soH
I
e n-i
-50-
-100-
S 8 § S
§= - !E §E §E
ISO-, o
li 50-
I
O
-50 H
-100-
Ramped
8 §
§E §E
E3
Figure 10-48. Box and whisker plots of ANC distributions En 10-year intervals projected using ETD
for NE lakes, Priority Classes A and B.
10-126
-------
3fd Quarts. »
(15 x kiMrqwrtto Rangt)-
MOuwtto
HtQuvtte
1* Quart*
(1.6 » M^Mrtto Rang*)"
"M* to wcied ndranw mlu*
Constant
2OO-
150-
8"
Si 50-
o—
-50-
-itxa—
0
£
«••
M
Hfe
^H
•1
•1
O
£
•
Ml
«•
4
••
•*
»
ss
••ft
(•••I
^
•ft
g£
•i
^^
••
^
9
£
•i
•i
•
s
SE
4W*
•HI
Ramped
200-
1SO-
I so-
d&^^
o —
-50—
*• f *- p- »•
^^
•i
^
•i
di
^
^ ^
^ _.
^
•h
•i
•i
-rt-
4» ™
•1
•1
!*•
>•
^
'
••IP ^^
•1
rfl
^
4^
^
^^ W
H
H
^
M
^
^
Figure 10-49. Box and whisker plots of ANC distributions in 10-year intervals projected using
ILWAS for NE lakes, Priority Classes A and B.
10-127
-------
MQwrtb*
(15.lrt«rq
3rdOiartM
in
ittQuwtto
(1.S x tanputDo Rtnge)"
-tkatj •xeMdwtnmvtlui
Constant
ISO
100-
-, SE
o-
-50-
-100-
1SO-1
IM-
50-
-50-
-100-
Ramped
s
Figure 10-50. Box and whisker plots of ANC distributions in 10-year Intervals projected using
MAGIC for NE fakes, Priority Classes A and B.
10-128
-------
250 -i
200-
150H
50-
o
DC
(1J5 x InUrquKlOa Rang*}"
an)Ouw«b
KtOuwfl*
UtOuottt
"fMIOI
Constant
&
DC
Z50 —
200-
f?
150—
I
*„
§ 100-
50-
r
^^
L-
^^
^v
j
^^
r
^
^
^H
VI
•
^H
4 ^^
r f
^B
•»
^^ ^^"
w. ^_
- «
•" •—
r— ^
_ —
__ —
Ramped
s
SE
Figure 10-51. Box and whisker plots of sulfate distributions in 10-year intervals projected using
ETD for NE lakes, Priority Classes A and B.
10-129
-------
3rt Quad* •»
-------
250-1
200 H
1
50 H
3rdQu«rH«*
(15 X frrtarqiMrtte Raflga)**
andOuwlto
(1.5 x Intvquutte Rang»)~
O 4BBOMKI •XtPMIWVUtM
Constant
s
5
Ramped
s
§E
8
250-
200-
f-
I™'
A
§"100-
50-
^
^^
^
^
^
» 4
4*
^*
^ _
^
r
•
•
•
•
•
MB
4^
^
8
Figure 10-53. Box and whisker plots of sulfate distributions in 10-year intervals projected using
MAGIC for HE lakes, Priority Classes A and B.
10-131
-------
3rd Quart!* +
(1.5 x Inaiquartte Rang*)"
Mian
Median
lit Quart!
(1.5 x InMrquwlB* R«ng>)~
~Ne«to !»Md •xtiwn* nkt
Constant
O Q (
y—
7-
6-
5-
A—
f
(
MV
^
VMBp
^
•BB
•••
n
••
)•*
••
^
i^
4V
•Ml
1
«
Ml
*
«l
•••
MM
»
mm
mm
J
mm
turn
mm
ml
Ramped
O •*•
§£ g
Figure 10-54. Box and whisker plots of pH distributions in 10-year intervals projected using ETD
for NE lakes, Priority Classes A and B.
10-132
-------
3idOwnfl* +
(1.5 x IntofquaRfe Rang*)**
MQuudto
MMH
ItiOutftfli
1«Qu«rtJto
(1£ x Imwqwitl* Ring*)**
"Not to noMd •xtrwm v»Ju»
Constant
o
CC
Ramped
£ §E
DC
9 8
DC GC
>- >
Figure 10-55. Box and whisker plots of pH distributions in 10-year intervals projected using ILWAS
for NE lakes, Priority Classes A and B.
10-133
-------
7-
5-
o
§E
8-1
7-
5-
3rd Quartto +
(1.5 «ItiMquaiMt Rang§)~
SnfQutnlb
Mun
Median
ItfQuangi
(13 x imuquaitg* Ring*)-
MtdwmmrakN
Constant
oc
9
>-
8
i
Ramped
§E SE
§
i
Figure 10-56. Box and whisker plots of pH distributions In 10-year intervals projected using MAGIC
for NE lakes, Priority Classes A and B.
10-134
-------
Northeast Lakes
Priority Class A - B
Model-ETD
Deposition = Constant
200
2 100-
0>
J3
-40 -15 10 35
60
85
110 135 160
Northeast Lakes
Priority Class A - B
Model = ETD
Deposition = Ramped 30% Decrease
200
-15
10
35 60 85
ANCOieqL-')
110 135 160
ETD Year 10
Year 50 Ramped
Figure 10-57. ETD ANC population distributions at year 10 and year 50 for current and decreased
deposition.
10-135
-------
200-
150-
CO
° 100-
o
.Q
50-
200-
150
M
2
3
° 100-
50
Northeast Lakes
Priority Class A - B
Model = ILWAS
Deposition = Constant
-15
10
35 60
ANC(|ieqL-')
85
110 135 160
D ILWAS Y«ar 10
B ILWAS Year 50
Northeast Lakes
Priority Class A - 8
Model = ILWAS
Deposition = Ramped 30% Decrease
-40 -15
10 35 60
ANC(neqL ')
85 110 135 160
D ILWAS Year 10
Q Year SO Ramped
Figure 10-58. ILWAS ANC population distributions at year 10 and year 50 for current and
decreased deposition.
10-136
-------
200
150-
° 100-
o>
a
n
2
50-
-40
200
150
r 100-
I
50
Northeast Lakes
Priority Class A > B
Model = Magic
Deposition = Constant
10 35 60 85
ANCOieqL-')
110 135 160
O MAGIC Year tO
H MAGIC Year 50
Northeast Lakes
Priority Class A - B
Model = Magic
Deposition * Ramped 30% Decrease
-40 -15 10 35 60 85
ANCfeieqL-1)
110 135 160
Q MAGIC Year 10
Z Year 50 Ramped
Figure 10-59. MAGIC ANC population distributions at year 10 and year 50 for current and
decreased deposition.
10-137
-------
Northeast Lakes
Priority Class A - B
Model = ETD
Deposition = Constant
2001
150
2 1001
3
50
30 40 50 60 70 80 90 100110120130140150160170180190200210220230240250260270
Northeast Lakes
Priority Class A - B
Model = ETD
Deposition = Ramped 30% Decrease
200
150
® 100
50
30.40 50 60 70 80 90100110120130140150160170180190200210220230240250260270
Figure 10-60. ETD sulfate population distributions at year 10 and year 50 for current and
decreased deposition.
10-138
-------
Northeast Lakes
Priority Class A - B
Model * ILWAS
Deposition = Constant
200
150
I
2 1001
50
30 40 50 60 70 80 90100110120130140150160170180190200210220230240250260270
Northeast Lakes
Priority Class A - B
Model = ILWAS
Deposition = Ramped 30% Decrease
1201
100
I 80
CO
_i
o 6o
|
| 40-
20
D
s a
30 40 50 60 70 80 90100110120130140150160170180190200210220230240250260270
Figure 10-61. ILWAS suifate population distributions at year 10 and year so for current and
decreased deposition.
10-139
-------
Northeast Lakes
Priority Class A - B
Model = Magic
Deposition = Constant
200
150
o 100
50
H
30 40 50 60 70 80 90 100110120130140150160170180190200210220230240250260270
Northeast Lakes
Priority Class A - B
Model = Magic
Deposition = Ramped 30% Decrease
200
150
2 too-
50'
JLa
30 40 50 60 70 80 90100110120130140150160170180190200210220230240250260270
Figure 10-62. MAGIC sulfate population distributions at year 10 and year 50 for current and
decreased deposition.
10-140
-------
10.11.2 Southern Blue Ridge Province
10.11.2.1 Target Population Projections Using MAGIC
An estimated 1323 streams in the SBRP target population were simulated using MAGIC. This target
population included both disturbed and undisturbed watersheds based on chloride concentrations; all
watersheds had positive sulfur retention. Three streams (which had NSS Pilot Survey ANC > 400
L"1 ) were excluded subsequently from this target population, although they were simulated by MAGIC.
The MAGIC projections indicated that ANC concentrations in these three systems essentially did not
change over the 200-year simulation. Including these streams in the discussion distorts the scales of the
ANC figures because two of these streams had ANC concentrations > 1000 /ueq L"1. These projections
apply only to streams in the SBRP target population and do not necessarily represent southeastern
stream responses.
10.11.2.1.1 Deposition scenarios -
There were significant changes in projected ANC and sulfate concentrations and in pH over the
200-year period assuming both current deposition and a 20 percent deposition increase (Figures 10-63
and 10-64). The 200-year time frame was selected to assess changes in surface water chemistry as the
watersheds approach sulfate steady state. The time frame is for comparative purposes only and does
not represent expected changes over this time frame. Median ANC was projected to decrease from 124
to 78 (-46) A*eq L*1 over 200 years under current deposition, and from 124 to 59 (-65) Meq L'1 assuming
increased deposition (Table 10-18). This decrease is greater than the uncertainty bounds on the
projections. The median sulfate concentration was projected to increase from 37 to 111 (+74) /ieq L"
1 over the 200-year period under current deposition and from 37 to 133 (+96) /ieq L"1 over the 200-
year period assuming increased deposition (Table 10-18). This increase also is greater than the
uncertainty bounds about the projections. The median pH was projected to decrease from 7.0 to 6.75
over the 200-year period under current deposition and from 7.0 to 6.6 with increased deposition. The
lower quartile pH, however, was projected to decrease from 6.75 to 6.2 over 200 years under current
deposition and from 6.75 to 5.3 under increased deposition. A decrease also was projected for median
10-141
-------
S8RP Stream Reaches
Model - MAGIC
Priority Class = A - E
Year = 20
O 0.8
O
tX
O
0.6
ys 0.4
re
0.0"—
-100
10
O 0.8
2 0.6
_«
3
0.4
Simulation Year 0
Constant Deposition
Ramp Deposition
100
ANC
200
L-i)
300
400
0.0
SBRP Stream Reaches
Model = MAGEC
Priority Class = A - E
Year = 20
Simulation Year 0
Constant Deposition
••••• Ramp Deposition
too 200
[SO*2-] (jieq ,L-i)
300
Year • 50
Year
50
1.0
O 0.8
O
Q.
E 0.6
Q.
ffl
I0"4,
E
o
0.0'—
-100
1.0
O 0.8
0.6
Simulation Year 0
.... Constant Deposition
Ramp Deposition
0 100
ANC
200
300
400
O
£
o
V= 0.4
•3
E
0.0
Simulation Year 0
Constant Deposition
Ramp Deposition
100 200
ISO.4-] (|ieq L-i)
300
Figure 10-63. MAGIC ANC and sulfate projections for SBRP streams, Priority Classes A - E, at year
20, year SO, year 100, and year 200 under current and increased deposition. (Continued).
10-142
-------
1.0
O 0.8
2 0.6
OJ
i
0.0
-100
88RP Stream Reaches
Model - MAGIC
Priority Class • A - E
Year - 100
•Simulation Year 0
— Constant Deposition
Ramp Deposition
o
1.0
0.8
o
S 0.6
-------
I.O
O 0.8
O
Q.
8 0.6
Q.
0)
7 0.4
3
£
O °-2
0.0
SBRP Stream Reaches
Model = MAGIC
Priority Class • A - E
Year = 20
Year 0
Constant
Ramp
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
pH
Year = 50
1.0
O 08
U»9
O
Q.
£ 0.6
Q.
-------
SBRP Stream Reaches
Model = MAGIC
Priority Class = A - E
Year « 100
1.0
O 0.8
V.
O
2 0.6
Q.
O
= 0.4
f£
i
3
O °-2
0.0
Year 0
Constant
Ramp
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
PH
Year - 200
1.0r
O 0.8
O
Q.
2 0.6
= 0.4
03
O
0.2
0.0
Year 0
- - - - Constant
Ramp
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
pH
Figure 10-64. (Continued).
10-145
-------
Table 10-18. Descriptive Statistics of Projected ANC, Sulfate, and Percent Sulfur Retention, and
Calcium and Magnesium for SBRP Streams in Priority Classes A - E Using MAGIC for Both Current
and Increased Deposition
Model
Std.
Mean Dev. Min.
P_25
Median
P_75
Max.
Current Deposition
Maaic All.
YrO
Yr20
YrSO
Yr 100
Yr200
Maaic All.
YrO
Yr20
YrSO
Yr 100
Yr200
MAGIC All
YrO
Yr20
YrSO
YMOO
Yr200
Maaic All.
YrO
Yr20
YrSO
Yr 100
Yr200
Maaic All.
YrO
Yr20
YrSO
Yr 100
Yr200
ANC
139.33
130.51
118.64
101.41
78.38
so,2'
47.88
60.43
77.71
98.36
112.08
. DH
6.87
6.68
6.11
5.68
5.52
% S Retention
59.31
48.54
34.11
18.08
6.99
Ca + Ma
131.18
135.50
139.17
140.76
129.39
93.87
96.34
99.13
96.71
85.30
26.14
28.77
31.32
29.42
28.72
0.28
0.38
0.59
0.72
0.79
20.92
22.65
23.61
17.30
7.69
72.91
72.90
74.12
79.41
74.10
20.45
9.97
-3.26
-12.78
-18.92
11.99
13.26
15.92
25.67
70.15
6.23
5.92
5.20
4.77
4.61
23.61
17.81
5.42
0.60
-2.45
49.82
53.44
57.17
56.69
49.14
70.88
63.43
52.36
39.80
18.97
28.68
35.23
57.15
84.66
87.03
6.76
6.71
6.63
6.50
6.19
35.16
27.83
20.99
10.00
0.82
85.05
86.19
102.38
93.71
87.46
123.62
122.46
112.19
100.25
77.86
37.23
53.67
75.34
96.92
111.17
6.99
6.99
6.96
6.91
6.80
64.88
48.07
27.10
14.49
5.60
115.04
121.00
119.35
108.40
104.92
156.21
143.68
124.58
111.02
90.90
68.28
76.05
99.41
113.32
122.94
7.10
7.06
7.01
6.95
6.86
78.41
70.26
50.52
18.38
10.44
139.53
144.57
161.74
154.51
132.83
510.12
507.41
509.44
465.94
370.95
98.63
118.29
143.86
173.07
208.82
7.60
7.59
7.59
7.55
7.46
90.71
88.56
85.37
76.41
27.64
370.33
370.03
382.33
437.51
385.24
continued
10-146
-------
Table 10-18. (Continued)
Model
Mean
Std.
Dev.
Min.
P 25 Median P 75 Max.
20% Increase in Deposition
Maaic All.
YrO
Yr20
YrSO
Yr 100
Yr200
Maaic All.
YrO
Yr20
YrSO
Yr 100
Yr200
MAGIC All
YrO
Yr20
YrSO
Yr 100
YrSOO
Magic AH.
YrO
Yr20
YrSO
Yr 100
Yr200
Maaic All.
YrO
Yr20
YrSO
Yr 100
Yr 200
ANC
139.33
128.20
111.09
87.43
59.50
SO*
47.88
64.39
93.71
122.59
136.50
, pH
6.87
6.61
5.79
5.53
5.28
% S Retention
59.31
51.68
33.84
15.38
5.39
Ca + Ma
131.18
136.88
145.18
146.87
129.72
93.87
96.36
99.79
95.36
81.58
26.14
30.91
37.91
37.30
32.97
0.28
0.41
0.70
0.76
0.91
20.92
21.30
23.71
16.41
6.04
72.91
73.28
75.35
84.14
74.73
20.45
7.77
-10.23
-17.68
-21.76
11.99
14.55
18.85
31.98
96.87
6.23
5.83
4.86
4.63
4.52
23.61
25.01
4.13
-0.19
-2.46
49.82
53.49
59.71
59.59
48.22
70.88
61.97
45.58
29.06
-0.32
28.68
36.96
67.72
102.28
104.64
6.76
6.70
6.57
6.37
5.42
35.16
31.84
21.28
5.86
0.38
85.05
86.08
104.34
98.10
84.41
123.62
120.68
104.55
82.73
59.21
37.23
56.82
91.81
124.67
133.40
6.99
6.99
6.93
6.82
6.68
64.88
51.49
25.97
11.38
5.15
115.04
122.13
123.19
109.13
105.85
156.21
141.46
114.84
91.96
72.97
68.28
80.76
118.18
145.87
147.53
7.10
7.05
6.97
6.88
6.77
78.41
72.72
51.14
17.70
8.77
139.53
145.44
170.85
174.21
135.25
510.12
506.52
506.81
443.43
343.54
98.63
127.78
175.00
228.00
250.60
7.60
7.59
7.59
7.54
7.43
90.71
89.01
85.56
75.50
16.74
370.33
373.24
394.03
472.24
387.73
10-147
-------
calcium plus magnesium concentration, with a projected decrease from 115 to 105 (-10) /ieq L"1 for both
current and increased deposition.
Median sulfur retention for the SBRP watersheds at year 0 is about 65 percent. Median sulfur
retention projected after 50 years was about 26 percent and after 200 years was about 5 percent for both
current and increased deposition. The lower and upper quartile values ranged from less than 1 to about
10 percent after 200 years for both deposition scenarios, indicating the watersheds were approaching
sulfate steady state.
Projections of the number of streams that might become acidic after 50, 100, and 200 years
assuming current deposition were 129 (9 percent), 159 (11 percent), and 203 (14 percent), respectively.
Projections of the number of streams that might become acidic after 50, 100, and 200 years assuming
a 20 percent deposition increase were 159 (11 percent), 159 (11 percent), and 337 (24 percent),
respectively. For these estimates, the three streams with ANC > 400 /xeq L"1 were included in the target
population, which represented 1429 streams.
10.11.2.1.2 Rate of change of ANC, sulfate, and pH over 200 years -
The projected change in ANC and sulfate concentrations and in pH over the 200-year period for
both current and increased deposition are shown in box and whisker plots (Figures 10-65 through 10-
67). The projected rates change in median ANC over 200 years assuming current and increased
deposition were -0.23 /ieq L'1 yr*1 and -0.32 /xeq L*1 yr~1, respectively. The relative changes in median
ANC projected to occur for the first 50 years, from 50 to 100 years, and from 100 to 200 years were -
11, -12, and -23 jueq L"1, respectively. These projections represent a relatively constant linear decrease
in ANC over time assuming current deposition. Assuming a 20 percent deposition increase, the projected
changes in median ANC for the first 50 years, from 50 to 100 years, and from 100 to 200 years were
-19, -21, and -23 jiteq L*1, respectively, indicating a constant relatively linear decrease for the first 100
years and then a slower rate of change over the next 100 years.
10-148
-------
MOutttt
(1£xbt«qui
-t« to «at»*d *0nim ratu*
Constant
o S 8 g ? 8
§E §E §E §E £ £
300-
fT 200~
^Kf*
J.100-
o
z n
< o
-100-'
r-
E
p -,-
"I •• • • ™ • l^
mf^ i.-
Ramped
0 £ 8 g § S
0 0
•^- CM
tr oc
— •—
nj_
€3 ^^J
^^% ^^J
300-
cc
f 200-
f 100-
z 0-
_-
-------
MOwnfe*
(1-Sxlrt.iqu.rtfl.R.ng.r
MQlitttl*
Constant
g
S 8
200-,
85
~ 150-
lj
f 100—
o-
w so-
rt—
i
^m
^m
^
^
Ml
••
^
M
^VH
_]
^Pi»
IB
•d
^
^H
^H
w
^
E
3
p
«•*
^
Ramped
: 8
£
200—
•^
3 inn
<,
CO «
so
n—
r
l_
n
J
j
i^H»
••••
^^
•HHI|
1
«••
Ml
r
1 4
l_
"1
-
_l
•••
•••
'
•ii^
••
*m
A^
^
r*v
^B
^
•••
1^
••
Figure 10-66. Box and whisker plots of sulfate distributions in 10-year intervals projected using
MAGIC for SBRP streams, Priority Classes A - E, for current and increased deposition.
10-150
-------
1A0UWU*
IrtQuntl*
Constant
8—I
7-
5—
s
£
S
§E
S 8 i
CH C]
C]
7—
1.6-
5—
Ramped
o 9 8 8
£ §E §E §=
Figure 10-67. Box and whisker plots of pH distributions in 10-year intervals projected using MAGIC
for SBRP streams, Priority Classes A - E, for current and increased deposition.
10-151
-------
The projected change in median sulfate concentration varied over the 200-year simulation period
with a relatively linear increase during the first 50 years, asymptotically approaching the 200-year sulfate
concentration, 111 jieq L'1 (Table 10-18). The increase for the first 50 years was from 37 to 75(+38)
Meq L"1, for 50 to 100 years from 75 to 97 (+22) /Lteq L"1, and for 100 to 200 years from 97 to 111(+14)
Meq L"1 under current deposition. The median sulfate projected for increased deposition was an increase
from 37 to 92 (+55) jieq L'1 for the first 50 years; for 50 to 100 years, the increase was from 92 to 125
(+33) /Jeq L*1; and for 100 to 200 years, the increase was from 125 to 133 (+8) Meq L*1.
Median pH values were relatively unchanged over the first 50 years under either current or
increased deposition and changed about -0.1 units over 100 years (Table 10-18). Over 200 years, the
median pH changed -0.25 units under current deposition and -0.4 units with Increased deposition. Lower
quartile pH values were projected to change about -0.15 units in 50 years under either deposition
scenario. The lower quartile pHs changed -0.5 units in 100 years under current deposition and -0.6 units
with increased deposition. After 200 years, the lower quartile pH values were projected to change by
-0.8 units under current deposition and -1 .7 units under increased deposition.
Projected median calcium plus magnesium concentrations increased from 115 to about 123
L"1 during the first 50 years and then decreased to about 108 to 110 peq L'1 by year 100, with a further
decrease to 105 Meq L*1 at year 200 under both current and increased deposition.
There was a differential change projected among streams based on their initial ANC concentrations.
Streams with higher initial ANC (based on NSS Pilot Survey data) were projected to have a greater
change in ANC than streams with lower initial ANC. This result is illustrated by the change in the
frequency intervals of streams in different ANC categories in the histograms (Figures 10-68 and 10-69)
and Table 10-15. The projected changes for ANC concentrations in streams with initial ANC between 25
and 100 jteq I-'1 and between 100 and 400 peq L'1 were -14 /Jeq L*1 versus -24 peq L'1 over a 40-
10-152
-------
SBRP Stream Reaches
Priority Class A -E
Model = Magic
Deposition = Constant
500
400-
W
(0
2 300-
03
"o
1 200-
E
Z
100-
ng
i
i
rf8
1
^
V
\
SSSSSSSSSS3S3
4M
I
r-n
••m
I
I
\
8
I
m
^rif^rb In ni™
-40-15 10 35 60 85110135160210235285310335360410460510
ANC(jieqL-')
Q MAGIC Yeaf 10
H MAGIC Year 50
SBRP Stream Reaches
Priority Class A - E
Model = Magic
Deposition = Ramped 20% Increase
CO
o
o
500-
400
300
200
100
-40-15 10 35 60 85110135160210235285310335360410460510
ANCftieqL-')
D MAGIC Year 10
B Year 50 Ramped
Figure 10-68. MAGIC ANC population distributions at year 10 and year 50 for current and
increased deposition, SBRP streams, Priority Classes A - E.
10-153
-------
SBRP Stream Reaches
Priority Class A -E
Model SB Magic
Deposition = Constant
500-
400
0)
i
|> 300-
8)
"5
IOT
/D
% 200-
e
3
z
100-
.
o-
n
^H
P
|
|
^
S
1
—
^r-0
II ffjl
1 it
1
|
10 20 30 40 50 60 70
fS(
lov
r
\
%
I
i
\
i
h
\
i
\
g
\
\
i
i
*
ri i
g «
1 |n |
i i \ m i
• • • * ' '
80 90 100110120130140150160170180
•ja-t, ._,, D MAGIC Year 10
4JuieqL ; a MAGIC Year SO
SBRP Stream Reaches
Priority Class A - E
Model = Magic
Deposition = Ramped 20% Increase
500
400
2 300
to
^ 200
100-
S3
J
10 20 30 40 50 60 70 80 90 100110120130140150160170180
D MAGIC Year 10
0 Year 50 Ramped
Figure 10-69. MAGIC sulfate population distributions at year 10 and year 50 for current and
increased deposition, SBRP streams, Priority Classes A - E.
10-154
-------
year period under current deposition. The projected changes for ANC concentrations in streams with
initial ANC between 25 and 100 Meq I'1 and 100 to 400 fj,eq L"1 were -21 versus -34 jueq L*1 over a 40-
year period under increased deposition.
10.11.2.2 Restricted Target Population Projections Using ILWAS and MAGIC
An estimated 567 streams in the target population were represented in simulations using both
ILWAS and MAGIC. These streams were considered undisturbed based on the chloride concentrations
and had NSS Pilot Survey ANC concentrations less than 200 /*eq L*1. All the watersheds had positive
sulfur retention.
10.11.2.2.1 Deposition scenarios -
ILWAS and MAGIC projected similar changes in ANC, sulfate, and pH after 50 years. Changes
projected for streams with lower initial ANC concentrations using MAGIC were greater than those
projected using ILWAS (Figures 10-70 through 10-72). The ILWAS model performed 50-year rather than
200-year simulations, because of time and computational restrictions, and comparisons are therefore
made only for this 50-year period.
Median ANC concentrations using the ILWAS model were projected to decrease from 87.4 to 72.4
Meq L"1 (-15.0 /ieq L"1 ) under current deposition and from 87.4 to 71.8 jueq L"1 (-15.2 jiteq L"1) for
increased deposition (Table 10-19). Median ANC using MAGIC was projected to decrease from 118.1 to
85.5 (-32.6) fJ,eq L*1 for current deposition and from 118.1 to 80.1 (-38.0) Ateq L*1 for increased deposition
for the 50-year simulation period. Differences between the change projected by the two models were 17.6
L*1 at current deposition and 22.8 M&q L'1 at Increased deposition.
Median sulfate concentrations using the ILWAS model were projected to increase from 25.0 to 58.9
(+33.9) jieq L"1 for current deposition and from 25.0 to 69.1 (+44.1) /Lteq L*1 for increased deposition
10-155
-------
10
OJ
hAi
I
0.6
0.4
0.0
SBRP Stream Reaches
Priority Class - A 4 B
Deposition - Constant
Year - 0
0 100 200 100
ANC dieq L-<)
400
SBRP Stream Reaches
Priority Class - A & B
Deposition « Ramp 20% Increase
Year - 0
0.8
Q.
£
ae
0,4
OJ>
-100
0 tQO 200 300 400
ANC
-------
tOr
SBRP Stream Reaches
Priority Class - A & B
Deposition • Constant
Year - 0
WO 900
{S0.*j (fieq
300
SBRP Stream Reaches
Priority Class - A & B
Deposition - Ramp 20% Increase
Year = 0
0 WO 200 300
[StVJ (tieq L-')
Deposition - Constant
Year - 20
WO 200 300
[S04*l (|ieq Li)
Deposition • Ramp 20% Increase
Year - 20
300
Deposition - Constant
Year - 50
[SO,*-] Qieq L-0
300
Deposition » Ramp 20 Increase
Year - SO
0 100 200 900
[SO,*] Qieq L-I)
Figure 10-71. Comparison of ILWAS and MAGIC projections for sulfate concentration at years 0,
20, and 50 for SBRP streams, Priority Classes A and B, under current and increased deposition.
10-157
-------
S8RP Stream Reaches
Priority Class - A & B
Deposition * Constant
Year - 0
to
0.1
tut
0.4
o
4.0 *6 5.0
PH
8,5 7.0 7J 10
SBRP Stream Reaches
Priority .Class - A & B
Deposition - Ramp 20% Increase
Year - 0
tOr
o as
I
a
02 •
OJ>
Pha$t 1
MAGIC
....*,*,*.. H tt/Afi
itn no
<0 44 SJJ
6.0
PH
74 7.5 6.0
Deposition « Constant
Year - 20
0.6
0.4
4J>
U Ul
pH
7J> 73 «JO
Deposition » Ramp 20% Increase
Year - 20
to
§ OJB
0.6
OJ>
44 4J U
8.0
PH
70)
«-0
Deposition - Constant
Year - 50
to
0.4
4JI 4J 8.0
U
PH
7.0 IS
Deposition - Ramp 20% Increase
Year - 50
to
Figure 10-72. Comparison of ILWAS and MAGIC projections tor pH at years 0, 20, and 50 tor
SBRP streams, Priority Classes A and B, under current and increased deposition.
10-158
-------
Table. 10-19. Descriptive Statistics of Projected ANC, Sulfate, Percent Sulfur
Retention, and Calcium Plus Magnesium for SBRP Streams in Priority Classes A
and B Using ILWAS and MAGIC for Both Current and increased Deposition
Model Mean
ANC
Model Year 0
ILWAS 97.64
MAGIC 109.01
Model Year 20
ILWAS 90.50
MAGIC 99.53
Model Year 50
ILWAS 79.48
MAGIC 86.89
Model Year 0
ILWAS 31.36
MAGIC 48.75
Model Year 20
ILWAS 47.17
MAGIC 61.17
Model Year 50
ILWAS 71.02
MAGIC 78.54
QH
Model Year 0
ILWAS 6.82
MAGIC 6.82
Model Year 20
ILWAS 6.77
MAGIC 6.62
Model Year 50
ILWAS 6.64
MAGIC 6.05
% S Retention
Model Year 0
ILWAS 73.63
MAGIC 58.12
Model Year 20
ILWAS 60.00
MAGIC 47.32
Model Year 50
ILWAS 39.55
MAGIC 32.64
Std.
Dev.
37.09
45.79
33.45
47.94
31.20
49.08
16.48
26.14
17.90
28.29
23.05
31.00
0.23
0.23
0.23
0.34
0.28
0.58
'
11.00
20.57
10.79
21.68
13.54
22.30
Min.
Current
22.10
20.45
20.84
9.97
18.98
-3.26
11.55
11.99
28.97
13.26
42.26
15.92
6.32
6.23
6.27
5.92
6.10
5.20
37.12
23.61
28.85
17.81
19.56
5.42
P_25
Deposition
83.37
70.21
78.20
62.46
53.18
52.36
20.65
29.51
35.14
43.29
54.96
69.48
6.71
6.76
6.63
6.71
6.44
6.63
64.47
35.16
55.13
27.83
22.43
20.99
Median
87.38
118.08
84.77
99.47
72.45
85.49
25.02
37.23
40.32
53.67
58.88
75.34
6.96
6.99
6.91
6.91
6.84
6.85
80.02
64.88
64.50
48.07
46.02
24.49
P_75
•117.50
151.52
104.40
142.22
98.60
123.63
40.11
68.28
53.87
76.05
88.12
94.86
7.09
7.09
7.05
7.06
6.95
7.00
80.32
77.25
66.00
66.83
48.73
29.52
Max.
159.10
208.36
144.60
210.11
126.30
215.47
72.76
98.63
93.42
118.29
118.00
143.86
7.27
7.23
7.27
7.24
7.23
7.25
89.22
89.21
81.85
87.81
73.83
85.37
continued
10-159
-------
Table 10-19 continued
Model Mean
Ca + Mq
Model Year 0
ILWA8 88.07
MAGIC 107.47
Model Year 20
ILWAS 89.63
MAGIC 111.46
Model Year 50
ILWAS 93.90
MAGIC 113.99
Std.
Dev.
26.92
35.96
26.36
35.15
26.15
33.40
Min.
39.43
49.82
40.52
53.44
43.69
57.17
P_25 Median P_75
63.71
85.05
61.74
86.19
73.09
88.05
82.32
115.04
97.86
121.00
95.67
114.77
104.35
126.65
104.63
127.97
113.29
134.02
Max.
127.93
190.55
132.22
185.51
143.19
180.53
20% Increase in Deposition
ANC
Model Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
Mode! Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
fit!
Model Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
0
97.64
109.01
20
90.47
97.27
50
78.89
79.18
0
31.36
48.75
20
47.80
65.14
50
81.84
94.64
0
6.82
6.82
20
6.77
6.55
50
6.60
5.72
37.09
45.79
33.39
48.19
31.77
49.43
16.48
26.14
18.00
30.44
29.12
37.70
0.23
0.23
0.23
0.38
0.30
0.69
22.10
20.45
20.84
7.77
18.91
-10.23
11.55
11.99
29.65
14.55
46.34
18.85
6.32
6.23
6.27
5.83
6.06
4.86
83.37
70.21
78.18
60.91
51.19
45.58
20.65
29.51
35.43
45.50
61.53
82.80
6.71
6.76
6.63
6.70
6.41
6.57
87.38
118.08
85.02
95.49
71.77
80.11
25.02
37.23
40.84
56.82
69.12
91.81
6.96
6.99
6.91
6.89
6.83
6.82
117.50
151.52
104.20
139.82
99.50
113.87
40.11
68.28
53.96
80.76
100.40
112.23
7.09
7.09
7.05
7.05
6.96
6.96
159.10
208.36
144.40
208.96
126.00
215.60
72.76
98.63
93.77
127.78
135.50
175.00
7.27
7.23
7.27
7.23
7.23
7.25
continued
10-160
-------
Table 10-19 continued
Model Mean
Std.
Oev.
Min.
P 25
Median
P 75 Max.
% S Retention
Model Year 0
ILWAS 73.63
MAGIC 58.12
Model Year 20
ILWAS 64.22
MAGIC 50.55
Model Year 50
ILWAS 41.98
MAGIC 32.42
Ca + MQ
Model Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
Model Year
ILWAS
MAGIC
0
88.07
107.47
20
89.99
112.88
50
101.20
119.76
11.00
20.57
9.64
20.40
15.15
22.50
26.92
35.96
26.45
36.03
27.06
35.17
37.12
23.61
36.16
25.01
17.61
4.13
39.43
49.82
41.07
53.49
52.26
59.71
64.47
35.16
60.34
31.84
25.77
22.10
63.71
85.05
61.48
86.08
79.44
88.49
80.02
64.88
68.23
51.49
48.34
25.01
82.32
115.04
98.91
122.13
103.38
120.44
80.32
77.25
69.75
69.24
50.38
29.14
104.35
126.65
105.06
130.45
122.12
147.24
89.22
89.21
83.97
88.20
77.14
85.56
127.93
190.55
132.77
189.10
148.80
186.86
10-161
-------
(Table 10-19). The median sulfate increases projected using MAGIC were from 37.2 to 75.3 (+38.1) jueq
I"1 for current deposition and from 37.2 to 91.8 (+54.6) jueq L'1 for increased deposition. Differences
between the changes projected using the two models were 4.2 /ueq L*1 for current deposition and 10.5
jueq L'1 for increased deposition.
Median pH values using the ILWAS model were projected to decrease from 7.0 to 6.8 (-0.2) for
current deposition and 7.0 to 6.8 (-0.2) for increased deposition (Table 10-19). The median pH values
projected using MAGIC decreased from 7.0 to 6.9 (-0.1) for current deposition and from 7.0 to 6.8 (-0.2)
for increased deposition.
Median calcium plus magnesium concentrations using the ILWAS model were projected to increase
from 82.3 to 95.7 (+13.4) fieq L"1 for current deposition and from 82.3 to 103.4 (+21.1) jieq L"1 for
increased deposition (Table 10-19). The median calcium plus magnesium concentrations using MAGIC
were projected to increase from 115 to 114.8 (-0.2) fjeq L*1 for current deposition and from 115 to 120.4
(+5.4) jiteq L*1 for increased deposition. Differences between the change projected using the two models
were 13.2 jieq L*1 for current deposition and 15.7 jLteq L*1 for increased deposition.
Watersheds in the SBRP had an estimated median sulfur retention of 46 percent for current
deposition and 48.3 percent for increased deposition (Table 10-19) after 50 years using ILWAS. Median
sulfur retention for SBRP watersheds using MAGIC was projected to vary from about 24.5 percent for
current deposition to 25 percent for increased deposition. '
None of the streams in the SBRP was projected to become acidic within 50 years using the ILWAS
model for either current or increased deposition. There were 129 (23 percent) streams that might become
acidic at current deposition levels within 50 years using MAGIC and an estimated 159 (28 percent) that
might become acidic for increased deposition levels within 50 years.
10-162
-------
10.11.2.2.2 Rate of change of ANC, sulfate, and pH over 50 years -
The change in median ANC and sulfate concentrations and in pH for streams in the SBRP are
shown in box and whisker plots (Figures 10-73 through 10-78). Median ANC was projected to change
by about -15 fj.eq I-'1 over the 50-year period using the ILWAS model for both current deposition and
an increase in deposition. MAGIC projected median changes in ANC of about -33 (ieq L"1 for current
deposition and about -38 fieq L*1 for increased deposition (Table 10-19). The change in ANC was
relatively small for the first 10 to 20 years and then decreased relatively linearly for the next 30 years.
Median sulfate concentrations, estimated from the ILWAS model, were projected to increase by
about 34 jueq L"1 over the 50-year period for current deposition and about 44 neq L*1 for Increased
deposition over the 50 years (Table 10-19). Using MAGIC, the median sulfate concentrations were
projected to increase by about 38 Meq L'1 for current deposition and about 55 jieq L*1 for increased
deposition. There was a relatively linear increase in sulfate concentrations over the 50-year period for
both models.
Median and lower quaitile pH values were projected to change less than 0.2 units for both ILWAS
MAGIC over 50 years for either deposition scenario.
There was an indication that the changes in ANC and sulfate were functions of the initial (NSS -
Pilot Survey ) ANC using the ILWAS model (Table 10-19). A larger increase in sulfate concentrations and
a larger decrease in ANC in the lower ANC groups (i.e., 25 < ANC < 100 jLteq L'1) than in the higher
ANC groups (i.e., 100 < ANC < 400 /*eq L'1 ) were projected with the ILWAS model. Relatively similar
changes in ANC and sulfate among ANC groups were projected, however, with MAGIC. This result is
indicated in the number of streams that change frequency intervals for distributions of ANC and sulfate
concentration over the 40-year period (Figures 10-79 through 10-82).
10-163
-------
150-
S"
*.iooH
50-
o
§E
MOuirtl*
(UixMM
Sri Ouiitl*
1«Qiarti>
•*Notto«DMdnM
-------
MOuati*
IrtC
ttQuntl*
300-
250-
200-
Ii 160-
8T
3 100-
5 so-
0-
-50-
-100
o
i
Constant
S3?
3
i §£
Ramped
300-
2SO-
200-
150-
100-
50-
0-
-50-
-100
8
§E
?
SE
8
Figure 10-74. Box and whisker plots for ANC distributions in 10-year intervals projected using
MAGIC for SBRP streams, Priority Classes A and B, for current and increased deposition.
10-165
-------
SidQMrtlk*
(14 x MwqMMli
ItiQuvtib
120-
1OO—
^^^ fi^^^^
I «H
20—
0
o
8E
Constant
8 8
5?
Ramped
200-1
160-
A
8'
60-
ec
S
1C
Figure 10-75. Box and whisker plots for sulfate distributions in 10-year Intervals projected using
ILWAS for SBRP streams, Priority Classes A and B, for current and increased deposition.
10-166
-------
MQuatib
(t £ x M«q Ring*)"
"MKtoMDHrfttftwmtfWbi
200-1
150-
100—
*»
50-
o «£
SE SE
Constant
S 8 «
S
Ramped
200-1
~ 150-
1 100-
*,
O
« so-
o
5
___
•
M
^
^
^ M B * 10 " «
§E §E §E §E §= §E §E
^^
r
1:
• ^^
^^
_ r
1 L
d b
^^ ^^
r
T J
- L
J
v«
-, r
- b
^
F*W ^^
1 E
•1 L_
r
3 t
_i
3 E
««»
q
j
Figure 10-76. Box and whisker plots for sulfate distributions in 10-year intervals projected using
MAGIC for SBRP streams, Priority Classes A and B, for current and increased deposition.
10-167
-------
(1.5 « MMquBli* Rwg«r
1« Quirt*
1*t
» to «md Mdiwm
Constant
8-1
I 6-
S—
DC
Ramped
8-1
7-
5-
GC
8
55
s
S5
Rgure 10-77. Box and whisker plots for pH distributions in 10-year intervals projected using ILWAS
for SBRP streams, Priority Classes A and B, for current and increased deposition.
10-168
-------
MQuttfe
(ii
M
1«<
MOunUi
"*Mol to UDMd wtmnv i
Constant
8—1
7—
1.6-
8
cc
-------
SBRP Stream Reaches
Priority Class A -B
Model = ILWAS
Deposition - Constant
200
150-
-------
SBRP Stream Reaches
Priority Class A-B
Model = Magic
Deposition = Constant
200i
-40 -15 10 35 60
110 135 160 185 210
D MAGIC Year 10
ft MAGIC Year 50
SBRP Stream Reaches
Priority Class A - B
Model = Magic
Deposition = Ramped 20% Increase
200
150
I
50-
-!_
-40 -15 10 35 60 85 110 135 160 185 210
a MAGIC YaaMO
B Year SO Ramped
Figure 10-80. MAGIC ANC population distributions at year 10 and year 50 for current and
increased deposition, SBRP Priority Class A and B streams.
10-171
-------
SBRP Stream Reaches
Priority Class A -B
Model = ILWAS
Deposition = Constant
sou-
250-
CO
1 200-
£
55
•5 150-
<5
XI
1 100-
z
50-
o-
,n
^•H
nog
«r
|
r
1
%
w
w
vL
\
y
|
I
10 20 30 40 50 60 70 80 90 100 110 120 130 140
[SO^OieqL") 0 ILWAS YearlO
1 4JVflOM ' a ILWAS Year 50
SBRP Stream Reaches
Priority Class A - B
Model = ILWAS
Deposition = Ramped 20% Increase
300
250-
-------
SBRP Stream Reaches
Priority Class A -8
Model = Magic
Deposition = Constant
300
250
I 200-1
•5 150
100-
50-
EL
i
10 20 30 40 50 60 70 80 90 100 110 120 130 140
O MAGIC Year 10
a MAGIC Year SO
SBRP Stream Reaches
Priority Class A - 8
Model = Magic
Deposition = Ramped 20% Increase
300
250
| 200
£
CO
o 150
1
i 100
50-
IP.
m
i
10 20 30 40 50 60 70 80 90 100 110 120 130 140
O MAGIC Year 10
a Year 50 Ramped
Figure 10-82. MAGIC sulfate population distributions at year 10 and year SO for current and
increased deposition, SBRP Priority Class A and B streams.
10-173
-------
10.11.3 Regional Comparisons
This section focuses on regional comparisons among the aquatic systems in the NE and the SBRP.
Although the representative northeastern systems are lakes and the SBRP systems are streams, it is
watershed processes that control projected changes in ANC and sulfate. Comparisons of relationships
between ANC and sulfate in these systems, and of changes in pH and calcium plus magnesium with
changes in sulfate, can reveal similarities and differences in these processes between the regions.
10.11.3.1 Northeastern Projections of Sulfate Steady State
All three models projected that northeastern lakes would be at sulfate steady state within 50 years
at current levels of deposition (Figure 10-83). To examine sulfate steady state in the NE, projected sulfate
concentrations are compared with steady-state sulfate concentrations computed using current deposition
and mass balance. A 1:1 line indicates perfect agreement between the two values. These sulfate steady-
state projections are consistent with the percent sulfur retention of northeastern watersheds presented
in Tables 10-14, 10-16 and 10-17. With a 30 percent reduction, the projected sulfate values fall below
the 1:1 line, indicating a reduction in lake sulfate concentrations within a 50-year period compared to
the sulfate concentrations projected for current levels of deposition (Figure 10-84). The watershed sulfur
retention values calculated on the basis of sulfur input/output indicate the watersheds are near zero sulfur
retention (i.e., near sulfate steady state), after 50 years with a 30 percent deposition decrease. The
estimated time to sulfate steady state in the NE is less than 50 years for both current and decreased
deposition.
Comparisons of projected sulfate concentrations among models indicates excellent agreement
among all models for both current and decreased deposition (Figure 10-85). Fewer data for ILWAS
comparisons than for MAGIC and ETD are shown, because only 28 lakes were simulated. The 1:1
relationship among models, however, is evident.
10-174
-------
300
NE Lakes
Priority Class - A & B
Model = MAGIC
Deposition » Constant
*o wo 200 SOD
Steady State [SO«*-] dieq L-<)
300
m
NE Lakes
Priority Class • A & B
Model - ETD
Deposition - Constant
0 100 200 300
Steady State [SO4*1 (jieq L-<)
aOOr
Priority Class • A & B
Model » ILWAS
Deposition • Constant
~0 100 200 300
Steady State (SO.*-] (neq L-<)
300
g
«. 200
•KM •
Priority Class - A - E
Model - ETD
Deposition • Constant
0 100 200 300
Steady State [SO,*-] (jieq L-<)
Priority Class - A - E
Model « MAGIC
Deposition - Constant
300
0 100 200 300
Steady State [SCVl <*ieq L"0
Priority Class - A * I
Model - MAGIC
Deposition - Constant
300
10
0 100 200 300
Steady State ISO41 (^eq L-'»
Figure 10-83. Comparison of projected sulfate versus sulfate steady-state concentrations using
ETD, ILWAS, and MAGIC for NE takes.
10-175
-------
NE Lakes
Priority Class - A & B
Model - MAGIC
Deposition - Ramp 30% Decrease
300
o
1C
„ 200
«8
I-
8
0 100 200 300
Steady State [SCV3 (neq L-i)
NE Lakes
Priority Class • A & B
Model *• ETO
Deposition - Ramp 30% Decrease
300
2
o
U7
_ soo
100
o'
«
0 100 200
Steady Slate [SCVl
300
L-')
Priority Class - A & B
Model - ILWAS
Deposition • Ramp 30% Decrease
300
2
e
to
J. wo
*"
0 WO 200 300
Steady State [SO^J (jieq L-i)
Priority Class - A - E
Model = ETD
Deposition - Ramp 30% Decrease
300
0 100 200 300
Steady State [SO4*1 (|ieq L-i)
Priority Class - A - E
Model - MAGIC
Deposition • Ramp 30% Decrease
soo
o
u>
„ 200
wo
Priority Class - A - I
Model * MAGIC
Deposition - Ramp 30% Decrease
SOOr
0 100 200 SOO
Steady State (SO4*J (jteq L*0
0 WO 200 300
Steady State [SO,"] (neq L-i)
Figure 10-84. Comparison of projected sulfate concentrations under decreased deposition with the
current sulfate steady-state concentrations using ETD, ILWAS, and MAGIC for NE lakes.
10-176
-------
NE Lakes
Priority Class - A - E
Deposition • Constant
Calculated [SO«*1 at SO Years
NE Lakes
Priority Class - A - E
Deposition - Ramp 30% Decrease
Calculated [SO,*j at SO Years
300r
TMO
'0 SO WO 150 200 230 300
ETD ISO4*-] (jieq L-i)
'0 SO 100 ISO 200 250 300
ETD tSO4*1 dieq L-I)
Priority Class - A & B
Deposition « Constant
Calculated IS
-------
10.11.3.2 Southern Blue Ridge Province Projections of Sulfate Steady State
Projections of sulfate steady state using MAGIC indicate sulfate steady state might be reached in
the SBRP within 200 years under current and increased deposition (Figure 10-86). The 1:1 line on the
figure indicates agreement between the projected and steady-state sulfate concentrations under current
deposition, consistent with the projections of watershed sulfur retention presented in Table 10-19. The
relationship between projected sulfate concentrations assuming a 20 percent Increase In sulfate
deposition indicates these sulfate concentrations lie above the 1:1 line for current deposition, because of
the increased sulfate loading and greater sulfate steady-state concentrations. The estimated time to
sulfate steady state in the SBRP is about 200 years, compared to less than 50 years in the NE.
10.11.3.3 ANC and Base Cation Dynamics -
All three models projected changes in ANC, sulfate, and pH. Only ILWAS and MAGIC, however,
projected changes in base cations. Relationships between changes in ANC and sulfate concentrations
and between changes in pH, calcium plus magnesium, and sulfate concentrations are examined in the
following sections.
10.11.3.3.1 Northeast-
Comparisons of projected ANC concentrations among models for northeastern watersheds after SO
years are shown in Figure 10-87. The 1:1 line indicates excellent agreement among model projections.
The comparisons for ILWAS contain only about 25 data points so the relationships are not as apparent.
The changes in ANC concentrations as functions of change in sulfate concentrations are shown
in Figure 10-88 for all three models. For current deposition, the relationships are not apparent because
the changes in ANC and sulfate concentrations were projected to be quite small. A negative trend with
decreased deposition is apparent for MAGIC and ILWAS because of greater changes in ANC and sulfate
concentrations. Given the uncertainty in the projections, however, the indicated trend Is not significant.
10-178
-------
300
o
o
200
cr
o 100
O
CO
SBHP Stream Reaches
Model = MAGIC
Deposition = Constant
Year = 100
0 100 200
Steady State ISO,*]
300
SBRP Stream Reaches
Model = MAGIC
Deposition = Ramp 20% Increase
Year » 100
300
2
o
o
200
I100
IT
6
CO
0 100 200 300
Steady State [SO/-] (jieq L-<)
300
M
O
O
-------
400
T 300
er
1
o
(9
S
200
100
-too
NE Lakes
Priority Class = A - E
Deposition * Constant
ANC at 50 Years
•100 0 WO 200 300
ETD ANC (|ieq L-')
400
NE Lakes
Priority Class = A • E
Deposition » Ramp 30% Decrease
ANC at 50 Years
400r
T SOD
*!00 0 100 200 300
ETD ANC (fieq L-i)
Priority Class - A & B
Deposition « Constant
ANC at 50 Years
•100
-100 0 100 200
ILWAS ANC
300
L-1)
Priority Class - A & 8
Deposition " Ramp 30% Decrease
ANC at 50 Years
400r
-MO 0 • 100 200 300 400
ILWAS ANC (peq L-»)
400,
-300
O
Z 100
Q
Ul 0
-WO
Priority Class - A & B
Deposition « Constant
ANC at 50 Years
-TOO 0 WO 300 300 400
ILWAS ANC
-------
20
15
o
-5
•IS
NE Lakes
Priority Class - A - E
Modal - MAGIC
Deposition - Constant
-SO -20 -tt 0 tt 20 10 40 SO
A ISO,*] (|ieq L-<)
NE Lakes
Priority Class = A • E
Model « MAGIC
Deposition - Ramp 30% Decrease
SO
25
20
a
a o
O
< o
< -s
•10
•is
-«0 -60 -40 -20
A [SO,*] <|ieq L-t)
20
r
^ -10
•15
-20
Priority Class - A - E
Model - ETD
Deposition • Constant
-30 -20 -« 0 10 20 30 40 50
A ISO,*! dieq L-1)
Priority Class - A - E
Model - ETD
Deposition - Ramp 30% Decrease
25
20
T
-J is
J*
»» s
.•*.
e°
•to
•CO -40 -20 0
A [S0,*1 Oteq L-t)
II
-15
•20
Priority Class - A & 8
Model - ILWAS
Deposition * Constant
-90 -W -» 0 W 20 SO 40 80
A ISO,*] Qieq LI>
Priority Class - A & B
Model - ILWAS
Deposition - Ramp 30% Decrease
25
c « a
o • «
<
< -s
-W -40 -20
A [S
-------
The pH - ANC relationship for each of the models is compared in Figure 10-89. There is good
agreement between the ETD and MAGIC relationships but greater scatter in the ILWAS pH - ANC
relationship. The ILWAS ANC - pH relationship is modified by seasonal changes in the pCO2 function
and organic acid production/decomposition. Comparisons of projected pH values between models are
shown in Rgure 10-90. There is greater scatter about the 1:1 line at the lower pH projections for
comparisons between all three model pairs with greater convergence on the 1:1 line at higher pH
values.
Comparison of changes in calcium plus magnesium concentrations as a function of changes in
sulfate concentrations is illustrated in Figure 10-91 for MAGIC and ILWAS. Minimal changes in calcium
plus magnesium and sulfate under current deposition resulted in a grouping of lakes in the upper
quadrant of the graph about the (0,0) point. The relationship between projected changes in calcium plus
magnesium and sulfate concentrations under decreased deposition, however, was relatively linear for both
MAGIC and ILWAS.
The projected rate of change for ANC and calcium plus magnesium, although small for the NE, Is
continuous and does not appear to asymptotically approach steady-state concentrations. This result is
illustrated by a plot of the median ANC and calcium plus magnesium concentrations over time for 100
years using the MAGIC results under both current and decreased deposition (Figure 10-92). For current
deposition, median ANC remains relatively constant for the first 20 years and then decreases. Median
calcium plus magnesium concentrations, however, decrease over the entire 100-year period, although the
rate of change slowly decreases. This rate of change is not significant considering the uncertainty in the
projections, but the trend is apparent.
The projected change in ANC assuming decreased deposition approaches a peak in SO years and
then declines slightly over the next 50 years (Figure 10-92). The projected change in calcium plus
magnesium decreases more rapidly over the 100-year period with decreased deposition than under
current deposition. The change over the last 50 years was less than during the first 50 years.
10-182
-------
8.0
7.5
7.0
6.5
1.6-0
5.5
5.0
4.5
4.0
NE Lakes
Ail Models
pH vs. ANC
"o&o'o
o MAGIC
v ETD
• ILWAS
-100 0 100 200 300 400
ANC
8.0
7.5
7.0
6.5
9
5.5
5.0
4.5
SBRP Stream Reaches
All Models
pH vs. ANC
-100
o MAGIC
• ILWAS
0 100
ANC
200 300
L-1)
400
Figure 10-89. Comparison of pH - ANC relationship for each of the models.
10-183
-------
NE Lakes
Deposition - Constant
Model pH at SO Years
NE Lakes
Deposition - Ramp 30% Decrease
Model pH at SO Years
SJ) U 10 tJS 7.0 7£
ETD pH
Deposition • Constant
Deposition - Ramp 30% Decrease
U AA U
B.WAS pK
6JD U 7-0 7JS
ILWAS pH
Deposition - Constant
Deposition - Ramp 30% Decrease
7A-
4J> 4J U> U «JB
LWAS pH
Figure 10-90. Comparison of projected pH values between models for NE lakes after 50 years
under current and decreased deposition.
10-184
-------
NE Lakes
Priority Class ° A - E
Model = MAGIC
Deposition = Constant
1U
T* o
S"-10
T
O»
2-30
T
|-40
"-50
-an
o
.$
-80 -60 -40 -20 0
A [SO,*! (ueq L*0
20
NE Lakes
Priority Class = A - E
Model = MAGIC
Deposition = Ramp 30% Decrease
10r
CT
O -10
S-20
&
O
2-30
V
o
-so
-60
0
o
-80 -60 -40 -20 0
A [SO*2-] (jieq L-1)
20
10
flf-10
-30
o
-50
-60
Priority Class • A & B
Model ° ILWAS
Deposition ° Constant
-80 -«0 -40 -20
A ISO4»-] (fieq
20
Priority Class - A & B
Model « ILWAS
Deposition » Ramp 30% Decrease
5 -10
£5 -20
*i»
2^-30
%-«
O
-60
-BO -60 -40 -20 0
A [SO4*] (}ieq L-I)
20
Figure 10-91. Comparison of projected changes in calcium and magnesium versus changes In
suifate using ILWAS and MAGIC for NE lakes.
10-185
-------
138
1
iao
,NE Lakes
Priority Class = A - I
Model - MAGIC
Deposition - Constant
0 W80W40SOW70W90WO
Simulation Year
ME Lakes
Priority Class « A - I
Model <• MAGIC
Deposition - Ramp 30% Decrease
Simulation Year
Deposition • Constant
200
s
TIM
170
0 1020304OSOCOTDaOMtOO
Simulation Year
Deposition - Ramp 30% Decrease
200r
1
I"
.
o
•no
0 10 20 SO 40 M «0 TO 80 M WO
Simulation Year
Deposition - Constant
120
•?tio
M
.M
i
! ro
W
Deposition • flamp 30% Decrease
•ttOr
Simulation Year
0 M20S040$0«070«OM100
Simulation Year
Figure 10-92. Change In median ANC, calcium and magnesium, and sulfate concentrations
projected for NE lakes using MAGIC under current and decreased deposition.
10-186
-------
The median sulfate concentration decreases from year 0 and asymptotically approaches a steady-
state value by year 50 under current deposition. Median sulfate concentrations were projected to
decrease linearly by 22 /ieq L*1 by year 30 and asymptotically approach steady state by year 50 under
decreased deposition.
Median pH is projected to change less than 0.1 unit over 100 years regardless of the deposition
scenario. However, if the change in pH is compared with the original calibrated pH at year 0, all three
models indicate the greatest change in pH occurs in lakes with initial pH values between about 5.0 and
6.5 (Figure 10-93). Under current deposition, those lakes with pH values between 5.0 and 6.0 were
projected using ILWAS and MAGIC to have the greatest decrease in pH. The ETD model also projected
these lakes might experience the greatest change, but in both the positive and negative directions. Under
decreased deposition, all three models projected lakes with initial pH values between 5.0 and 6.0 might
have a net Increase in pH of from 0.1 to 1.0 pH units.
10.11.3.3.2 Southern Blue Ridge Province -
Comparisons of model-projected ANC, sulfate concentration, and pH for S8RP watersheds after 50
years are shown in Figure 10-94. The 1:1 line Indicates an apparent relationship among model
projections, but there are relatively few points for inter-model comparison as well as considerable scatter
about the 1:1 line.
The changes in ANC as functions of change in sulfate concentrations are shown in Figure 10-95
for both ILWAS and MAGIC. A similar figure (Figure 10-96) illustrates the MAGIC projections for all 32
streams simulated In the SBRP, not just the comparable 14 ILWAS watersheds. The projected changes
in ANC concentrations were negatively correlated with the projected changes in sulfate concentrations
for both current and increased deposition. The relationships between the change in ANC and change
in sulfate, computed using a weighted regression for MAGIC were AANC = -2.8 - 0.372 ASO42' (r2 =
0.28) for current deposition and AANC = -1.05 - 0.441 ASO42" (r2 = 0.42) for increased deposition. The
changes in calcium plus magnesium concentrations as functions of change in sulfate concentrations are
10-187
-------
NE Lakes
Model - MAGIC
Deposition - Constant
. 0.4
0.0
-0.4
-04
*J> *J 5* U •£ «£ 7JB 7*
Year 0 Model pH
NE Lakes
Model - MAGIC
Deposition » Ramp 30% Decrease
12
>
4.0 4Jt U IS W» U TJt 74
Year 0 Model pH
NE Lakes
Model a ETD
Deposition - Ramp 30% Decrease
5.
•^
0.0
-0.4
-u
4J &0 SJS «LO (W T.O 7J
Year 0 Model pH
OJ
, 0:4
0.0
-0.4
-OJ
NE Lakes
Model - ILWAS
Deposition - Constant
ota
4JB 44 6J) U U *8 7J>
Year 0 Model pH
NE Lakes
Model • ILWAS
Deposition • Ramp 30% Decrease
Ur o
OJ
04
•0.4
-OS
M M «J) «.* 7.0 7.S
Year 0 Model pH
Figure 10-93. Comparison of the change in pH after 50 years as a function of the initial calibrated
pH for MAGIC, ETD and ILWAS on northeastern lakes.
10-188
-------
400
T. 300
3-200
< wo
2
o
< 0
-too
SBRP Stream Reaches
Priority Class - A & B
ANC at SO Years
Deposition - Constant
SBRP Stream Reaches
Priority Class - A & B
ANC at 50 Years
Deposition - Ramp 20% Increase
400r
-tOO 0 MO 200 300 400
ILWAS ANC (|ieq LI)
-WO1'—
-104 0 tOO 200 300 400
ILWAS ANC (jieq L-')
300
^250
o-
§200
i-
(3
SO
ISO4»-I at SO Years
Deposition « Constant
0 SO WO tSO ZOO 250 400
ILWAS ISO.*] (|«q L-n
ISO.1-] at SO Years
Deposition « Ramp 20% Increase
300
0 SO MO ISO 200 250 300
ILWAS ISO.1
1.0
«•
"
u
4.6
4J>
pH at SO Years
Deposition - Constant
H
ph at 50 Years
Deposition • Ramp 20% Increase
4J
U U SJ) 0.5 7.0 7J IdO
ILWAS pH
Figure 10-94. Comparisons of projected ANC and suJfate concentrations and pH between ILWAS
and MAGIC after so years for SBRP streams.
10-189
-------
SBRP Stream Reaches
Priority Class • A & B
Model = MAGIC
Deposition = Constant
0 20 40 60 80 100
A ISO,*] (jieq L-t)
SBRP Stream Reaches
Priority Class = A & B
Model = MAGIC
Deposition = Ramp 20% Increase
10
0
ZJ-10
«•»
O^
-3- -20
o
<-30
<
-40
-so
10
o
-: °: =
o T
r -1 -10
o
*«>> o o ° 3
.3-20
0 < -30
0 •<
-40
•
o
' ° * 0
0 °
0
o
o
0 0°
o
-
o
1 1 1 1 I
0 20 40 60 80
A [SO*2"] (jieq L-1)
100
Priority Class = A & B
Model - ILWAS
Deposition o Constant
20 40 60 80
A ISO,*] (jieq L-i)
100
Priority Class - A & B
Model = ILWAS
Deposition = Ramp 20% Increase
10
0
II -10
cr
o
<-30
•40
-so
10
*
°o 0
0
• . «a%° 0o» ° ^'10
••••:. g-
3-20
0 0
<-30
-40
o
I 1 1 1 1— — i -SO
•
0
0
0
0 o o
0 0
o
0
•
1 1 I 1 i
20 40 60 80
A ISO4*1 (|ieq
100
Rgure 10-95. Comparison of projected AANC and Asulfate relationships In SBRP Priority Class A
and B streams using ILWAS and MAGIC.
10-190
-------
SBRP Stream Reaches
Priority Class » A - E
Model = MAGIC
Deposition = Constant
10
0
c-
-J -10
tr
o
-5-20
O
<-30
t
I -10
r
i
t-20
I
•
;-30
I
-40
-50
20 40 60 80 100
A [SO4*-] (|ieq L-i)
70
_. SO
o-
o
A 30
10
•^pp
I
'a
O
-10
-30
-50
Priority Class • A - E
Model - MAGIC
Deposition = Constant
0
o
00°
O g^"
0°°°
o 8
o «
0 20 40 60 80
A [S04a'J (ueq L-1)
100
Priority Class = A - E
Model - MAGIC
Deposition = Ramp 20% Increase
70r
o
50!
Q>
-30
«
O
"-30
-SO
00
0
o &
00 0
00%
o o
00
0 20 40 60 80 100
A [SO4*] (jieq L-i)
Figure 10-96. Comparison of projected AANC and Asulfate relationships and A(calcium and
magnesium) and Asulfate relationships for SBRP Priority Class A - E streams using MAGIC.
10-191
-------
shown in Figure 10-97 for both the ILWAS and MAGIC results under current and increased deposition.
Linear regression models assume no error in the independent variable with all the error assumed for the
dependent variable. Therefore, a structural regression model is required to compute the slope of the
regression line because a structural regression model accounts for error in both the independent and
dependent variables. The structural regression model, however, requires additional analyses, which are
ongoing and are expected to be available in September 1989. Computing the slope of the relationship
of calcium and magnesium versus sulfate using linear regression to estimate an "P factor (Henrickson,
1982), is not recommended.
»
Median ANC concentrations were relatively constant for the first 20 years under current deposition,
and then were projected to decrease linearly over the remainder of the 200-year period (Figure 10-98).
The rate of ANC decrease from year 10 to year 100 was greater under increased deposition (i.e., -0.47
peq L"1 yr"1) than for current deposition (i.e., -0.28 Meq L"1 yr"1). From year 100 to year 200, however,
the rate of change in ANC was similar for both deposition scenarios.
Median calcium plus magnesium concentrations were projected to increase until about year 40 and
then decrease for the rest of the simulation for both deposition scenarios (Figure 10-98). The rates of
change in median calcium plus magnesium concentrations from year 50 to year 100 for current and
increased deposition were -0.22 and -0.28 jueq L'1 yr"1, respectively. The rates of change in calcium plus
magnesium concentrations from year 100 to year 200 were -0.03 /ieq L'1 yr"1 for both current and
increased deposition.
Median sulfate concentrations were projected to increase at rates of 0.76 Meq I-"1 y~1 for the first
50 years, 0.43 jiteq L"1 yr"1 from year 50 to year 100, and 0.14 jueq L"1 yr"1 from year 100 to year 200
under current deposition (Figure 10-98). Under increased deposition, the projected rates of change in
median sulfate concentrations were 1.1 Meq L'1 yr"1 for the first 50 years, 0.66 /jeq L'1 yr"1 from year 50
10-192
-------
SBRP Stream Reaches
Priority Class = A & B
Model - MAGIC
Deposition = Constant
70
ll SO
^
a
30
y
F°
"-30
' -SO
o° o
O o o
0 6 „
o o
0 20 40 60 80 100
A [SO,4"] {jieq L-I)
SBRP Stream Reaches
Priority Class = A & B
Model m MAGIC
Deposition = Ramp 20% Increase
70
j so
cr
a
^30
"
co
o
-50
.00
o o
20 40 60 80
A [SCv8-] {yeq L-I)
100
ro
1 so
cr
o>
^.30
A
O
-50
Priority Class = A & B
Model « ILWAS
Deposition = Constant
00
0
00
~o
o o
0 20 40 SO 80
A IS
-------
SBRP Stream Reaches
Priority Class - A - E
Model - MAGIC
Deposition - Constant
tso
s
JVM
575
so
50 WO ISO
Simulation Year
200
SBRP Stream Reaches
Priority Class - A - E
Model - MAGIC
Deposition - Ramp 20% Increase
ISOr
100
Simulation Year
Deposition • Constant
Tno
B
50 WO 150
Simulation Year
200
Deposition • Ramp 20% Increase
ttOr
50 100 ISO
Simulation Year
Deposition - Constant
so MO iso
Simulation Year
200
Deposition - Ramp 20% Increase
ttOr
80 100 1SO
Simulation Year
200
Figure 10*98. Change in median ANC, calcium and magnesium, and sulfate concentrations
projected for SBRP streams under current and increased deposition using MAGIC.
10-194
-------
to year 100, and 0.09 /ieq L"1 yr"1 from year 100 to year 200. Both deposition scenarios resulted in
median sulfate concentrations near sulfate steady state after 200 years.
About 44 percent (669 streams) of the DORP streams in the SBRP had pH values below 7.0 with
17 percent (262 streams) having pH less than 6.75. Comparing the projected change in pH versus the
initial pH at year 0, however, indicates that streams with initial pH values less than 6.75 might decrease
between -0.5 and -1.0 units within 50 years under current deposition and might have greater than -1.0
unit decrease under increased deposition (Figure 10-99). By year 200, streams with pH less than 7.0
might experience pH decreases between -0.25 and -0.5 under increased deposition with some streams
projected to have greater than a -2.0 unit decrease.
10.12 DISCUSSION
10.12.1 Future Projections of Chances in Acid-Base Surface Water Chemistry
The Level III Analyses used typical year deposition scenarios to examine the potential effects of
alternative deposition levels on future changes in surface water chemistry. The typical year, as discussed
in Section 5.6, represents the average meteorology for a 30-year period of record and the average
deposition for a 3- to 7-year period of record adjusted for the average meteorological year. Deposition
was then estimated for each of the watersheds considered in the Level III Analyses. The typical year
scenario enabled each modelling group to use the same input and provided a common basis for
comparing changes in surface water chemistry as functions of comparable deposition among all the
models. The intent was not to forecast future meteorological or deposition conditions, but rather to have
a common basis for comparison among model results. Comparable watershed morphometry, physical
and chemical soils data, and surface water chemistry data also were provided to each of the modelling
groups, enabling them to assess and contrast the different model formulations and projections. These
models integrate much of our knowledge on how watershed processes control surface water acidification,
and comparing the output from these models, in part, provides a test of how well we understand these
processes. There are different hypotheses on how these processes operate and different philosophies on
how to integrate this information in the models (Eary et al., I989; Jenne et al., 1989). These results are
10-195
-------
SBHP Stream Reaches
Modal - MAGIC
Deposition - Constant
Year - 50
to
O.S
0.0
.-03
1 -to
as
-z,o
«J> U 7.0 7J OLD
Simulation Year 0 pH
SBRP Stream Reaches
Model - MAGIC
Deposition - Ramp 20% Increase
Year - 50
to
0.5
0.0
""
-ts
-20
'"«
U 7J> 7.5
Simulation Year 0 pH
Deposition » Constant
Year - 100
lOr
0.5
0.0
t
-to
•ts
-s.o
•*•&
Simulation Year 0 pH
Deposition B Ramp 20% Increase
Year - 100
to
OS
. -O.S
1 -u
-IS
-2.0 >
Simulation Year 0 pH
Deposition • Constant
Year - 200
to-
0.0
•&S
-10
-ts
-8.0
74
Simulation Year 0 pH
Deposition • Ramp 20% Increase
Year - 200
to
OS
0.0
.-OS
'-to
-ts
-to
o o o
"fc
U 7JO 73 1.0
Simulation Year 0 pH
Figure 10-99. Comparison of the change In pH after 200 years as a function of the Initial calibrated
pH for MAGIC on SBRP streams, Priority Classes A - E.
10-196
-------
not intended, and should not be interpreted, as forecasts of conditions that might be expected over the
next 50 to 200 years.
10.12.2 Rate of Future Change
The Panel on Processes of Lake Acidification raised questions on the extent of surface water
acidification, the processes that control changes in surface water chemistry (including surface water
acidification and chemical improvement), and the rate at which these processes occur. The extent of
acidic and low ANC surface waters in the United States was addressed through the NSWS. The
processes that control changes In surface water chemistry were discussed in Section 3 and summarized
In Galloway et al. (I983a), Church and Turner (1986), Reuss and Johnson (1986), and Martin (1986).
The DORP was Initiated because scientists did not concur on how watershed processes control the
rate and magnitude of surface water acidification and how to project such changes in surface water
chemistry. A primary area of disagreement among scientists on the Panel was whether future ANC
decreases would be gradual over a period of centuries or perceptible over years to decades, i.e., they
disagreed about the rate at which acidification might occur. The rates at which changes in sulfate
adsorption and base cation supply and surface water acidification and chemical improvement might occur
in northeastern lakes and SBRP streams are discussed below.
10.12.2.1 Northeast
Changes that might occur in the NE over the next 100 years (summarized in Figure 10-92) are
consistent with various conceptual models of surface water acidification (Galloway et al., I983a; NAS,
1984; Church and Turner, 1986; Cosby et al., 1985a,b,c; Reuss and Johnson, 1986).
Sulfate deposition in the NE has declined since the 1970s concurrent with declining sulfur emissions
in the NE (OTA, 1984; Kulp, 1987). The decline in sulfate concentrations at the start of the projections
for the NE under current deposition (Figure 10-92) reflects this deposition decrease as the watersheds
approach a sulfate steady-state concentration that is lower than it was in the 1970s. The relatively
10-197
-------
constant ANC concentrations under current deposition for the first 20 years of the projection occurred
primarily because the decline in sulfate concentrations of about 8 neq L"1 was compensated by a decline
of about 8 /*eq L"1 in calcium plus magnesium concentrations. Sulfate concentrations asymptotically
approached steady state after 20 years, changing by about 2 to 3 fj,eq L."1 over the next 80 years. A
continual depletion of about 8 jueq L*1 in base cation concentrations (calcium plus magnesium) was
projected during this 80-year period as sulfate approached steady state, however, which resulted in the
continual decline In ANC of about 4 /xeq L*1 over this same 80-year period. These results are consistent
with observations made In Plastic Lake, Ontario, Canada where ANC concentrations continued to
decrease following a reduction in sulfate deposition even though sulfate concentrations remained relatively
constant In the lake (Dillon et al., 1987). The ANC decrease in Plastic Lake was attributed to depletion
of the available pool of base cations in the watershed (Dillon et al., 1987), although no soil measurements
were made. A depletion of the pod of available soil base cations was projected for the northeastern
watersheds using both ILWAS and MAGIC and resulted in similar ANC decreases in the northeastern
lakes.
All three models projected that northeastern watersheds might be at or near sulfate steady state
within 50 years assuming either current or decreased deposition. All three models projected decreased
ANC concentrations over 50 years and that additional lakes might become acidic, because of the slow
but continual decrease in base cation and ANC concentrations. The lakes currently not acidic that might
become acidic over the next 50 to 100 years represented about 3 percent of the 3227 lakes in the MAGIC
target population. When compared with the ELS-I target population of 7157 (many of which have ANC
concentrations exceeding 400 /ieq L'1), this additional percentage of acidic lakes represents less than
1 percent of the population. The ELS-I target population, however, included only lakes larger than 4 ha.
Ongoing analyses of small lakes indicates that the ratio of smaller acidic lakes (< 4 ha) to acidic lakes
larger than 4 ha is about 2:1 (Sullivan et al., submitted). Considering these small lakes might Increase the
projected percentage of acidic lakes over the next 50 years to 2 percent. The models, however, support
10-198
-------
the hypothesis that future ANC decreases in the NE will be gradual over a period of decades to centuries
rather than years to decades.
Following the 30 percent decrease in sulfate deposition beginning in year 10, there was a rapid
increase In projected ANC over the next 40 years (Figure 10-92). This 11 peq L"1 increase in ANC
occurred because the concurrent projected decrease in median sulfate concentrations of about 22 ^eq
L'1 occurred with a projected decrease in median base cation concentrations (calcium plus magnesium)
of about 11 Meq L'1. This rapid increase in ANC probably occurred because the watersheds were initially
near sulfate steady state. A rapid increase in ANC might not be expected if the systems are not at or
near sulfate steady state (Cosby et al., 1985a,b,c). All three models projected this rapid increase in ANC
following the 30 percent decrease In sulfate deposition. Even though the watersheds were nearly at
sulfate steady state within 50 years under decreased deposition, there was a continued decrease in base
cations projected from year 50 to year 100, which resulted in a small but continued decrease in ANC
concentrations.
Although there was no apparent relationship between the rates of change in ANC and suffate and
the initial ELS-I ANC concentration, the projected rate of change under current deposition in the NE was
small. If the majority of the watersheds are near sulfate steady state, then most of these systems might
be expected to respond relatively quickly to changes in sulfate concentration regardless of the initial ANC.
Projections for all three models indicated that as many as 125 currently acidic lakes might
chemically improve (increase in ANC) in 50 years assuming a 30 percent deposition decrease. This
estimate represents about 77 percent of the estimated 162 currently acidic DDRP target population lakes,
but only about 4 percent of the 3277 lakes in the MAGIC target population. The number of lakes
estimated to chemically improve was moderated by the continued decrease in ANC from year 50 to year
100: after year 100, the estimated number had decreased to 113 (70 percent).
10-199
-------
Differences among model projections were more apparent for Priority Class A and B lakes for three
reasons. First, the sample size for this priority class is small and are available for comparison. Second,
this priority class includes low ANC systems, which have the greatest variability in terms both of ANC
measurements (Unthurst et al., 1986a) and model calibration. The ILWAS and MAGIC models are
calibrated on base cations and acid anions and ANC is a computed, not a calibrated value (I.e., ANC =
sum base cations - sum acid anions). ELS-I field measurements for many lakes indicate cation or anion
deficits that reflect the accepted sampling and measurement error in the analysis. The models, however,
require charge balance so the calculated ANC concentrations following calibration might not equal the
measured ANC in the lake or stream. This difference between calibrated and measured ANC values for
the models was generally greatest at the low ANC concentrations where the relative measurement errors
also are greater. The differences between models, however, are well within the uncertainty bounds about
the projections. Third, MAGIC performs hindcasts as part of its calibration/projection exercise and, thus,
simulates the declining sulfur deposition levels over the past 10 years. These declining sulfur deposition
levels continue to exhibit a cumulative effect over the first 10-20 years of the projections. ILWAS and ETD
assume historically deposition values are the same as current deposition values and calibrate to them,
which also contributes to the differences among models.
The change in pH projected using MAGIC might be underestimated because the initial or calibrated
ANC concentrations at year 0 were greater than the ELS-I ANC concentrations. Because of the pH - ANC
relationship, the unit change in pH for each unit change in ANC decreases as the ANC increases (i.e.,
at higher ANC concentrations, pH changes are less). To assess this possible underestimate in pH
change, the change in ANC projected using MAGIC was added to the ELS-I ANC, and a derived pH was
estimated using the pH - ANC relationship incorporated in MAGIC (Figure 10-100). The change in the
derived pH is similar to that in the modelled pH for current deposition, although the maximum change
is greater. Under decreased deposition, the estimated change in pH is greater with the derived rather
than modelled pH values, but only for a few lakes (Figure 10-100). Because the changes in ANC are
both small and not influenced by the initial ANC, the change in pH does not appear to be greatly
underestimated.
10-200
-------
0.8
CO
o
in 0.0
-0.4
-0.8
NE Lakes
Model = MAGIC
Deposition = Constant
2.0
1.6
.
o Model pH
& Derived pH
04
4 *V Q V ^Q
* 6 A a
/>*«
'**•
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5
Simulation Year 0 pH
NE Lakes
Model = MAGIC
Deposition = Ramp 30% Decrease
2.0
1.6
I1"2
0
A
Model pH
Derived pH
-
0.8
c
U5 0.0
-0.4
-0.8
4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5
Simulation Year 0 pH
Figure 10-100. Comparison of projected MAGIC change in pH versus derived pH after 50 years
for NE lakes.
10-201
-------
10.12.1.2. Southern Blue Ridge Province
Projected changes in surface water chemistry that might occur in the SBRP were shown in Figure
10-98. ILWAS and MAGIC projected similar changes in ANC, calcium plus magnesium, and suifate over
the 50-year period. MAGIC projections for the SBRP, however, suggest that substantial changes also
occur between 50 and 200 years. This discussion, therefore, focuses on the MAGIC projections.
For the first 50 years, both models projected a decrease in ANC and an increase in base cation
and suifate concentrations for both deposition scenarios. The decrease in ANC concentrations over the
first 20 years was slight, but a relatively constant linear decrease in ANC after 20 years was projected.
Under current deposition, suifate concentrations increased linearly for the first 50 years from about 37 to
75 /Lieq L"1 while base cations increased from about 110 to 123 /ueq L*1 by year 30. Increased suifate
concentrations were compensated by increased base cation transport from the watershed and relatively
little change in ANC for the first 20 to 30 years. However, when base cations began to decrease, ANC
concentrations also decreased from about 122 to 100 /ieq L*1 from year 20 to year 100, respectively.
Over the interval from year 30 to year 100, suifate concentrations increased by about 35 Jieq L"1, base
cations declined by about 15 jueq L*1, and ANC decreased by about 20 jueq I-'1, projections which are
consistent with charge balance requirements and current understanding of soil processes (Reuss and
Johnson, 1986; see Sections 3 and 9). Although the rates of suifate increase and base cation decrease
changed from year 100 to year 200 compared with year 50 to year 100, the ratio or relationship between
increased suifate concentrations and decreased base cations remained relatively constant because the
rate of change in ANC concentrations was relatively linear from year 30 to year 200. The SBRP
watersheds were asymptotically approaching suifate steady state by year 200, and median watershed
sulfur retention had declined to about 5 percent.
The two models differed in the projected number of streams that might become acidic within 50
years under current deposition. The ILWAS model projected no acidic streams while MAGIC projected
130 streams that might become acidic in 50 years assuming current deposition. The estimate of 130
10-202
-------
acidic stream reaches, however, is derived from differences In the projections for one SBRP stream with
a relatively large weight. This stream's ANC decreased from an initial concentration of about 20 /ieq L"
1 to 3 jueq L"1 within 50 years. Given the uncertainty in the projections, 130 is probably the maximum
estimated number of streams that might become acidic within 50 years. MAGIC projections also indicated
additional streams might become acidic over the next 200 years In the SBRP, with between 12 and 15
percent of the systems potentially becoming acidic by 100 years and 200 years, respectively, under
current deposition.
Changes in surface water chemistry projected for the SBRP under increased deposition showed
similar patterns to those projected under current deposition (Figure 10-98). The rate at which sulfate
asymptotically approached the steady-state concentrations with Increased deposition was greater than that
under current deposition because of the change in sulfate loading during the first 100 years. The rate
of increase in stream sulfate concentrations during the initial phase of approaching steady state is nearly
linear and becomes asymptotic as the soil solution sulfate concentration approaches the steady-state
sulfate concentration. Higher loadings with the 20 percent increased sulfur deposition scenario resulted
in the SBRP soils approaching the new sulfate steady state more quickly on the linear portion of the
curve. The rate of change in sulfate from year 100 to year 200 under increased deposition was less than
under current deposition because the Increased loading over the first 100 years resulted in the watersheds
being nearer to sulfate steady state. This increased sulfate loading also resulted in greater base cation
depletion rates over the first 100 years. The rate of change in base cations from year 100 to year 200
under increased deposition was slightly greater than under current deposition. Because the rate of
change in sulfate under increased deposition was less and the rate of change in base cations was greater
than under current deposition, there was a slight decrease in the rate of change In ANC concentrations
from year 100 to year 200.
10-203
-------
The increased deposition and more rapid increase toward sulfate steady state resulted in a larger
number of streams that might become acidic by year 200. The estimated numbers of streams that might
become acidic by year 100 and year 200 were 159 (11 percent) and 337 (24 percent) streams.
The models also support the hypothesis that future ANC decreases in the SBRP will be gradual
over the period of decades to centuries rather than occur over years to decades. Streams in the SBRP
might experience a slow but steady decline in ANC over the next 200 years assuming constant or
increased deposition. The stream population in the SBRP typically had higher initial ANC concentration
than streams in other geographic regions of the Southeast. Thirty percent of the DDRP SBRP stream
population had ANC concentrations between 25 and 100 peq L"1, and 70 percent of the stream reaches
had ANC > 100 jueq L*1. Extrapolating results from the SBRP to the population of other streams in the
Southeast, therefore, might not be appropriate because the proportion of streams with lower initial ANC
concentrations in other southeastern regions is greater than in the SBRP (Kaufmann et ah, 1988). in
addition, the projected changes in pH in the SBRP stream accompanying these changes in ANC might
range from -0.5 to -1.0 over 50 years and up to -2.0 pH unit changes over 200 years. Other southeastern
streams with lower current ANC might exhibit even greater pH changes within 50 years than projected
for SBRP streams.
10.12.3 Uncertainties and Implications for Future Changes In Surface Water Acid-Base Chemistry
Uncertainty is defined as Intrinsic variability plus error. Intrinsic variability represents the natural
variability or noise in the systems that cannot be reduced. The components of error include
measurement error, sampling error, model structural error, prediction error, and population estimation error
(Beck, 1987). The uncertainty analyses conducted for the Level III models quantitatively estimated many
of these error components (although the total error was not partitioned into its respective components)
and incorporated this error in the confidence bounds around the model projections. Unknown or poorly
understood processes, however, are more difficult to estimate quantitatively but can be qualitatively
10-204
-------
discussed. The implications of these processes on estimates of future change in ANC and pH are listed
in Table 10-20.
10.12.3.1 Deposition Inputs
Analyses were performed to determine the effect of deposition input uncertainty on the model
projections (Section 10.10.2). These uncertainty estimates were used to establish confidence Intervals
about the model projections In Appendix C. Analyses indicated the Input uncertainty contributed about
half of the total uncertainty in the projections with the other half an'sing from parameter uncertainty.
Uncertainty in dry deposition, particularly of base cations, is certainly a major contributor to deposition
input uncertainty. The approaches used to estimate the deposition inputs, however, were reasonable,
based on Input from the deposition modelers, conversations with technical experts on dry and wet
deposition, analyses of existing data, and conventional theory. In part, underestimates or overestimates
in anion or cation deposition inputs are compensated by increasing or decreasing mineral weathering
rates, respectively, of the anion or cation species to match observed surface water chemistry. Watershed
exchange pools are tightly coupled with deposition inputs.
This tight coupling of declining base cation concentrations to declining surface water sulfate
concentrations was recently reported for Hubbard Brook (Driscoll et al., 1989b). Two mechanisms were
indicated that can contribute to this coupling: (1) atmospheric deposition of base cations and (2) release
of base cations from mineral weathering or watershed pools of exchangeable base cations (Driscoll et
al., 1989b).
For the Level III projections the typical year deposition/precipitation scenario was repeated each
year for 50 years, so annual atmospheric deposition was constant for the 50-year period (with daily
meteorological variations). For the 30 percent deposition decrease, only sulfate concentrations were
reduced in deposition with charge balance maintained by adjusting hydrogen ion concentration. Base
cation concentrations were not decreased in either deposition scenario. For these projections, surface
10-205
-------
Table 10-20. Effects of Critical Assumptions on Projected Rates of Change.
Assumptions Resulting in Under- Assumptions Resulting in Over-
Estimates of ANC and oH Changes Estimates of ANC and pH Change
1. Mineral weathering overestimated 1. Mineral weathering underestimated
2. Nitrate assimilation overestimated 2. Organic acids buffer surface water chemistry
3. Total sulfur deposition underestimated 3. Total sulfur deposition overestimated
4. Calibrated ANC greater than observed 4. Calibrated ANC less than observed
5. Watershed land use changed 5. Watershed land use changes
6. Episodic acidification of surface waters 6. Desorption Is not the reverse of adsorption-
hysteresis-related delays in change
7. Biotic uptake/assimilation reducing 7. Weathering and sulfate adsorption increased by
available base cation pool decreased soil pH
8. Effects at distribution extremes over-
smoothed through aggregation
10-206
-------
Table 10-20. Effects of Critical Assumptions on Projected Rates of Change.
Assumptions Resulting in Under- Assumptions Resulting in Over-
Estimates of ANC and pH Changes Estimates of ANC and oH Change
1. Mineral weathering overestimated 1. Mineral weathering underestimated
2. Nitrate assimilation overestimated 2. Organic acids buffer surface water chemistry
3. Total sulfur deposition underestimated 3. Total sulfur deposition overestimated
4. Calibrated ANC greater than observed 4. Calibrated ANC less than observed
5. Watershed land use changed 5. Watershed land use changes
6. Episodic acidification of surface waters 6. Desorptlon is not the reverse of adsorption -
hystersis-related delays in change
7. Blotic uptake/assimilation reducing 7. Weathering and sulfate adsorption increased by
available base cation pool decreased soil pH
8. Effects at distribution extremes over-
smoothed through aggregation
10-206
-------
water base cation concentrations were tightly coupled with sulfate concentrations through the depletion
of soil exchangeable base cations. Depletion of soil exchangeable base cations occurred because sulfate
moved through the watersheds as a mobile anion. Under decreased deposition, the reduction In sulfate
concentration was compensated by soil base cations and a subsequent increase in ANC. These patterns
were consistent with those observed at Hubbard Brook.
While the projected changes in surface water sulfate concentrations are consistent with the
depletion of watershed pools of base cations, these processes cannot be decoupled from atmospheric
processes in natural watersheds. Atmospheric deposition of base cations clearly is an important process
that must be investigated in assessing the effects of sulfate deposition on surface water chemistry. The
calibrated models used in the Level III Analyses represent an excellent opportunity for evaluating different
hypotheses related to atmospheric deposition and watershed processes. Simulation experimentation on
different hypotheses represents one of the most Important uses of watershed models.
The deposition inputs, indeed, might be highly Inaccurate. The intent, however, was not to forecast
but rather to project the effects of alternative sulfur deposition scenarios on future changes in surface
water acid-base chemistry. Additional analyses are being proposed as part of the 1990 NAPAP Integrated
Assessment but It is likely that this Issue will remain beyond 1990.
10.12.3.2 Watershed Processes
Each of the three models has different formulations and different data requirements. If the three
models provide similar projections for similar reasons, however, greater confidence can be placed in the
conclusions. Questions remain, however, as to whether the models incorporate the key watershed
processes affecting surface water acidification and how important the model formulations, operational
assumptions, and parameter selection are on the long-term projections.
The key watershed processes incorporated in each model were listed in Table 10-1 and are
discussed in detail in Eary et at. (1989) and Jenne et al. (1989). All three models focus on the effects
10-207
-------
of sulfur deposition on surface water acidification. Each model considers total deposition acidity,
including nitrate, but the nitrogen dynamic formulations included in each model, including ILWAS, are
rudimentary. Because most of eastern forested watersheds are nitrogen-limited (Likens et al., 1977;
Swank and Crossley, 1988), nitrogen inputs are effectively removed from the soil complex. Deposition
inputs of nitrate are about twice the ammonium inputs for the eastern United States (Kulp, 1987).
Although nitrification has an acidifying effect (Lee and Schnoor, 1988), nitrate assimilation has an alkalizing
effect (Lee and Schnoor, 1988). Nitrate concentrations are low in receiving lakes and streams, indicating
nitrate is not moving as a mobile anlon. Median nitrate concentrations measured during the ELS-I for
northeastern lakes were less than 1 /ieq L'1. Median nitrate concentrations for SBRP streams were about
10 j*eq L*1. This does not preclude soil acidification, however, because biotic processes might influence
surface water chemistry. The assumption that nitrogen is not a primary contributor to chronic surface
water acidification and, therefore, that nitrogen dynamics do not have to be explicitly modelled represents
a limitation of the models, rather than a short-coming in the DORP design. Nitrate also might be an
important component of episodic acidification. The DORP, however, is not addressing episodic
acidification.
Changes in soil pH might influence mineral weathering rates and sulfate adsorption capacities. Plot
experiments have indicated these processes can be affected by decreased soil solution pH. Although
these effects might occur, median soil solution pH were projected to change less than 0.1 units In the
NE and less than 0.2 units in the SBRP.
One of the operational assumptions of the Level III Analyses was that the relationship of organic
acids to other chemical species would remain constant. Krug and Frink (1983) hypothesized that
reversing surface water acidification by strong mineral acids could result in increased dissociation of
humic acids and mobility of organic acids and, therefore, return naturally acidic lakes to their original
state. The historical acidic status of the currently acidic lakes is unknown, so the estimated chemical
improvement of the 125 currently acidic lakes might be liberal. Historical reconstruction of water
10-208
-------
chemistry for Adirondack Lakes should be available in the fall of 1989 and might be compared with the
DDRP projections of chemical improvement for the same lakes.
Mineral weathering is critical for all long-term projections, but is the process about which little
information can be obtained. The mineral weathering parameters are calibration parameters but are not
completely unconstrained. The range over which these parameters can vary while maintaining reasonable
ranges for other, better characterized parameters (e.g., selectivity coefficients) and still match observed
surface water chemistry constituent concentrations (e.g., silica, calcium, sodium, and other base cation
concentrations) is bounded. All three models yield similar long-term projections, even though ETD and
ILWAS use a fractional order weathering formulation based on hydrogen ion and MAGIC uses a zero
order weathering formulation. Long-term projections, however, are sensitive to the mineral weathering
parameters in all three models. The sensitivity of the MAGIC and ETD models to changes in the mineral
weathering parameters was identified in Table 10-10. Although mineral weathering rates cannot be
unequivocally estimated, the model formulations and mass balance approaches used in the models might
be analogous to the mass balance approaches used to estimate weathering in watershed studies (Velbel,
1986b; Paces 1973).
Data aggregation might result in underestimates of change in the tails or extremes of the
distributions. Soil horizon physical and chemical attributes are averaged (weighted) to Master horizons,
Master horizons aggregated to sampling classes, and sampling class attributes aggregated to the
watershed values, which are used for model calibration. This averaging or aggregation process will
preserve the central tendency in watershed attribute, and subsequent projected effects, but will reduce
the variability or extremes in the distribution of soil horizons through watershed attributes. While these
extremes represent a small proportion of the target population, the changes in these watersheds might
be underestimated so the changes in ANC or pH might be greater than projected.
Although data are not available for model confirmations of long-term projections, short-term
calibration and confirmation studies on Woods Lake, Panther Lake, and Clear Pond indicate the RMSEs
10-209
-------
among the models and the observed standard errors of the data were similar. Identical data were
provided to each of the modelling groups in performing the projections; a consistent, methodological
approach was used for the sensitivity analyses and the long-term projections; and uncertainty analyses
were performed for all three models. The rates of change for different constituents were comparable
among models and the processes controlling changes in surface water chemistry under different
deposition scenarios and among regions were similar among and between models. Even through there
are differences In model structure, process formulations, and temporal and spatial scales, the model
projections were remarkably similar. Regardless, long-term projections can be confirmed only with long-
term periods of record (Simons and Lam, 1980), which do not exist. Moreover, this study does not
establish the adequacy of the formulations representing important watershed processes, the procedures
for spatial aggregation of data, or the calibration approaches for long-term acidification projections.
10.13 CONCLUSIONS FROM LEVEL III ANALYSES
Conclusions from the Level III Analyses follow:
* All three models produced comparable results for the northeastern watersheds. ILWAS and
MAGIC produced comparable but more variable results for the SBRP.
• All three models projected minimal changes in ANC and sulfate concentrations and pH for
lakes in the NE over the next 50 years at current deposition rates. The median changes in
ANC, sulfate, and pH over the next 50 years were -1 to -5 /ieq L , <0.1 pH units, -0.1 to -
5 jiteq L"1, respectively, each of which is within the projection error of the respective analyses.
• ETD and MAGIC projected about 3 percent and 5 percent, respectively, of the lakes in Priority
Classes A - E that are currently not acidic might become acidic within 50 years at current
deposition and 2 and 3 percent, respectively, at decreased deposition. ETD estimated about
22 and 46 percent of the currently acidic lakes in Priority Classes A - E might chemically
Improve (i.e., increase in ANC) in 50 years for current and decreased deposition, respectively.
MAGIC estimated about 39 percent and 77 percent, respectively, of the currently acidic lakes
might improve in 50 years for the entire target population.
• All three models projected reduced lake sulfate and increased ANC concentrations and pH
with a 30 percent reduction in deposition. The median changes in sulfate, ANC, and pH,
respectively, were -23 to -28 /ieq L , +6 to +10 /ieq L , and 0 to +0.5 pH units over 50
years.
• MAGIC and ILWAS projections of changes in ANC, sulfate concentrations, and pH for SBRP
streams over 50 years were similar but there was considerable scatter in the comparisons
because of the small sample size.
10-210
-------
For current deposition, MAGIC projections in the SBRP indicated the change in median sulfate
after 50, 100, and 200 years was 38, 60, and 74 ^eq L , respectively. The changes in
median ANC after 50, 100, and 200 years were -11, -23, and -46 peq L , respectively. The
median percent sulfur retention at 0 years and after 50, 100, and 200 years was 65 percent
and 27 percent, 15 percent, and 6 percent, respectively. The changes in median pH after
50, 100, and 200 years were -0.04, -0.09 and -0.20, respectively.
• The percentage of SBRP stream reaches that might become acidic after 50, 100, and 200
years was < 9, 11, and 14 percent, respectively, for current deposition and 11, 11, and 24
percent for Increased deposition.
• With a 20 percent increase in deposition, MAGIC projections for the SBRP indicated the
1 changes in median sulfate concentrations after 50,100, and 200 years, respectively, were 55,
87, and 96 peq L . The changes in median ANC after 50, 100, and 200 years, respectively,
T
were -19, -41, and -64 /ieq L . The changes in median pH after 50, 100, and 200 years,
respectively, were -0.07, -0.12, and -0.32.
Based on the Level III projections, lakes in the NE might not change significantly over the
next 50 years with current deposition.
Acidic lakes in the NE might improve chemically with a 30 percent reduction in deposition
assuming organic acid relationships with other chemical constituents remain constant,
although some lakes might continue to acidify.
Streams in the SBRP might experience a slow but steady decline in ANC and a linear
increase in sulfate concentration over the next 50 years assuming current or increased
deposition. About 10 percent of the SBRP streams might become acidic within 50 years.
The stream population in the SBRP typically had higher initial ANCs than streams in other
geographic regions of the Southeast. Thirty percent of the population had ANC
concentrations between 25 and 100 ueq L , and 70 percent of the stream reaches had ANC
en in
population of other streams in the Southeast.
> 100 /ieq L'1. Care should be taken in extrapolating results from the SBRP to the
10-211
-------
SECTION 11
SUMMARY OF RESULTS
11.1 RETENTION OF ATMOSPHERICALLY DEPOSITED SULFUR
11.1.1 Current Retention
On average, watersheds in the Northeast have sulfur budgets near steady state, with negligible net
retention of atmospherically deposited sulfur (Section 7). A small proportion of northeastern watersheds
have significant net retention, which appears to be controlled by reduction In wetlands or within lakes.
In contrast, net retention in stream systems of the Southern Blue Ridge Province is high, averaging about
75 percent These observations are qualitatively consistent with theory (Galloway et al., 1983a; MAS,
1984) and with site-specific budgets summarized by Rochelle et al. (1987).
The Mid-Appalachian Region is a zone of transition between the NE and SBRP in terms of observed
current sulfur retention. Because of the similarities between soils in this region and the SBRP, it is likely
that this region at one time retained as much of the elevated sulfur deposition as is now evident in the
SBRP (i.e., 70-80 percent). It is also likely that continued high sulfur deposition is bringing soils near
steady state, leading to reduced sulfur retention, perhaps very dramatically in the westernmost area
(Subregion 2Cn of the National Stream Survey, which now has median percent sulfur retention of only
3 percent) (Plate 11-1), and has led to the low ANC and acidic stream reaches (excluding stream reaches
affected by acid mine drainage) identified there by the National Stream Survey (Kaufmann et al., 1988).
The Mid-Appalachian Region is the subject of additional in-depth soil sampling and analyses now
underway within the DDRP.
Results of the sulfur input-output analyses are consistent with results of Level I regression analyses
summarized in Section 8. Regression analyses indicate that in the NE, sulfate concentrations are more
highly correlated with sulfur deposition than with any watershed characteristic, as would be expected for
systems at or near steady state (i.e., systems where sulfur input equals output). Additionally in the NE,
11-1
-------
NSWS SUBREGIONS
MEDIAN I SULFUR RETENTION
AND WET SULFATE DEPOSITION
2,25
MEDIAN PERCENT
SULFUR RETENTION
H 0 - 20
20 - 40
40 - 60
Average Annual
Wet Sulfete % 2-
Deposi-tion (g m~* yr~')* '3.00-
60 - 80
80 - 100
3.25
Eastern Ukt Sum;
2.00'
-2.25
yedion
Suirtg.on I Rdention
1A
IB
1C
ID
IE
-14
8
-7
-9
-12
2-00
Naltoiiol Strtti Suney
led inn
I Uttdtien
3
40
34
50
75
70
2Cn
2B«
3B
n
2Ai
U
Deposition for 1980 - 1984
(A. 01 senp Personal Communication)
-------
Plate 11-1. Sulfur retention and wet sulfate deposition for National Surface Water Survey
subregions in the eastern United States.
11-2
-------
percent watershed sulfur retention is correlated with the extent of wetlands and wet soils on watersheds
(Section 8.5). This provides empirical support for the hypothesis that, to the limited extent sulfur retention
is observed In NE watersheds, reduction in wetlands is the principal retention mechanism.
In the SBRP, sulfate concentrations are correlated principally with edaphic factors. Sulfate
concentrations are relatively high in watersheds with high proportions of shallow soils and in catchments
having soils with low adsorption capacity. Similarly, percent sulfur retention increases with soil depth and
with sulfate adsorption capacity of soils. In both the NE and SBRP, watershed disturbance (e.g., mining
activity) is associated with elevated surface water sulfate concentrations.
11.1.2 Projected Retention
Using deposition scenarios described in Section 5.6, projections were made of future sulfur retention
in the NE and SBRP using both a single factor (Level II) adsorption model (Section 9.2) and the three
integrated models discussed in Section 10. For the sake of consistency, projections presented graphically
in this section are from the Level III MAGIC model. Because different target populations were modelled
by the four models (i.e., Level II and three Level III models) and because the projected results vary
t
somewhat among those populations, compan'sons will be discussed qualitatively.
In the NE, median lake sulfate concentrations are already very close to steady state. For the
scenario of constant deposition, all of the models thus projected only small changes in median sulfate
concentration, and projected those changes to occur relatively rapidly (10-20 year lags). Among the
Level III models, MAGIC and ETD project smail decreases in median sulfate concentration during the next
20 to 50 years, whereas ILWAS projects very small increases. Slight (3-5 percent) positive sulfur
retention is projected by all three models by year 50, with in-lake reduction as the principal retention
mechanism. The differences in the direction of changes for sulfate concentration result from differences
in target lake populations, in process representation by the models, and in calibration procedures;
absolute differences among projections are minor and are relatively unimportant. For the scenario of
decreased sulfur deposition, the models consistently project substantial decreases in median lake sulfate
11-3
-------
concentration by year 50. MAGIC and ETD project decreases in median sulfate of about 40 jueq L'1 In
50 years; ILWAS projects a somewhat slower decrease and a smaller, but still significant decrease of 21
peq L"1 in median lake sulfate.
Changes projected by the Level II sulfate model are very similar to those projected by MAGIC and
ETD. The Level II model projects only a small median decrease (7 /*eq L*1) in sulfate concentration by
year 20 for the constant deposition scenario, and a decrease in median sulfate of 40 jueq L*1 by year 50
for the decreased sulfur deposition scenan'o. The principal difference in projections between the Level
II and III models is that the Level II model projects all watersheds to eventually reach exactly steady state,
rather than the small positive sulfur retention projected by Level III models. This results from differences
in the processes considered by the models; the Level II model considers only sulfate sorption by soils,
whereas the Level ill models Include in-lake reduction, which accounts for the slightly positive retention
at long time intervals.
Projected changes in sulfate concentrations for SBRP surface waters occur much more slowly than
in the NE, and are much larger in magnitude. Median sulfate retention in SBRP watersheds is currently
about 75 percent, but retention is projected to decrease sharply over the next several decades (Plate 11-
2). Results were available for the Level II model (Section 9) and two of the Level III models (MAGIC and
ILWAS) (Section 10); all three models projected generally similar changes for sulfate in the SBRP. For
the constant deposition scenario, the two integrated models project increases in median stream sulfate
of roughly 15 jueq L'1 in the next 20 years and about 40 jueq L*1 in 50 years; median percent retention
is projected to decrease by about 40 percent over the 50-year period. For the increased deposition
scenario, slightly larger Increases in median sulfate concentration, of slightly over 50 /ieq L"1, are
projected by year 50. The Level II model projects somewhat faster increases for sulfate, with increases
of 31 and 56 ^eq L"1 in median sulfate concentration at 20 and 50 years, respectively. The Level II model
and MAGIC both project that rates of increase in sulfate concentration will decrease by year 100 as SBRP
-------
Plate 11-2. Changes in sulfur retention in the Southern Blue Ridge Province as projected by
MAGIC for constant sulfur deposition.
11-5
-------
% SULFUR RETENTION
Model = MAGIC
Deposition = Constant
___ Jrd Quorlile * .,
(1.5 » Interquorlile Rongt)
}r
-------
watersheds approach steady state (ILWAS projections were not made beyond 50 years) (Section 10).
The cumulative increases projected for median sulfate at 100 and 200 years are 60 and 74 //eq L*1 for
MAGIC and 66 and 81 jueq L"1 for Level II. The differences among the models at 20 and 50 years are
attributable to differences in hydrologic routing in the models and to assumptions about the chemistry
of deep subsoils. The 20- and 50-year projections occur during the period when the models project the
most rapid changes in sulfate concentration, and can be regarded as a measure of uncertainty in the
projections. In terms of the most important aspects of sulfur dynamics, the three models are consistent.
All project that under the deposition scenarios simulated, the delayed response phase of SBRP
watersheds would end for sulfate, and that there would be substantial increases in sulfate concentration
in the next 20 to 50 years. Such changes would be accompanied by decreases In surface water ANC
to a degree dependent upon the relative leaching of acids and base cations from watershed soils.
The results of the various sulfate analyses are all internally consistent. Level II projections of base
year sulfate in watersheds of the NE and SBRP are consistent with, and provide a mechanistic explanation
for, analyses by Rochelle and Church (1987), summarized in Section 7.3, showing watersheds in the
northeastern United States to be at or near steady state for sulfur and watersheds in the SBRP to have
high net sulfur retention. The very short sulfate response times projected for the NE are also consistent
with results of regression analyses in Sections 7 and 8, which indicate that deposition is the principal
control on surface water sulfate in the NE, and that significant sulfur retention (where observed), is
probably attributable to sulfate reduction in lakes and/or wetlands rather than to sorption. Similarly, the
long response times predicted by dynamic models for the SBRP are consistent with results of the Level
I regression analyses, which found sulfate concentration and percent sulfur retention to be correlated with
soil variables directly affecting adsorption capacity of soils (i.e., soil thickness and isotherm parameters).
11.2 BASE CATION SUPPLY
11.2.1 Current Control
Base cations are supplied from watersheds to surface waters by two processes acting in concert.
The initial source is mineral weathering, which is a slow process that supplies base cations to the soil
11-6
-------
exchange complex. Equilibrium between the exchange complex and soil water (and thus waters delivered
to lakes and streams) is reached quickly. It is generally accepted that weathering rates are likely to
change negligibly or increase only slightly due to the effects of acidic deposition since only slight
decreases in soil pH are likely. If weathering supplies base cations to surface waters at rates equal to
or greater than rates of acid anion deposition, then systems would be relatively "protected". If weathering
rates are low and cation exchange dominates base cation supply rates, then the rate of depletion of base
cations from the exchange complex becomes an important determinant of rates of surface water
acidification. Our analyses indicate that surface water ANCs >100 /ieq L'' cannot be explained by the
cation exchange model of Reuss and Johnson (1986); thus, ANC generation appears to be dominated
by weathering in these systems and they, presumably, are relatively protected against loss of ANC
(Section 9). Surface waters with ANCs < 100 jueq L'1 are likely controlled by a mix of weathering and
cation exchange. The exact proportion of the mix is difficult to determine.
11.2.2 Future Effects
Single factor base cation analyses, using the models of Reuss and Johnson (1986) and of Bloom
and Grigal (1985), were developed as a "worst-case" analysis by (1) considering only processes occurring
in the top 1.5-2 meters of the regoiith and (2) setting mineral weathering rates to zero (i.e., assuming
that the supply of base cations was totally controlled by cation exchange). This analysis indicated that
depletion of base cations from the exchange complex would occur under the sulfur deposition scenarios
simulated. The effect on surface water ANCs was initially slight but was not negligible. The magnitude
of soil base cation depletion was projected to accelerate in the future. At current levels of deposition,
about 15 percent of the lakes in the ELS target population are potentially susceptible to significant
depletion of exchangeable cations and, thus, depletion of associated surface water ANCs. The greatest
portion of such changes is projected to occur on a time scale of about 50 years. In the SBRP, a greater
percentage of systems are projected to be susceptible to adverse effects, but at longer time scales (i.e.,
about 100 years) than northeastern systems.
11-7
-------
Any effects of base cation depletion would be superimposed upon effects resulting from changes
in sulfate mobility in soils. The combined effects were simulated using the Level III watershed models
and are summarized in the next section.
11.3 INTEGRATED EFFECTS ON SURFACE WATER ANC
The three Level III watershed models (Section 1.3.4) were used to project the integrated watershed
and surface water responses to the sulfur deposition scenarios. Results among the models were
remarkably comparable. For example, within modelling Priority Classes A and B in the NE (Section 10)
and for the decreased sulfur deposition scenario, the MAGIC, ETD, and ILWAS models project changes
(at 50 years) In the median target population ANC for ANC groups <0 and 0 - 25 Meq L'1 within 2 jieq
L'1 (5 - 7 /ieq L'1) and 3 yeq L'1 (10 - 13 /*eq L'1), respectively. For the ANC group 25 - 100 /ieq L"1
the ILWAS and MAGIC models project increases in median ANC within 1 Meq L*1 (5.4 • 6.3 A*eq L'1).
Increases in the median ANC of this group (25-100 jueq L'1) under these conditions projected by the
ETD model are quite a bit greater (ke., -14 /ieq L"1 vs. ~ 6 /neq L"1).
The greatest disagreement among the model projections (at 50 years) is for the increased sulfur
deposition scenario in the SBRP. For modelling Priority Classes A and B and ANC group 100-400 Meq
L"1, the ILWAS model projects a decrease in median ANC of 7 jieq L"1, whereas the MAGIC model
projects a decrease of 24 /ieq L"1. Otherwise, comparative results among the models are remarkably
uniform, especially among the lower ANC groups of systems that are of the greatest concern.
Results from MAGIC are presented here because this model was successfully calibrated to the
largest number of watershed systems in the two regions (i.e., 123 of the 145 DDRP sample watersheds,
representing a target population of 3,227 systems in the NE; and 30 of the 35 DDRP sample watersheds;
representing a target population of 1,323 stream reaches in the SBRP).
11-8
-------
As discussed in Section 10, the watershed modelling analyses make use of watershed soil
representations as aggregated from the DDRP Soil Survey, Because of the focus of the DDRP on
regional characteristics and responses, soils data were gathered and aggregated so as to capture the
most important central tendencies of the study systems. As a result, extremes of individual watershed
responses probably are not fully captured in the analyses presented here (see discussion in Sections 8
and 10). Those systems that are projected to respond to the greatest extent or most quickly to current
or altered levels of sulfur deposition might, in fact, be expected to respond even more extensively or
more quickly than indicated here. This should be kept in mind when reviewing the simulation results
presented in this Section.
11.3.1 Northeast Lakes
Results of the projections for both deposition scenarios are presented in a couple of ways. Plate
11-3 and Table 11-1 illustrate the projected change in the median ANC at 50 years for lakes classified
into four ANC groups (i.e., <0 /ieq L'1, 0 • 25 /ueq L"1, 25-100 jueq L'1, and 100-400 /ueq L*1). These
projections indicate a general, very slight decline in ANC over the 50-year period under the current
deposition scenario and an Increase of roughly 5-15 jtieq L"1 in ANC for ail groups under the decreased
sulfur deposition scenario. Plates 11-4 and 11-5 illustrate the overall projected ANCs for the target
population at 20, 50 and 100 years for the constant and decreased deposition scenarios, respectively.
Table 11-2 presents the population estimates (with 95 percent confidence intervals) of northeastern
lakes having values of ANC <0 /ieq L'1 and <50 jueq L*1 at 20 and 50 years as projected by the MAGIC
model for the two deposition scenarios. The ANC = 0 /xeq L*1 value is used to define acidic systems,
and the ANC value of 50 jieq L'1 (for index values as sampled in the NSWS, see Section 5.3) has
recently been suggested as useful in approximating the level at or below which systems are susceptible
to severe episodic acidification (i.e., brief periods of ANC depression to very low or negative values)
(Eshleman, 1988) with consequent adverse effects on biota. It is extremely important to keep in mind
that these values only serve as indices in an otherwise smooth continuum of surface water chemistry
11-9
-------
Plate 11-3. Changes in median ANC of northeastern lakes at 50 years as projected by MAGIC (see
Section 1.3.4 for definition of the deposition scenarios used).
11-10
-------
CHANGE IN MEDIAN ANC
Year 10 to Year 50
Model = MAGIC
-------
Table 11-1. Weighted Median Projected Change in ANC at 50 Years for Northeastern
DDRP Lakes
ANC Group ( uea L'1
<0 0-25 25-100 KXWOO
Target 162 398 1054 1612
Population
Change in Median (yeq L*1) -2 -2 -1 -3
(deposition = constant)3
Change in Median (Meq L'1) 5 10 10 15
(deposition = decreased)
See Section 1.3.4 for definition of the deposition scenarios used.
11-11
-------
Plate 11-4. ANCs of northeastern lakes versus time, as projected by MAGIC for constant sulfur
deposition.
11-13
-------
ANC vs, TIME
Model = MAGIC; Deposition = Constant
ANC Group (s) = All
Maximum
3rd Quorlile t
(1.5 x Interquartile Range)"
jfd Ouortile
Mean
Uedion
1st Ounriile
1st Ouortile -
(!.5 x Interquartile Range)"
Minimum
Not to exceed extreme value.
'YEAS 0 • Phot. I NSIS Sanpli
-------
Plate 11-5. ANCs of northeastern lakes versus time, as projected by MAGIC for decreased sulfur
deposition.
11-13
-------
ANC vs, TIME
Model = MAGIC; Deposition = Decreased
ANC Group(s) = Al I
Maximum
3rd Quorliie +
(1.5 x Interquartile Range)"
3rd Quortile
Mean
Median
1st Ouortile
Is) Ouartile -
(1.5 x Interquartile Rouge)"
Minimum
No) to exceed extreme value
._. . >
i*sS- Bja^df Sl&iJ^ "*• *S- f AS1
TEAR 0 - Phti* I NSIS Sempli
-------
Table 11-2. Lakes in the NE Projected to Have ANC Values <0 and <50 /ieq L'1
for Constant and Decreased Sulfur Deposition*1**
Time from
Present (yr)
0 #°
20 #
50 #
Constant
ANC <0
162d
5
161 (134)
5(4)
186 (143)
5(4)
Deposition
ANC <50
880d
27
648 (246)
20 (8)
648(246)
20 (8)
Decreased
ANC <0
162d
5
136 (124)
4(4)
87 (100)
3(3)
Deposition
ANC <50
880d
27
621 (242)
19 (18)
586 (237)
18(7)
a Projections are based on 123 lake/watersheds successfully calibrated by MAGIC.
b Sea Section 1.3.4 for definition of the deposition scenarios used.
c # is the number of takes; % is percent of the target population of 3,227 lakes; () indicate 95 percent confidence
estimates.
d Indicates estimate from NSWS Phase I sample for the same 123 lakes; target population * 3,227 lakes
11-14
-------
conditions and responses to acidic deposition. It is also important to remember that adverse biological
effects occur at higher ANCs (i.e., greater than 50 jueq L"1) in systems that previously (i.e., prior to the
advent of acidic deposition) were adapted to more circumneutral conditions (Schindler, 1988).
As indicated in Table 11-2, under the constant deposition scenario, the number of lakes with ANC
<0 /ieq L"1 increases at 50 years whereas the number of lakes with ANC <50 Meq L'1 decreases. For
the scenario of decreased sulfur deposition, a marked decrease is projected in the number of systems
with ANC <0 and ANC <50 jieq L'1. Plate 11-6 shows the changes in pH for northeastern lakes at 50
years as projected by MAGIC. MAGIC projects the greatest change for the lowest ANC group. For
this group the change projected by ILWAS is virtually identical to that projected by MAGIC. Projections
by ILWAS and ETD for the higher ANC groups are somewhat greater than projections by MAGIC (see
Section 10).
Model projections indicate a mixed response of northeastern lake systems at current levels of sulfur
deposition. Slight decreases in median ANCs are projected for all ANC groups, along with a slight
increase In the number of systems with ANC < 0 peq L*1. The number of systems having ANC < 50
jieq L*1 (and thus potentially susceptible to episodic acidification), however, is projected to decrease.
Projected responses to decreased sulfur deposition show a clearer pattern; MAGIC projects surface water
ANCs to increase and the number of lakes with ANC <0 j/eq L'1 and ANC <50 peq L*1 to decrease.
Such a response would be consistent qualitatively with reported changes In the chemistry of lakes near
Sudbury, Ontario, following reductions of sulfur dioxide emissions from the Sudbury smelter (Dillon et al.,
1986; Hutchinson and Havas, 1986; Keller and Prtbaldo, 1986).
Because of the highly organic nature of some soils in the NE, the exact nature of chemical
"recover/ of northeastern lakes is uncertain. To our knowledge, there are no field studies in that region
that carefully document such a situation over a sufficient time period to cast much light upon this subject.
As discussed in Section 1, it has been hypothesized that leaching of organic acids could be controlled
11-15
-------
Plate 11-6. Changes in median pH of northeastern lakes at 50 years as projected by MAGIC (see
Section 1.3.4 for definition of the deposition scenarios used).
11-16
-------
CHANGE IN MEDIAN pH
Year 10 to Year 50
Model = MAGIC
-------
by changes in soil water pH (e.g., as caused by acidic deposition) and that this, in turn, could have
important effects on surface water pH values (Krug and Prink, 1983; Krug, 1989). In this hypothesis, a
decrease in precipitation acidity would result in an increase in leaching of organic acids to surface
waters, partially offsetting (i.e., toward lower pH) pH increases associated with the 'Improved" chemical
quality of the atmospheric deposition. Recently, Wright et al. (1988) noted such an effect in a stream
catchment in Norway where acidic deposition was excluded and reconstituted, more circumneutral waters
were substituted as "rain". The catchment studied by Wright et al. (1988) has extremely thin, organic soils
and, thus, is a site almost ideally suited to the observation of such an effect. Wright et al. (1988) noted
that in other areas of Norway having soils of a more mineral nature (and probably much more similar to
soils of the type found on DORP northeastern study sites) the potential for enhanced mobilization of
organic anions would likely be much suppressed and minor relative to the effects of decreasing sulfur
deposition.
Even if there was an appreciable increase in organic acid leaching as a response to reduced
deposition acidity, the net effect would be beneficial to aquatic biota inasmuch as it would most likely
be accompanied by reductions in surface water concentrations of inorganic monomeric aluminum, which
is highly toxic to fish (Baker and Schofield, 1982).
Thus, although the exact chemical response of the northeastern DORP systems is unknown,
projections indicate' an improvement in surface water quality as a consequence of reduced sulfur
deposition in the region.
11.3.2 Southern Blue Ridge Province
Plate 11-7 and Table 11-3 illustrate the projected changes (MAGIC model) in median ANC at 50
years for stream reaches in the SBRP. The MAGIC model used in this analysis was successfully
calibrated to 32 of the 35 ODRP SBRP stream reach watersheds. Two stream reaches had ANC >
1000 Meq L1 and were dropped from this presentation. The remaining 30 stream reaches had ANC >
25 Meq L'1 and < 400 jieq L'1 and represent a target population of 1,323 stream reaches in the SBRP.
11-17
-------
Plate 11-7. Changes in median ANC of Southern Blue Ridge Province stream reaches at 50 years
as projected by MAGIC (see Section 1.3.4 for definition of the deposition used).
11-18
-------
CHANGE IN MEDIAN ANC
Year 10 to Year 50
Model = MAGIC
S » Deposition
Constant .
Depju-i-HYn
-------
Table 11-3. Weighted Median Projected Change in ANC at 50 Years for DDRP SBRP Stream
Reaches
ANC Group (Meq L-1)
25-100 100-400
Target
Population
Median Change
(Meq L'1)
407
-14
916
-24
(deposition = constant)3
Median Change (JLieq L*1) -20 -34
(deposition = increased)
aSee Section 1.3.4 for definition of the deposition scenarios used.
11-19
-------
The projected changes in median ANCs have been computed for the same ANC groups (25 -100 jieq
L'1 and 100-400 /ieq L'1 ) as for the NE (Plate 11-3). Plates 11-8 and 11-9 illustrate the overall
projected ANCs for the target population at 20, 50, 100, and 200 years for the current and increased
deposition scenarios, respectively.
Table 11-4 presents the population estimates (with 95 percent confidence intervals) of SBRP stream
reaches having ANC < 0 jueq L'1 and <50 Meq L*1 at 20 and 50 years as projected by the MAGIC model
for the two deposition scenarios. The 95 percent confidence intervals about these projections are broad
but understandable, given the low number of systems available for simulation (30) and the Inherent
uncertainties involved in such a complex simulation of environmental response.
Plates 11-10 and 11-11 show decreases in pH of SBRP stream reaches as projected by MAGIC
and ILWAS, respectively, for the increased sulfur deposition scenario. Changes projected by the two
models are highly comparable.
Model projections for the SBRP stream reaches indicate decreased surface water quality under
scenarios of either current or increasing sulfur deposition. Due to the fact that soils in this region are
much less organic in nature than those in the NE (e.g., wetlands in the SBRP are virtually non-existent;
mean stream DOC at lower nodes was < 1 mg L*1), these model projections are uncomplicated by
potential effects of organic acid leaching. Model projections for the increased sulfur deposition scenario
indicate the potential for about one-quarter of the target population of stream reaches in the SBRP to
reach an ANC of < 50 jieq L'1 in 50 years, and thus to have the potential to reach an ANC -0 ^eq L'
1, during storm event episodes (Eshleman, 1988). As noted in Sections 9 and 10, responses to changes
in sulfur deposition levels in the SBRP are projected to be slower than those in the NE; i.e., there is a
considerable lag in the response of the systems due to the storage of sulfur in the soils. The result is
that there is a delay not only in the acidification of surface waters In the region, but also in any potential
recovery if sulfur deposition were to be decreased.
11-20
-------
Plate 11-8. ANCs of Southern Blue Ridge Province stream reaches versus time, as projected by
MAGIC for constant sulfur deposition (see Section 1.3.4 for definition of the deposition scenarios
used).
11-21
-------
ANC vs, TIME
Model - MAGIC; Deposition - Constant
ANC Group(s) = <400 ueq L'1
3rd Quutile 4 .,
(1.5 i intetqyarlilc Range)
Jrd Quorlile
Ilian
Miditn
Is! Quortile
111 ttudrlile -
(1.5 i (nleiquorlile Koagt}
0 « NSS Sinpli
' Not lo tutted ettreme volge.
-------
Plate 11-9. ANCs of Southern Blue Ridge Province stream reaches versus time, as projected by
MAGIC for increased sulfur deposition (see Section 1.3.4 for definition of the deposition scenarios
used).
11-22
-------
ANC vs, TIME
Model = MAGIC; Deposition = Increased
ANC Group (s) • <400 ueq I'1
SBRP Study Am
'• D O
>— CM
~ >
JUO —
.-
200 -
100-
0 -
1 ft A
1 UU
....
*-
r
A
If
i«*
^m
i
•
1 m
i (§i
is
— . *-• '"
•^i
S IE
M m
IB
i'
. -1"*
——I
•w
H
B
^
i^*1"
^' *
'TEAR 0 * NSS
3(J Owlile +
(i.S i latctqygrlilc Songe)
3rd Ouorltle
1sl Ouorifle
111 Qugrlile -
I Isl Qugrlile -
(1.5 i Interquartile Range)
" Hot to eicced eit'eme
-------
Table 11-4. SBRP Stream Reaches Projected to Have ANC Values <0 and <50 jueq L"1
for Constant and Increased Sulfur Deposition**
Time from
Present (yr)
0 #c
20 #
50 #
Constant
ANC <0
Od
0
0
0
129 (195)
10 (15)
Deposition
ANC <50
3d
0.2
187 (228)
14(17)
203 (236)
15 (18)
Increased
ANC <0
Od
0
0
0
159 (213)
12 (16)
Deposition
ANC <50
3d
0.2
187 (228)
14 (17)
340 (286)
26 (22)
* Projections are based on 30 stream/watersheds successfully calibrated by MAGIC.
b See Section 1.3.4 for definition of the deposition scenarios used.
0 # is the number of lakes; % is percent of the target population of 1,323 stream reaches; {) indicate 95 percent
confidence estimates.
d Indicates estimate from Pilot Stream suivey sample for the same 30 streams; target population = 1,323 stream
reaches
11-23
-------
Plate 11-10. Changes in pH of SBRP stream reaches as projected by MAGIC (see Section 1.3.4
for definition of the deposition scenarios used).
11-24
-------
pH vs. TIME
Model = MAGIC; Deposition ^Increased
ANC Group(s) * <400 ueq I'1
3rd Quortlle 4
(1.5 i Interquartile Range)"
itt Ouorlile
Hedion
1st Out/tile
I 1st fluorlik -
(1.5 i tntengagrlite doegt)"
'TEAR 0 - Uotfil Ytor 0
' Not to exceed etlreme volte.
-------
Plate 11-11. Changes in pH of SBRP stream reaches as projected by ILWAS (see Section 1.3.4
for definition of the deposition scenarios used).
11-25
-------
pH vs. TIME
Model = ILWAS; Deposition = Increased
ANC Group(s) = <400 ueq L'1
3rd Ouorlile *
(1.5 i Interquartile Range) '
3rd
Kerfion
1st Oucrtile
Isl Quirlile -
(1.5 i litientiurtile Rggge)"
•TEAR o » Ho4
-------
Projections of stream water quality response for the DDRP SBRP target population clearly indicate
future adverse effects of sulfur deposition at increased or current levels.
11.4 SUMMARY DISCUSSION
The NE is currently at sulfur steady state and sulfate concentrations in surface waters would
respond relatively rapidly to decreases in sulfur deposition. Associated with these changes would be
increases in surface water ANC. Continued sulfur deposition at current levels is gradually depleting the
cation exchange pool in northeastern soils with consequent decreases in surface water ANC. Such
changes are relatively slow and minor, however, relative to direct effects of increased anion mobility in
watersheds on surface water chemistry.
Watersheds in the SBRP are currently retaining nearly three-quarters of the atmospherically
deposited sulfur on the average but soils are projected as becoming more saturated with regard to
sulfur. Sulfate concentrations are projected to be increasing in the surface waters of the region. A
»
marked increased in stream sulfate concentrations response is projected over the next 50 years at either
current or increased levels of sulfur deposition, as are decreases in stream water ANC. Superimposed
upon this effect is a relatively minor acidification effect of base cation depletion.
Results from all level of DDRP analyses are (1) consistent internally, (2) consistent with theory
(Galloway et al., 1983a) and (3) consistent with recent observations of lakes monitored during changing
sulfur deposition regimes (Dillon et al., 1986; Hutchinson and Havas, 1986; Keller and Pitbaldo, 1986).
11-26
-------
SECTION 12
REFERENCES
Abrahamsen, G. 1980. Acid precipitation, plant nutrients, and forest growth, pp. 58-63. In: D. Drabl0s,
and A. Totlan, eds. Ecological Impact of Acid Precipitation: Proceedings of an International
Conference, Sandefjord, March 11-14. SNSF Project, Oslo-As, Norway.
Adams, F., and Z. Rawajfih. 1977. Basaluminite and alunite: A possible cause of sulfate retention by acid
soils. Soil Sci. Soc. Am. J. 41:686-692.
Akaike, H. 1969. Fitting autoregressive models for prediction. Ann. Institute Statist. Math. 21:243-247.
Aimer, B., W. Dickson, C. Ekstrom, and E. Homstrom. 1978. Sulfur pollution and the aquatic ecosystem,
pp. 271-311. In: J.O. Nriagu, ed. Sulfur in the Environment, Part II: Ecological Impacts. John Wiley
& Sons, Inc., New York, NY.
Altshuller, A.P., and R.A. Linthurst. 1984. The Acidic Deposition Phenomenon and Its Effects: Critical
Assessment Review Papers. EPA/600/8-83/016bf. U.S. Environmental Protection Agency,
Washington, DC.
Appleby, P.G., and F. Oldfield. 1978. The calculation of lead-210 dates assuming a constant rate of
supply of unsupported Pb-210 to the sediment. Catena 5:1-8.
April, R., and R. Newton. 1985. Influence of geology on lake acidification in the ILWAS watersheds.
Water. Aor. Soil Pollut. 26:373-386.
Arnold, R.W. 1977. Clean brush approach achieves better concepts in soil survey, pp. 61-92. In: Quality
Conference Proceedings, Bergams, New York, NY.
Arnold, R.W. 1980. Graphical Solution of Binomial Confidence Limits in Soil Survey. Northeastern Soil
Survey Work Planning Conference Report.
Arora, J.S., P.B. Thanadar, and C.H. Tseng. 1985. User's Manual for Program IDESIGN, Version 3.4 for
PRIME Computers. Tech. Report No. ODL 85.10. Optimal Design Laboratory, College of Engineering,
University of Iowa, IA.
Asbury, C.E., F.A. Vertucci, M.D. Mattson, and G.E. Likens. 1989. Acidification of Adirondack lakes.
Environ. Sci. Technol. 23:362-365.
Bache, B.W. 1983. The role of soil in determining surface water composition. Water Sci. Technol. 15:33-45.
Backes, C.A., and E. Tipping. 1987. Aluminum complexation by an aquatic humic fraction under acidic
conditions. Water Res. 2t;211-216.
Bard, Y. 1974. Nonlinear Parameter Estimation. Academic Press. N.Y. 341 pp.
Baker, J.P., and C.L Schofield. 1982. Aluminum toxicity to fish in acidic waters. Water, Air, Soil Pollut.
18:289-309.
Baker, LA., P.L Brezonik, E.S. Edgerton, and R.W. Ogburn, III. 1985. Sediment acid neutralization in
softwater lakes. Water, Air, Soil Pollut. 25:215-230.
12-t
-------
Baker, LA., P.L Brezonik, and E.S. Edgerton. 1986a. Sources and sinks of ions in a softwater acidic
lake in Rorida. Water Resour. Res. 22:715-722.
Baker, LA., P.L Brezonik, and C.D. Pollman. 1986b. Model of internal alkalinity generation: Sulfate
retention component. Water, Air, Soil Pollut. 31:89-94.
Baker, LA., C.D. Pollman, and J.M. Eilers. 1988. Alkalinity regulation in softwater Florida lakes. Water
Resour. Res. 24:1069-1082.
Barrow, N.J. 1967. Studies on the adsorption of sulfate by soils. Soil Sci. 104:342-349.
Barrow, N.J., and T.C. Shaw. 1977. The slow reactions between soil and anions: Effect of time and
temperature of contact between an adsorbing soil and sulfate. Soil Sci. 124:347-354.
Barrow, N.J., K. Spencer, and W.M. McArthur. 1969. Effects of rainfall and parent material on the ability
of soils to adsorb sulfate. Soil Sci. 108:120-126.
Bartz, J.K., S.K. Drouse, K.A. Cappo, M.L Papp, G.A. Raab, LJ. Blume, M.A. Stapanian, F.C. Gamer,
and D.S. Coffey. 1987. Direct/Delayed Response Project: Quality Assurance Plan for Soil Sampling,
Preparation and Analysis. EPA/600/8-87/021. U.S. Environmental Protection Agency, Washington,
DC. 419 pp.
Bayley, S.E., R.S. Behr, and C.A. Kelly. 1986. Retention and release of S from a freshwater wetland.
Water, Air, Soil Pollut. 31:101 -114.
Bayley, S.E., D.H. Vftt, R.W. Newbury, K.G. Beaty, R. Behr, and C. Miller. 1987. Experimental acidification
of a Sphagnum-dominated peatland: First year results. Can. J. Fish. Aquat. Sci. 44(Suppl.
1): 194-205.
Beamish, R.J., and H.H. Harvey. 1972. Acidification of the La Cloche Mountain lakes, Ontario, and
resulting fish mortalities. J. Fish. Res. Board Can. 29:1131-1143.
Beck, M.B. 1987. Water quality modeling: A review of the analysis of uncertainty. Water Resour. Res.
23:1393-1442.
Belsley, DA, E, Kuh, and R.E. Welsch. 1980. Regression Diagnostics. Wiley-lnterscience, New York, NY.
Best, M.D., LW. Creelman, S.K. Drouse, and D.J. Chaloud. 1986. National Surface Water Survey: Eastern
Lake Survey (Phase I), Analytical Methods Manual. EPA-600/4-86/011. U.S. Environmental Protection
Agency, Las Vegas, NV.
Bettany, J.R., J.W.B. Stewart, and E.H. Halstead. 1973. Sulfur fractions and carbon, nitrogen and sulfur
relationships in grasslands, forest and associated transitional soils. Soil Sci. Soc. Am. J. 37:915-918.
Beven, K.J. 1986. Runoff production and flood frequency in catchments of order: An alternative approach.
pp. 107-132. In: V.K. Gupta, I. Rodriguez-lturbe, and E.F. Woods, eds. Scale Problems in Hydrology.
D. Reidel Publ. Co., Dordrecht, The Netherlands.
Beven, K.J., and M.J. Kirkby. 1979. A physically based, variable constant contributing area model of
basin hydrology. Hydro). Sci. Bull. 24:43-69.
Billings, M.P. 1980. Geologic Map of New Hampshire. U.S. Geological Survey, Washington, DC. (scale
1:250,000).
Birks, H.J.B. 1985. Recent and future mathematical developments in quantitative palaeoecology.
Palaeogeogr., Palaeoclimat., Palaeoecol. 50:107-147.
12-2
-------
Bloom, P.P., and D.F. Grigal. 1985. Modeling soil response to acidic deposition in nonsulfate adsorbing
soils. J. Environ. Qual. 14:489-495.
Bloom, P.R., M.B. McBride, and R.M. Weaver. 1979a. Aluminum organic matter in acid soils: Buffering
and solution aluminum activity. Soil Sci. Soc. Amer. J. 43:810-813.
Bloom, P.R., M.B. McBride, and R.M. Weaver. 1979b. Aluminum organic matter in acid soils: Salt-
extractable aluminum. Soil Sci. Soc. Amer. J. 43:813-815.
Bogucki, D.J., G.K. Gruendling, and E.B. Allen. 1987. Final Report: Evaluating Accuracy of Photo
Interpretation for Land Use, Wetland, Beaver, and Stream Data Collected for Direct/Delayed
Response Project (DDRP) Watersheds. Rep. Center for Earth and Environmental Science, State
University of New York, Pittsburgh, NY. (Unpub.). 191 pp.
Bogucki, D.J., G.K. Gruendling, E.B. Allen, K.B. Adams, and M.M. Remillard. 1986. Photointerpretation
of historical (1948-1985) beaver activity in the Adirondacks. pp. 299-308. In: Proceedings of the
ASPRS-ACSM Fall Convention. September 28-October 3, Anchorage, AK.
Bolan, N.S., D.R. Scotter, J.K. Syers, and R.W. Tillman. 1986. The effect of adsorption on sulfate leaching.
Soil Sci. Soc. Am. J. 50:1419-1424.
Bornemisza, E., and R. Llanos. 1967. Sulfate movement, adsorption, and desorption in three Costa Rican
soils. Soil Sci. Soc. Am. Proc. 31:356-360.
Brakke, D.F., D.H. Landers, and J.M. Eilers. 1988. Chemical and physical characteristics of lakes in the
northeastern United States. Environ. Sci. Techno). 22:155-163.
Brezonik, P.L, LA. Baker, and T.E. Perry. 1987. Mechanisms of alkalinity generation in acid-sensitive
softwater lakes, pp. 229-262. In: R. Hites and S.J. Eisenreich, eds. Sources and Fates of Aquatic
Pollutants. Advances in Chemistry, Series 216. American Chemical Society, Washington, DC.
Bricker, O.P., A.E. Godfrey, and E.T. Cleaves. 1968. Mineral-water interaction during the chemical
weathering of silicates, pp. 128-142. In: Advances in Chemistry, Series 73. American Chemical
Society, Washington, DC.
Brown, P.M. 1985. Geologic Map of North Carolina. Dept. Natural Resources and Community
Development, Dtv. Land Resources, Raleigh, NC. (scale 1:500,000).
Buol, S.W., F.D. Hole, and R.J. McCracken. 1980. Soil Genesis and Classification. Iowa State University
Press, Ames, IA.
Buso, D.C., S.W. Bailey, S.F. Baird, J.W. Hornbeck, and C.W. Martin. 1985. Watershed Interactions
Affecting Pond Acidification. Research Report No. 62. New Hampshire Water Resource Research
Center, Durham, NH.
Byers, G.E., R.D. Van Remortel, J.E. Teberg, M.J. Mian, C.J. Palmer, M.L Papp, W.H. Cole, A.D. Tansey,
D.L Cassell, and P.W. Shaffer. 1989. Direct/Delayed Response Project: Quality Assurance Report
for Physical and Chemical Analyses of Soils from the Northeastern United States.
EPA/600/X-88/136. U.S. Environmental Protection Agency, Environ. Monitoring Systems Lab., Las
Vegas, NV. 216 pp.
Calles, C.M. 1983. Dissolved inorganic substances. Hydrobiologia 101:13-27.
Camburn, K.E., and J.C. Kingston. 1986. The genus Melosira from soft-water lakes with special reference
to northern Michigan, Wisconsin, and Minnesota, pp. 17-34. In: J.P. Smol, R.W. Battarbee, R.B.
Davis, and J. Merilainen, eds. Diatoms and Lake Acidity. Dr. W. Junk, Dordrecht, The Netherlands.
12-3
-------
Campbell. W.G., and M.R. Church. 1989. EPA uses GIS to study lake and stream acidification. Federal
Digital Cartography Newsletter 9:1-2.
Campbell, W.G., M.R. Church, G.D. Bishop, D.C. Mortenson, and S.M. Pierson. In Press. The role for a
Geographic Information System in a large environmental project. Internal. J. GIS.
Cappo, K.A., LJ. Blume, G.A. Raab, J.K. Bartz, and J.L Engels. 1987. Analytical Methods Manual for
the Direct/Delayed Response Project Soil Survey. EPA/600/8-87/020. U.S. Environmental Protection
Agency, Environ. Monitoring Systems Lab., Las Vegas, NV.
Cardwell, D.H., R.B. Erwin, and H.P. Woodward. 1968. Geologic Map of West Virginia. West Virginia
Geological and Economic Survey.
Cariston, C.W. 1963. Drainage Density and Streamflow. U.S. Geological Survey Professional Papers
422-C, Washington, DC. 8 pp.
Chan, W.H., D.B. Orr, and R.J. Vet. 1982a. The Acid Precipitation in Ontario Study (APIOS). An Overview:
The Cumulative Wet/Dry Deposition Network. Report #ARB-15-82-ARSP, Ontario Ministry of the
Environment.
Chan. W.H., D.B. Orr, and RJ. Vet. 1982b. The Acid Precipitation in Ontario Study (APIOS). An Overview:
The Event Wet/Dry Deposition Network. Report #ARB-11-82-ARSP, Ontario Ministry of the
Environment.
Chang, M., J.D. McCullough, and A.B. Granillo. 1983. Effects of land use and topography on some water
quality variables in forested east Texas. Water Resour. Bull. 19:191-196.
Chang, J.S., R.A. Brest, I.S.A. Isaksen. S. Modronich, P. Middleton, W.R. Stockwell, and CJ. Walcek.
1987. A three-dimensional Eulerian acid deposition model: Physical concepts and formulation.
J. Geophys. Res. 92:14681-14700.
Chao, T.T., M.E. Harward, and S.C. Fang. I962a. Movement of S,- tagged sulfate through soil columns.
Soil Sci. Soc. Am. Proc. 26:27-32.
Chao, T.T., M.E. Harward, and S.C. Fang. 1962b. Adsorption and desorption by soils. Soil Sci. Soc. Am.
Proc. 26:234-237.
Chao, T.T., M.E. Harward, and S.C. Fang. 19643. Anionic effects on sulfate adsorption by soils. Soil Sci.
Soc. Am. Proc. 28:581-583.
Chao, T.T., M.E. Harward, and S.C. Fang. 1964b. Iron or aluminum coatings in relation to sulfate
adsorption characteristics of soils. Soil Sci. Soc. Am. Proc. 28:632-635.
Chen, C.W., J.D. Dean, S.A. Gherini, and R.A. Goldstein. 1982. Acid Rain Model: Hydrologic Module. Jour.
Env. Engr. ASCE 108:302-318.
Chen, C.W., S.A. Gherini, R.J.M. Hudson, and J.D. Dean. 1984. The Integrated Lake-Watershed
Acidification Study, Volume 2: Hydrologic Analysis. C.W. Chen, ed. Rep. No. EA-3221. Electric
Power Research Institute, Palo Alto, CA.
Chen, C.W., S. Gherini, R.J.M. Hudson, and J.D. Dean. 1983a. The Integrated Lake/Watershed
Acidification Study, Volume 1: Model and Procedures. Rep. No. EA-3221. Electric Power Research
Institute, Palo Alto, CA.
Chen, C.W., S.A. Gherini, J.D. Dean, R.J.M. Hudson, and R.A. Goldstein. 1983b. Modeling of Precipitation
Series, Volume 9. Ann Arbor Sciences, Butterworth Publishers, Boston, MA. 175 pp.
12-4
-------
Chen, C.W., S.A. Gherini, R.K. Munson, LE. Gomez, and D. Donkers. 1988. Sensitivity of Meander Lake
to acid deposition. J. Environ. Engineer. 114.
Chou, L, and R. Wollast. 1985. Steady-state kinetics and dissolution mechanisms of aibite. Am. J. Sci.
285:963-993.
Christophersen, N., and R.F. Wright. 1981. Sulfate flux and a model for sulfate concentrations in
streamwater at Birkenes, a smalt forested catchment in southernmost Norway. Water Resour. Res.
17:377-389.
Christophersen, N., H.M. Seip, and R.F. Wright. 1982. A model for streamwater chemistry at Birkenes,
Norway. Water Resour. Res. 18:977-996.
Church, M.R. In Press. Predicting the future long-term effects of acidic deposition on surface water
chemistry: The Direct/Delayed Response Project. EOS, Transactions, American Geophysical Union.
Church, M.R., and R.S. Turner. 1986. Factors Affecting the Long-term Response of Surface Waters to
Acidic Deposition: State of the Science. EPA/600/3-86/025. NTIS PB 86 178 188-AS. U.S.
Environmental Protection Agency, Corvallis, OR. 274 pp.
Church, M.R., P.W. Shaffer, K.N. Eshleman, and B.P. Rochelle. In Review. Potential effects of sulfur
deposition on stream chemistry in the Southern Blue Ridge Province. Water, Air, Soil Pollut.
Clark, T.L, R.L Dennis, and S.K. Seilkop. 1989. Re-examination of Interim Estimates of Annual Sulfur Dry
Deposition Across the Eastern United States. EPA/600/4-89/026. U.S. Environmental Protection
Agency, Atmospheric Research and Exposure Assessment Laboratory, Research Triangle Park, NC,
34pp.
Clayton, J.L 1986. An estimate of plagioclase weathering rate in the Idaho batholith based upon chemical
transport rates, pp. 453-466. In: S.M. Colman and D.P. Dethier, eds. Rates of Chemical Weathering
of Rocks and Minerals. Academic Press, Orlando, FL
Cleaves, E.T., A.E. Godfrey, and Q.P. Bricker. 1970. Geochemical balance of a small watershed and its
geomorphic implications. Geol- Soc. Am. Bull. 81:3015-3032.
Cochran, W.G. 1977. Sampling Techniques, pp. 259-261. John Wiley & Sons, Inc., New York, NY.
Coffey, D.S., J.J. Lee, J.K. Bartz, R.D. Van Remortel, M.L Papp, and G.R. Holdren. I987a. Direct Delayed
Response Project: Field Operations and Quality Assurance Report for Soil Sampling in the Southern
Blue Ridge Province of the United States, Volume I: Sampling. EPA/600/4-87/041. U.S.
Environmental Protection Agency, Environ. Monitoring Systems Lab., Las Vegas, NV. 205 pp.
Coffey, D.S., J.J. Lee, D.A. Lammers, M.G. Johnson, and G.R. Holdren. 1987b. Direct/Delayed Response
Project Northeast Field Sampling Report, Volume I: Field Sampling. U.S. Environmental Protection
Agency, Environ. Monitoring Systems Lab., Las Vegas, NV. 199 pp.
Comeau, P.L, and D.J. Bellamy. 1986. An ecological interpretation of the chemistry of mire waters from
selected sites in eastern Canada. Can. J. Bot. 64:2576-2581.
Cook, R.B., C.A. Kelly, D.W. Schindler, and M.A. Turner. 1986. Mechanisms of hydrogen ion neutralization
in an experimentally acidified lake. Umnol. Oceanogr. 31:134-148.
Cook, R.B., M.L Jones, D.R. Marmorek, J.W. Elwood, J.L Malanchuk, R.S. Turner, and J.P. Smol. 1988.
The Effects of Acidic Deposition on Aquatic Resources in Canada: An Analysis of Past, Present, and
Future Effects. Oak Ridge National Laboratory, Environ. Sci. Div., Oak Ridge, TN.
12-5
-------
Cooper, J.R., J.W. Gillam, and T.C. Jacobs. 1986. Riparian areas as a control of nonpoint pollutants.
pp. 166-192. In: D.L Correll, ed Watershed Research Perspectives. Smithsonian Institution Press,
Washington, DC.
Cosby, B.J., G.M. Hornberger, J.N. Galloway, and R.F. Wright. 1985a. Modeling the effects of acid
deposition: Assessment of a lumped parameter model of soil water and streamwater chemistry.
Water Resour. Res. 21:51-63.
Cosby, B.J., G.M. Hornberger, J.N. Galloway, and R.F. Wright. I985b. Time scales of catchment
acidification: A quantitative model for estimating freshwater acidification. Environ. Sci. Technol.
19:1144-1149.
Cosby, B.J., G.M. Wright, G.M. Hornberger, and J.N. Galloway. 1985c. Modeling the effects of acid
deposition: Estimation of long-term water quality responses in a small forested catchment. Water
Resour. Res. 21:1591-1601.
Cosby, B.J., G.M. Hornberger, E.B. Rastetter, J.N. Galloway, and R.F. Wright. 1986a. Estimating catchment
water quality response to acid deposition using mathematical models of soil ion exchange
processes. Geoderma 38:77-95.
Cosby, B.J., G.M. Hornberger, R.F. Wright, and J.N. Galloway. 1986b. Modeling the effects of acid
deposition: Control of long-term sulfate dynamics by soil sulfate adsorption. Water Resour. Res.
22:1283-1292.
Cosby, B.J., P.G. Whftehead, and R. Neale. 1986c. A preliminary model of long-term change in stream
acidity in southwest Scotland. J. Hydrol. 84:381-401.
Cosby, B.J., G.M. Hornberger, P.F. Ryan, and D.M. Wolock. 1989. MAGIC/DDRP Final Report. Vol 1:
Models, calibration, results, uncertainty analysis, QA/QC. Internal Report. EPA Environmental
Research Laboratory-Corvallils. Corvallis, OR.
Cosby, B.J., G.M. Hornberger, and R.F. Wright. In Press. A regional model of surface water acidification
In southern Norway: Calibration and validation using survey data. In: J. Kamari, ed. Proceedings
of the IIASA-IMGW Task Force Meeting on Environmental Impact Models to Assess Regional
Acidification. D. Reidel Publ. Co.
Couto, W., D.J. Lathwell, and D.R. Bouldin. 1979. Sulfate sorption by oxisols and an alfisol of the tropics.
SoH Scl. 127:108-116.
Cowardin, L.M., V. Carter, F.C. Golet, and E.T. LaRoe. 1979. Classification of Wetlands and Deepwater
Habitats of the United States. FWS/OBS-79-31. Biological Services Program, U.S. Dept. of the
Interior, Fish and Wildlife Service, Washington, DC. 131 pp.
Cozzarelli, I.M., J.S. Herman, and R.A. Pamell, Jr. 1987. The mobilization of aluminum in a natural soil
system: Effects of hydrologic pathways. Water Resour. Res. 23:859-874.
Creasey, J., A.C. Edwards, J.M. Reid, D.A. MacLeod, and M.S. Cresser. 1986. The use of catchment
studies for assessing chemical weathering rates in two contrasting upland areas in northeast
Scotland, pp. 467-502. In: S.M. Colman and D.P. Dethier, eds. Rates of Chemical Weathering of
Rocks and Minerals. Academic Press, Orlando, FL
Cronan, C.S., and G.R. Aiken. 1985. Chemistry and transport of humic substances in forested watersheds
of the Adirondack Park, New York, NY. Geochim. Cosmochim. Acta 49:1697-1705.
Cronan, C.S., W.A. Reiners, R.C. Reynolds, Jr., and G.E. Lang. 1978. Forest floor leaching: Contributions
from mineral, organic and carbonic acids in New Hampshire subalpine forests. Science 200:309-311.
12-6
-------
Cronan, C.S., J.C. Conlan, and S. Skibinski. 1987. Forest vegetation in relation to surface water chemistry
in the North Branch of the Moose River, Adirondack Park, NY. Biogeochemistry 3:121-128.
Dasch, J.M. 1987. Measurement of dry deposition to surfaces in deciduous and pine canopies. Environ.
Pollut. 44:261-277.
David, M.B., and M.J. Mitchell. 1985. Sulfur constituents and cycling in waters, seston, and sediments
of an oligotrophic lake. Umnol. Oceanogr. 30:1196-1207.
David, M.B., M.J. Mitchell, and J.P. Nakas. 1982. Organic and inorganic sulfur constituents of a forest
soil and their relationship to microbial activity. Soil Sci. Soc. Am. J. 46:847-852.
David, M.B., M.J. Mitchell, and S.C. Schindler. 1984. Dynamics of organic and inorganic sulfur constituents
in hardwood forest soils, pp. 221-246. In: E.L Stone, ed. Forest Soils and Treatment Impacts:
Proceedings of the Sixth North American Forest Soils Conference, June 1983. Dept. Forestry,
Fisheries, and Wildlife, University of Tennessee, Knoxville.
Davis, J.A. 1982. Adsorption of natural dissolved organic matter at the oxide/water interface. Geochim.
Cosmochim. Acta 46:2381-2393.
Davis, G.F., J.J. Whipple, S.A. Gherini, C.W. Chen, R.A. Goldstein, P.W.H. Chan, and R.K. Munson. 1986.
Big Moose Basin: Simulation of response to acidic deposition. Biogeochemistry 3:141-161.
De Grosbois, E., P.J. Dillon, H.M. Seip, and R. Seip. 1986. Modelling hydrology and sulfate concentrations
in small catchments in central Ontario. Water, Air, Soil Pollut. 31:45-57.
Dethier, D.P. 1986. Weathering rates and the chemical flux from catchments in the Pacific Northwest,
U.S.A., pp. 503-530. In: S.M. Colman and D.P. Dethier, eds. Rates of Chemical Weathering of Rocks
and Minerals. Academic Press, Orlando, FL.
Dillon, P.J., D.S. Jeffries, and W.A. Scheider. 1982. The use of calibrated lakes and watersheds for
estimating atmospheric deposition near a large point source. Water, Air, Soil Pollut. 18:241-258.
Dillon, P.J., R.A. Reid, and R. Girard. 1986. Changes in the chemistry of lakes near Sudbury, Ontario,
following reductions of SO2 emissions. Water, Air, Soil Pollut. 31:59-65.
Dillon, P.J., R.A. Reid, and E. de Grosbois. 1987. The rate of acidification of aquatic ecosystems in
Ontario, Canada. Nature 329:45-48.
Dingman, S.L 1981. Elevation: A major influence on the hydrology of New Hampshire and Vermont,
USA. Hydro). Sci. Bull. 26:399-413.
Dise, N.B. 1984. A synoptic survey of headwater streams in Shenandoah National Park, VA, to evaluate
sensitivity to acidification by acid deposition. M.S. Thesis. Dept. Environmental Sciences, University
of Virginia, Charlottesville, VA. 135 pp.
Doll, C.G., W.M. Cady, J.B. Thompson, Jr., and M.P. Billings. 1961. Centennial Map of Vermont, (scale
1:250,000).
Doner, H.E., and W.C. Lynn. 1977. Carbonate, halide, sulfate, and sulfide minerals, pp. 75-98. In: J.B.
Dixon and S.B. Weed, eds. Minerals in Soil Environments. Soil Science Society of America,
Madison, Wl.
12-7
-------
Drabl0s, D., and I. Sevaldrud. 1980. Lake acidification, fish damage, and utilization of outfields: A
comparative survey of six highland areas, southeastern Norway, pp. 354-355. In: D. Drabl0s and
A. Tollan, eds. Ecological Impact of Acid Precipitation: Proceedings of an International Conference,
Sandefjord, March 11-14. SNSF Project, Oslo-As, Norway.
Drabl0s, D., and A. Tollan. 1980. Ecological Impact of Acid Precipitation. Proceedings of an International
Conference, Sandefjord, March 11-14. SNSF Project, Oslo-As, Norway.
Draper, N.R., and H. Smith. 1981. Applied Regression Analysis. John Wiley & Sons, Inc., New York. NY.
Drever, J.I. 1982. The Geochemistry of Natural Waters. Prentice-Hall, Inc., Englewood Cliffs, NJ. 388 pp.
Driscoll, C.T. 1980. Chemical characterization of some dilute acidified lakes and streams in the Adirondack
Region of New York State. Ph.D. Dissertation. Cornell University, Ithaca, NY.
Driscoll, C.T., and R.M. Newton. 1985. Chemical characteristics of Adirondack lakes. Environ. Sci. Technol.
19:1018-1024.
Driscoll, C.T., J.P. Baker, J.J. Bisogni, and C.L Schofield. 1980. Effect of aluminum speciation on fish
in dilute acidified waters. Nature 284:161-164.
Driscoll, C.T., C.P. Yatsko, and F.J. Unangst. 1987a. Longitudinal and temporal trends in the water
chemistry of the North Branch of the Moose River. Biogeochemistry 3:37-61.
Driscoll, C.T., B.J. Wyskowski, C.C. Consenting and M.E. Smith. 1987b. Processes regulating temporal
and longitudinal variations in the chemistry of a low-order woodland stream in the Adirondack
Region of New York. Biogeochemistry 3:225-241.
Driscoll, C.T., N.M. Johnson, G.E. Likens, and M.C. Feller. 1988. Effects of acidic deposition on the
chemistry of headwater streams: A comparison between Hubbard Brook, New Hampshire, and
Jamieson Creek, British Columbia. Water Resour. Res. 24:195-200.
Driscoll, C.T., R.D. Fuller, and W.P. Schecher. 1989a. The role of organic acids in the acidification of
surface waters on the eastern U.S. Water, Air, Soil Pollut. 43:21-40.
Driscoll, C.T., G.E. Likens, LO. Hedin, J.S. Eaton, and F.H. Bormann. 1989b. Changes in the chemistry
of surface waters: 25-year results at the Hubbard Brook Experimental Forest, NH. Environ. Sci.
Technol. 23:137-143.
DuMouchei, W.H., and G.J. Duncan. 1983. Using sample survey weights in multiple regression analyses
of stratified samples. J. Am. Statist. Assoc. 78:535-547.
Dunne, T., and LB. Leopold. 1978. Water in Environmental Planning. W.H. Freeman and Company, New
York, NY. 818 pp.
Eary, L.E., E.A. Jenne, LW. Vail, and D.C. Gin/in. 1989. Numerical models for predicting watershed
acidification. Arch. Environ. Contam. Toxicol. 18:29-53.
Eder, B.K., and R.L Dennis. In Revision. On the development of an inference technique used in the
estimation of dry deposition of Mg*. Ca+, Na+, and K+. Atmospheric Research and Exposure
Assessment Laboratory, U.S. Environmental Protection Agency, Research Triangle Park, NC.
Eder, B.K., D.H. Coventry, T.L Clark, and C.E. Bollinger. 1986. RELMAP: A Regional Lagrangian Model
of Air Pollution User's Guide. EPA/600/8-86/013. NTIS PB 86-171 394/AS. U.S. Environmental
Protection Agency, Research Triangle Park, NC.
12-8
-------
Eilers, J.M., G.E. Glass, K.E. Webster, and J.A. Rogalla. 1983. Hydrologic control of lake susceptibility to
acidification. Can. J. Fish. Aquat. Sci. 40:1896-1904.
Eilers, J.M., P. Kanciruk, R.A. McCord, W.S. Overton, L Hook, DJ. BHck, D.F. Brakke, P.E. Kellar, M.S.
DeHaan, M.E. Silverstein, and O.H. Landers. 1987. Characten'stics of Lakes in the Western United
States, Volume II: Data Compendium for Selected Physical and Chemical Variables.
EPA/600/3-86/054b. U.S. Environmental Protection Agency, Washington, DC.
Eilers, J.M., D.H. Landers, and D.F. Brakke. 1988. Chemical and physical characteristics of lakes in the
southeastern United States. Environ. Sci. Technol. 22:164-172.
Elassal, A.A., and V.M. Caruso. 1983. Digital elevation models. U.S. Geological Survey Circular 895-B.
40 pp.
Engstrom, D.R. 1987. Influence of vegetation and hydrology on the humus budgets of Labrador lakes.
Can. J. Fish. Aquat. Sci. 44:1306-1314.
Environmental Systems Research Institute. 1986. ARC/INFO User's Manual, Version 3.2. Redlands, CA.
Eralp, A.E., and M.B. Tomson. 1978. pH averaging. J. Water Pollut. Control Fed. 389-392.
Eshleman, K.N. 1988. Predicting regional episodic acidification of surface waters using empirical models.
Water Resour. Res. 24:1118-1126.
Eshleman, K.N., and H.F. Hemond. 1985. The role of organic acids in the acid-base status of surface
waters at Bickford watershed, Massachusetts. Water Resour. Res. 21:1503-1510.
Eshleman, K.N., and P.R. Kaufmann. 1988. Assessing the regional effects of sulfur deposition on surface
water chemistry: The Southern Blue Ridge. Environ. Sci. Technol. 22:685-690.
Evangelou, V.P., and R.E. Phillips. 1987. Sensitivity analysis on the comparison between the Gapon and
Vansdow exchange coefficients. Soil Sci. Soc. Am. J. 51:1473-1479.
Eyre. F.H. 1980. Forest cover types of the United States and Canada. Society of American Foresters,
Washington, DC.
Falkengren-Grerup, U., N. Linnermark, and G. Tyler. 1987. Changes in acidity and cation pools of south
Swedish soils between 1949 and 1985. Ctemosphere 16:2239-2248.
Fay, J.A., D. Golomb, and S. Kumar. 1986. Modeling of the 1900-1980 trend of precipitation acidity at
Hubbard Brook, New Hampshire. Atmos. Environ. 20:1825-1828.
Fernandez, I.J. 1985a. Acid deposition and forest soils: Potential impacts and sensitivity, pp. 223-239. In:
D.D. Adams and W. Page, eds. Acid Deposition - Environmental, Economic, and Policy Issues.
Plenum Publ. Corp., New York, NY.
Fernandez, I.J. I985b. Potential effects of atmospheric deposition on forest soils, pp. 237-250. In:
Symposium on the Effects of Air Pollutants on Forest Ecosystems. The Acid Rain Foundation, St.
Paul, MN.
Fernandez, I.J., and P.A. Kosian. 1987. Soil air carbon dioxide concentrations in a New England spruce-fir
forest. Soil Sci. Soc. Am. J. 51:261-263.
Fisher, R.A. 1985. The logic of inductive inference (with discussion). J. Royal Stat. Soc. 98:39-54.
12-9
-------
Fitzgerald, J.W., and D.W. Johnson. 1982. Transformations of sulphate in forested and agricultural lands,
pp. 411-426. In: A.I. Moore, ed. Sulphur, Volume 1. British Sulphur Corp., London.
Fitzgerald, J.W., and M.E. Watwood. 1987. Sulfur retention in hardwood forest litter and soil: Mechanisms
and environmental parameters, pp. 119-126. In: NAPAP Aquatic Effects Task Group VI Peer Review
Project Summaries, Volume I, May 17-23. New Orleans, LA.
Fitzgerald, J.W., T.C. Strickland, and W.T. Swank. 1982. Metabolic fate of inorganic sulphate in soil
samples from undisturbed and managed forest ecosystems. Soil Bid. Biochem. 14:529-536.
Retcher, R., and M.J.D. Powell. 1963. A rapid convergent descent method for minimization. Computer
J. 6:163-168.
Forsberg, C., G. Moriing, and R.G. Wetzel. 1985. Indications of the capacity for rapid reversibility of lake
acidification. Ambio 14:164-6.
Foss, W.M. 1953a. After the storm. NY State Conserv. 7:2-3.
Foss, W.M. 1953b. Salvage Report (12). NY State Conserv. 7:4-5.
Freney, J.R. 1961. Some observations of the nature of organic sulphur compounds in soil. Aust. J. Agric.
Res. 12:424-432.
Freund, R.J., and R.C. Uttell. 1986. SAS System for Regression. SAS Institute, Inc., Gary, NC. 167 pp.
Fry, B. 1986. Stable sulfur isotopic distributions and sulfate reduction in lake sediments of the Adirondack
Mountains, New York. Biogeochemistry 2:329-343.
Fuller, W.A. 1987. Measurement Error Models. John Wiley & Sons, Inc., New York, NY. 440 pp.
Fuller, R.D., M.B. David, and C.T. Driscoll. 1985. Sulfate adsorption relationships in forested Spodosols
of the northeastern USA. Soil Set. Soc. Am. J. 49:1034-1040.
Fuller, R.O., C.T. Driscoll, S.C. Schindler, and M.J. Mitchell. 1986a. A simulation model of sulfur
transformations in forested Spodosols. Biogeochemistry 2:313-328.
Fuller, R.D., M.J. Mitchell, H.R. Krouse, BJ. Wyskowski, and C.T. Driscoll. 1986b. Stable sulfur isotope
ratios as a tool for interpreting ecosystem sulfur dynamics. Water, Air, Soil Pollut. 28:163-171.
Fuller, R.D., C.T. Driscotl, G.B. Lawrence, and S.C. Nodvin. 1987. Processes regulating sulphate flux after
whole-tree harvesting. Nature 325:707-710.
Furrer, G., J. Westall, and P. Sollins. 1989. The study of soil chemistry through quasi-steady-state models:
I. Mathematical definition of model. Geochim. Cosmochim. Acta. 53:595-601.
Gaines, Jr., G.J., and H.C. Thomas. 1953. Adsorption studies on clay minerals. II. A formulation of the
thermodynamics of exchange adsorption. J. Chem. Phys. 21:714-718.
Galloway, J.N., C.L Schofield, G.R. Hendrey, N.E. Peters, and A.H. Johannes. 1980. Sources of acidity
in three lakes acidified during snowmelt, pp. 264-265. In: D. Drabl0s and A. Tollan, eds. Ecological
Impact of Acid Precipitation, Proceedings of an International Conference, Sandefjord, March 11-14.
SNSF Project, Oslo-As, Norway.
Galloway, J.N., G.E. Likens, W.C. Keene, and J.M. Miller. 1982. The composition of precipitation in remote
areas of the world. J. Geophys. Res. 87:8771 -8787.
t2-tO
-------
Galloway, J.N., S.A. Norton, and M.R. Church. 1983a. Freshwater acidification from atmospheric deposition
of sulfuric acid: A conceptual model. Environ. Sci. Technol. 17:541-545.
Galloway, J.N., C.L Schofield, N.E. Peters, G.R. Hendrey, and E.R. Altwicker. 1983b. Effect of atmospheric
sulfur on the composition of three Adirondack lakes. Can. J. Fish. Aquat. Sci. 40:799-806.
Galloway, J.N., G.E. Likens, and M.E. Hawley. 1984. Acid precipitation: Natural versus anthropogenic
components. Science 226:829-831.
Galloway, J.N., G.R. Hendrey, C.L Schofield, N.E. Peters, and A.H. Johannes. 1987. Processes and
causes of lake acidification during spring snowmelt in the west-central Adirondack Mountains, New
York. Can. J. Fish. Aquat. Sci. 44:1595-1602.
Gapon, Y.N. 1933. On the theory of exchange adsorption in soils. J. Gen. Chem. USSR 3:144-160.
Garrels, R.M., and F.T. Mackenzie. 1967. Origin of the chemical compositions of some springs and lakes.
pp. 222-242. In: Advances in Chemistry, Series 73. American Chemical Society, Washington, DC.
Garrison, et al. 1987. Application of the ILWAS model to the Northern Great Lakes States. Lake Reserv.
Manage.
Gatz, D.F., W.R. Barnard, and G.J. Stensland. 1986. The role of alkaline materials in precipitation
chemistry: A brief review of the issues. Water, Air, Soil Pollut. 30:245-251.
George, T.H. 1986. Aerial verification of polygonal resource maps: A low-cost approach to accuracy
assessment. Photogram. Engin. Remote Sensing 52:839-846.
Georgakakos, K.P., G.M. Valle-Filho, N.P. Nikolaidis, and J.L Schnoor. In Press. Lake acidification studies:
The role of input uncertainty in long-term predictions. Water Resour. Res.
Gherini, S.A., L Mok, R.J. Hudson, G.F. Davis, C.W. Chen, and R.A. Goldstein. 1985. The ILWAS model:
Formulation and application. Water, Air, Soil Pollut. 26:425-459.
Gherini, S.A., R.A. Munson, E. Altwicker, R. April, C. Chen, N. Clesceri, C. Cronan, C. Driscoll, R.J.
Johannes, R. Newton, N. Peters, and C. Schofield. 1989. Regional Integrated Lake-Watershed Study
(RILWAS): Summary of Major Findings. EPRI RP-2174-1. Electric Power Research Institute, Palo
Alto, CA.
Gilbert, D.A., T.H. Sagraves, M.M. Lang, R.K. Munson, and S.A. Gherini. 1988. Blue Lake Acidification
Study. Pacific Gas and Electric.
Glover, G.M., and A.H. Webb. 1979. Weak and strong acids in the surface waters of the Tovdal region
in southern Norway. Water Res. 18:781-783.
Gobran, G.R., and E. Bosatta. 1988. Cation depletion rate as a measure of soil sensitivity to acidic
deposition: Theory. Ecol. Model. 40:25-36.
Goldstein, R.A., and S.A. Gherini, eds. 1984. The Integrated Lake-Watershed Acidification Study. Volume
4: Summary of Major Results. EA-3221, Research Project 1109-5.
Goldstein, R.A., S.A. Gherini, C.W. Chen, L Mok, and R.J.M. Hudson. 1984. Integrated acidification study
(ILWAS): A mechanistic ecosystem analysis. Phil. Trans. R. Soc. Lond. 3058:409-425.
Goldstein, R.A., C.W. Chen, and S.A. Gherini. 1985. Integrated lake-watershed acidification study:
Summary. Water, Air, Soil Pollut. 26:327-337.
12-11
-------
Goldstein, R.A., S.A. Gherini, C.T. Driscoll, R. Aprfl, C.L Schofleld, and C.W. Chen. 1987. Lake-watershed
acidification in the North Branch of the Moose River: Introduction. Biogeochemistry 3:5-20.
Gorham, E. 1955. On the acidity and salinity of rain. Geochim. Cosmochim. Acta 7:231-239.
Gorham, E., and N.E. Detenbeck. 1986. Sulphate in bog waters: A comparison of ion chromatography
with Mackereth's cation-exchange technique and a revision of earlier views on cause of bog acidity.
Ecology 74:899-903.
Gorham, E., P.M. Vitousek, and W.A. Reiners. 1979. The regulation of chemical budgets over the course
of terrestrial ecosystem succession. Ann. Rev. Ecology Systematics 10:53-84.
Gorham, E., F.B. Martin, and J.T. LJtzau. 1984. Acid rain: Ionic correlations in the eastern United States.
Science 225:407-409.
Gorham, E., S.J. Eisenreich, J. Ford, and M.V. Santelmann. 1985. The chemistry of bog waters, pp.
339-363. In: W. Stumm, ed. Chemical Processes in Lakes. Wiley-lnterscience, New York, NY.
Gorham, E., J.K. Underwood, F.B. Martin, and J.G. Ogden, III. 1986. Natural and anthropogenic causes
of lake acidification in Nova Scotia. Nature 324:451 -453.
Graczyk, D J., W.A. Gebert, W.R. Krug, and G.J. Allord. 1988. Maps of Runoff in the Northeastern Region
and the Southern Blue Ridge Province of the United States During Selected Periods in 1983-85.
I U.S. Geological Survey Open-File Report 87-106, Washington, DC. 2 plates. 8 pp.
Greb, S., R.K. Munson, S.A. Gherini, LE. Gomez, C.W. Chen, P.J. Garrison, and D.R. Knauer. 1987.
RILWAS-Wisconsin: Simulation of the Effects of Acid Deposition on Crystal, Vandercook, and Little
Rock Lakes. WDNR Technical Bulletin.
Grennfelt, P. 1987. Deposition processes for acidifying compounds. Environ. Technol. Letters 8:515-527.
Groterud, O. 1984. A conceptual model of lake acidification. Verh. Internal. Verein. Umnol. 22:686-691.
Gschwandtner, G., K.C. Gschwandtner, and K. Eldridge. 1985. Historic emissions of sulfur and nitrogen
oxides in the United States from 1900 to 1980. EPA/600/7-85/009a. U.S. Environmental Protection
Agency, Research Triangle Park, NC. 105 pp.
Haines, T.A. 1981. Acidic precipitation and its consequences for aquatic ecosystems: A review. Trans.
Am. Fish. Soc. 110:669-707.
Haines, T.A., and J.J. Akielaszek. 1983. A regional survey of the chemistry of headwater lakes and
streams in New England: Vulnerability to acidification. In: Air Pollution and Acid Rain. FWS/OBS-80
#15. U.S. Dept. Interior, Fish and Wildlife Service, Washington, DC. 141 pp.
Hall, R.J., G.E. Likens, S.B. Fiance, and G.R. Hendrey. 1980. Experimental acidification of a stream in
Hubbard Brook Experimental Forest, New Hampshire. Ecology 61:976-989.
Hammond, E.H. 1964. Classes of land surface form in the forty-eight states, Annals of the Assoc. of
American Geographers, 54(1) map supplement No. 4.
Haque, I., and D. Walmsley. 1973. Adsorption and desorption of sulfate in some soils of the West Indies.
Geoderma 9:269-278.
Hardeman, W.D. 1966. Geological Map of Tennessee. State of Tennessee Dept. Conservation, Division
of Geology, (scale 1:250,000).
12-12
-------
Haren, M.F., and R.D. Van Remortel. 1987. Direct/Delayed Response Project: Field Operations and
Quality Assurance Report for Soil Sampting and Preparation in the Southern Blue Ridge Province
of the United States, Volume II: Preparation. EPA/600/4-87/041. U.S. Environmental Protection
Agency, Environ. Monitoring Systems Lab., Las Vegas, NV. 69 pp.
Harvey, H.W. 1969. The Chemistry and Fertility of Sea Waters. Cambridge University Press. NY. 240 pp.
Hasan, S.M., R.L Fox, and C.C. Boyd. 1970. Solubility and availability of sorbed sulfate in Hawaiian
soils. Soil Sci. Soc. Am. Proc. 34:897-901.
Hayden, N.J. 1987. Suifate Retention and Release in Six Eastern Soils. M.S. Thesis. Dept. Civil and
Environ. Engineering, Michigan State University, East Lansing, Ml. 165 pp.
Hedin, LO., G.E. Ukens, and F.H. Bormann. 1987. Decrease in precipitation acidity resulting from
decreased sulfate concentration. Nature 325:244.
Heigeson, H.C., W.M. Murphy, and P. Aagaard. 1984. Thermodynamic and kinetic constraints on reaction
rates among minerals and aqueous solutions: II. Rate constants, effective surface area and the
hydrolysis of feldspar. Geochim. Cosmochim. Acta 48:2405-2432.
Heivey, J.D., and S.H. Kunkle. 1986. Input-Output Budgets of Selected Nutrients on an Experimental
Watershed Near Parsons, West Virginia. USDA Forest Service, Northeastern Forest Experiment
Station. Broomall, PA. 7 pp.
Hemond, H.F. 1980. Biogeochemistry of Thoreau's Bog, Concord, Massachusetts. Ecological Monographs
50:507-526.
Hemond, H.F., and K.N. Eshleman. 1984. Neutralization of acid deposition by nitrate retention at Bickford
Watershed, Massachusetts. Water Resour. Res. 20:1718-1724.
Hendrey, G.R., J.N. Galloway, S.A. Norton, C.L SchofieJd, P.W. Shaffer, and D.A. Burns. 1980. Geological
and Hydrochemical Sensitivity of the Eastern United States to Acid Precipitation. EPA/600/3-80/024.
U.S. Environmental Protection Agency, Washington, DC. 100 pp.
Henriksen, A. 1979. A simple approach for identifying and measuring acidification of freshwater. Nature
278:542-545.
Henriksen, A. 1980. Acidification of freshwater - a large scale titration, pp. 68-74. In: D. Drabl0s and A.
Tollan, eds. Ecological Impact of Acid Precipitation: Proceedings of an International Conference,
Sandefjord, March 11-14. SNFS Project, Oslo-As, Norway.
Henriksen, A. 1984. Changes in base cation concentrations due to freshwater acidification. Verh. Internal.
Verein. LJmnol. 22:692-698.
Henriksen, A., and D.F. Brakke. 1988. Increasing contributions of nitrogen to the acidity of surface
waters in Norway. Water, Air, Soil Pollut. 42:183.
i
Henriksen, A., and R.F. Wright. 1977. Effects of acid precipitation on a small acid lake in southern Norway
Nordic Hydrology 8:1-10.
Henriksen, A., D.F. Brakke, and S.A. Norton. 1988. Total organic carbon concentrations in acidic lakes
in southern Norway. Environ. Sci. Technol.
Herczeg, A.L, and R.H. Hesslein. 1984. Determination of hydrogen ion concentration in softwater lakes
using carbon dioxide equilibria. Geochim. Cosmochim. Acta 48:837-845.
12-13
-------
Herfihy. AT., A.L Mills, G.M. Hornberger, and A.E. Bruckner. 1987. The importance of sediment sulfate
reduction to the sulfate budget of an impoundment receiving acid mine drainage. Water Resour.
Res. 23:287-292.
Hewlett, J.D., and A.R. Hibbert. 1967. Factors affecting the response of small watersheds to precipitation
in humid areas, pp. 275-289. In: W.E. Sopper and H.W. Lull, eds. Symposium on Forest Hydrology.
Pergammon Press, New York, NY.
Hicks, B.B., R.P. Hosker, Jr., and J.D. Womack. 1986. Comparisons of wet and dry deposition derived
from the first year of trial dry deposition monitoring, pp. 486-490. In: Symposium on Acid Rain: I.
Sources and Atmospheric Processes Presented before the Division of Petroleum Chemistry, Inc.
American Chemical Society, April 13-18, New York, NY.
HilJman, D.C.J., J.F. Potter, and S.J. Simon. 1986. National Surface Water Survey: Eastern Lake Survey
(Phase I). Analytical Methods Manual. EPA/600/4-86/009. U.S. Environmental Protection Agency,
Las Vegas, NV.
Hingston, F.L, R.J. Atkinson, A.M. Posner, and J.P. Quirk. 1967. Specific adsorption of anions. Nature
215:1459-1461.
Hingston, F.J., A.M. Posner, and J.P. Quirk. 1972. Anion adsorption by geothite and gibbsite: I. The role
of the proton in determining adsorption envelopes. Soil Sci. 23:177-192.
Holdren, G.R., and C.I. LJff. in Preparation. An expanded version of Reuss' model for cation exchange
equilibria: A user's guide. U.S. Environmental Protection Agency, Corvallis, OR.
Holdren, G.R., and P.M. Speyer. 1985. pH dependent changes in the rates and stoichiometry of dissol
ution of an alkali feldspar at room temperature. Am. J. Sci. 285:994-1026.
Holdren, G.R., T.M. Brunelle, G. Matisoff, and M. Wahlen. 1984. Timing the increase in atmospheric
sulphur deposition in the Adirondack Mountains. Nature 311:245-248.
Holdren, G.R., C.I. Liff, and O.L Cassell. 1989. Cation exchange models and the prediction of soil
buffering capacity. In: R. Bassett and D. Melchior, eds. Chemical Modeling in Aqueous Systems II.
ACS Symposium Series, Washington, DC.
Hornbeck, J.W., and C.A. Federer. 1985. Estimating the buffer capacity of forest soils. J. Forestry
83:690-691.
Hornberger, G.M., K.J. Beven, B.J. Cosby, and D.E. Sappington. 1985. Shenandoah watershed study:
Calibration of a topography-based, variable contributing area hydrological model of a small forested
catchment. Water Resour. Res. 21:1841-1850.
Hornberger, G.M., B.J. Cosby, and J.N. Galloway. 1986. Modelling the effects of acid deposition:
Uncertainty and spatial variability in estimation of long-term sulfate dynamics in a region. Water
Resour. Res. 22:1293-1302.
Hornberger, G.M., B.J. Cosby, and R.F. Wright. 19873. Analysis of historical surface water acidification
in southern Norway using a regionalized conceptual model (MAGIC), pp. 127-132. In: M.B. Beck,
ed. Systems Analysis in Water Quality Management. Pergammon Press, New York, NY.
Hornberger, G.M., B.J. Cosby, and R.F. Wright. 1987b. A regional model of surface water acidification in
southern Norway: Uncertainty in the long-term hindcasts and forecasts. In: J. Kamari, ed.
Proceedings of the 11 ASA-IMG W Task Force Meeting on Environmental Impact Models to Assess
Regional Acidification. D. Reidel Publ. Co.
12-14
-------
Hosker, R.P., and J.D. Womack. 1986. Simple meteorological and chemical filterpack monitoring system
for estimating dry deposition of gaseous pollutants, pp. 23-29. In: Proceedings of the Fifth Annual
National Symposium on Recent Advances in Pollutant Monitoring of Ambient Air and Stationary
Sources, January. EPA/600/9-85/029. Atmospheric Turbulence and Diffusion Division,
NOAA/ERL/Air Resources Laboratory, Oak Ridge, TN.
Houghton, C., and FA Rose. 1976. Liberation of sulfate from suffate esters by soils. Appl. Environ.
Microbiol. 31:969-976.
Huckabee, J.W., C.P. Goodyear, and R.D. Jones. 1975. Acid rock in the Great Smokies: Unanticipated
impact on aquatic biota of road construction in regions of sulfide mineralization. Trans. Am. Fish.
SOC. 1040:677-684.
Huckabee, J.W., J.S. Mattice, LF. Pitelka, D.B. Porcella, and R.A. Goldstein. 1989. An assessment of
the ecological effects of acidic deposition. Arch. Environ. Contam. Toxicol. 18:3-27.
Huete, A.R., and J.G. McColl. 1984. Soil cation leaching by "acid rain" with varying nitrate-to-sulfate
ratios. J. Environ. Qua!. 13:366-371.
Hunsaker, C.T., S.W. Christensen, J.J. Beauchamp, R.J. Olson, R.S. Turner, and J.L Malanchuk. 1986a.
Empirical Relationships Between Watershed Attributes and Headwater Lake Chemistry in the
Adirondack Region. Environ. Sci. Div. Pub. No. 2884. ORNL/TM-9838. Oak Ridge National
Laboratory, Oak Ridge, TN. 123 pp.
Hunsaker, C.T.. J.L Malanchuk, R.J. Olson, S.W. Christensen, and R.S. Turner. 198Gb. Adirondack
headwater lake chemistry relationships with watershed characteristics. Water, Air, Soil Pollut.
31:79-88.
Hurlbut, Jr., C.S., and C. Klein. 1977. Manual of Mineralogy. John Wiley & Sons, Inc., New York, NY. 532
pp.
Husar, R.B. 1985. Manmade SOX and NOX emissions and trends for eastern North America 48 pp.
Husar, R.B. 1986. Emissions of sulfur dioxide and nitrogen oxides and trends for eastern North America.
pp. 48*92. In: Committee on Monitoring and Assessment of Trends in Acid Deposition, ed. Acid
Deposition: Long-Term Trends. National Academy Press, Washington, DC.
Hutchinson, T.C., and M. Havas. 1986. Recovery of previously acidified lakes near Coniston, Canada,
following reductions in atmospheric sulphur and metal emissions. Water, Air, Soil Pollut. 28:319-
333.
Hynes, H.B.N. 1975. Edgardo Baldi memorial lecture: The stream and its valley. Verh. Intemat. Verein.
Umnol. 19:1.
Isachsen, Y.W., and D.W. Fisher. 1970. Geologic Map of New York, Adirondack, Hudson-Mohawk and
Lower Hudson Sheets. New York State Museum and Science Service, Map and Chart Series No.
15. Albany, NY. (scale 1:250,000).
James, B.R., and S.J. Riha. 1986. pH buffering on forest soil organic horizons: Relevance to acid
precipitation. J. Environ. Qua). 15:229-234.
Jeffries, D.S., R.G. Semkin, R. Neuteuther, and M. Seymour. 1986. Influence of atmospheric deposition
on lake mass balances in the Turkey Lakes watershed, central Ontario. Water, Air, Soil Pollut.
30:1033-1044.
12-15
-------
Jeffries, D.S., J.R.M. Kelso, and I.K. Morrison. 1988. Physical, chemical, and biological characteristics of
the Turkey Lakes watershed, central Ontario, Canada. Can. Spec. Pub). Fish. Aquat. Sci. 45:3-13.
Jenne, E.A., LE. Eary, LW. Vail, D.C. Girvin, A.M. Liebetrau, LF. Hibler, T.B. Miley, and M.J. Monsour.
1989. An Evaluation and Analysis of Three Dynamic Watershed Acidification Codes (MAGIC, ETO,
and ILWAS). PNL-6687/UC-11. Richland, WA.
Johannes, A.H., and E.R. Altwicker. 1980. Atmospheric inputs to three Adirondack lake watersheds, pp.
256-257. In: D. Drabl0s and A. Tollan, eds. Ecological Impact of Acid Precipitation: Proceedings of
an International Conference, Sandefjord, March 11-14. SNSF Project, Oslo-As, Norway.
Johannes, A.H., J.N. Galloway, and D.E. Troutman. 1980. Snow pack storage and ion release, pp.
260-261. In: D. Drabl0s and A. Tollan, eds. Ecological Conference Impact of Acid Precipitation:
Proceedings of an International Conference, Sandefjord, March 11-14. SNSF Project, Oslo-As,
Norway.
Johannes, A.H., E.R. Altwicker, and N.L Clesceri. 1981. Characterization of acid precipitation in the
Adirondack Region. EPRI EA-1826. Electric Power Research Institute, Palo Alto, CA.
Johannes, A.H., E.R. Altwicker, and N.L Clesceri. 1984. Atmospheric Inputs to the ILWAS Watersheds in
the Integrated Lake-Watershed Acidification Study, Volume 4: Summary of Major Results. EPRI
EA-3221. Tetra Tech, Inc., Palo Alto, CA.
Johannes, A.H., E.R. Altwicker, and N.L Clesceri. 1985. The integrated lake-watershed acidification study:
Atmospheric inputs. Water, Air, Soil Pollut. 26:339-353.
Johnson, D.W. 1984. Sulfur cycling in forests. Biogeochemistry 1:29-43.
Johnson, D.W., and O.W. Cole. 1980. Anion mobility in soils: Relevance to nutrient transport from forest
ecosystems. Environ. Internal. 3:79-90.
Johnson, D.W., and G.S. Henderson. 1979. Sulfate adsorption and sulfur fractions In a highly weathered
soil under a mixed deciduous forest. Soil Sci. 128:34-40.
Johnson, D.W., and J.O. Reuss. 1984. Soil-mediated effects of atmospherically deposited sulphur and
nitrogen. Phil. Trans. R. Soc. Lond. 3058:383-392.
Johnson, D.W., and D.E. Todd. 1983. Relationships among iron, aluminum, carbon, and sulfate in a
variety of forest soils. Soil Sci. Soc. Am. J. 47:792-800.
Johnson, D.W., and D.E. Todd. 1987. Nutrient export by leaching and whole-tree harvesting in a loblolly
pine and mixed oak forest. Plant and Soil 102:99-109.
Johnson, D.W., J.W. Hornbeck, J.M. Kelly, W.T. Swank, and D.E. Todd. 1980. Regional patterns of soil
sulfate accumulation: Relevance to ecosystem sulfur budgets, pp. 507-519. In: D.S. Shriner, C.R.
Richmond and S.E. Lindberg, eds. Atmospheric Sulfur Deposition: Environmental Impact and Health
Effects. Ann Arbor Science, Ann Arbor, Ml.
Johnson, D.W., J. Turner, and J.M. Kelly. 1982a. The effects of acid rain on forest nutrient status. Water
Resour. Res. 18:449-461.
Johnson, D.W., G.S. Henderson, D.D. Huff, S.E. Lindberg, D.D. Richter, D.S. Shriner, D.E. Todd, and J.
Turner. I982b. Cycling of organic and inorganic sulfur in a chestnut oak forest. Oecologia
54:141-148.
12-16
-------
Johnson, D.W., G.S. Henderson, and D.E. Todd. I988a. Changes in nutrient distribution in forests and
soils of Walker Branch watershed, Tennessee, over an eleven-year period. Biogeochemistry
5:275-293.
Johnson, M.G. 1986. Clay mineralogy and chemistry of selected Adirondack Spodosols. Ph.D.
Dissertation. Cornell University, Ithaca, NY. (Diss. Abstr. 86-7348) 152 pp.
Johnson, M.G., P.W. Shaffer, D.L Stevens, K.W. Thornton, and R.S. Turner. 1988b. Predicting and
Forecasting Surface Water Acidification: A Plan to Assess Data Aggregation Effects. U.S.
Environmental Protection Agency, Environ. Res. Lab., Corvallis, OR. 93 pp.
Johnson, N.M., C.T. Driscoll, J.S. Eaton, G.E. Likens, and W.H. McDowell. 1981. "Acid rain," dissolved
aluminum and chemical weathering at the Hubbard Brook Experimental Forest, New Hampshire.
Geochim. Cosmochim. Acta 45:1421-1437.
Johnson, N.M., G.E. Likens, M.C. Feller, and C.T. Driscoll. 1984. Acid rain and soil chemistry. Science
225:1424-1425.
Johnson, R.A., and D.W. Wichern. 1982. Applied Multivariate Statistical Analysis. Prentice Hall, Inc.,
Englewood Cliffs, NJ. 594 pp.
Johnston, C.A., G.D. Bubenzer, G.B. Lee, F.W. Madison, and J.R. McHenry. 1984. Nutrient trapping by
sediment deposition in a seasonally flooded lakeside wetland. J. Environ. Qual. 13(2}:283-290.
Johnston, C.A., N.E. Detenbeck, J.P. Bonde, and G.J. Niemi. 1988. Geographic information systems for
cumulative impact assessment. Photogram. Engin. Remote Sensing 54:1609-1615.
Jones, K.E., G.K. Morris, C.R. Myers, B.T. Rhyne, and J.A. Watts. 1986. User's Guide for the Soil
Description Program: Southern Blue Ridge Province Soil Sampling Survey. Oak Ridge National
Laboratory, Computing and Telecomm. Div., Oak Ridge, TN. (Draft).
Kahl, J.S., and M. Scott 1987. The aquatic chemistry of Maine's high elevation lakes: Results from the
HELM project. In: Proceedings of the NALMS meeting, November. Orlando, FL
Kahl, J.S.. S.A. Norton, and J.S. Williams. 1984. Geologic aspects of acid deposition, pp. 23-35. In: O.P.
Bricker, ed. Ann Arbor Science, Ann Arbor, Ml.
Kanciruk, P., J.M. Eilers, R.A. McCord, D.H. Landers, D.F. Brakke, and R.A. Unthurst. 1986a.
Characteristics of Lakes in the Eastern United States, Volume III: Data Compendium of Site
Characteristics and Chemical Variables. EPA/600/4-86/007c. U.S. Environmental Protection Agency,
Washington, DC. 439 pp.
Kanciruk, P., M. Gentry, R. McCord, L Hook, J.M. Eilers, and M.D. Best. 1986b. National Surface Water
Survey: Eastern Lake Survey (Phase I), Data Base Dictionary. Environ. Services Div. Pub. No. 2778.
ORNL/TM-10307. Oak Ridge National Laboratory. Oak Ridge, TN. 77 pp.
Kaufmann, P.R., AT. Herlihy, J.W. Elwood, M.E. Mitch, W.S. Overton, M.J. Sale, J.J. Messer, K.A. Cougar,
D.V. Peck, K.H. Reckhow, A.J. Kinney. S.J. Christie, D.D. Brown, C.A. Hagley, and H.I. Jager. 1988.
Chemical Characteristics of Streams in the Mid-Atlantic and Southeastern United States, Volume I:
Population Descriptions and Physico-Chemical Relationships. EPA/600/3-88/021 a. U.S.
Environmental Protection Agency, Washington, DC. 397 pp.
Keller, W., and J.R. Pitbaldo. 1986. Water quality changes in Sudbury area lakes: A comparison of
synoptic surveys in 1974-1976 and 1981-1983. Water, Air and Soil Pollut. 29:285-296.
12-17
-------
Kelly, C.A., and J.W.M. Rudd. 1984. Epilimnetic sulfate reduction and its relationship to lake acidification.
Biogeochemistry 1:63-77.
Kelly, C.A., J.W.M. Rudd, R.B. Cook, and D.W. Schindler. 1982. The potential importance of bacterial
processes in regulating rate of lake acidification. LJmnol. Oceanogr. 27:868-882.
Kelly, C.A., J.W.M. Rudd, R.H. Hesslein, D.W. Schindler, P.J. Dillon, C.T. Driscoll, S.A. Gherini, and R.E.
Hecky. 1987. Prediction of biological acid neutralization in acid-sensitive lakes. Biogeochemistry
3:129-140.
Kerekes, J., S. Beauchamp, R. Tordon, C. Tremblay, and T. Pollock. 1986. Organic versus anthropogenic
acidity in tributaries of the Kejimkujik watersheds in western Nova Scotia. Water, Air, Soil Pollut.
31:165-173.
Khanna, P.K., J. Prenzel, K.J. Meiwes, B. Ulrich, and E. Matzner. 1987. Dynamics of sulfate retention by
acid forest soils In an acidic deposition environment. Soil Sci. Soc. Am. J. 51:446-452.
King, P.B., R.B. Neuman, and J.B. Hadley. 1968. Geology of the Great Smoky Mountains National Park,
Tennessee and North Carolina. Geological Survey Professional Paper 587. U.S. Government Printing
Office, Washington, DC.
Knox, C.E., and T.J. Nordenson. 1955. Average annual runoff and precipitation in the New England-New
York area. U.S. Geological Survey Hydrologic Investigations Atlas HA-7. U.S. Government Printing
Office, Washington, DC. 6 pp.
Kramer, J.R. 1986. Re-evaluation of trends in surface water alkalinity in New York and New Hampshire.
Dept. Geology, McMaster University, Hamilton, Ontario, Canada.
Kramer, J.R., and S.S. Davies. 1988. Estimation of non-carbonate protolytes for selected lakes in the
Eastern Lake Survey. Environ. Sci. Technol. 22:182-184.
Kramer, J.R., and A. Tessier. 1978. Acidification of aquatic systems: A critique of chemical approaches.
Environ. Sci. Technol. 16:606-615.
Kramer, J.R., A.W. Andren, R.A. Smith, A.H. Johnson, R.B. Alexander, and G. Oehlert. 1986. Streams and
lakes, pp. 231-299. In: Committee on Monitoring and Assessment of Trends in Acid Deposition, ed.
Acid Deposition: Long-Term Trends. National Academy Press, Washington, DC.
Krug, E.C. 1987. Acid deposition/watershed interactions: The importance of humic substances and
biological processes. AAAS Annual Meeting, Manuscript.
Krug, E.C. 1989. Assessment of the theory and hypotheses of the acidification of watersheds. Illinois State
Water Survey Division, SWS Contract Report 457. 252 pp.
Krug, E.C., and C.R. Frink. 1983. Acid rain on acid soil: A new perspective. Science 221:520-525.
Krug, E.C., P.J. Isaacson, and C.R. Frink. 1985. Appraisal of some current hypotheses describing
acidification of watersheds. J. Air Pollut. Contr. Assoc. 35:109-114.
Krug, W.R., W.A. Gebert, D.J. Graczyk, D.L Stevens, B.P. Rochelle, and M.R. Church. In Press. Runoff
Maps for the Northeastern, Southeastern, and Mid-Atlantic United States for 1951-1980. U.S.
Geological Survey Water Resour. Invest. Report No. 88-4094.
Kulp, L 1987. Interim Assessment, The Causes and Effects of Acidic Deposition, Volume I: Executive
Summary. National Acid Precipitation Assessment Program, Washington, DC.
12-18
-------
Lam, D.C.L, A.G. Bobba, D.S. Jeffries, and D. Craig. 1988. Modelling stream chemistry for the Turkey
Lakes Watershed: Comparison with 1981-84 data. Can. J. Fish, and Aquatic Sci. 45:72-80.
Lammers, D.A., D.L Cassell, J.J. Lee, O.L Stevens, W.G. Campbell, and M.G. Johnson. 1987a. Field
Operations and Quality Assurance/Quality Control for Direct/Delayed Response Project Soil
Mapping Activities in the Southern Blue Ridge Region. EPA/600/3-88/016. NTIS PB88 195 722/AS.
U.S. Environmental Protection Agency, Corvallis, OR. 122 pp.
Lammers, D.A., D.L Cassell, J.J. Lee, D.L Stevens, R.S. Turner, W.G. Campbell, and M.G. Johnson.
1987b. Field Operations and Quality Assurance/Quality Control for Direct/Delayed Response Project
Soil Mapping Activities in the Northeast Region. EPA/600/3-87/017. U.S. Environmental Protection
Agency, Washington, DC. 127 pp.
Lammers, DA, D.L Stevens, and M.G. Johnson. 1987c. Soil mapping data management system. In:
Agronomy Abstracts, American Society of Agronomy, Atlanta, November 29 - December 4, 1987.
Lammers, D.A., J.J. Lee, M.G. Johnson, and S.W. Buol. In Review. Classifying soils for acidic deposition
aquatic effects: A scheme for the Southern Blue Ridge Province.
Landers, D.H., M.B. David, and M.J. Mitchell. 1983. Analysis of organic and inorganic sulfur constituents
in sediments, soils and water. Intern. J. Environ. Anal. Chem. 14:245-256.
Landers, D.H., J.M. Eilers, D.F. Brakke, W.S. Overton, P.E. Kellar, M.E. Silverstein, R.D. Schonbrod, R.E.
Crowe, R.A. U'nthurst, J.M. Omemik, S.A. Teague, and E.P. Meier. 1987. Characteristics of Lakes
in the Western United States, Volume I: Population Descriptions and Physico-Chemical Relationships.
EPA/600/3-86/054a. U.S. Environmental Protection Agency, Washington, DC. 176 pp.
Landers, D.H., W.S. Overton, R.A. Linthurst, and D.F. Brakke. 1988. Eastern Lake Survey: Regional
estimates of lake chemistry. Environ. Sci. Technol. 22:128-135.
Langmuir, D. 1981. The power exchange function: A general model for metal adsorption onto geological
materials, pp. 1-18. In: P.H. Tewari, ed. Adsorption from Aqueous Solutions. Plenum Press, New
York, NY.
Lasaga, A.C. 1984a. Chemical kinetics of water-rock interactions. J. Geophys. Res. 89:4009-4025.
Lasaga, A.C. 1984b. Kinetics of silicate dissolution, pp. 269-274. In: Fourth International Congress on
Water/Rock Interactions. Edmonton, Canada.
Lau, W.M., and S.J. Mainwaring. 1985. The determination of soil sensitivity to acid deposition. Water, Air,
Soil Pollut. 25:451-464.
Lawrence, G.B., and C.T. Driscolf. 1988. Aluminum chemistry downstream of a whole-tree-harvested
watershed. Environ. Sci. Technol. 22:1293-1299.
LaZerte, B.D., and P.J. Dillon. 1984. Relative importance of anthropogenic versus natural sources of
acidity in lakes and streams of central Ontario. Can. J. Fish. Aquat. Sci. 41:1664-1677.
Lee, J.J., DA Lammers, M.G. Johnson, M.R. Church, D.L Stevens, D.S. Coffey, R.S. Turner, LJ. Blurhe,
L.H. Liegel, and G.R. Holdren. I989a. Watershed surveys to support an assessment of the regional
effect of acidic deposition on surface water chemistry. Environ. Manage. 13:95-108.
Lee, J.J., D.A. Lammers, D.L Stevens, K.W. Thornton, and K.A. Wheeler. 1989b. Classifying soils for
acidic deposition aquatic effects: A scheme for the northeastern U.S. Soil Sci. Soc. Am. J. 53:1153-
1163.
12-19
-------
Lee, J.J., O.R. Marmorek, K.W. Thornton, D.L Stevens, and DA Larimers. 1989c. Direct/Delayed
Response Project: Definition of Soil Sampling Classes and Selection of Sampling Sites for the
Northeast. EPA/600/3-89/041. U.S. Environmental Protection Agency, Environmental Research
Laboratory, Corvallis, OR.
Lee, S. 1987. Uncertainty Analysis for Long-Term Acidification of Lakes in Northeastern USA. Ph.D. Thesis.
University of Iowa, Iowa City.
Lee, S., and J.L Schnoor. 1988. Reactions that modify chemistry in lakes of the NSWS. Environ. Sci.
Technol. 22:190-197.
Lee, S.J., K.P. Georgakakos, and J.L Schnoor. An uncertainty analysis of long-term lake alkalinity
predictions. Water Resour Res. (Submitted).
Lepisto, A., P.O. Whitehead, C. Neal, and BJ. Cosby. 1988. Modeling effects of acid deposition: Estim
atlon of long-term water quality responses in forested catchments in Finland. Nordic Hydrology
19:99-120.
Levine, E.R., and E.J. Ciolkosz. 1986. A computer simulation model for soil genesis applications. Soil Sci.
Soc. Am. J. 50:661-667.
Levine, E.R., and E.J. Ciolkosz. 1988. Computer simulation of soil sensitivity to acid rain. Soil Sci. Soc.
Am. J. 52:209-215.
Lewis, Jr., W.M., and M.C. Grant. 1979. Change in the output of ions from a watershed as a result of the
acidification of precipitation. Ecology 60:1093-1097.
Liege), L.H., M.R. Church, D.L Cassell, W.G. Campbell, D.J. Bogucki, G.K. Gruendling, and E.B. Allen. In
Review. Using color infrared aerial photography in regional acid precipitation research. Photogram.
Engin. Remote Sensing.
LJesko, I., K. Zotter, and F. Szakal. 1987. Clarification of surface waters affected by acid rain, pp. 391-398.
In: R. Perry, R.M. Harrison, J.N.B. Bell and J.N. Lester, eds. Acid Rain: Scientific and Technical
Advances. Selper, Ltd., Ealing, London.
Likens, G.E., F.H. Bormann, R.S. Pierce, J.S. Eaton, and N.M. Johnson. 1977. Biogeochemistry of a
Forested Ecosystem. Springer-Verlag, Inc., New York, NY. 146 pp.
Lin, J.C., and J.L Schnoor. 1986. Acid precipitation model for seepage lakes. J. Environ. Engineer.
112:677-694.
Un, J.C., J.L Schnoor, and G.E. Glass. 1987. Ion budgets in a seepage lake. pp. 209-227. In: R.A. Heits
and S.J. Eisenreich, eds. Sources and Fates of Aquatic Pollutants. Advances in Chemistry, Series
216, American Chemical Society, Washington, DC.
LJndberg, S.E., and C.T. Garten, Jr. 1988. Sources of sulphur in forest canopy throughfall. Nature
336:148-151.
Lindberg, S.E., G.M. Lovett, D.D. Richter, and D.W. Johnson. 1986. Atmospheric deposition and canopy
interactions of major ions in a forest. Science 231:141-145.
Lins, H.F. 1986. Recent patterns of sutfate variability in pristine streams. Atmos. Environ. 20:367-375.
12-20
-------
Linthurst, R.A., D.H. Landers, J.M. Eilers, D.F. Brakke, W.S. Overton, E.P. Meier, and R.E. Crowe. 1986a.
Characteristics of Lakes in the Eastern United States, Volume I: Population Descriptions and
Physico-Chemical Relationships. EPA/600/4-86/007a. U.S. Environmental Protection Agency,
Washington, DC. 136 pp.
Unthurst, R.A., D.H. Landers, J.M. Eilers, P.E. Kellar, D.F. Brakke, W.S. Overton, R.E. Crowe, E.P. Meier,
P. Kanciruk, and D.S. Jeffries. 1986b. Regional chemical characteristics of lakes in North America:
Part II - Eastern United States. Water, Air, Soil Pollut. 31:577-591.
Loucks, O.L, G.E. Glass, J.A. Sorensen, B.W. Liukkonen, J. Allert, and G. Rapp, Jr. 1986. Role of
precipitation chemistry versus other watershed properties in Wisconsin lake acidification. Water, Air,
Soil Pollut 31:67-77.
Lull, H.W., and W.E. Sopper. 1966. Factors that influence streamflow in the Northeast. Water Resour. Res.
2:371-379.
Lynch, D.D., and N.B. Dise. 1985. Sensitivity of Stream Basins in Shenandoah National Park to Acid
Deposition. USGS Water Resources Investigation Report 85-4115. U.S. Geological Survey,
Washington, DC, and Dept. Environ. Sci., University of Virginia, Richmond, VA. 61 pp.
Madansky, A. 1988. Prescriptions for Working Statisticians. Springer-Verlag, Inc., New York, NY. 295 pp.
Mairs, D.F. 1967. Surface chloride distribution in Maine lakes. Water Resour. Res. 3:1090-1092.
Mansell, R.S., S.A. Bloom, H.M. Selim, and R.D. Rhue. 1986. Multispecies cation leaching during
continuous displacement of electrolyte solutions through soil columns. Geoderma 38:61-75.
Martin, H.C. 1986. Acidic Precipitation. Part 1, 2. In: Proceedings of the International Symposium on
Acidic Precipitation, Muskoka, Ontario. D. Reidel Publ. Co., Boston, MA.
Mast, M.A., and J.I. Drever. 1987. The effect of oxalate on the dissolution rate of oligoclase and tremolite.
Geochim. Cosmochim. Acta 51:2559-2568.
Matalas, N.C., and B. Jacobs. 1964. A Correlation Procedure for Augmenting Hydrology Data. U.S.
Geological Survey Professional Paper 434, Washington, DC. 7 pp.
May, H., P.A. Helmke, and M.L Jackson. 1979. Gibbsite solubility and thermodynamic properties of
hydroxy-aluminum ions in aqueous solutions at 25 C. Geochim. Cosmochim. Acta 43:861-868.
McFee, W.W. 1980. Sensitivity of Soil Regions to Acid Precipitation. EPA/600/3-60/013. U.S.
Environmental Protection Agency, Washington, DC. 178 pp.
Messer, J.J., C.W. Ariss, J.R. Baker, S.K. Drouse, K.N. Eshleman, P.R. Kaufmann, RA Linthurst, J.M.
Omernik, W.S. Overton, M.J. Sale, R.D. Schonbrod, S.M. Stambaugh, and J.R. Tuschafl, Jr. 1986a.
National Stream Survey Phase I, Pilot Survey. EPA/600/4-86/026. U.S. Environmental Protection
Agency, Washington, DC. .179 pp.
Messer, J.J., D.H. Landers, R.A. Linthurst, and W.S. Overton. 1986b. Critical design and interpretive
aspects of the National Surface Water Survey. Lake Reserv. Manage. 30:463-469.
Miles, C.E. 1980. Geologic Map of Pennsylvania. Dept. Environmental Resources, Harrisburg, PA. (scale
1:250,000).
Mitchell, M.J., D.H. Landers, and D.F. Brodowski. 1981. Sulfur constituents of sediments and their
relationship to lake acidification. Water, Air, Soil Pollut. 16:351-359.
12-21
-------
Mitchell, M.J., D.H. Landers, D.F. Brodowski, G.B. Lawrence, and M.B. David. 1984. Organic and inorganic
sulfur constituents of the sediments of three New York lakes: Effect of site, sediment depth, and
season. Water, Air, Soil Pollut 21:231-245.
Mitchell, M.J., M.B. David, and A.J. Uutala. 1985. Sulfur distribution in lake sediment profiles as an index
of deposition^ patterns. Hydrobiofogia 121:121-127.
Mitchell, M.J., M.B. David, D.G. Maynard, and S.A. Telang. 1986. Sulfur constituents in soils and streams
of a watershed in the Rocky Mountains of Alberta. Can. J. For. Res. 16:315-320.
Mitchell, M.J., S.C. Schindler, J.S. Owen, and S.A. Norton. 1988. Comparison of sulfur concentrations
within lake sediment profiles. Hydrobiologia 157:219-229.
Mohnen, V.A. 1988. The challenge of acid rain. Scientific American 259:30-38.
Morgan, M.D., and R.E. Good. 1988. Stream chemistry in the New Jersey pinelands: The Influence of
precipitation and watershed disturbance. Water Resour. Res. 24:1091-1100.
Mortenson, D.C. 1989a. Geographic Information System Documentation of Watershed Data for
Direct/Delayed Response Project - Northeast Database. ERL-COR-519. EPA/600/3-89/001 and
Project Summary. U.S. Environmental Protection Agency, Corvallis, OR. 35 pp.
Mortenson, D.C. 1989b. Geographic Information System Documentation of Watershed Data for
Direct/Delayed Response Project - Southern Blue Ridge Province Database. ERL-COR-535.
EPA/600/3-89/002 and Project Summary. U.S. Environmental Protection Agency, Corvallis, OR. 25
pp.
Mulholland, P.J., and E.J. Kuenzler. 1979. Organic carbon export from upland and forested wetland
watersheds. Limnol. Oceanogr. 24:960-966.
Munson, R.K., S.A. Gherini, M.M. Lang, LE. Gomez, C.W. Chen, R.A. Goldstein, and D.R. Knauer. 1987.
ILWAS model applications: Response of various surface waters to deposition acidity, pp. 301-308.
In: R. Perry, R.M. Harrison, J.N.B. Bell, and J.N. Lester, eds. Acid Rain: Scientific and Technical
Advances. Selper Ltd., Ealing, London.
Murdoch, P.S., N.E. Peters, and R.M. Newton. 1984. The integrated lake-watershed acidification study. In:
S.A. Gherini and C.W. Chen, eds. Volume 2: Hydrologic Analysis. EA-3321.
Musgrove, T.J., P.G. Whitehead, B.J, Cosby, G.M. Hornberger, and R.F. Wright. 1987. Regional modeling
of catchments In the Galloway region in southwest Scotland. In: J. Kamari, ed. Proceedings of the
IIASA-IMGW Task Force Meeting on Environmental Impact Models to Assess Regional Acidification.
D. Reidel Publ. Co.
Naiman, R.J., J.M. Melillo, and J.E Hobbie. 1986. Ecosystem alterations of boreal forest streams by
beaver (Castor canadensis). Ecology 67:1254-1269.
Nair, D.R. 1984. Multiple Regression Analysis of Factors Affecting Alkalinity of Lakes in Northeastern
United States. M.S. Thesis. University of Iowa, Iowa City.
National Academy of Sciences. 1984. Acid deposition: Processes of lake acidification. Summary of a
Discussion. National Research Council Commission on Physical Sciences, Mathematics, and
Resources. Environmental Studies Board, Panel on Processes of Lake Acidification. National
Academy Press, Washington, DC. 11 pp.
12-22
-------
National Academy of Sciences. 1986. Acid Deposition: Long-Term Trends. National Research Council
Commission on Physical Sciences, Mathematics, and Resources. Environmental Studies Board.
National Academy Press, Washington, DC. 506 pp.
Neal, C., P.O. Whitehead, R. Neale, and B.J. Cosby. 1986. Modeling the effect of acid deposition and
conifer afforestation on stream acidity in the British uplands. J. Hydrol. 86:15-26.
Neary, B.P., and P.J. Dillon. 1988. Effects of sulphur deposition on lake-water chemistry in Ontario,
Canada Nature 333:340-343.
Neller, J.R. 1959. Extractable sulfate - sulfur in soils of Florida In relation to amount of clay in the profile.
Soil Sci. Soc. Am. J. 23:346-348.
Newton, R.M., and F.H. April. 1982. Surficial geologic controls on the sensitivity of two Adirondack lakes
to acidification. N. E. Environ. Sci. 1:143-150.
Newton, R.M., J. Weintraub, and R. April. 1987. The relationship between surface water chemistry and
geology in the North Branch of the Moose River. Biogeochemistry 3:21-35.
Nikolaidis, N. 1987. Modeling the direct versus delayed response of surface waters to acid deposition in
northeastern United States. Ph.D. Thesis. University of Iowa, Iowa City. 288 pp.
Nikolaidis, N.P., H. Rajaram, J.L Schnoor, and K.P. Georgakakos. 1988. A generalized soft water
acidification model. Water Resour. Res. 24:1983-1996.
Nikolaidis, N.P., J.L Schnoor, and K.P. Georgakakos. 1989. Modeling of long-term lake alkalinity
responses to acid deposition. J. Water Pollut. Control Fed. 61:188-199.
Nilsson, S.I., H.G. Miller, and J.D. Miller. 1982. Forest growth as a possible cause of soil and water
acidification: An examination of the concepts. Oikos. 39:40-49.
Nodvin, S.C., C.T. Driscoll, and G.E. Likens. 1986. The effect of pH on sulfate adsorption by a forest soil.
Soil Sci. 142:69-75.
Nordstrom, D.K. 1982. The effects of sulfate on aluminum concentrations in natural waters: Some stability
relations in the system AI2O3 - SO3 - H2O at 298 K. Geochim. Cosmochim. Acta 46:681-692.
Norton, S.A., J.J. Akielaszek, T.A. Halnes, K.J. Stromborg, and J.R. Longcore. 1982. Bedrock geologic
control of sensitivity of aquatic ecosystems in the United States to acidic deposition, pp. 1-13. In:
National Atmospheric Deposition Program.
Norton, S.A., M.J. Mitchell, J.S. Kahl, and G.F. Brewer. 1988. In-lake alkalinity generation by sulfate
reduction - a paleolimndogical assessment. Water, Air, Soil Pollut.
Nyborg, M. 1978. Sulfur pollution and soil, pp. 359-390. In: J.O. Nriagu, ed. Sulfur in the Environment,
Part II: Ecological Impacts. Wiley-lnterscience, New York, NY.
Nye, P.M., and DJ. Greenland. 1960. The Soil Under Shifting Cultivation. Commonwealth Bureau of Soils
Tech. Comm. No. 51. Commonwealth Agricultural Bureaux, Farnham Royal, Bucks.
Office of Technology Assessment. 1984. Acid Rain and Transported Air Pollutants: Implications for Public
Policy. OTA Report 0-204. U.S. Government Printing Office, Washington, DC. 323 pp.
Oliver, B.G., E.M. Thurman, and R.K. Malcolm. 1983. The contribution of humic substances to the acidity
of colored natural waters. Geochim. Cosmochim. Acta 47:2031-2035.
12-23
-------
Olsen, A.R., and C.R. Watson. 1984. Acid Deposition Annual Data Summaries: 1980, 1981, 1982.
EPA/600/7084/097. U.S. Environmental Protection Agency, Washington, DC.
Olsen, A.R., E.C. Voldner, D.S. Bigelow, W.H. Chan, T.L Clark, M.A. Lusis, P.K. Misra, and R.J. Vet. In
Press. Unified wet deposition data summaries for North America: Data summary procedures and
results for 1980-1986. Atmos. Environ.
Omernik, J.M., and A.J. Kinney. 1985. Total Alkalinity of Surface Waters: A Map of the New England and
New York Region. EPA/600/D-84/216. U.S. Environmental Protection Agency, Corvallis, OR.
Omemik, J.M., and C.F. Powers. 1983. Total alkalinity of surface waters - a national map. Ann. Assoc.
Am. Geogr. 73:133-136.
Osberg, P.H., A.M. Hussey, Jr., and G.M. Boone, eds. 1985. Bedrock Geologic Map of Maine. Dept.
Conservation, State of Maine, Augusta, (scale 1:500,000).
Osborne, LL, and M.J. Wiley. 1988. Empirical relationships between land use/cover and stream water
quality in an agricultural watershed. J. Environ. Manage. 26:9-27.
Overstreet, W.C., and H. Bell, III. 1965. Geologic Map of the Crystalline Rocks of South Carolina, (scale
1:250,000).
Overton, W.S. 1987. Phase II Analysis Plan, National Lake Survey - Working Draft. Technical Report No.
115, Department of Statistics, Oregon State University, Corvallis.
Overton, W.S., P. Kanciruk, LA. Hook, J.M. Eiiers, D.H. Landers, D.F. Brakke, DJ. Blick, Jr., R.A.
Unthurst, M.D. DeHaan, and J.M. Omernik. 1986. Characteristics of Lakes in the Eastern United
States, Volume II: Lakes Sampled and Descriptive Statistics for Physical and Chemical Variables.
EPA/600/4-86/007b. U.S. Environmental Protection Agency, Washington, DC.
Paces, T. 1973. Steady-state kinetics and equilibrium between ground water and granitic rock. Geochim.
Cosmochim. Acta 37:2641-2663.
Papp, M.L, and R.D. Van Remortel. 1987. Direct/Delayed Response Project: Field Operations and Quality
Assurance Report for Soil Sampling and Preparation in the Northeastern United States, Volume II:
Sample Preparation. EPA/600/4-87/030. U.S. Environmental Protection Agency, Environ. Monitoring
Systems Lab., Las Vegas, NV. 142 pp.
Parfitt, R.L, and R.S.C. Smart. 1978. The mechanism of sulfate adsorption on iron oxides. Soil Sci. Soc.
Am. Proc. 42:48-50.
Peters, N.E., and C.T. Driscoll. 1987. Hydrogeologic controls of surface-water chemistry in the Adirondack
Region of New York State. Biogeochemistry 3:163-180.
Peters, N.E., and P.S. Murdoch. 1985. Hydrogeologic comparison of an acidic-lake basin with a
neutral-lake basin in the west-central Adirondack Mountains, New York. Water, Air, Soil Pollut.
26:387-402.
Pickering, S.M., and J.B. Murray. 1976. Geologic Map of Georgia. Georgia Dept. Natural Resources"
Geologic and Water Resources Division, Georgia Geological Survey, Atlanta, (scale 1:500,000).
Pritchett, W.L., and R.F. Fisher. 1987. Properties and Management of Forest Soils. John Wiley & Sons,
Inc., New York, NY.
Quinn, A.W. 1971. Bedrock Geologic Map of Rhode Island. U.S. Geological Survey Bull. No. 1295, Plate
1. (scale 1:125,000).
12-24
-------
Rajan, S.S.S. 1978. Sulfate adsorbed on hydrous alumina, ligands displaced, and changes in surface
charge. Soil Sci. Soc. Am. J. 42:39-44.
Rajan, S.S.S. 1979. Adsorption and desorption of sulfate and charge relationships in aliophanic clays. Soil
Sci. Soc. Am. J. 43:65-69.
Rapp, G., J.D. Allert, B.W. Liukkonen, J.A. Use, O.L Loucks, and G.E. Glass. 1985. Acidic deposition and
watershed characteristics in relation to lake chemistry In northeastern Minnesota. Environ. Internal.
11:425-440.
Rapp, Jr., G., B.W. Liukkonen, J.D. Allert, J.A. Sorensen, G.E. Glass, and O.L. Loucks. 1987. Geologic
and atmospheric input factors affecting watershed chemistry in upper Michigan. Environ. Geol.
Water Sci. 9:155-171.
Rascher, C.M., C.T. Driscoll, and N.E. Peters. 1987. Concentration and flux of solutes from snow and
forest floor during snowmelt in the west-central Adirondack Region of New York. Biogeochemtstry
3:209-224.
Rawls, W.J., D.L Bakensiek, and K.E. Saxton. 1982. Estimation of soil water properties. Trans. ASAE
25:1316-1320, 1328.
Reuss, J.O. 1983. implications of the calcium-aluminum exchange system for the effect of acid
precipitation in soils. J. Environ, dual. 14:26-31.
Reuss, J.O., and D.W. Johnson. 1985. Effect of soil processes on the acidification of water by acid
deposition. J. Environ. Qua). 14:26-31.
Reuss, J.O., and D.W. Johnson. 1986. Acid Deposition and the Acidification of Soils and Waters.
Ecological Studies Volume 59. Springer-Vertag, Inc., New York, NY.
Reuss, J.O., N. Christophersen, and KM. Seip. 1986. A critique of models for freshwater and soil
acidification. Water, Air. Soil Pollut. 30:909-930.
Reuss, J.O., B.J. Cosby, and R.F. Wright. 1987. Chemical processes governing soil and water acidification.
Nature 329:27-32.
Reuss, J.O., P.M. Walthall, and R.W.E. Hooper. 1988. The relationship among aluminum solubility,
calcium-aluminum exchange, and pH in selected forest soils. Soil Sci. Soc. Am. J.
Reynolds, B., C. Neal. M. Homung, S. Hughes, and P.A. Stevens. 1988. Impact of afforestation on the
soil solution chemistry of Stagnopodzols in mid-Wales. Water, Air, Soil Pollut. 38:55-70.
Richter, D.D., P.J. Comer, K.S. King, H.S. Sawin, and D.S. Wright. 1988. Effects of low ionic strength
solutions on pH of acid forested soils. Soil Sci. Soc. Am. J. 52:261-264.
Ritchie, G.S.P., and P.J. Dolling. 1985. The role of organic matter in soil acidification. Aust. J. Soil Res.
23:569-576.,.
Rochelle, B.R., and M.R. Church. 1987. Regional patterns of sulfur retention in watersheds of the eastern
U.S. Water, Air, Soil Pollut. 36:61-73.
Rochelle, B.P., M.R. Church, and M.B. David. 1987. Sulfur retention at intensively studied sites in the U.S.
and Canada. Water, Air, Soil Pollut. 33:73-84.
Rochelle, B.P., M.R. Church, W.A. Gebert, D.J. Graczyk, and W.R. Krug. 1988. Relationship between
annual runoff and watershed area for the eastern U.S. Water Resour. Bull. 24:35-41.
12-25
-------
Rochelle, B.P., C.I. Lrff, W.G. Campbell, D.L Cassell, M.R. Church, and R.A. Nusz. In Press-a. Relating
map geomorphic/hydrologic parameters to surface water chemistry regionally: Relative to acidic
deposition. J. Hydrol.
Rochelle, B.P., D.L Stevens, and M.R. Church. In Press-b. Uncertainty analysis of runoff estimates from
a runoff contour map. Water Resour. Bull. 36:
Rodgers, J. 1985. Bedrock Geological Map of Connecticut. State of Connecticut, Natural Resources
Center, Dept. Environmental Protection, Hartford, CT. (scale 1:125,000).
Rogalla, J.A., P.L Brezonik, and G.E. Glass. 1986. Empirical models for lake acidification In the Upper
Great Lakes Region. Water, Air, Soil Pollut. 31:95-100.
Romesburg, H.C. 1984. Ouster analysis for researchers. Lifetime Learning Publications, Belmont, CA.
344 pp.
Romesburg, H.C., and K. Marshall. 1984. User's manual for CLUSTER/CLUSTID computer programs for
hierarchical cluster analysis. Lifetime Learning Publications, Belmont, CA. 89 pp.
Rosenbrock, H.H. 1960. An automatic method for finding the greatest or least value of a function.
Comput. J. 3:175-184.
Rudd, J.W.M., ed. 1987. Acidification of the Moose River system in the Adirondack Mountains of New
York State. Biogeochemistry 3:1-296.
Rudd, J.W.M., C.A. Kelly, and A. Furutani. 1986. The role of sulfate reduction in long term accumulation
of organic and inorganic sulfur in lake sediments. Limnol. Oceanogr. 36:1281-1291.
Rudd, J.W.M.. C.A. Kelly, V. St. Louis, R.H. Hesslein, A. Furutani, and M.H. Holoka. 1986. Microbial
consumption of nitric and sulfuric acids in north temperate lakes. Limnol. Oceanogr. 31:1267-1280.
Rudd, J.W.M., C.A. Kelly, D.W. Schindler, and M.S. Turner. 1988. Disruption of the nitrogen cycle in
acidified lakes. Science 240:1515-1517.
Rustad, S., N. Christophersen, and H.M. Seip. 1986. Model for streamwater chemistry of a tributary to
Harp Lake, Ontario. Can. J. Fish. Aquat. Sci. 43:625-633.
Sale, M.J., P.R. Kaufmann, H.I. Jager, J.M. Coe, K.A. Cougan, A.J. Kinney, M.E. Mitch, and W.S. Overton.
1988. Chemical Characteristics of Streams in the Mid-Atlantic and Southeastern United States,
Volume II: Streams Sampled, Descriptive Statistics, and Compendium of Physical and Chemical
Data. EPA/600/3-88/021 b. U.S. Environmental Protection Agency, Washington, DC. 597 pp.
Sanders, F.E., and P.B.H. Tinker. 1975. Adsorption of sulphate by a sandy loam soil (calcic cambisol).
Geoderma 13:317-324.
SAS Institute, Inc. 1985. SAS User's Guide: Statistics, Version 5 Edition. SAS Institute, Inc., Gary, NC.
956 pp.
SAS Institute, Inc. 1987. SAS/STAT Guide for Personal Computers, Version 6 Edition. SAS Institute, Inc.,
Cary, NC. 1028 pp.
SAS Institute, inc. 1988. SAS/STAT User's Guide, Release to 6.03 Edition. SAS Institute, Inc., Cary, NC.
Sawa, T. 1978. Information criteria for discriminating among alternative regression models. Econometrics
46:1273-1282.
12-26
-------
Schafran, G.C., and C.T. Driscoll. 1987a. Comparison of terrestrial and hypolimnetic sediment generation
of acid neutralizing capacity for an acidic Adirondack lake. Environ. Sci. Techno). 21:988-993.
Schafran, G.C., and C.T. Driscoll. I987b. Spatial and temporal variations in aluminum chemistry of a dilute
acidic lake. Biogeochemistry 3:105-119.
Schecher, W.D., and C.T. Driscoll. 1987. An evaluation of uncertainty associated with aluminum equilibrium
calculations. Water Resour. Res. 23:525*534.
Schindler, D.W. 1986. The significance of in-lake production of alkalinity. Water, Air, Soil Pollut. 30:931-946.
Schindler, D.W. 1988. The effects of acid rain on freshwater ecosystems. Science 239:149-157.
Schindler, D.W., M.A. Turner, M.P. Stainton, and G.A. LJnsey. 1986b. Natural sources of acid neutralizing
capacity in low-alkalinity lakes of the Precambrian Shield. Science 232:844-847.
Schindler, D.W., S.E.M. Kasian, and R.H. Hesslein. 1988. Biological damage to lakes of the midwestern
and northeastern USA from acid rain. Biosclence.
Schindler, S.C., M.J. Mitchell, T.J. Scott, R.D. Fuller, and C.T. Driscoll. 19863. Incorporation of 35S-sulfate
into inorganic and organic constituents of two forest soils. Soil Sci. Soc. Am. J. 40:457-462.
Schmoyer, D.D., G.J. Morris, A.E. Osborne-Lee, J.A. Watts, J.C. Goyert, and R.S. Turner. In Review.
User's Guide for the Soil Chemistry System. Oak Ridge National Laboratory, Computing and
Telecomm. Div., Oak Ridge, TN.
Schnabel, R.R. 1985. Nitrate concentrations in a small stream as affected by chemical and hydrologic
interactions in the riparian zone, pp. 263-282. In: D.L Correll, ed. Watershed Research Perspectives.
Smithsonian Institution Press, Washington, DC.
Schnrtzer, M. 1980. Effect of low pH on the chemical structure and reaction of humic substances, pp.
203-222. In: T.C. Hutchlnson and M. Havas, eds. NATO Conference Series 1: Ecology, Volume 4.
Plenum Press, New York, NY.
Schnoor, J.L, and W. Stumm. 1985. Acidification of aquatic and terrestrial systems, pp. 311-338. In: W.
Stumm, ed. Chemical Processes in Lakes. Wiley-lnterscience, New York, NY.
Schnoor, J.L, and W. Stumm. 1986. The role of chemical weathering in the neutralization of acidic
deposition. Schweiz. Z. Hydro!. 48:171-195.
Schnoor, J.L, W.D. Palmer, Jr., and G.E. Glass. 1985. Modeling impacts of acid precipitation for
northeastern Minnesota, pp. 155-173. In: J.L Schnoor, ed. Modeling of Total Acid Precipitation
Impacts. Ann Arbor Sciences, Butterworth Publishers, Boston, MA.
Schnoor, J.L, S. Lee, and N.P. Nikolaidis. 1986a. Lake resources at risk to acidic deposition in the
eastern United States. Water, Air, Soil Pollut. 31:1091-1101.
Schnoor, J.L, N.P. Nikolaidis, and G.E. Glass. 1986b. Lake resources at risk to acidic deposition in the
Upper Midwest. J. Water Pollut. Control Fed. 58:139-148.
Schofield, C.L 1976. Acidification of Adirondack Lakes by Atmospheric Precipitation: Extent and
Magnitude of the Problem. Final report, Project F-28-R. New York State Dept. Environ. Conserv.,
Albany, NY.
12-27
-------
Schofield, C.L, J.N. Galloway, and G.R. Hendrey. 1985. Surface water chemistry in the ILWAS basins.
Water, Air, Soil Pollut. 26:403-423.
Schott, J., and R.A. Berner. 1984. Dissolution mechanisms of pyroxenes and olivines during weathering.
pp. 35-54. In: J.I. Drever, ed. The Chemistry of Weathering. D. Reidel Publ. Co., Boston, MA.
Science Advisory Board. 1988. SAB-EEC Resolutions on the Use of Mathematical Models by EPA for
Regulatory Assessment and Decision-making. Memorandum to L Thomas, Administrator.
Environmental Protection Agency.
Searle, S.R. 1987. Unear Models for Unbalanced Data. John Wiley & Sons, Inc., New York, NY.
Seilkop, S.K., and P.L Finkeistein. 1986. Acid precipitation patterns and trends in eastern North America,
1980-84. J. dim. Appl. Meteorol. 26:980-994.
Seip, H.M. 1980. Acidification of freshwaters - sources and mechanisms, pp. 358-366. In: D. Drabl0s and
A. Tollan, eds. Ecological Impact of Acid Precipitation: Proceedings of an International Conference,
Sandefjord, March 11-14. SNSF Project, Oslo-As, Norway.
Shaffer, P.W., and M.R. Church. 1989. Terrestrial and in-lake contributions to the alkalinity budgets of
drainage lakes: An assessment of regional differences. Can. J. Fish. Aquat. Sci. 46:509-515.
Shaffer, P.W., R.P. Hooper, K.N. Eshleman, and M.R. Church. 1988. Watershed vs. in-lake alkalinity
generation: A comparison of rates using input-output studies. Water, Air, Soil Pollut. 39:263-273.
Shilts, W.W. 1981. Sensitivity of Bedrock to Acid Precipitation: Modification by Glacial Processes. Paper
81-44. Geological Survey of Canada, Ottawa.
Shreider, Y.A. 1988. The Monte Carlo Method. Pergammon Press, New York, NY.
Shriner, D.W., and G.S. Henderson. 1978. Sulfur distribution and cycling in a deciduous forest watershed.
J. Environ. Qual. 7:392-397.
Shriner, D.S., C.R. Richmond, and S.E. LJndberg. 1980. Atmospheric sulfur deposition, pp. 335-343. In:
Environmental Impact and Health Effects. The Butterworth Group, Boston, MA.
Simmons, C.E. 1976. Sediment Characteristics of Streams in the Eastern Piedmont and Western Coastal
Plain Regions of North Carolina. Water-Supply Paper 1798-0. U.S. Geological Survey, Washington,
DC. 32 pp.
Simons, T.J., and D.C.L Lam. 1980. Some limitations of water quality models for large lakes: A case
study of Lake Ontario. Water Resour. Res. 16:105-116.
Singh, B.R. 1984. Sulfate sorption by acid forest soils: 3. Desorption of sulfate from adsorbed surfaces
as a function of time desorbing ion, pH, and amount of adsorption. Soil Sci. 138:346-353.
Small, M.J., and M.C. Sutton. 1986. A direct distribution model for regional aquatic acidification. Water
Resour. Res. 22:1749-1758.
Smith. R.A. 1972. Air and Rain: The Beginnings of a Chemical Climatology. Longmans, Green, London.
Smith, R.A., and R.B. Alexander. 1983. Evidence for Acid-Precipitation-Induced Trends in Stream Chemistry
at Hydrologic Bench-Mark Stations. Geological Survey Circular 910. U.S. Geological Survey
Alexandria, VA. 12 pp.
Smith, R., and R. Alexander. 1986. Correlations between stream sulphate and regional SO2 emissions.
12-28
-------
Smith, R.A., R.B. Alexander, and M.Q. Wolman. 1987. Water-quality trends in the nation's rivers. Science
235:1607-1616.
Snedecor, G.W., and W.G. Cochran. 1967. Statistical Methods. Iowa State University Press, Ames. 593
pp.
Soil Survey Staff. 1975. Soil Taxonomy: A Basic System of Soil Classification for Making and Interpreting
Soil Surveys. USDA Soil Conservation Service, Agricultural Handbook No. 436. U.S. Government
Printing Office, Washington, DC. 754 pp.
Soil Survey Staff. 1981. Soil Survey Manual. USOA Soil Conservation Service, Agricultural Handbook No.
18. U.S. Government Printing Office, Washington, DC.
Soil Survey Staff. 1983. National Soils Handbook. USDA Soil Conservation Service, Agricultural Handbook
No. 430. U.S. Government Printing Office, Washington, DC.
Soil Survey Staff. 1984. Procedures for Collecting Soil Samples and Methods of Analysis for Soil Survey.
USDA Soil Conservation Service, SSIR 1. U.S. Government Printing Office, Washington, DC. 68 pp.
Sokal, R.R., and F.J. Rohlf. 1969. Biometry. W.H. Freeman and Co., San Francisco, CA.
Solomon, K.P., and T.E. Cerling. 1987. The annual carbon dioxide cycle in a montane soil: Observations,
modeling, and implications for weathering. Water Resour. Res. 25:2257-2265.
Sposito, G. 1977. The Gapon and Vanselow selectivity coefficients. Soil Sci. Soc. Am. J. 41:1205-1206.
Sposito. G. 1981. Cation exchange in soils: An historical and theoretical perspective, pp. 13-31. In:
Chemistry in the Soil Environment. ASA Special Publication, No. 40.
Sposito, G., and S.V. Mattigod. 1977. On the chemical foundation of the sodium adsorption ratio. Soil Sci.
Soc. Am. J. 41:323-329.
Steers, C.A.. and B.F. Hajek. 1969. Determination of map unit composition by a random selection of
transects. Soil Sci. Soc. Am. J. 43:156-160.
Stensland, G.J., and R.G. Semonin. 1982. Another interpretation of the pH trend in the United States. Bull.
Am. Meteorol. Soc. 63:1277-1293.
Stevenson, F.J. 1982. Humus Chemistry: Genesis, Composition, Reactions. Wiley-lnterscience, New York.
NY. 443 pp.
Strickland, T.C., and J.W. Fitzgerald. 1984. Formation and mineralization of organic sulfur in forest soils.
Biogeochemistry 1:79-95.
Strickland, T.C., J;W. Fitzgerald, J.T. Ash, and W.T. Swank. 1987. Organic sulfur transformations and sulfur
pool sizes in soil and litter from a Southern Appalachian hardwood forest. Soil Sci. 143:453-458.
Stuanes, A.O. 1984. A simple extraction as an indicator of a soil's sensitivity to acid precipitation. Acta
Agric. Scand. 34:113-127.
Stumm, W., and J.J. Morgan. 1981. Aquatic Chemistry: An Introduction Emphasizing Chemical Equilibria
in Natural Waters, 2nd Edition. Wiley-lnterscience, New York, NY. 780 pp.
Stumm, W., L Sigg, and J.L Schnoor. 1987. Aquatic chemistry of acid deposition. Environ. Sci. Technol.
21:8-13.
12-29
-------
Sullivan, T.J., C.T. Driscoll, J.M. Eilers, and D.H. Landers. 1988a. Evaluation of the role of sea salt inputs
in the long-term acidification of coastal New England lakes. Environ. Sci. Techno). 22:185-190.
Sullivan, T.J., J.M. Eilers, M.R. Church, D.J. Blick, K.N. Eshleman, D.H. Landers, and M.S. DeHaan. 1988b.
Atmospheric wet sulphate deposition and lakewater chemistry. Nature 331:607-609.
Sullivan, T.J., C.T. Driscoll, S.A. Gherini, R.K. Munson, R.B. Cook, D.F. Charles, and C.P. Yatsko. 1989.
Influence of aqueous aluminium and organic acids on measurement of acid neutralizing capacity
in surface waters. Nature 338:408-410.
Summers, P.W., V.C. Bowersox, and G.J. Stensland. 1986. The geographical distribution and temporal
variation of acidic deposition in eastern North America. Water, Air, Soil Pollut. 31:523-535.
Sverdrup, H., and P. Wafvinge. 1988. Weathering of primary silicate minerals in the natural soil
environment in relation to a chemical weathering model. Water, Air, Soil Pollut. 38:387-408.
Swank, W.T., and D.A. Crossley, Jr., eds. 1988. Ecological Studies. Volume 66: Forest Hydrology and
Ecology at Coweeta. Springer-Verlag, New York, NY. 469 pp.
Swank, W.T., and J.B. Waide. 1988. Characterization of baseline precipitation and stream chemistry and
nutrient budgets for control watersheds, pp. 57-79. In: W.T. Swank and D.A. Crossley, Jr., eds.
Ecological Studies, Volume 66: Forest Hydrology and Ecology at Coweeta. Springer-Verlag, New
York, NY.
Swank, W.T., J.W. Fitzgerald, and J.T. Ash. 1984. Microbial transformations of sulfate in forest soils.
Science 223:182-184.
Switzer, G.L, and LE. Nelson. 1972. Nutrient accumulation and cycling in loblolly pine (Pinus taeda L)
plantation ecosystems: The first twenty years. Soil Science Society of America Proceedings 36:143-
147.
Tang, A.J.S., W.H. Chan, D.H.S. Chung, and M.A. Lusis. 1986. The Acid Precipitation in Ontario Study:
Precipitation and Air Concentration and Wet and Dry Deposition Fields of Pollutants In Ontario,
1983. Report #ARB-OQ8-86-AQM, Ontario Ministry of the Environment.
Taylor, C.H., and D.C. Pierson. 1985. The effect of a small wetland on runoff response during spring
snowmelt. Atmosphere-Ocean 23:137-154.
Tenbus, F.J. 1987. Hydrological characteristics of the alluvial infill at the White Oak Run catchment,
Shenandoah National Park. M.S. Thesis. Dept. Environmental Sciences, University of Virginia,
Charlottesville, VA.
Thompson, M.E. 1987. Comparison of excess sulfate yields and median pH values of rivers in Nova
Scotia and Newfoundland, 1971-1973 and 1982-1984. Water, Air, Soil Pollut. 35:19-26.
Thompson, M.E., and M.S. Mutton. 1985. Sulfate in lakes of eastern Canada: Calculated yields compared
with measured wet and dry deposition. Water, Air, Soil Pollut. 24:77.
Thornton, K., J.P. Baker, D. Marmorek, D. Bernard, M.L Jones, P.J. McNamee, C. Wedeles, and K.N.
Eshleman. 1987. Episodic Response Project Draft Research Plan. U.S. Environmental Protection
Agency, Office of Research and Develop., Washington, DC.
Tiedmann, A.R., T.M. Quigley, and T.D. Anderson. 1988. Effects of timber harvest on stream chemistry
and dissolved nutrient losses in northeast Oregon. Forest Sci. 34:344-358.
12-30
-------
Tisdale, S.L, and W.L Nelson. 1975. Soil Fertility and Fertilizers. Macmillan Publ. Co.. New York, NY.
Tomovic, R. 1963. Sensitivity Analysis of Dynamic Systems. McGraw-Hill Book Co., New York, NY.
Turchenek, LW., S.A. Abboud, C.J. Thomas, R.J. Fessenden, and N. Holowaychuk. 1987. The Acid
Deposition Research Program: Biophysical Research, Effects of Acid Deposition on Soils in Alberta.
Kananaskis Centre for Environmental Research, University of Calgary, Alberta, Canada. 202 pp.
Turchenek, LW., SA Abboud, and N. Holowaychuk. 1988. Modeling soil response to acid deposition in
Alberta. Commun. Soil Sci. Plant Anal. 19:805-818.
Turk, J.T. 1984. An Evaluation of Trends in the Acidity of Precipitation and the Related Acidification of
Surface Water in North America. U.S. Geological Survey Water-Supply Paper 2249, Washington, DC.
Turk, J.T., and D.C. Campbell. 1987. Estimates of acidification of lakes in the Mt. Zirkel Wilderness Area,
Colorado. Water Resour. Res. 23:1757-1761.
Turner, J., and M.J. Lambert. 1986. Effects of forest harvesting nutrient removals on soil nutrient reserves.
Oecologia (Berlin) 70:140-148.
Turner, J.D., D.W. Johnson, and M.J. Lambert. 1980. Sulfur cycling in a Douglas-fir forest and its
modification by nitrogen application. Oecol. Plant 15:27-35.
Turner, R.S., A.H. Johnson, and D. Wang. 1985. Biogeochemistry of aluminum in McDonalds Branch
watershed, New Jersey Pine Barrens. J. Environ. Qual. 14:314-323.
Turner, R.S., C.C. Brandt, D.D. Schmoyer, J.C. Goyert, K.D. Van Hoesen, LJ. Allison, G.R. Holdren, P.W.
Shaffer, M.G. Johnson, D.A. Lammers, J.J. Lee, M.R. Church, LJ. Blume, and M.L. Papp. In Review.
Direct/Delayed Response Project: Data Base Users' Guide. ORNL/TM-10369. Oak Ridge National
Laboratory, Oak Ridge, TN.
Ulrich, B. 1983. Soil acidity and its relation to acid deposition. In: B. Ulrich, and J. Pankrath, eds. Effects
of Accumulation of Air Pollutants in Forest Ecosystems. D. Reidel Publ. Co., Boston, MA.
Ulrich, B., and E. Matzner. 1985. Anthropogenic and natural acidification in terrestrial ecosystems.
Experientia 42:344-350.
U.S. Department of Agriculture. 1981. Land Resource Regions and Major Land Resource Area of the
United States. Agriculture Handbook 296. Soil Conservation Service, Washington, DC. 156 pp.
U.S. Department of Interior. 1977. National Handbook of Recommended Methods for Water-Data
Acquisition. U.S. Geological Survey, Reston, VA.
U.S. Geological Survey. 1969. Water Resources Data, Virginia, Water Year 1968. U.S. Geological Survey
Water Report VA-68-1.
U.S. Geological Survey. 1970. Water Resources Data, Virginia, Water Year 1969. U.S. Geological Survey
Water Report VA-69-1.
Valentin!, J.L, and S.A. Gherini. 1986. Integrated Lake-Watershed Acidification Study: Volume 5. Database
Documentation. EPRI EA-3321. Palo Alto, CA.
Van Breemen, N., J. Mulder, and C.T. Driscoll. 1983. Acidification and alkalinization of soils. Plant and Soil
75:283-308.
VanLoon, G.W. 1984. Acid rain and soil. Can. J. Physiol. Pharmacol. 62:991-997.
12-31
-------
Van Miegroet, H., and D.W. Cole. 1984. The impact of nitrification on soil acidification and cation leaching
in a Red Alder ecosystem. J. Environ, dual, 13:586-590.
Van Remortel, R.D., G.E. Byers, J.E. Teberg, M.J. Miah, C.J. Palmer, M.L Papp, M.H. Bartling, A.D.
Tansey, D.L Cassell, and P.W. Shaffer. 1988. Direct/Delayed Response Project: Quality Assurance
Report for Physical and Chemical Analyses of Soils from the Southern Blue Ridge Province of the
United States. EPA/600/8-88/100. U.S. Environmental Protection Agency, Environ. Monitoring
Systems Lab., Las Vegas, NV. 263 pp.
Vanselow, A.P. 1932. Equilibria of the base-exchange reactions of bentonites, permutites, soil colloids, and
zeolites. Soil Sci. 33:95-113.
Velbel, MA 1985. Geochemical mass balances and weathering rates in forested watersheds of the
Southern Blue Ridge. Am. J. Sci. 285:904-930.
Velbel, M.A. 1986a. The mathematical basis for determining rates of geochemical and geomorphic
processes in small forested watersheds by mass balance: Examples and implications, pp. 439-452.
In: S.M. Colman and D.P. Dethier, eds. Rates of Chemical Weathering of Rocks and Minerals.
Academic Press, Orlando, FL
Velbel, M.A. 1986b. Influence of surface area, surface characteristics, and solution composition on feldspar
weathering rates, pp. 615-634. In: J.A. Davis and K.F. Hayes, eds. Geochemical Processes at
Mineral Surfaces. American Chemical Society Symposium Series No. 323, Washington, DC.
Velleman, P.F., and D.C. Hoaglin. 1981. Applications, Basics, and Computing of Exploratory Data Analysis.
Duxbury Press, Boston. 354 pp.
Vong, R., S. Cline, G. Reams, J. Bernert, D. Charles, J. Gibson, T. Haas, J. Moore, R. Husar, A. Olsen,
J. Simpson, and S. Seilkop. 1989. Regional Analysis of Wet Deposition for Effects Research.
EPA/600/3-89/030. U.S. Environmental Protection Agency, Environmental Research Laboratory,
Corvallis, OR. 57 pp.
Vorst, P., and T.C. Bell. 1977. Catchment geomorphology and its hydrologic significance: A review. Aust.
Water Resour. Council, Dept. Natural Resources. Representative Basin Program Report Series No.
2, Australia Government Publishing Service.
Vreeland, J.L 1983. The role of shallow groundwater in the hydrogeochemicai response of the White Oak
Run catchment, Shenandoah National Park. M.S. Thesis. Dept. Environmental Sciences, University
of Virginia, Chariottesville, VA.
Wampler, S.J., and A.R. Olsen. 1987. Spatial estimation of annual wet acid deposition using supplemental
precipitation data, pp. 248-251. In: Tenth Conference on Probability and Statistics, Oct. 4-5,
Edmonton, Atla., Canada. American Meteorological Society, Boston, MA.
Weaver, G.T., P.K. Khanna, and F. Beese. 1985. Retention and transport of sulfate in slightly acid forest
soil. Soil Sci. Soc. Am. J. 49:746-750.
Webb, J.R. 1987a. Retention of atmospheric sulfate by catchments in Shenandoah National Park, Virginia.
M.S. Thesis. Dept. Environmental Sciences, University of Virginia, Chariottesville, VA.
Webb, J.R. 1987b. Virginia trout stream sensitivity study. April, Synoptic Survey, Study Handbook. Dept.
Environmental Sciences, University of Virginia, Chariottesville, VA.
Weider, R.K., and G.E. Lang. 1988. Cycling of inorganic and organic sulfur in peat from Big Run Bog,
West Virginia. Biogeochemistry 5:221-242.
12-32
-------
Wetzel, R.G. 1975. Limnology. W.B. Saunders Co., Philadelphia, PA. 743 pp.
Whitehead, P.G., S. Bird, M. Hornung, BJ. Cosby, C. Neal, and P. Paricos. In Press-a. Stream
acidification trends in the Welsh uplands: A modeling study of the Uynbrianne catchments. J.
Hydrology.
Whitehead, P.G., B. Reynolds, M. Homung, B.J. Cosby, C. Neal, and P. Paricos. In Press-b. Modeling
long-term stream acidification trends in the upland Wales at Plynlimon. Process. Hydrology.
Wollast, R. 1967. Kinetics of the alteration of K-feldspar in buffered solutions at low temperature. Geochim.
Cosmochim. Acta 31:635-648.
Wolock, D.M. 1988. Topographic and soil hydraulic control of flow paths and soil contact time: Effects
on surface water acidification. Ph.D. Thesis. (Unpub).
Wolock, D.M., G.M. Homberger, K.J. Seven, and W.G. Campbell. 1989. The relationship of catchment
topography and soil hydraulic characteristics to lake alkalinity in the northeastern United States.
Water Resour. Res. 25:829-837.
Woodruff, J.F., and J.D. Hewlett. 1970. Predicting and mapping the average hydrologic response for the
eastern United States. Water Resour. Res. 6:1312-1326.
Wright, R.F. 1982. A model for streamwater chemistry at Birkenes, Norway. Water Resour. Res.
18:977-996.
Wright, R.F. 1988. Acidification of lakes in the eastern United States and southern Norway: A comparison.
Environ. Sci. Technol. 22:178-182.
Wright, R.F., and B.J. Cosby. 1987. Use of a process oriented model to predict acidification of a
manipulated catchment in Norway. Atmos. Environ. 21:727-730.
Wright, R.F., and A. Henriksen. 1978. Chemistry of small Norwegian lakes, with special reference to acid
precipitation. Limnol. Oceanogr. 23:487-498.
Wright, R.F., and A. Henriksen. 1983. Restoration of Norwegian lakes by reduction in sulphur deposition.
Nature 305:422-424.
Wright, R.F., B.J. Cosby, G.M. Hornberger, and J.N. Galloway. 1986. Comparison of paleolimnological with
MAGIC model reconstructions of water acidification. Water, Air, Soil Pollut. 30:367-380.
Wright, R.F., E. Lotse, and A. Semb. 1988. Reversibility of acidification shown by whole-catchment
experiments. Nature 334:670-675.
Yuretich, R.F., and G.L Batchelder. 1988. Hydrochemical cycling and chemical denudation in the Fort
River watershed, central Massachusetts: An appraisal of mass-balance studies. Water Resour. Res.
24:105-114.
Zen, E-an, ed. 1983. Bedrock Geologic Map of Massachusetts. U.S. Geological Survey, Washington, DC.
(scale 1:250,000).
12-33
-------
SECTION 13
GLOSSARY
13.1 ABBREVIATIONS AND SYMBOLS
13.1.1 Abbreviations
ADS
AERP
ANC
AREAL-RTP
CDF
Cl
CIR
Acid Deposition System
Aquatic Effects Research Program
Acid neutralizing capacity
Atmospheric Research and Exposure Assessment Laboratory - Research Triangle
Park, an EPA laboratory
Cumulative distribution function
Confidence interval
Color infrared photography
DDRP
DEM
DOC
DQO
ELS-I
EMSL-LV
EPA
EPRI
ERL-C
ERP
ETD
GIS
IBM PC
ILWAS
IQR
LAI
LTA
MAGIC
MLRA
NAS
NADP/NTN
NAPAP
NCDC
NE
NHAP
NOAA
NRC
NSS-I
Direct/Delayed Response Project
Digital elevation models
Dissolved organic carbon
Data quality objective
Eastern Lake Survey-Phase I
USEPA Environmental Monitoring and Systems Laboratory - Las Vegas
U.S. Environmental Protection Agency
Electric Power Research Institute
USEPA Environmental Research Laboratory - Corvallis
Episodic Response Project
Enhanced Trickle Down Model
Geographic Information System
International Business Machines Corporation - personal computer
Integrated Lake/Watershed Acidification Study
Interquartile range
Leaf area index
Long-term annual average deposition
Model for Acidification of Groundwater in Catchments
Major land resource areas
National Academy of Sciences
National Acid Deposition Program/National Trends Network
National Acid Precipitation Assessment Program
National Climatic Data Center
Northeast Region
National high altitude photography
National Oceanographic and Atmospheric Administration
National Research Council
National Stream Survey-Phase I
13-1
-------
NSWS
National Surface Water Survey
ORNL
OTA
PCA
PNL
QA
QC
RADM
RELMAP
RCC
R1LWAS
RMSE
RSD
SAB
SAS
SBR
SBRP
SCS
SE
SOBC
SOEBC
SUNY-P
TMY
UMW
UDDC
USDA
USOOI
USFS
USGS
UTM
WA
WBA
Oak Ridge National Laboratory, Tennessee
Office of Technolgy Assessment
Principal component analysis
Battelle-Pacific Northwest Laboratories
Quality assurance
Quality control
Regional Acid Deposition Model
Regional Lagrangian Model of Air Pollution
Regional Coordinator/Correlator
Regional Integrated Lake/Watershed Acidification Study
Root mean square error
Relative standard deviation
Science Advisory Board
Statistical Analysis System
Southern Blue Ridge
Southern Blue Ridge Province
Soil Conservation Service
Standard error
Sum of base cations
Sum of exchangeable base cations
State University of New York, Pittsburgh
Typical meterdogical year
Upper Midwest
Unified Deposition Database Committee
U.S. Department of Agriculture
U.S. Department of Interior
U.S. Forest Service
U.S. Geological Survey
Universal Transverse Mercator
Watershed area
Watershed Based Aggregation
13.1.2 Symbols
2As
2Bn
2Cn
2X
3A
3B
A
AC_BaCI
AH
AL
Al AO
Southern Blue Ridge subregion (NSS Pilot Survey)
Valley and Ridge subregion (NSS Pilot Survey)
Northern Appalachians subregion (NSS Pilot Survey)
Southern Appalachians subregion (NSS Pilot Survey)
Piedmont subregion (NSS Pilot Survey)
Mid-Atlantic Plain subregion (NSS Pilot Survey)
acid that is leached out of the soil
barium chloride triethanolamine exchangeable acidity
area of all open water bodies in drainage basin, in kilometers squared
area of primary lake, in kilometers squared
aluminum, acid oxalate extractable
13-2
-------
Al CD
AfPYP
AP+
ALPOT
ANN_AVG
AVG EL
B CENT
8~LEN
B_PER!M
B SHAPE
B~WIDTH
B3 Cl
C
C_TOT
Ca+Mg-DRY
Ca+Mg-WET
Ca Cl
CaCI2
CEC Cl
CO2
COMPACT
DDENSITY
ELONG
FRAG
aluminum, citrate dithionite extractable
aluminum, pyrophosphate extractable
aluminum ion
aluminum potential (pH - 1/spAI)
flow-weighted annual average sulfate concentration
average elevation; (MAX ELEV + MIN_ELEV)/2, in meters
total watershed area, in kilometers squared
drainage basin centroid expressed as an X,Y coordinate
length of drainage basin: air-line distance from basin outlet to farthest upper point
basin, in kilometers
the length of the line which defines the surface divide of the drainage basin, in
kilometers
basin shape ratio; B_LEN2/WS AREA
average basin width; WS_AREA7B_LEN, in kilometers
base saturation calculated from "unbuffered 1N ammonium chloride CELod
exchangeable bases
correct ion factor for the decrease in acidity due to the protonation of bicarbonate
carbon total
the annual loading of Ca plus Mg in dry deposition
the annual loading of Ca plus Mg in wet deposition
exchangeable calcium in unbuffered 1N ammonium chloride
calcium ion
calcium chloride
unbuffered 1N ammonium chloride cation exchange capacity
chloride ion
carbon dioxide
compactness ratio; ratio of perimeter of basin to the perimeter of a circle with
equal area; (PERIM)/(2 * (* AJ5)
drainage density; TOTSTRM/WS_AREA
elongation ratio; (4 * WS AREA)/L BEN
fragments > 2 mm diameter
M+
" total
H20 WS
H2O~
H2S04
H5up
ha
HC03-
H-DRY
H-WET
I
IND_AVG
INT
hydrogen ion
total effective acidity (H+ + NH/ - NO3")
ratio of open water bodies area to total watershed area; H20_AREA/ws_area
water
sulfuric acid
the percent of a watershed covered by bedrock with sensitivity codes of 5 and
6
hectare (2.47 acres or ten thousand square meters)
bicarbonate ion
annual hydrogen ion loading in dry deposition
annual hydrogen ion loading in wet deposition
amount of effective acidity in deposition
flow-weighted average sulfate concentration for the index sample time frame
(spring or fall)
total length of intermittent streams as defined from USGS topographic maps of
aerial photos, in kilometers
13-3
-------
K
K+
K_CI
Keq ha'1
kg
km
kso4
L CENT
L_PERiM
UMEPOT
ln(a/TanB)
ln(a/KbTanB)
LTA-rbc
LTA-zbc
M PATH
M04
MAX EL
MAX~REL
mg
Mg Q
Mg7*
MINEL
Na Cl
NE"CMPON
NECMPOS
NEIDLGD
NH
Ol-r
PC02
PER DD
PERlMRAT
PERIN
PH 01 M
PH~H2O
hydraulic conductivity
potassium ion
exchangeable potassium in unbuffered IN ammonium chloride
KiloequivaJent per hectare
kilogram
kilometer
sulfate mass transfer coefficient (m yr*)
primary lake centroid expressed as X,Y coordinates
perimeter of primary basin lake, in kilometers
lime potential (pH - 1/apCa)
an index of flowpath partitioning used in the TOPMODEL hydrologic model
an index of flowpath partitioning used in the TOPMODEL hydrologic model
long-term annual average, reduced dry base cation
long-term annual average, zero dry base cation
estimate of mean flowpath, in meters
miscellaneous land area mapped as quarry pits <
elevation at approximately highest point, in meters
maximum relief; MAX_EL£V - MIN_ELEV, in meters
microequivalents per liter, unit of concentration /
milligram /
exchangeable magnesium in unbuffered 1N ammonium chloride
magnesium ion
elevation of primary lake, in meters
PEL RAT
ROTUND
RTB
sodium ion
exchangeable sodium in unbuffered 1N ammonium chloride
data file with soil and miscellaneous area components of map units for the DDRP
Northeast region j
map unit composition data file for the DDRP Northeast region
identification legend data file for the DDRP Northeast region
ammonium
nitrate
hydroxide ion
partial pressure of carbon dioxide
drainage density calculated from perennial streams only; PERIN/WS_AREA
ratio of the lake perimeter to the watershed perimeter; Lake Perimeters/B_PERIM
total perennial stream length as defined from USGS topographic maps and aerial
photos, in kilometers 1
pH (0.01 M CaCI2)
pH (deionized water) |
runoff estimate (length time"1) j
Average annual runoff; interpolated to each site from Krug et al. (in press) runoff
map, in centimeters
correlation coefficient
coefficient of determination, the proportion of variability explained by a regression
model
relief ratio; (MAX ELEV-MIN ELEV)/B LEN
rotundity ratio; (B~J_EN)2/(4~* WS_ARlA)
lake retention time, in years
13-4
-------
s
S04 XIN
SBfJCI
SE_MP_CM
SE MP UN
SECMPNT
SEDBMNT
SiO2
SO4_B2
S04 EMX
SO4~H2O
S04 P04
S04~SLP
SO/"
SO4-DRY
SO4-WET
[S042"]ss
SOILDEN
S
STRMORDER -
SUB_BAS(n) -
Sum~(AI)
'aq
w
THKA
TOT_DD
TOTSTRM
V
w
V6
WA:LA
WM PATH
WS~AREA
WS~LA
sum of base cations
zero net adsorption concentration
sum of base cations as measured in unbuffered 1N ammonium chloride
dry sulfur deposition (mass length"2 time"1)
map unit composition data file for the DDRP Southern Blue Ridge region
map unit identification legend data file for the DDRP Southern Blue Ridge region
data file with soil and miscellaneous area components of map units for the DDRP
Southern Blue Ridge region
Southern Blue Ridge Mapping Database Management System
silicon dioxide
half saturation constant
adsorption asymptote
sulfate, water extractable
sulfate, phosphate extractable
zero net adsorption, slope
sulfate
annual loading of sulfate in dry deposition
annual loading of sulfate in wet deposition
steady state sulfate concentration
soil bulk density
surface water sulfur (mass length"3)
maximum stream order (Morton) of streams in the watershed (aerial photos used
to aid in reducing cooling problems between 7.5 and 15 minute maps)
area of each subcatchment in the drainage basin, in kilometers squared
wet sulfur deposition (mass length'2 time'1)
thickness adjusted for FRAG
estimated drainage density based on crenulations
identified on topographic map
total stream length; combination of perennial and intermittent, in kilometers
hydraulic residence time, in years
hydrologic retention time, in years
volume of primary lake, 106m3
watershed area to lake area ratio
estimate of weighted mean flowpath, in meters
total watershed area, in kilometers squared
ratio of the total watershed area to the area of the primary lake
13-5
-------
13.2 DEFINITIONS
ACCURACY - the difference between the approximate solution obtained using a numerical model and
the exact solution of the governing equations (or a known standard concentration), divided by the exact
solution (or known standard concentration).
ACID ANION - negatively charged ion that combines with hydrogen ion to form an acid.
ACID CATION - hydrogen ion or other metal that can hydrolyze water to produce hydrogen ions, e.g.
Al, Mn, Fe.
ACID CRYSTALLINE - in the Southern Blue Ridge Province, rocks or bedrock containing HIV clays.
ACID DEPOSITION SYSTEM (ADS) - a national database of precipitation amount and chemistry
maintained at Battelle-Pacific Northwest Laboratories.
ACID MINE DRAINAGE - runoff with high concentration of metals, sulfate, and acidity resulting from the
OXIDATION of sulfide minerals that have been exposed to air and water (usually from mining activities).
ACID NEUTRALIZING CAPACITY - the total acid-combining capacity of a water sample determined by
titratlon with a strong acid to a preselected equivalence point pH: an integrated measure of the ability
of an aqueous solution to neutralize strong acid inputs. Acid neutralizing capacity includes strong bases
(e.g., hydroxide) as well as weak bases (e.g., borates, carbonates, dissociated organic acids, alumino-
hydroxyl complexes).
ACIDIC DEPOSITION • rain, snow, or dry fallout containing high concentrations of sulfuric acid, nitric
acid, or hydrochloric acid, usually produced by atmospheric transformation of the by-products of fossil
fuel combustion (power plants, smelters, autos, etc.). Precipitation with a pH of less than 5.0 is generally
considered to be unnaturally acidic, i.e., altered by ANTHROPOGENIC activities.
ACIDIC EPISODE * an episode in a water body in which ACIDIFICATION of SURFACE WATER to an
ACID NEUTRALIZING CAPACITY less than or equal to 0 /ieq L"1 occurs.
ACIDIC LAKE OR STREAM - an aquatic system with an ACID NEUTRALIZING CAPACITY less than or
equal to 0 Meq L .
ACIDIFICATION - any temporary or permanent loss of ACID NEUTRALIZING CAPACITY in water or BASE
SATURATION in soil by natural or ANTHROPOGENIC processes.
ACIDIFIED - a natural water that has experienced any temporary or permanent loss of ACID
NEUTRALIZING CAPACITY or a soil that has experienced a reduction in BASE SATURATION.
ACTIVITY COEFFICIENTS • empirically derived coefficients used to transform concentration data to salt
or ion activities.
~i
ADJUSTED R2 • the standard R2 of regression analysis, modified to balance increasing the R2 against
increasing the number of explanatory variables.
AFFORESTATION - the natural process through which non-forested lands become forested.
AGGRADING FORESTS - forests in which there is a net annual accumulation of biomass.
AGGREGATION - a method for statistically reducing a set of data to a single calculated or index value
for each parameter (e.g., a weighted average).
13-6
-------
AKAIKE'S INFORMATION CRITERION - a criterion for selecting one of a sequence of regression models,
based on formulae from information theory.
ALFISOLS - in Soil Taxonomy, the ORDER of mineral soils with an argillic horizon with at least 35 percent
base saturation.
ALIASING - occurrence of an apparent shift in frequency of a periodic phenomenon. It arises as the
consequence of the choice of discrete space or time sampling points to represent a continuous process.
The choice may introduce a spurious periodic solution or mask a real periodic phenomenon.
ALKALINITY - the titratable base of a sample containing hydroxide, carbonate, and bicarbonate ions, i.e.,
the equivalents of acid required to neutralize the basic carbonate components.
ALKALINITY MAP CLASS - a geographic area defined by the expected ALKALINITY of SURFACE
WATERS (does not necessarily reflect measured alkalinity); used as a STRATIFICATION FACTOR in ELS-
I design.
ALLOPHANE - an amorphous to cryptocrystalline alminosilicate mineral, commonly thought to be a pre-
cursor phase to kaolinite.
ALUMINUM BUFFERING - a chemical process in which hydrogen ion activities are buffered by the
precipitation/dissolution of aluminum hydroxides.
ALUMINUM BUFFER RANGE - pH 4.2 - 2.8
AMPHOTERIC - a substance capable of acting as either an acid or a base; positively charged at high
pH and with an OH' functional group at low pH.
ANAEROBIC - without free oxygen (e.g., hypolimnetic lake waters, sediments, or poorly drained soils).
ANALYTE - a chemical species that is measured in a water sample.
ANALYTICAL CHARACTERIZATION * physical and chemical properties of soils measured in the
laboratory.
ANALYTICAL DUPLICATE - a QUALITY CONTROL sample made by splitting a sample.
•f
ANION - a negatively charged ion.
ANION CATION BALANCE - a method of assessing whether all CATIONS and ANIONS have been
accounted for and measured accurately; in an electrically neutral solution, such as water, the total charge
of positive ions (cations) equals the total charge of negative ions (anions).
ANION EXCHANGE/ADSORPTION - a reversible process occurring in soil in which ANIONS are
adsorbed and released.
ANTHROPOGENIC - of, relating to, derived from, or caused by human activities or actions.
APPARENT SOLUBILITY PRODUCT - an approximate form of an equilibrium constant calculated using
solution concentration data instead of activities.
AQUEOUS SPECIES - any dissolved ionic or nonionic chemical entity.
AQUIC - a moisture regime of soils in which a water table and reducing conditions occur near the
surface.
13-7
-------
AQUIFERS - below-ground stratum capable of producing water as from wells or springs.
AQUO LIGAND - a water molecule held to Fe or Al in a clay edge or hydrous oxide by ligand exchange.
ARC - represents line features and borders of area features. One line feature may be made up of many
arcs. The arc is the line between two nodes.
ARC/INFO - a commercial geographic information system (GIS) software used to automate, manipulate,
analyze, and display geographic data in digital form.
ATTRIBUTE - the class, characteristics or other properties associated with a specific feature, area on a
map, lake or stream.
AVAILABLE TRANSECT - a transect identified to represent a map unit and listed for random selection.
BASE CATION - a nonprotolytic CATION that does not affect ACID NEUTRALIZING CAPACITY; consists
principally of calcium, magnesium, sodium, and potassium.
BASE CATION EXCHANGE - the process by which BASE CATIONS (Ca2+, Mg2+, Na+, K+) are
adsorbed or released from negatively charged sites on soil particles from or to, respectively, soil solutions.
Such exchange processes are instrumental in determining pH of soil solutions.
BASE CATION SUPPLY - (1) the pool of BASE CATIONS (Ca2+, Mg2*, Na*. K+) in a soH available for
exchange with hydrogen ions (H ). The base cation pool is determined by the CATION EXCHANGE
CAPACITY of the soil and the percentage of exchange sites occupied by BASE CATIONS.
BASE SATURATION - the percentage of total soil CATION EXCHANGE CAPACITY that is occupied by
exchangeable cations other than hydrogen and aluminum, i.e., the base cations Ca , Mg , Na+, and
K+.
BEDROCK - solid rock exposed at the surface of the earth or overlain by unconsolidated material.
BEDROCK GEOLOGY - the science of the physical and chemical nature and composition of solid rock
at or near the earth's surface.
BEDROCK UTHOLOGY - see UTHOLOGY.
BEDROCK SENSITIVITY SCORES - a six point scale, developed for DDRP, designed to distinguish the
relative reactivities of different lithologies.
BEDROCK UNITS - the smallest homogenous entity depicted on a bedrock map.
BIAS - a systematic error in a method caused by artifacts or idiosyncracy of the measurement system.
BIOMASS - the quantity of paniculate organic matter in units or weight or mass.
BIOMASS ACCRETION - net accumulation of plant mass in a growing, or aggrading, ecosystem; also
refers to net accumulation of an individual nutrient associated with accumulation of biomass.
BLOOM-GRIGAL MODEL - a numerical model used to investigate the evolution of soil exchange
characteristics under various H* ion deposition SCENARIOS. The code is based on mass balance
consideration with empirical functions used to describe the pH-base saturation relationships.
BONFERRONI INEQUALITY - an inequality from probability theory that is used to carry out multiple
simultaneous statistical comparisons.
13-8
-------
BOXPLOT - a graph of data with a box drawn from the 25th percentile to the 75th percentile; lines
extending from the box as far as the data extend to a distance of at most 1.5 times the INTERQUARTILE
RANGE, and more extreme observations marked individually.
BUFFERING CAPACITY - the quantity of acid or base that can be added to a water sample with little
change in pH.
BULK DENSITY - the integrated density of a volume of soil, including solid matter, soil solutions, voids,
roots, etc.
Ca/AI EXCHANGE REACTION - the reaction describing the distribution of Ca and Al between the soil
exchange complex and the soil solution.
CALCITE • a mineral with the formula CaC03. A carbonate mineral.
CALIBRATION - process of checking, adjusting, or standardizing operating characteristics of instruments
and model appurtenances on a physical model or coefficients in a mathematical model with empirical data
of known quality. The process of evaluating the scale readings of an instrument with a known standard
in terms of the physical quantity to be measured.
CALIBRATION BLANKS - a zero-concentration QUALITY CONTROL standard that contains only the
matrix of the CALIBRATION standard.
CAPACITY FACTOR - a chemical property of a system defined as a function of the quantity or size of
that system.
CAPACITY-LIMITED PROCESS - A mechanism (e.g., sulfate adsorption or cation exchange) for which
the long-term ability to supply or consume cations or anions is constrained by the size of a watershed
pool or capacity (e.g., pool of exchangeable bases and sulfate adsorption capacity) rather than by
reaction kinetics.
CARBON-BONDED SULFUR - a reduced form of organic sulfur, characterized by C-S bonds.
CARBONIC ACID - a weak acid, HgCO-j, formed by dissolution of carbon dioxide in water. Dissociation
of carbonic acid (to H+ and HCO^) and subsequent consumption of H+ by exchange or weathering
reactions generates ANC in the form of bicarbonate ions.
CATCHMENT - see WATERSHED.
CATION - a positively charged ion.
CATION DEPLETION • a process through which base cations on a soil exchange site are progressively
replaced by ACID CATIONS at rates higher than those expected during normal pedogenesis.
CATION EXCHANGE - a reversible process occurring in soil sediment in which ACIDIC CATIONS (e.g.,
hydrogen ions) are adsorbed and BASE CATIONS are released.
CATION EXCHANGE CAPACITY • the sum total of exchangeable cations that a soil can absorb.
CATION (OR ANION) LEACHING - movement of cations (or anions) out of soil, in conjunction with
mobile anions in soil solution.
CATION RETENTION - the physcial, biological, and geochemical processes by which cations in
watersheds are held, retained, or prevented from reaching receiving SURFACE WATERS.
CHRONIC ACIDIFICATION - see LONG-TERM ACIDIFICATION.
13-9
-------
CIRCUMNEUTRAL - close to neutrality with respect to pH (pH = 7); in natural waters, pH 6 - 8.
CLAY - a soil separate consisting of particles with an equivalent diameter less than 0.002 mm; also a soil
textural class containing >. 40 percent clay-sized material, < 45 percent sand and < 40 percent silt.
CLAY MINERALS - any of a series of sheet silicate minerals believed to form in a soil or low-temperature
diagenetic environment.
CLOSED LAKES - a lake with a surface water inlet but no surface water outlet.
CLUSTER ANALYSIS - a multivariate classification technique for identifying similar (or dissimilar) groups
of observations.
COARSE PARTICLE DRY DEPOSITION - atmospheric DRY DEPOSITION of particles greater than 2
microns in effective diameter.
COLLINEAR - see MULTICOLUNEARITY.
COMBINATION BUFFER - land area surrounding a lake including area within a 40-foot contour area
around perennial streams, and area around contiguous wetlands.
COMPLEX - a map unit consisting of two or more dissimilar soil components or miscellaneous areas
occurring in a regularly repeating pattern.
COMPONENTS - see MAJOR COMPONENTS, MINOR COMPONENTS, and MAP UNIT COMPOSITION.
CONSOCIATION - a map unit dominated by a single soil taxon (or miscellaneous area) and similar soils.
CONTOUR LINE - a line connecting the points on the land surface that have the same elevation.
CONVERGENCE - state of tending to a unique solution. A given scheme is convergent if an increasingly
finer computational grid leads to a more accurate approximation of the unique solution. Note that a
numerical method may sometimes converge on a wrong solution.
COOK'S D - a regression statistic designed to indicate LEVERAGE POINTS.
COVERAGE - a digital analog of a single map sheet; forms the basic unit of data storage in ARC/INFO.
CUMULATIVE DISTRIBUTIVE FUNCTION - a function, F(x), such that for any reference value X, F(x)
is the estimated proportion of individuals (lakes, streams, estuaries, coastal waters) in the population
having a value x <_ X.
DARCY'S LAW - An equation to predict the flux of water through a porous medium, of the form Q =
K * A * S, where Q = lateral water flux, K = saturated hydraulic conductivity, A = cross sectional area,
and S = hydraulic gradient.
DATABASE FILE - a collection of records that share the same format.
DEPOSITIONAL FLUXES • the mass transfer rate to the earth's surface of any of a number of chemical
species.
DEPTH TO BEDROCK - depth to solid, fixed, unweathered rock underlying soils.
DEPTH TO A SLOWLY PERMEABLE OR IMPERMEABLE LAYER - depth to a layer in soils or
underlying soils that restricts the downward flow of water (e.g., bedrock dense till or fragipan).
13-10
-------
DETECTION LIMIT QC CHECK SAMPLE - a QUALITY CONTROL sample that contains the ANALYTE
of interest at two to three times the contract required detection limit.
DIAZO - a photocopy whose production involves the use of a coating of a diazo compound.
DIGITIZATION • the process of entering lines or points into a GEOGRAPHIC INFORMATION SYSTEM.
DIGITIZED COORDINATES - lines or points that have been entered into a GEOGRAPHIC INFORMATION
SYSTEM.
DISSIMILATORY REDUCTION - a process in which an oxidized chemical species (e.g., S04 - 8} is
utilized by an organism as an electron acceptor in the absence of free oxygen and released in a reduced
form (e.g., S ") rather than assimilated.
DISSOCIATION - separation of an acid into free H* and the conjugate base of that acid (e.g., H2CO3 •
-> H + HCO-"). or separation of a base into a free hydroxyi and the conjugate acid of the base (e.g.,
NH4OH -> NH/ + Orf).
DISSOLUTION RATES - the rate at which a mineral is transformed to other species or minerals in an
aqueous environment.
DISSOLVED ORGANIC CARBON - a measure of organic (nonorganic) fraction of carbon in a water
sample that is dissolved or unfilterabie.
DOLOMITE - a mineral with the chemical formula CaMg(COJ2. A carbonate mineral.
DOWNSTREAM REACH NODE - see LOWER NODE.
DRAINAGE - the frequency and duration of periods when the soil is free of saturation or partial saturation
and the depth to which saturation commonly occurs.
DRAINAGE BASIN - see WATERSHED.
DRAINAGE CLASS • any of the seven classes that characterize the frequency and duration of soil
saturation.
DRAINAGE LAKE - a lake with SURFACE WATER outlet(s) or with both inlets and outlets.
DRY DEPOSITION - for the purposes of DDRP analysis, atmospheric deposition of materials to
watersheds in any form other than rain or snow.
DRY DEPOSITION VELOCITY - an effective velocity used with airborne concentrations to compute dry
depositional flux of materials to surfaces or watersheds.
EIGENVALUE - the eigenvalues of a square matrix A are the roots c of the polynomial equation det(A-
cl) = 0, where det(.) is the determinant and I is an identity matrix.
ELECTRON ACCEPTOR - an oxidized (or at least partially oxidized) chemical species capable of
undergoing a reduction reaction by addition of an electron.
ELEVATIONAL BUFFER - land area around a lake bounded by a topographic contour.
ELS PHASE I LAKES - the population of lakes sampled during phase I of the Eastern Lake Survey of
the EPA's National Surface Water Survey.
13-11
-------
EMPIRICAL MODEL - representation of a real system by a mathematical description based on
experimental data rather than on general physical laws.
ENTISOLS - in Soil Taxonomy, the ORDER of mineral soils with no or very poorly developed genetic
horizons.
EPISODE - a short-term change in stream pH and ACID NEUTRALIZING CAPACITY during storm flows
or snowmelt runoff.
EQUIVALENT - unit of ionic concentration; the quantity of a substance that either gains or loses one
mole of protons or electrons.
ESTER SULFATE - an oxidized form of sulfur in soil organic matter, characterized by C-O-SO3 or N-O-
SO3 linkages.
EVAPORITE - a mineral formed from solution phase due to supersaturation and chemical precipitation
resulting from evapoconcentration of the solution; sulfate, chloride, and many carbonate minerals form
in this manner.
EVAPOTRANSPIRATiON (%ET) - the proportion of precipitation that is returned to the air through direct
evaporation or by transpiration of vegetation.
EXCHANGE POOL - the reservoir of BASE CATIONS in soils available to participate in exchange
reactions.
EXTENSIVE PARAMETERS - variables that depend on the size (extent) of the system.
EXCHANGE REACTIONS - any of a number of reactions that describe the partitioning of two chemical
species between a solution and soil exchange complex.
FELDSPARS - a group of tectosilicate minerals that are the most abundant group in the earth's crust.
FIELD REVIEW - a review of soil surveys made in the field by supervisory soil scientists to help field soil
scientists maintain standards that are both adequate for the objectives of the survey and consistent with
those of other surveys. Samples of the fieldwork are examined for soil identification, placement
of boundaries, and map detail in relation to survey objectives.
FINE PARTICLE DRY DEPOSITION - atmospheric DRY DEPOSITION of particles of size less than 2
microns in effective diameter.
FIRST-ORDER REACTION - a chemical reaction, the rate of which is proportional to the concentration
of the limiting reactant.
FOREST COVER TYPE - a descriptive classification of forestland based on present occupancy of an area
by tree species. (The term "vegetation" implies total forest community, whereas the focus here is on trees
defining type. Whenever the term "vegetation" is used in this report it should be construed as FOREST
COVER TYPE.)
FREUNDUCH ISOTHERM - an exponential adsorption isotherm of the form Ec = aCb, where: Ec =
concentration of adsorbed species (per unit mass adsorbent), C = dissolved concentration of species
being adsorbed, and a and b are derived coefficients.
FULVIC ACID - a family of naturally-occurring weak organic acids found in soils and surface waters; fulvic
acids are operationally defined as the acid-soluble (pH = 1.0) fraction of an alkali-soluble soil extract; pK
is roughly 3.5.
13-12
-------
GAINES THOMAS FORMULATION - a formulation used to describe exchange processes.
GAPON - a formulation used to describe exchange processes.
GENERIC BEDROCK TYPE - see GENERIC ROCK TYPE.
GENERIC ROCK TYPE - a general classification of different BEDROCK UNITS into groups according to
the primary UTHOLOGY.
GEOGRAPHIC INFORMATION SYSTEM (GIS) - a computerized system designed to store, process, and
analyze data.
GEOLOGY - see BEDROCK GEOLOGY.
GEOMORHPIC POSITION - the relative location in the landscape described by hillslope elements (cross
section view) and slope components (plane view), e.g., sideslope footslope.
GIBBSITE - a mineral with the chemical formula AI(OH)3.
GIS BUFFERS - land area surrounding a lake, stream, or wetland, delineated using a GIS. See
COMBINATION BUFFER and LINEAR BUFFER.
GLACIAL TILL - see TILL
GLACIOFLUVIAL - a material that has been deposited by glaciers and sorted by meftwater.
GLEY SOIL - a soil developed under conditions of poor drainage, characterized by oxygen depletion and
reduction of iron and other metals (Mn), resulting in gray colors and mottles.
GRAN ANALYSIS - a mathematical procedure used to determine the equivalence points of a TITRATION
CURVE for acid and base neutralizing capacity.
GREAT GROUP - in Soil Taxonomy, the level of classification just below SUBORDER, e.g., Haplorthods.
GROUNDWATER - water in the part of the ground that is completely saturated.
HETEROSCEDASTIC - referring to a statistical situation in which variances are not all equal.
HEURISTIC MODEL - representation of a real system by a mathematical description based on reasoned,
but unproven argument.
HINDCAST - to estimate some prior event or condition as a result of a rational study and analysis of
available pertinent current and historical data.
HISTIC SOILS - organic-rich soils.
HISTOSOLS - in Soil Taxonomy, the ORDER of soils formed from organic PARENT MATERIAL
HOMOSCEDASTIC - referring to a statistical situation in which the variances are all equal.
HORIZON - a horizontal layer of soil with distinct physical and/or chemical characteristics. Genetic
horizons are the result of soil-forming process.
HORNBLENDE - a common amphibole mineral with the approximate chemical formula
(Ca,Na)3(Mg,Fe,AI)5(Si,AI)8022(OH)2.
13-13
-------
HYDRAULIC HEAD - hydrostatic pressure created by a difference in height of water columns in different
portions of a connected aquifer.
HYDRAULIC RESIDENCE TIME - a measure of the average amount of time water is retained in a lake
basin. It can be defined on the basis of inflow/lake volume, represented as RT, or on the basis of
outflow/lake volume and represented as Tw. The two definitions yield similar values for fast flushing lakes,
but diverge substantially for long residence time SEEPAGE LAKES.
HYDROLOGIC CHARACTERISTICS mj
HYDROLOGIC FLOW PATHS - the distribution and circulation of water deposited by precipitation on the
surface of the land, in the soil, and underlying rocks within a WATERSHED.
HYDROLOGIC RETENTION TIME - see HYDRAULIC RESIDENCE TIME.
HYDROUS OXIDE - a collective term referring to any of a group of amorphous or crystalline species of
iron or aluminum that are partially or fully hydrated (e.g., MO(OH), M(OH).j).
HYPOLIMNION - in a thermally-stratified lake, the portion in a lake at depths below the thermocline; these
waters are isolated from reaeration at the surface and oxygen is likely to be depleted, leading to
mobilization of reduced chemical species.
IMMOBILIZATION REACTION - conversion of an inorganic form of a nutrient (especially S or N) to
organic matter.
IMPOUNDMENT - a man-made lake created by construction of a dam; also applied to natural lakes
whose level is controlled by a man-made spillway.
INCEPT1SOLS - in Soil Taxonomy, the ORDER of soils with at least one diagnostic horizon, but with no
horizon strongly enough developed to place them in another ORDER.
INCLUSIONS - see MINOR COMPONENTS.
INDEX OF CONTACT TIME - the theoretical maximum potential of contact between runoff and the soil
matrix. The index is calculated by dividing the soil water flow rate (obtained using Darcy's law) by
average annual runoff.
INDEX SAMPLE - in NE lakes, one sample per lake, used to represent chemical conditions on that lake.
In streams, any sample (or the average of one to three samples) collected at a stream NODE during the
SPRING BASEFLOW INDEX PERIOD, used to represent chemical conditions in the stream.
INFO • a database management system that stores, maintains, manipulates, and reports information
associated with geographic features in ARC/INFO.
INITIAL CONDITIONS - given values of DEPENDENT VARIABLES or relationship between dependent and
independent variables at the time of start-up of the computation.
IN-LAKE SULFUR RETENTION - net retention of sulfur within a lake, occurring principally by reduction
within sediments.
INTENSITY FACTOR - a variable with properties defined by concentration in solution, and therefore
independent of the quantity or size of the system.
INTENSIVE PARAMETERS - variables whose values are independent of the size or extent of the system,
e.g., temperature and pH.
13-14
-------
INTERQUARTILE RANGE - the difference between the 75th and the 25th percentiles.
IONIC STRENGTH - a measure of the interionic effect resulting from the electrical attraction and repulsion
between various ions. In very dilute solutions, ions behave independently of each other and the ionic
strength can be calculated from the measured concentrations of ANIONS and CATIONS
present in the solution. Units are moles per liter.
ISOTHERM - a linear or nonlinear function describing partitioning of an absorbent between solid and
sorbed phases. Such isotherms were originally used to characterize (nearly) ideal processes (e.g., the
Langmuir equation was developed to describe adsorption of a gas by a solid), but are often empirically
defined for adsorption of anions or organic compounds on soils because they provide a convenient
shorthand to describe partitioning.
KAOLINITE - a 2-lay clay mineral with the chemical formula AI2Si2O5(OH)4.
KINETIC MODELS - any of a family of numerical models that use kinetic considerations as the purifying
principle in describing natural processes.
KRIGING - a technique for spatial interpolation.
LABEL - represents point features or Is used to assign identification numbers to POLYGONS.
LAKE TYPE - a classification of lakes based on the presence or absence of inlets, outlets, and dams as
represented on LARGE-SCALE MAPS.
LAND COVER - see FOREST COVER TYPE.
LAND USE - the dominant use of an area of land (e.g., crop land).
LANDFORM SEGMENT - a small part of the local landform that is uniquely related to landscape
processes.
LANGMUIR ISOTHERM - a hyperbolic adsorption isotherm (used in this project for sulfate) of the form
Ec = (Bj * C)/(B2 + C). where: EC = net adsorbed sulfate, C = dissolved sulfate, and B- and B2 are
empirically derived coefficients. When appropriate, the isotherm can be "extended" by addition of a third
coefficient to describe a non-zero Y-intercept.
LARGE-SCALE MAPS -1:24,000,1:25,000, or 1:62,500 scale U.S. Geological Survey topographical maps.
LEACHING - the transport of a solute from the soil in the soil solution.
LEVERAGE POINT - a data point that strongly influences the parameter estimates in a regression.
UGAND EXCHANGE - a mechanism of bond formation between an oxyanion and a soil mineral bearing
hydroxyl groups. The exchange involves formation of inner sphere complexes of anions to Lewis acid
sites, following replacement of water from the Lewis acid site by the oxyanion.
LIMESTONE - a rock type consisting primarily of CALCITE.
LINEAR BUFFER - land area within a set distance of a lake or stream.
UTHOLOGY - the physical characteristics of a rock or mapped BEDROCK UNIT. Generally relates to
mode of formation, mineralogy, and texture.
13-15
-------
LJTTERFALL - fresh organic detritus, usually leaves, needles, twigs, etc., that compose the bulk of the
forest floor.
LOCAL LANDFORM - a subdivision of the regional landform that is the result of localized landscape
processes.
LONG-TERM ACIDIFICATION - a long-term partial or complete loss of ACID NEUTRALIZING CAPACITY
from a lake or stream.
LONG-TERM ANNUAL AVERAGE DEPOSITION (LTA) - a dataset of atmospheric deposition representing
atmospheric deposition during the early-to-mid 1980s for the purposes of the DDRP.
LOWER NODE - the downstream NODE of a STREAM REACH.
MAJOR LAND RESOURCE AREA - a geographic area characterized by a particular pattern of soils,
climate, water resources, and LAND USE.
MAJOR COMPONENTS - soil components or miscellaneous areas that are identified in the name of a
map unit.
MALLOWS' CP - a criterion for selecting one of a sequence of regression models.
MAP COMPILATION - the process of checking and measuring soil map unit data.
MAPPING PROTOCOLS - instructions that guide the field mapping and provide for quality control.
MAP SYMBOL - a symbol used on a map to identify map units.
MAPPING TASK LEADER - the person responsible for field mapping activities.
MAP UNIT - see SOIL MAP UNIT.
MAP UNIT COMPOSITION - the relative proportion (expressed in percent) of all soil components and
miscellaneous areas in a map unit.
MAP UNIT COMPOSITION FILE - a DATABASE FILE that contains all components and their relative
proportion for each map unit in the survey area (components are identified by an assigned code, i.e.,
SCODE).
MAP UNIT CORRELATION - see SOIL CORRELATION.
MAP UNIT DELINEATION - an area on a map uniquely identified with a symbol. A delineation of a soil
map has the same major components as identified and named in the map unit.
MAP UNIT NAME - the title of a map unit identified by the major soil components or miscellaneous areas
followed by appropriate phase terms.
MASS ACTION MODELS - any of a family of numerical models that use equilibrium-based principles as
the central unifying theme.
MASS BALANCE MODELS - any of a family of numerical models that use conservation-of-mass
principles as the central unifying theme.
13-16
-------
MASS TRANSFER COEFFICIENTS - a removal or rate constant used in models of in-lake alkalinity
generation (and elsewhere) to quantify the average removal rate of a reactant from solution. Specific
reference in this project is transfer from solution to sediment by all processes, including sedimentation
and reduction in sediments. In many systems, the mass transfer coefficient for sulfur is essentially a
diffusion constant for sulfate across the water-sediment interface; for nitrate a biological
uptake/sedimentation rate.
MASTER HORIZONS • the most coarsely based delineations within a pedon. Usually, A/E horizons
denote zones of net mass depletion, B horizons are zones of net accumulation, and C horizons indicate
minimal pedogenic evolution.
MATRIX SPIKE - a QUALITY CONTROL sample made by adding known quantity of an ANALYTE to a
sample aliquot.
MAX - the maximum sensitivity code observed on a WATERSHED.
MEAN - the weighted average of sensitivity codes for a WATERSHED.
MEDIAN (M) - the value of x such that the cumulative distribution function F(x) = 0.5; the 50th percentile.
METASEDIMENTARY - rocks or bedrock formed from metamorphic sedimentary rocks.
MICAS - a group of primary phylosilicate minerals, frequently including biotite, vermiculite, and muscovite.
MMID-APPALACHIAN REGION • one of the three geographic regions considered by the DDRP,
consisting of upland areas (subregions 2Bn and 2Cn) of the Mid-Atlantic region (MD, PA, VA, WV) defined
by the National Stream Survey.
MINERAL WEATHERING - dissolution of rocks and minerals by erosive forces.
MINERALIZATION - microbially-mediated conversion of nutrients from an organically bound (especially
N and S) to an inorganic form.
MINOR COMPONENTS • soil components or miscellaneous areas that are not identified in the name of
the map unit. Many areas of these components are too small to be delineated separately.
MISCELLANEOUS AREA - land areas that have no soil and thus support little or no vegetation without
major reclamation. Rock outcrop is an example.
MISCELLANEOUS LAND AREAS - see MISCELLANEOUS AREA.
MOBILE ANION - an anion that remains in solution and passes through a soil without significant delays
due to biological or chemical processes; also a model or paradigm for cation leaching from soils, based
on the premise that the rate of cation leaching from a soil is controlled by the sum of mobile anions
(which are regulated by a suite of more-or-less independent processes).
MONTE CARLO METHOD - technique of STOCHASTIC sampling or selection of random numbers to
generate synthetic data.
MOTTLING - spots or blotches of different color in a soil, including gray to black blotches in poorly
drained soils due to presence of reduced iron and other metals.
MULTICOLLINEARITY - when one of the explanatory variables can be reproduced as a linear
combination of the other explanatory variables. In such a case, the usual regression estimates cannot
be computed.
13-17
-------
NITROGEN TRANSFORMATION - biochemical processes through which nitrogen deposited in an
environment is converted to other forms.
NODE - the points identifying either an upstream or downstream end of a REACH.
NONPARAMETR1C - referring to a statistical procedure that does not make the classical distributional
assumptions.
NON-SILICATE IRON AND ALUMINUM • soil iron and/or aluminum occurring in the soil as an
amorphous or a (hydrous) oxide phase rather than as an ion incorporated within a silicate mineral lattice.
OFFICIAL SOIL SERIES DESCRIPTION - a record of the definitions of a soil series and other relevant
information about each series. These definitions are the framework within which most of the detailed
information about soils of the United States is identified with soils at specific places. These definitions
also provide the principal medium through which detailed information about the soil and its behavior at
one place is projected to similar soils at other places.
ORDER - in Soil Taxonomy, the highest level of classification, e.g., SPODOSOLS.
ORGANIC ACID - organic compound possessing an acidic functional group; includes fulvic and humic
acids.
ORGANIC ANION - an organic molecule with a negative net ionic charge.
ORGANIC 'BLOCKING" - a reduction in the sulfate (or other anion) adsorption capacity of a soil resulting
from preferential sorption of organic acids by the soil.
ORGANIC HORIZONS - any identifiable soil horizon containing in excess of 20 percent organic matter
by weight.
OUTLIER • observation not typical of the population from which the sample is drawn.
OXIDATION - loss of electrons by a chemical species, changing it from a lower to a higher oxidation
state (e.g., Fe2* to Fe3* or S_2 to S+6, with intermediates).
PARAMETER - (1) a characteristic factor that remains at a constant value during the analysis, or (2) a
quantity that describes a statistical population attribute.
PARENT MATERIAL - the material from which soils were formed.
PARTIAL PRESSURE - the percentage of a gaseous sample that is composed of one particular
component.
PEDON - the smallest block of soil that contains all the characteristics of a soil (usually about 1 m2); a
soD individual.
PERCENT COARSE FRAGMENTS - the percentage of soil, by volume, that is composed of rock
fragments unable to pass through a 2-mm sieve.
PERMEABILITY - the ease with which gases, liquids, or plant roots penetrate or pass through a bulk
mass of soil or a layer of soil.
pH • the negative logarithm of the hydrogen ion activity. The pH scale runs from 1 (most acidic) to 14
(most alkaline); a difference of 1 pH unit indicates a tenfold change in hydrogen activity.
POLYGON - represents area features.
13-18
-------
PRECISION - a measure of the capacity of a method to provide reproducible measurements of a
particular ANALYTE (often represented by variance).
PRIMARY MINERAL WEATHERING - the natural process by which thermodynamically unstable minerals
are converted to more stable phases under earth surface conditions.
PRINCIPAL COMPONENT ANALYSIS - a statistical analysis concerned with explaining the variance-
covariance structure through the use of PRINCIPAL COMPONENTS.
PRINCIPAL COMPONENTS - particular linear combinations of the original data, which geometrically
represent a new coordinate system with axes in the directions of maximum variability.
PROBABILITY SAMPLE - a sample in which each unit has a known probability of being selected.
QC CHECK SAMPLE - a QUALITY CONTROL sample that contains the ANALYTE of interest at a
concentration in the mid-calibration range.
QUALITY ASSURANCE - steps taken to ensure that a study is adequately planned and implemented to
provide data of known quality, and that adequate information is provided to determine the quality of the
database resulting from the study.
QUALITY CONTROL - steps taken during a study to ensure that data quality meets the minimum
standards established by the quality assurance plan.
QUARTILE - any of three values (Qv Q2, Q.J) that divide a population into four equal classes, each
containing one-fourth of the population.
QUARTZ - a crystalline form of silicon dioxide (SiO2).
QUARTZITES - a metamorphic rock-type composed of primarily QUARTZ.
QUINTILE • any of the four values (Q1 , Q2 , Q, , Q4) that divide a population into five equal classes,
each representing 20 percent of the population; used to provide additional values to compare
characteristics among popluations of lakes and streams.
RATE-LIMITED REACTION • a process (e.g., mineral weathering) for which the long-term ability to supply
reaction products (e.g., base cations) is constrained by reaction or transport kinetics.
RCC TRANSECTS - transects conducted by the Regional Coordinator/Correlator (RCC).
REACH - segments of the stream network represented as blue lines on 1:250,000-scale U.S. Geological
Survey maps. Each reach (segment) is defined as the length of stream between two blue-line
confluences. In the NSS-I, stream reaches were the sampling unit.
yREACTION ORDER - the relationship between the rate of a chemical reaction and the concentration
of a reaction substrate, defined by the value of the exponent of that substrate.
REACTIVITY SCALE - any of a number of relative scales designed to categorize the general
"weatherability of different LJTHOLOGIES.
REACTIVITY SCORE - see REACTIVITY SCALE.
REAGENT BLANK - a QUALITY CONTROL sample that contains all the reagents used and in the same
quantities used in preparing a soil sample for analysis.
13-19
-------
REDUCTION/OXIDATION - chemical reaction in which substances gain or lose electrons.
REGION - a major area of the conterminous United States where a substantial number of streams with
ALKALINITY less than 400 peq L can be found.
REGIONAL LANDFORM - physiographic areas that reflect a major land-shaping process over a long
period of time.
REGIONAL SOILS LEGEND - a correlated and controlled legend for an entire region (see SOIL
IDENITIFICATION LEGEND).
RELMAP - a source-receptor model designed to estimate dry deposition of sulfur; not used directly in
the DDRP.
REPORTS - relative to GIS activities, a format designed by the user for printing out information containing
the data files.
RESERVOIR - a body of water collected and stored for future use in a natural or artificial lake.
RESIDUAL - in regressions, the difference between the observed dependent variable and the value
predicted from the regression fit.
REUSS MODEL - a numerical model used to describe exchange processes in a soil environment.
RIPARIAN - a zone bounding and directly influenced by SURFACE WATERS.
ROBUST - a statistical procedure that is insensitive to the effect of OUTLIERS.
ROUNTINE TRANSECTS - transects conducted by field soil scientists responsible for the mapping.
RUSTY WEATHERING METASEDIMENTS rh
SALT EFFECT - the process by which hydrogen ions are displaced for the soil exchange complex by
BASE CATIONS (from neutral salts). The result is a short-term increase in the acidity of associated water.
SAMPLING CLASS - see SOIL SAMPLING CLASS.
SAMPLING CLASS CODE - a three-character code assigned to each SOIL SAMPLING CLASS.
SAMPLING CLASS COMPOSITION - the relative proportion of sampling classes in a map unit.
SAND - a soil separate between 0.05 and 2.0 mm in diameter; also a soil texture class containing at least
85 percent sand, and whose percentage of silt, plus 1.5 times the percent clay, does not exceed 15.
SATURATION INDEX - the ratio of the ion activity product (of dissolved ions) to the solubility product
for a solid phase; if the saturation index (SI) exceeds 1.0, the solution is supersaturated with respect to
that phase; if SI = 1.0, the solution is at equilibrium, if SI < 1.0, the solution is undersaturated with
respect to that solid phase.
SCENARIO - one possible deposition sequence following implementation of a control or mitigation
strategy and the subsequent effects associated with this deposition sequence.
SECONDARY MINERALS - any inorganic phase that forms as a direct product of dissolution or
transformation of another mineral.
SEEPAGE LAKE - a lake with no permanent SURFACE WATER inlets or outlets.
13-20
-------
SELECTIVITY COEFFICIENT - the apparent constant used to describe the partitioning of species in an
exchange reaction.
SENSITIVITY ANALYSIS - test of a model in which the value of a single variable or parameter is
changed, and the impact of this change on the DEPENDENT VARIABLE is observed.
SENSITIVITY CODES - see BEDROCK SENSITIVITY SCORES.
SIGNIFICANCE LEVEL - the conditional probability that a statistical test will lead to rejection of the null
hypothesis, given that the null hypothesis is true.
SILICA - the dissolved form of silicon dioxide (Si02).
SILT - a soil separate consisting of particles between 0.05 and 0.002 mm in equivalent diameter; also a
soil texture class containing at least 80 percent silt and < 12 percent clay.
SILVICULTURAL PRACTICES - forest management practices to increase wood yields: thinning, pruning,
fertilization, spraying with herbicides/insecticides, and irrigating.
SIMULATION - replication of the prototype using a model.
SKELETAL SOILS - soils with at least 35 percent rock fragments in the control section.
SLOPE PHASE - the slope gradient of a map unit or taxonomic unit expressed in percent.
SLOPE SHAPE ACROSS - shape of the surface parallel to the contours of the landscape {e.g. concave,
convex, plane).
SLOPE SHAPE DOWN - shape of the land surface at right angles to the contours of the landscape.
SMALL-SCALE MAP - 1:250,000-scale U.S. Geological Survey map.
SMECTITES - a family of 3-layer clay minerals.
SOIL - unconsolidated material on the surface of the earth that serves as a natural medium for the growth
of plants.
SOIL ACIDIFICATION - a process through which BASE CATIONS are removed from the soil and are
replaced by ACID CATIONS.
SOIL BUFFERING CAPACITY - the capacity of a soil to resist changes in pH with the addition of acids
to the system.
SOIL COMPONENT CODE - four-character code assigned to each soil or miscellaneous area component
of map units in a survey area. Codes were used to link data files.
SOIL COMPONENTS FILE - a computer data file that contains all the soil and miscellaneous area
components in a survey area and identified with a code (i.e., SCODE).
SOIL CORRELATION - the process of maintaining consistency in naming, classifying, and interpreting
kinds of soils and of the units delineated on maps.
SOIL EXCHANGE COMPLEX - all components of a soil that contribute to the absence of exchange
properties of that soil.
13-21
-------
SOIL FAMILY - next to the lowest category in Soil Taxonomy in which classes are separated mainly on
particle size, temperature, and mineralogy.
SOIL IDENTIFICATION LEGEND - a map legend that lists the symbols used to identify SOIL MAP UNITS
and the names of the map units.
SOIL LEGEND - see SOIL IDENTIFICATION LEGEND.
SOIL MAP UNIT - a collection of areas defined and named in terms of their soil components or
miscellaneous areas or both. Each map unit differs in some respect from all others in a survey area and
Is uniquely identified on a soil map.
SOIL SAMPLING CLASS - an arbitrary grouping of soils either known or expected to have similar
physical and/or chemical effects on drainage waters with respect to effects of acidic deposition.
SOIL SERIES - the most homogenous category in the taxonomy used in the United States. A group of
soils that have horizons similar in arrangement and in differentiating characteristics.
SOIL SOLUTIONS - those aqueous soils in contact with soils.
SOIL TAXONOMIC CLASS - the soil members within limits of ranges set by Soil Taxonomy. Taxonomic
units are members of the taxonomic class.
SOIL TAXONOMIC UNIT - a kind of soil described in terms of ranges in soil properties of the
polypedons referenced by the taxonomic unit in the survey area.
SOIL TEXTURE - the relative proportion by weight, of the several soil particle size classes finer than 2
mm in equivalent diameter (e.g., sandy loam).
SOIL TEXTURE MODIFIER - suitable adjectives added to soil texture classes when rock fragments
exceed about 15 percent by volume, for example, gravelly loam. The terms "very" and "extremely" are
used when rock fragments exceed about 35 and 60 percent by volume, respectively.
SOIL TRANSECT • a distance on the surface of the earth represented by a line on a map. Transects
can be straight, dogleg, or zigzag.
SOLID PHASE EXCHANGERS - those components of soils, primarily organic matter, clay minerals and
mineral oxides, that serve as the sites for exchange reactions.
SOLUM - soil layers that are affected by soil formation.
SPECIATION MODEL - a numerical model used to desribe the distribution of aqueous species among
various possible complexes and conpairs; usually for the purpose of estimating single ion activities.
SPECIFIC ADSORPTION - adsorption of sulfate by ligand exchange, often involving exchange of two
ligands and formation of a bridged (M-O-SO2-O-M) structure.
SPODIC HORIZONS - a soil horizon in which iron oxides, aluminum oxides, and organic matter have
accumulated from higher horizons.
SPODOSOLS - in Soil Taxonomy, the ORDER of mineral soils with well-developed SPODIC HORIZONS.
SPRING BASEFLOW INDEX PERIOD - a period of the year when streams are expected to exhibit
chemical characteristics most closely linked to ACIDIC DEPOSITION. The time period between snowmelt
and leafout (March 15 to May 15 in the NSS-I) when NSS-I stream reaches were visited coinciding with
expected periods of highest geochemical and assessment interest (i.e., low seasonal pH and
13-22
-------
SULFATE RETENTION - the physical, biological, and geochemical processes by which sulfate In
WATERSHEDS is held, retained, or prevented from reaching receiving SURFACE WATERS.
SULFIDE - an ion consisting of reduced sulfur (S2~) or a compound containing sulfide, e.g., hydrogen
sulfide (H2S) or the iron sulfide pyrite (FeS2).
SULFIDE OXIDATION - chemical reaction in which a sulfide loses electrons and assumes a higher
oxidation state; sulfate is the completely oxidized end product.
SULFITIC - containing sulfide minerals, usually pyrite.
SULFUR INPUT/OUTPUT BUDGET - an approach to describing sulfur mobility in a watershed by
comparing fluxes of sulfur to and from the watershed (as the difference between Input and output or as
a ratio).
SURFACE WATER - streams and lakes.
SURFACE WATER RUNOFF - precipitation that flows overland to reach SURFACE WATERS.
SURFICIAL GEOLOGY - characteristics of the earth's surface, especially consisting of unconsolidated
residual, colluvial, or glacial deposits lying on the BEDROCK.
SYNOPTIC - relating to or displaying conditions as they exist at a point In time over a broad area.
SYSTEMATIC ERROR - a consistent error introduced in the measuring process. Such error commonly
results in biased estimations.
TARGET POPULATION - a subset of a population explicitly defined by a given set of exclusion criteria
to which inferences are to be drawn from the sample attributes.
THERMODYNAMIC CONSTANTS - an empirically derived constant used to describe the relative
distribution of chemical species in a specified reaction when at equilibrium.
THROUGHFALL - precipitation that has interacted with a forest canopy, the chemistry of which is thus
modified from that of incident precipitation due to washoff of dry-deposited material and leaf exudates
as well as by ion exchange and uptake by leaf surfaces.
TICS • registration or geographic control points for a COVERAGE.
TILL - unstratified material deposited by glaciers.
TITRATION CURVES - a loci of points describing some solution property, usually pH, as a function of
the sequential addition of a strong acid (or base) to the system.
TOPMODEL - topographically based, variable source area hydrologic model.
TOPOGRAPHIC MAP - a map showing contours of surface elevation.
TRANSECT - see SOIL TRANSECT.
TRANSECTING - a field activity involving the collection of data at points along a designated line (see
TRANSECT POINTS).
TRANSECT POINTS - locations along a TRANSECT where data are collected.
13-24
-------
TRANSECT SEGMENT UNION - all transect stops in the same map unit on a WATERSHED.
TRANSECT STOPS - see TRANSECT POINTS.
TRANSFORMATION ERROR - calculates the residual mean square error of the digitized TIC locations
and the existing TICs.
TRAVERSING - a field activity that involves observation at uncontrolled representative locations in the
landscape.
TYPICAL YEAR (TY) DEPOSITION DATA - a dataset of atmospheric deposition developed within the
DDRP for specific use with the integrated watershed models.
UNCERTAINTY ANALYSIS • the process of partitioning modelling error or uncertainty to four sources:
intrinsic natural variability, prior assumptions/knowledge, model identification, and prediction error.
UNIVERSAL TRANSVERSE MERCATOR (UTM) PROJECTION - a standard map projection used by the
U.S. Geological Survey.
UPPER NODE - the upstream NODE of a STREAM REACH.
UPSTREAM REACH NODE - see UPPER NODE.
UTM COORDINATES - lines or points as represented in a UNIVERSAL TRANSVERSE MERCATOR
PROJECTION.
VALIDATION - comparison of model results with a set of prototype data not used for verification.
Comparison Includes the following: (1) using a dataset very similar to the verification data to determine
the validity of the model under conditions for which it was designed; (2) using a dataset quite different
from the verification data to determine the validity of the model under conditions for which it was not
designed but could possibly be used; and (3) using post-construction prototype data to determine the
validity of the predictions based on model results.
VANSELOW EXCHANGE FORMULATION - a formulation used to describe soil exchange reactions.
VARIABLE - a quantity that may assume any one of a set of values during the analysis.
VARIABLE SOURCE AREA - A topographically convergent, low transmissivrty area within a watershed
that tends to produce saturation excess overland flow during storm runoff periods.
VEGETATION - see FOREST COVER TYPE.
VERIFICATION - check of the behavior of an adjusted model against a set of prototype conditions.
WATERSHED - the geographic area from which SURFACE WATER drains into a particular lake or point
along a stream.
WATERSHED STEADY STATE - a condition in which inputs of a constituent to a WATERSHED equal
outputs.
WATERSHED SULFUR RETENTION - retention of sulfur by any of a number of mechanisms within a
WATERSHED.
WEATHERED BEDROCK • soft or partly consolidated BEDROCK that can be dug with a spade.
13-25
-------
WEATHERING - physical and chemical changes produced in rocks at or near the earth's surface by
atmospheric agents with essentially no transport of the altered materials.
WEIGHT - the inverse of a sample's inclusion probability; each sample site represents this number of sites
in the TARGET POPULATION.
WET DEPOSITION - for the purposes of the DORP, atmospheric deposition of materials via rain or snow.
WETLAND - an area, generally with hydric soils, that is saturated, flooded, or ponded long enough during
the growing season to develop anaerobic conditions in the upper soil horizons and that is capable of
supporting the growth of hydrophrtic vegetation.
ZERO ORDER - reaction rate set to some constant value, not affected by other factors.
ZERO-ORDER REACTION - a chemical reaction, the rate of which is independent of reactanl
concentration.
13-26
------- |