Appendix
o
•en
CVl
Future Effects of Long-Term Sulfur Deposition
on Surface ^at^Ghemis^; ,
in me Northeast and Southern Blue Ridge Province
of the
;;R: Church, K. W. Thornton- R^W. Shaffer, %; V^RS^ S; P
ta. R. Holdren, M. G. Johnsbri; d. Jv4^ R^^wir; I3t L. Glssfeii; :
D. A. Lammers, W; G. Campbell, C. t&ff;C; G;:Btaif^ L. H. ^e$el7
G. D. Bishop, D^C.
A Contribution to the
National Acid
U.S.
Office of Research and
Environmental ResearcfctMfcratb^S^^
-------
NOTICE
The information in this document has been fended 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 xfo
Plates ; xxvii
Contributors xxix
Acknowledgments xxxl
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 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 ofjtegults 4-8
4.4.5 Use of 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 thej^grtheastJjegipji 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 Ridae 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 Ridoe 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 SAMPUNG 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 PJiysjcal, 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 Levei 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 Hvdroloaic 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 Van'able 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 Lavrt 8-13
8.3.1.1 Introduction 8-13
8.3.1.2 Results and Discussion 8-18
8.3.2 Geomorphic/HvdroloQic Parameters 8-21
8.3.2.1 Introduction 8-21
8.3.2.2 Results and Discussion '.. 8-22
8.3.3 TOPMODEL 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 pH 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 Plus Ma. and pH 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 SumrnanLand 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 Mo and pH 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 Ma. 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 Mo fSOBCl ANC. and oH 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 (SOBC), 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 (SOBCV ANC. and oH 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
viil
-------
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 Model 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-Griaal and Reuss Models 9-185
9.3.5 Summary and Conclusions 9-196
10 LEVEL III 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 1 . 10-6
10.2.2 Integrated Lake-Watershed Acidification Study (ILWAS1 Model 10-7
10.2.3 Model of Acidification of Groundwater in Catchments (MAGICS 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 Ridoe 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 DDRP 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 Sensith/itv 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 MAGJC 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 Chances 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 III 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 Ridge 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 /Jteq L"1 for Constant
1-2
and Decreased Sulfur Deposition 1-19
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 Reclassrfication 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 (TY) 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-lake 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 Fall/Spring Row-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 Row 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
Geomorphic/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
xiil
-------
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 ln(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 Matn'x 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
xiv
-------
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 Soli 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 (CAMG16) 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
(S0416) 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 [SO4 ] 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 Suifate 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 Reg ton 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 Clear 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 /aeq L"1 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 j*eq 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 cycle 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 Meq L 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 Row 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 [SO/"]^ 7-28
7-6. Comparison of percent sulfur retention calculated using (A) modified-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 calculation 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 soil 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 fake
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 i) 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% i) 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% i) 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% t) 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 1 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 plots 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 suifate 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 suifate 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 ODRP classification of lake hydrologic type - Subregion 16 . . 5-33
5-4 Rnal DDRP classification of lake hydrologic type - Subregion 1C 5-34
5-5 Final DDRP classification of lake hydrologic type - Subregion 1D 5-35
5-6 Rnal 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 suifate deposition for the DDRP NE study sites 5-184
5-22 Pattern of typical year suifate deposition for the DDRP study sites in Subregion 1A. ... 5-185
5-23 Pattern of typical year suifate deposition for the DDRP study sites in Subregion 1B. ... 5-186
5-24 Pattern of typical year suifate deposition for the DDRP study sites in Subregion 1C. ... 5-187
5-25 Pattern of typical year suifate deposition for the DDRP study sites in Subregion 1D. ... 5-188
5-26 Pattern of typical year suifate deposition for the DDRP study sites in Subregion 1E. ... 5-189
5-27 Pattern of typical year suifate deposition for the DDRP SBRP study sites 5-190
5-28 Pattern of LTA suifate deposition for the DDRP NE study sites. 5-193
5-29 Pattern of LTA suifate deposition for the DDRP study sites in Subregion 1A 5-194
5-30 Pattern of LTA suifate deposition for the DDRP study sites in Subregion 1B 5-195
5-31 Pattern of LTA suifate deposition for the DDRP study sites in Subregion 1C. 5-196
5-32 Pattern of LTA suifate deposition for the DDRP study sites in Subregion 1D 5-197
5-33 Pattern of LTA suifate deposition for the DDRP study sites in Subregion 1E 5-198
5-34 Pattern of LTA suifate 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 sulfur 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 Changes in pH of SBRP stream reaches as projected by ILWAS 11-25
xxviii
-------
PRIMARY CONTRIBUTORS TO THE DDRP 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.
Sect/on 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, NSf Technology Services Corp.
C. J. Palmer, NSI Technology Services Corp.
M. L Papp, Lockheed Engineering and Sciences Co.
B. P. Rocnelle, 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 Gassed, NSi Technology Services Corp.
D. L Stevens, Eastern Oregon State University
M. B. Adams, Automated Systems Group, Inc.
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
0. D. Schmoyer, Martin Marietta Energy Systems. Inc.
P. W. Shaffer, NSI Technology Services Corp.
D. A. Woff, Martr'n Marietta Energy Systems, me.
Section 9: Level II Single-Factor Time Estimates1
G. R. Holdren, NSI Technology Services Corp.
M. G. Johnson, NSI Technology Services Corp.
C. I. LJff, 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. Homberger. 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, Unrversfty 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 thissection 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 ail the
others who helped us, but who are not named here, we also extend our sincere thanks.
William Ruckleshaus led the way in calling for the initiation of the ODRP and Lee Thomas showed
a continued and very patient interest in seeing that ft was compfeted properly. We thank them for their
foresight and leadership.
Courtney Riordan 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 U'nthurst, 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-Corvaffis (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 Omernik (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 Lipscomb, George Martin (Pennsylvania); Dave Van
Houten (Vermont); Talbert Gerald, Bob Wilkes (Georgia); Horace Smith, Andy Goodwin (North Carolina);
Darwin Newton, David Lewis (Tennessee); Miles 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 DORP activities by Milt Meyer, Ken Hinkley, and Dick Arnold of the 80S 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 Litzke 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 DORP 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
Stan 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 Hornberger, Pat
Ryan and David Wolock (University of Virginia), Jerry Schnoor, Tom Lee, Nikdaos Nikolaidis, 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 suifate 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 Ofsen, 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 dark of the EPA's Atmospheric and Exposure Assessment
Laboratory-Research Triangle Park and Steve Se/tkop of Analytical Services, incorporated, provided key
Information on estimates of atmospheric dry deposition. Steve Lindberg 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.
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 Mamnorek, 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 Ippoliti (NSI), Amy Vickland (USDA Forest Service), Una 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
-------
APPENDIX
SUPPLEMENTAL LEVEL II! RESULTS
AND INFORMATION
A1: MODEL CALIBRATION/CONFIRMATION REPORTS
A2: WATERSHEDS SIMULATED BY ETD, ILWAS, AND MAGIC
A3: UNCERTAINTY ESTIMATES AND CONFIDENCE BOUNDS FOR MODEL
PROJECTIONS
-------
APPENDIX A.1 MODEL CALIBRATION/CONFIRMATION REPORTS
A.1-1 Enhanced Trickle-Down (ETD
A.1-2 Integrated Lake-Watershed Acidification Study (ILWAS)
A. 1-3 Model of Acidification of Groundwater in Catchments (MAGIC)
-------
APPENDIX A.1-1
Enhanced Trickle-Down (ETD)
-------
Technical Report Ho. CEE-ARRG-86.03
Modeling Short and Long Term Impacts
of Acid Precipitation Using the
Enhanced Trickle-Down Model:
Lake Woods Case Study
by:
Nikolaos P. Nlkolaidis
{Constantine P. Georgakakos
Jerald L. Schnoor
Civil and Environmental Engineering
The University of Iowa
Iowa City, Iowa 52242
First Draft: November 1986
-------
MODELING SHORT AMD LONG TERM IMPACTS OF ACID PRECIPITATION USING THE ENHANCED
TRICKLE-DOWf MODEL: LAKE VTOODS CASE STUDY
INTRODUCTION
Public awareness of the effect of acid deposition on the terrestrial and
aquatic environment has imposed pressure on the governments of European and
North American countries to control the emissions of sulfur and nitrogen
oxides. However, before any threshold limit be established a critical
question concerning the response of aquatic ecosystems to acidification should
be answered. The case is whether further decreases in alkalinity of surface
waters will occur if present emission loadings remain constant. It is the
purpose of this research to aid in answering the above question by application
of a mathematical model to lakes in the Adirondack Mountains of New York,
where most of the effects of acid rain in the United States have been
documented.
In order to examine the behavior of aquatic responses to acid rain, let
us define those waters whose alkalinity changes further with time as "direct
response** systems. Figure 1 is from a 1981 National Research Council Panel on
"Processes of Lake Acidification," which discussed direct response and delayed
response systems. A direct response system is expected to respond to changes
in acid deposition over a relatively short period of time, depending on the
hydraulic detention time of the lake. In this case, the rate of
neutralization in the watershed Is a kinetically-controlled process, and
changes In acid deposition may modify the rate of acid neutralization but
-------
would not exhaust the capacity of the watershed for neutralization. A delayed
response watershed, on the other hand, would gradually lose its capacity for
neutralizing acid deposition, and the alkalinity of its waters would gradually
decrease over time (decades to centuries). There are two possible causes of a
delayed response: (1) a gradual depletion of mineral bases or base cations
from soil exchange sites, and/or (2} a gradual exhaustion of sulfate
adsorption sites in the soil. The two processes are related to the ordinate
and abscissa of Figure 1. If a wateshed has a low ability to supply bases and
a low ability to retain sulfate (S0j,~2 from HgSO^), then it is a direct
(quick) response system.
Figure 2 gives a more temporal picture of the difference between direct
and delayed response systems. If watersheds are of the "protected" type of
Figure 1, there will be no response to a step function increase in acid
deposition (panel B., Figure 2). These watersheds are capable of neutralizing
any observed amount of acid deposition. Moderately sensitive lakes may
respond with significant decreases in alkalinity. However, they will retain
some acid neutralizing capacity (ANC). These lakes are shown in panel C of
Figure 2 and are the "direct11 response lakes. Depending on their watershed
characteristics (water flows only through shallow acidic soil zones, etc.),
these lakes may becone acidic after 1000 years (delayed-1) or some may have a
much faster response (delayed-2), depending on when their soil ANC or sulfate-
sorption capacity becomes exhausted. Extremely sensitive lakes (panel D) have
already become acidified due to acid deposition.
Mathematical models can be used to estimate the alkalinity response of
lakes over time. The Trickle-Down model has been used to assess resources-at-
risk to acid deposition in the Northeast and Upper Midwest. A steady state
version of the model was first reported by Stumm and Schnoor (1983), and a
-------
;time variable version of the model was given by Schnoor, Palmer and Glass ^x
^X'-- ^
(1981) for Omaday Lake and Filson Creek, Minnesota. The time variable model
has been applied to two seepage lakes in Wisconsin (Lakes Clara and
Vandercook) by Lin (1985). The steady state version has been used to estimate
the number of lakes the would become acidic under various acid loading
scenarios for approximately 1400 lakes across the depositional gradient
(Schnoor, Lee, Nikolaidis and Nair, 1986). Being simple and yet based on the
principle of continuity for alkalinity, the model has proven to be useful in
acid precipitation assessments. The model, as it is used by Lin (1985) is
suitable for seepage lakes only. Several modifications were required to
generalize the model for every surface water hydrologic system.
The Enhanced Trickle-Down model (ETD) is a combination of four submodels
(Nikolaidis et. al., 1986). It simulates the hydrology, alkalinity, sulfate
and chloride of a catchment simultaneously. The construction of the
hydrologic submodel equations was done in such a way that each flow component
is multiplied by a correction factor that reflects the lumped nature of the
model. The alkalinity and sulfate submodels are sets of mass balance
equations on each compartment. The mass balance equations for alkalinity and
sulfate include the predominant processes of ion exchange, chemical
weathering, sulfate sorption, sulfate reduction and iron oxidation as reaction
term.
The scope of this report is to calibrate, verify, and .perform a
sensitivity analysis on the parameters' and examine the long term response to
various acid loadings of Lake Woods, using the ETD model. Lake Woods is a 2.1
km2 forested watershed located in the Adirondack Park region of Kew York
State. Lake Woods is considered acidic, with a mean outlet pH of 4.7 and mean
outlet alkalinity of -10 ueq/L. THe lake receives an annual precipitation of
-------
J..2 m and the mean annual temperature Is .5° C. The vegetation of the ..
»••'
watershed is dominated by Sugar Maple, Beech, Yellow Birch and Red Spruce.
The watershed is underlain by granitic bedrock. The surficial geology is
comprised by thin but variable thickness glacial till. The soils are
predominately Beeket Spodosols.
MODEL CALIBRATION
PROCEDURE
The calibration of ETD for Lake Woods was achieved by decoupling the
hydrologic, sulfate and alkalinity submodels. Since there is a coupling
between hydrology and chemistry, the hydrologic parameters were calibrated
first. When a good fit between the simulated and the observed outflow was
achieved, then the hydrologic parameters were fixed and the sulfate reaction
parameters were calibrated. Finally, with fixed hydrologic and sulfate
parameters the weathering rates were adjusted so a good fit would be
established for alkalinity. The calibration of each submodel was achieved by
using a standardized optimization package, IDESIGN (Arora et. al., 1985) and a
trial and error procedure. The ranges of parameters for calibration were
input to IDESIGN, which was connected with the ETD model. At each iteration
(complete two year simulation) the cost function was evaluated. The cost
function is of the form:
m 2
COST - S (S.. - 0.)
t-1 fc fc
where: St - simulation value of model output variable at time, t;
Ot • observed value of model output variable at time, t; and
at - total number of time steps of observed data.
-------
The model output variables that participated in the cost function were:
outflow, alkalinity and sulfate. IDESIGN performs minimization of the cost
function using Fletcher-Reeves algorithm (gradient method). Bounds known
a priori on physical quantities and parameters were included as constraints in
the optimization process. However, in all gradient methods when the objective
function is of the least-squares type, assume that the residuals are
homoscedastic, independent and sufficiently small to assume their normality.
When this is the case, the maximum likelihood method justifies the use of the
least squares objective functions (Isabel et. al., 1986). In our case the
residuals are not homoscedastic, the thus the results of the optimization by
IDESIGN tend to bias the results towards the extreme values of the
residuals. On the other hand, IDESIGN does bring the parameters within a
reasonable range from their optimal value. After this point, optimal
calibration was based on trial and error method. There were three main
guidelines used to establish the optimum value:
1) obtain closure of cumulative flow or mass during the whole period,
2) capture the seasonal variability of the state variables, and
3) capture the peaks and valleys of the daily flows and concentrations.
The calibration of the hydrologic parameters was established in six
steps. Model variables symbols were borrowed from Nikolaidis et. al., 1986.
1. Evaluation of QNET:
Based on the knowledge acquired from the field data evaluation, the
groundwater net import-export flow is evaluated as follows:
QNET -P- ZET-0± iS
where P - total precipitation input,
E ET - total evapotranspiration,
0 - outflow, and
-------
ffA S - change In storage over the whole calibration period.
2. Initial Estimate of KPAN3 and KPAN5:
By taking a large enough portion of the summer record where A Slake »
0.0 and assuming that change in storage of the terrestial compartments vs.
zero, one can establish a mass balance and select initial estimates for the
soil and lake evaporation correction coefficients.
3. Calibration of Lateral and Vertical Permeability Correction
Coefficients:
The lateral and vertical flows of the soil and unsaturated zone
compartments are fully functional only during the period that the ground is
not frozen. The frozen ground formulation included in the ETO model accounts
for the seasonal variation of the effective lateral and vertical
permeabilities. At this step, IDESIGN is allowed to calibrate the KLAT and
KPERC correction factors utilizing a non-frozen ground period.
4. Calibration of Snow Parameters:
Utilizing the full record, and fixing all the previously mentioned
parameters at the values established above, the snow parameters BETA, KAPPA
and KPAN2 are calibrated using IDESIGN.
5. Refinement of KPAN, KLAT and KPERC Parameters:
Setting the snow parameters, IDESIGN Is allowed to optimize and refine
the evaporation and permeability parameters.
6. Trial and Error Identification of Optimum:
At this point, a comprehensive evaluation of the results is performed.
The evaluation of the cumulative and dally flows would identify the bias in
the calibrated parameters due to objective function and optimization algorithm
formulation. Corrective action can be applied by. varying the hydrologic
parameters until aggrement between the simulation and field data is achieved.
-------
^Similarly, the calibration of sulfate parameters is being achieved by
allowing IDESIGN to calibrate the sulfate sorption and reduction
coefficients. Then trial and error adjustments are made. The calibration of
alkalinity model paramaters is performed in a manner similar to the sulfate
model parameters calibration.
INPUT DATA
The time series data of surface precipitation, evaporation and
temperature for the calibration and verification periods are presented in
Figures 3 through 8. Precipitation and temperature data were collected in the
WLW station and it was part of the ILWAS project. Evaporation measurements
were obtained by using van Buvel's combination method (Nikolaidis, 1986).
Figure 9 and 10 present the wet and dry daily sulfate and acidity loading
time series. An analysis of wet and dry deposition is presented in Table 1.
Here, one should notice that dry deposition of acidity is about 32% of the
total acidity and dry deposition of sulfate is about U2$ of the total sulfate
loading. On the average, 12 metric tons of sulfate are deposited on Lake
Woods' watershed per year, which roughly corresponds to 4 metric tons of
sulfur.
RESULTS
A list of the hydrologic and chemical watershed descriptors that were
input to the model are presented in Table 2. The optimum values of the
calibrated parameters after following the procedure described in previously
are presented in Table 3. Time-series input for precipitation, evaporation
-------
evaporation |nd temperature for the calibration period (9/78 through 8/80) are
shown in Figures 3, 4, and 5. Time series plots for wet and dry sulfate and
acidity loadings for both calibration and verification periods are presented
in Figures 9 and 10 respectively. Comparisons between lake outflow, chloride,
sulfate and alkalinity simulations and field data are presented in Figures 11
through 16.
A complete hydrologic budget for Lake Woods is shown in Table M. 61$ of
the total inflow to the lake is from the soil compartment. Direct
precipitation and snow contributions are 12* each. Finally, H% of the total
inflow to the lake is from the unsaturated zone. The I/Q ratio (total
precipitation input to outflow ratio) for Lake Woods is 1.65. As a final
comment, one could say that the watershed of Lake Woods is very flashy, since
75$ of the total inflow is due either to direct precipitation or due to flow
through soil.
An alkalinity budget of Lake Woods is presented in Table 5. The majority
of the acidity input to the lake is coming through soil and total, direct to
the lake deposition with 69.2 and 18$ respectively. Wet precipitation
contributes 11.5$ and snowmelt contributes 12.5$ of the total acidity input.
According to the simulation results, 63.3$ of the neutralization of the total
acidity input to the watershed is due to the in-lake processes (weathering,
sulfate reduction and iron oxidation).
A sulfate budget is shown in Table 6. The majority of sulfate input is
through the soil and total, direct to the lake deposition with 66$ and 12.9$
respectively. Wet precipitation and snowmelt contribute about 7.1$ each. The
net sulfate reduction in the lake sediments is 11.2$ of the total input to the
lake.
-------
7 presents the MSB evaluation.for the calibration period for
discharge, cumulative discharge, chloride, alkalinity, and sulfate.
VERIFICATION
The verification simulations of Lake Woods were performed by using one
more year of input field data time series not included in the calibration
period. The time-series input for precipitation, evaporation and temperature
for the verification period (9/80-8/81) are shown in Figures 6, 7, and 8. The
wet and dry deposition time-series of sulfate and acidity were part of Figures
9 and 10 respectively. The simulation results between lake outflow, chloride,
sulfate and alkalinity for the verification period and field data are
presented in Figures 17 through 22. The results are comparable with the
calibration results except for the fact that the model underestimates the
total cumulative outflow. This is due to the fact that a 25 mm/day flow was
measured for a few days the first week of March 1981 where the temperature was
about 10°C and there was snowcover. Examining the amount of evaporation
during the same period, one can see that it is rather high (~5~7 mm/day).
This indicates that instead of melting this water, the model evaporates it.
It is obvious that such a condition is not taken under consideration. Since
it was observed during the verification period, no changes were made in the
model formulation. Table 8 presents the MSB evaluation for the verification
period.
SENSITIVITY ANALYSIS
The sensitivity of lake outflow to each of the hydrologic parameters was
-------
10
Checked by varying the parameters 10% above and 10J below their optimum values
using the calibration period record. The percent change of the mean square
error (USE) of the outflow is presented in Table 9. The most sensitive
parameters were: the snow parameters BETA, KAPPA and KPAN2, the lake
evaporation correction factor, KPAN5 and the soil lateral and vertical
permeability correction coefficients.
The sensitivity of outflow alkalinity to each of the chemical parameters
was similarly checked. Table 10 contains the percent change in MSE, The most
sensitive parameters were: the lake compartment weathering constants KH5 and
K05 and the lake sulfate reduction reaction rate k. The alkalinity MSE
sensitivity results were expected to be low because Woods Lake has a flashy
hydrology so the detention time of the water in the watershed is small and
there is not enough time for weathering to occur.
LONG-TERM SIMULATION
In order to perform long term simulations, the existing 3 year record was
input repeatedly for 17 times, which extended the simulation period to 51
years. Three simulation runs were made: one for present loading, one for
half loading and one for double loading. In order to establish the half and
double loadings both wet and dry sulfate and alkalinity inputs were halved or
doubled respectively. Figures 23 and 21 show, the projection of lake
alkalinity and sulfate for the next 50 years under three different loading
scenarios. The prediction indicates that Lake Woods is a direct response
system and that it would only take a few years for the lake to respond to a
decrease in acid deposition loading. The lake currently has reached steady
state with a mean alkalinity of about -10 ueq/L and a mean sulfate of 130
-------
11
ftieq/L. If the loading was to be halved, the mean alkalinity would be 2!
yeq/L and the mean sulfate concentration 65 ueq/L. On the other hand, if
the loading was to be doubled, the mean alkalinity would be -40 ueq/L and the
mean sulfate concentration 260
-------
12
REFERENCES
Arora, J. S., Thanedar, P. B. and Tseng, C. H. (1985). User's manual for
program IDESIGN, version 3.4 for PRIME computers. Optimal Design Laboratory,
College of Engineering, University of Iowa, Technical Report No. ODL 85.10.
Isabel, B. and Villeneuve, J. P. (1986). Importance of the convergence
criterion in the automatic calibration of hydrological models. Water
Resources Research. Vol. 22, No. 10, pp. 1367-1370.
Lin, J. C. (1985). Modeling aluminum and alkalinity concentrations in
watershed receiving acid deposition. Ph.D. Thesis, University of Iowa, Iowa
City, 235 pp.
Nlkolaidis, N. P., Rajaram, H., Schnoor, J. L. and Georgakakos, K. P.
(1986). Enhanced Trickle-Down model description. Civil and Environmental
Engineering, University of Iowa, Technical Report No. CEE-ARRG-86.01.
Schnoor, J. L., Lee, S. J., Nikolaidis, N. P. and Nair, D. R. (1985). Lake
resources at risk to acidic deposition in eastern United States Submitted to
Hater, Air and Soil Pollution.
Schnoor, J. L., Palmer, W. D., Jr. and Glass, G. E. (198*0. In "Modeling of
total acid precipitation impacts"; Schnoor, J. L. (ed), Butterworth
Publishers, Woburn, MA.
Stumm, W., Furrer, G. and Kunz, B. (1983). The role of surface coordination
in precipitation and dissolution of mineral phases. Croat. Chem. Acta. 56, 4
593-611.
Stumm, W. and Schnoor, J. L. (1983). Naturwissenschaften, 70, 216. (in
German).
-------
13
TABLE 1. LAKE WOODS
ANALYSIS OF WET AND DRY DEPOSITION
ACIDITY (eq/ha • yr) SULFATE (eq/ha • yr)
Period:
9/78-8-79
8/79-8/80
9.80-8/81
Average
Wet
865.6
719.4
781.0
788.7
Dry
382. 4
3^2.3
332.3
352.3
Wet
820.0
617.0
671.0
702.7
Dry
5^7.6
157. 4
487.0
497.3
Total average Acidity - 1141 eq/ha yr
Dry Deposition - 30.9$ total Deposition
Total average Sulfate - 1200 eq/ha yr
Dry Deposition - 41.5% total Deposition
Sulfate Loading (Average)
Wet - 33.73 kg/ha/yr
Dry - 23.87 kg/ha/yr
Total - 57.60 kg/ha/yr
On the average, 12 metric tons of sulfate are deposited on Lake Woods'
watershed (or 4 metric tons of sulfur).
-------
TABLE 2. List of watershed descriptors used for model calibration.
GENERAL WATERSHED CHARACTERISTICS:
AQUATIC AREA -
TERRESTRIAL AREA -
CHARACTERISTIC DISTANCE -
DEPTH TO BEDROCK -
PARTIAL PRESSURE OF ATM C02 -
0.2300E+06 SQUARE METERS
0.1840E+07 SQUARE METERS
0.2625E+03 METERS
2.3000 METERS
0.0003 ATMOSPHERES
SURFACE WATER PC02 IS 1.50 TIMES SATURATED PC02
SOIL COMPARTMENT CHARACTERISTICS:
POROSITY -
DEPTH OF SOIL LAYER -
SUM OF BASES -
SOIL DENSITY -
.2700
0.3684 METERS
106.1000 EQUIVALENTS/KILOGRAM
1009.0000 KILOGRAMS/CO. METER
UNSATURATED ZONE COMPARTMENT CHARACTERISTICS:
POROSITY -
TRANSPIRATION COEFFICIENT -
BARE-GROUND FROST COEFFICIENT -
REDUCTION IN FROST COEFFICIENT -
DAILY THAW RATE -
INITIAL FROST INDEX -
LIMITING FROST INDEX -
THAW COEFFICIENT -
BULK DENSITY -
SURFACE WATER BODIES CHARACTERISTICS:
STREAM BED ELEVATION AT OUTFLOW -
0.2000
0.0010
0.1000
0.0800
0.1200
0.0000
-3-0000
0.2000
1620.0000 KILOGRAMS/CU.
INCHES/DAY
DEGREES C
DEGREES C
DEGREES C
METER
11.9000 METERS
GROUNDWATER COMPARTMENT CHARACTERISTICS:
POROSITY - 0.2000
-------
TABLE 3. List of optimum values of the calibrated parameters,
a) Hydrologic Parameters b)
-Snows
BETA - 0.6795
KAPPA - 1.1423 in/day/°C
KPAN2 - 1.0066
-Evaporation:
KPAH3 - 1.5471
KPAN4 - 1.6381
-Lateral and Vertical Flows: c)
KLAT3 - 242.810
Alkalinity Parameters
RE3 - 6.4E-8 m3/eq/day
KH4 - 9.0E-1 meq/mVday
KH5 • 8.5E-2 meq/m2/day
KH6 - 1 .OE-2 meq/nr/day
KOU » 1.1E-2 meq/nr/day
K05 - 7.9E-2 raeq/ra2/day
K06 - 1.1E-3 meq/m2/day
Sulfate Parameters
CF - 2.0217
KLAT4 - 14.791
KPERC3 - 1.3117E-2
KPERC4 - 1.1711E-2
-Groundwater:
01 - 0.5963
FRAX - 0.4804
ALF1 - 6.2375E-4
ALF2 - 4.4043E-1
eq/m
KP3 « 6.0E-5
KP4 - 4.90E-6
K - 2.033E-3 1/day
KP6 - 4.90E-7 £3^8
eg/nr
-------
.TABLE 3." "MONTHLY ALKALINITY BUDGET OP LAKE WOODS
INPUT i
INPUT 2
INPUT 3
INPUT 4
INPUT S .^.INPUT 6
OUT 1
INPUT 7
OUT 2
REACTION
STORAGE
-6.
-0.
•0.
-0.
-0.
-0.
-0.
-c.
-0.
-0.
-c.
-0.
-•3.
-0.
-O.
-0.
-0.
0.
-c.
-0,
-0.
-0.
-.;-.
• •;.
2422E+04
1602E+04
5029E+03
3001E+03
1904E+03
2469E+02
1843c-K"»
SSOAE-f-OJ
•SC34£f04
6480E+03
1174E+04
3O76E+04
AS'l'/E-t-OS
1843E-H54
1939E+04
44B1E+03
1116E+02
OOOOEt-00
3480E+03
1662E+Q4
1507E+04
I 18OE+04
22B5c«-G4
^E-C-l
O.C0MOE+OO1
O.OOOOE-i-00
-0.2830E+03
-0.2779E+04
-0.1429E+04
-0. 18SQE+04
-0.1162E+05
-0. 206£E*O4
O.COOO5-OO
O.OQOOE+OO
O.OOCOEfOO
O.OOOOE+OO
O.OOOCiE»-Oi5
O.OOOCE-OO
-0. 1¥4«E' O4
-0. 18B7E-»-O4
-0.9072E+03
.-0.7174E+03
-O. &449E*O4
O.OOOOE-vOO
O.OOOOE+OO
O.OOOOE^OO
O.OOOOE-»-CO
O.CCOOE^O
-0. 1626E+03
-O. 1332E+03
-0.2793E*04
-0.3422E+04
-0:3377E+04
-0. 673BE-M33
-0. 123BE-f03
-0. 101SE-K>3
-O. 14S3E*O3
-O.6329E*O4
-0.3S34E+C4
-O. 1943E+03
-0. !O93E-»-O3
-O. 1 l&SE+OS
-0. 14455*03
-O.S388E+04
— 0.9143£-*-O3
-0.8231E+02
-0.4706£*O4
.-0.1064E+03
*-Os4491E+04 "
-0. 1282E+OS
-0. I777E+OS
-0. 1219E+05
0. 4234E+02
0. 4596E+02
0.1612E+03
0.1323E*03
0. 123AE+03
0. 1389E+03
O.7297E»02
0. 1294E+03
0.111CE+03
0. H29E*03
O.1412E*O3
-O.a033E*O2
O.1O9QE+03
O. 3832E+03
O. 32O3E+O3
O. 1037E*04
0. 170OE+04
0. 1847E+04
0. 1968E*04
0. 1300E+04
0.1433E+04
O.S568E+-03
0.64O7E+03
0.4229E+03
-O.6581E-I-O3
-0.3393E*Ol
-0.6299E-01
-0.19S2E-H50
-0.3271E*00
-0.7992E-04,4r
-0. 13BSE+O2 s
-O.3899E+01 t
-O.3630E+01 §
-0. 4342E.-03 1
-0.249CE+O1
-0.4B33E*02
-0. 2161£fO2
-0.32S6E+OC
-O.77B2E*01
-0. 1115E+OO
O. OOOOE+00
O. COOOE<-CO
-O. 1003E+OO
-0.3098E+01
-O.3038E*Ol ';
-0.3069E+01
-0.61Q3E+02
-0.23S7E+OO
0.79i4E*02
O.8021E+02
0.7643E1-02
O.8076E+O2
O.B231E*02
0.7323E+O2
O.B467E+02
0.7888E+02
0.8070E+02
0.7797E+O2
C.842CE+O2
O.B848E4-02
O.B317E+O2
0.3639E-»-02
O-BSliE*O2
O.8917E+02
0.9104E-t-02
O.8425E+02
0.9&&9E+02
O.9423E-»-02
:.0. I026E+03
O. 1046E*O3
0.1131E+O3
O. 1 IS6S+O3
-O.3304E-Ot
-0. 4S23E-O1
-0.3133E-01
-O.S408E-01
-0.8884E-01
-O.8309E-01
-O. 1&86E+00
-O.208SE+OO
-O. 164SE-I-00
-O.US9E+00
-0.3417S-01
-0.226&E-01
-O. I647E-01
O. 1298E-Ci
— O. J673S— Oi
-0.7438E-01
-O.8332E-01
-0.7J06E-O1
-O.900SE-01
-0. 1183E+00
-oi"iZ331E-Oi
O. A008E-O2
-0. 1S4AE-O2
-O. i371E*O4
-O.S911E+03
-0. 4533E+03
-0.4128E-I-03
-0. 6987E+03
— 0.3232E+03
-0. 48O6E+03
-0.4802E+03
-0. IO32E+04
-0.9960E+03
-O.9333E+C3
-0. 1030E*04
-O.97B4E+O3
-O.a4lCE-i-C3
-0.7323E>03
-O.3494E+03
-O.29O2E+03
-0.3214E+Q3
-0.3130E+03
-O. 34S7E*03
-0.4720E+03
-O. 703BE+03
-O. 1296E-S-04
-O. 1 430E+04
-O. JOSOH-t-04
-O.8703E*O3
-O.2404E>O3
-0.9530E+03
-0.223OE*O4
-0.8fl44E-K)3
-0. i-033E*63 "
-0. 7234E+04
-O.3092E+04
-0.1063E+04
-0.1478E*03
-0.&931E-103
-0.7723E*03
0. 339CE-:-03
-0.6272E+03
-0. 1782E-«-04
-0. 1082E+04
-0.4157E+03
-O.3C31E-I-04
-0.4177E-»-04
-0.483&E+O3
-O. 6969E+O3
0.6133E*02
-O.B706E*O2
0.2049E*OS
O.S123E+04
0.6732E+03
O.S2&3E+03
0. 3SA4E-»-O3
O.3320E*04
O.2O18E+03
0.2S93E-»-03
0.2398E*03
0.2240E+03
O.2094E+C3
O» i^SlE^OS
O* i7o4a*OS
O. 4533E+O4
O. 1978E+C4
O.B470E+03
0.8340E*O2
0.762BE+O3
0. 1770E*03
0.2210E+OS
O. 1986E+O3
0. 1894E-rOS
0.i9:iS*03
O. 32272*03
0.3762E+04
0.1368E+04
-0.3274E-»-O4
-O.2712E-»-04
—0. J398E-K54
-O.1233E*OS
O. 1403E+O3
0. 1059Ef05
0.17O7E+CS
O.lSt IE-OS
-0- 19'66E-»-<34
0. 7575E+O4
0. 3S6SE «-«:,4
-O. 130BE+05
-0. 3387E-»-O4
0.1S96E+O4
0. 13O9E+O4
-0.3937E-I-O4
0. lOi>^E-«-O5
0. 1765E+OS
0.68l2E-t-04
-0.1778E-t-O4
0.4312E-0*
-O.3192E-I-OS -0.21&9EM56 C.1337E*C«5 -0.8661E+O3 0.2117E+04
-0.4160E+OS
0.6791E-«-03
ota: Unit: eq/month,
Input Is .Atmospheric compartment to Lake,
Input 2: Snow compartment to Lake,
Input ~: Soil compartment to Laka,
1'nput 4: UnsAtur>ited zone compartment ta Lake,
Input 3t Overland -flow to L*k0,
Input 6t Gruundwatwr compartmant to Laka,
Out Is Uiktt :u.i>p3''ta;aftt to iSraur.dwat-jr-,
.nput 7:Dry (jepcsition to the Lake,
Cut 2s Outflow from the Lake,
\\jactijr.: "nte-oal prcciuction in i*..ie Lake compart/r.snt,
Storage: Change in Lake stcrac.<*.
-------
TABLE 4. MONTHLV HyDROLOOlC PUOdET OF LAKE WOODS
INPUT 1
INPUT 2
tNPLT 3
INPUT 4
INPUT S
INPUT 6
OUT 1
our 2
OUT
STORAGE
0. 1312E-01
0. 1243E-OI
0.42046-02
0.42606-02
0. 3933E-02
0.42336-03
O. 124S6-V1
0. 97936-0::
0. 12306-01
0.3133E-02
0, 793O6-02
O. 19426-O1
O. 1600E-01
0. 1324E-OI
0. 1279E-O1
0. 39316-02
O. 36446-04
O.OOOOE+00
0.37336-02
0. 1I54E-01
0.78436-02
0. 1O61E-01
O. 17816-01
0. 77346-O2
0.0000E+00
0. 00006+00
0.1783E-O2
0. 10076-01
O.2636E-01
0.8484E-02
0.6S9 16-01
0. 1B20E-U1
0.01)006+00
0.00006+00
O.00O06+OO
0.00006+00
O.OOOOE+OO
O.OOOOE+OO
0. 1O77E-01
0.99O46-O2
0. 10836-01
0. 1703E-02
0.3300E-01
0.4031E-02
0.00006+00
O.WOOOE+OO
O.OOOOE+OO
0.00006+00
0.
0.
0.
0.
0.
0.
O.
0.
0.
0.
0.
O.
0.
t>.
0.
O.
0.
0.
•o.
0.
o.
0.
0.
o.
63746-01
55406-01
21946-01
2822E-C1
33386-01
11736-01
8029E-O1
606S6-O1
23276 -O»
22B6E-O1
21046-01
59846 -Ol
862SE-01
6279E— Ol
60376-01
34936-01
73416-02
9018E-03
. 3446E-O1
64716-01
2384E-01
3914S-O1
,74466-01
4004E-O1
0.S0246-02
0,47496-02
0.33006-02
0.3O366-02
0 . 26 1 66-02
0. 23386-02
0. 4O796-02
0.32536-02
0.47146-02
O.32SIE-02
0. 13876-02
0.30436-02
0.61906-02
0.80186-02
0.9889E-02
O. I213E-01
0. 1 1 126-01
O. 8006E-02
0. 7S8OE-02
O. 10396-01
O. I003E-01
O. 96OVE-02
0.847S6-02
O. 96236-02
0.
O.
O.
0.
0.
O.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
33676-02
3060E-04
63S6E-O6
1626E-OS
2067E-04
1 4S6E-08
23326-03
42926-04
12936-04
3333E-O8
151 46-04
38296-03
16776-02
42946-03
1093E-03
189BE-O3
96776-09
2984E-1 1
436££— OS
32376-04
26716-O4
4241E-04
,47436-03
1324E-C3
0. 1290E-02
0. I334E-02
O. 1280E-O2
0.13236-02
0.1323E-02
0. 11B7E-02
0.1320E-02
O. 1Z61E-0"
0. t'2936-02
0. 12426-02
0. 12796-02
O. 12986-O2
0. 12466-02
0. 12796-02
0.12346-02
0.12676-O2
0. 12666-02
0.11426-02
0.12736-02
0. 12336-02
0. 12636-02
0. 12316-02
0.12766-02
0.127-16-02
0.26976-03
0.2790E-03
0.26786-03
O. 27726-03
0.2758E-03
0.24B3E-03
0.2761E-03
(.'- 26376-03
0.27046-03
0. 2S97E-O3
0.26726-03
0.2713E-O3
0.26076-03
O. 26746-03
O. 2S80E-O3
0.26306-03
0.26476-03
0.23896-05
0. 2646E-63
0.237flE-OS
0.244&E-03
0.23736-03
0. 26666-03
0.26386-03
O. 13686-01
0. IBB7E-01
O. 12*76-01
O. 24S0E-O3
O. OOOOE+00
O.OOOOE+OO
0. 50986-03
0.66E9E-02
«>. '.20036-0!
0.2607E-01
0.23316-01
0. 17O3E-01
0. 1397E-OI
0. 1196E-O1
0.66326-02
0.3113E-O2
0.4S48E-O3
O.00g+O0
O.
Input It AtmoBpherlc compartmant to L«lc«,
Input Zl Snow caaip«rtin*nt to L«k»,
Input 3i Sail eompcrtinvnt to !.*!<•,
Input 4i Un»«tur«t»d zone compartment to Like,
Input S> Overland 41OM to U»k«,
Input 6« Groundwiitar eomp*rtmant to L«kc,
Out It L*km eonip«rtin«nt to Oroundwatur,
Out 2i Evaporation from th» L*k«t
Out 3l Outflow from th« Lak«,
Storagvt Chang* in Laka »tor»g«.
-------
INPUT 1
INPUT 2
..TABLE 6. MONTHLY SULFATE BUDGET OF LAKE WOODS
"•>'*"., " '
INPUT 3 INPUT 4 INPUT 3 INPUT & OUT 1 INPUT 7 OUT 2
REACTION STORAGE
0.
0.4671E+03
0.411SE+03
0.1091E+03
0.5187E+02
0.1657E+Q4'
0.7829£*C3
3.2727E+C4
0.OOOOE+CV
O.OOOOE+00
0.22&3E+03
0.403SE+04
O.I313E+04
0.1231E+04
0.:018E+OS
C.IA01E+04
0.1338E+04
0.2933E+04
0.6B86E+03
O.1394E+04
0.1357E+C4
O.3384E+03
O.8891E+01
0.OOOOE+00
0.2418E+03
0.1151E+04
0.1693E+04
0.1298E+04
0.230BE+04
0.1313E+04
C.OOOOE+00
O.OOOOE+00
' O.OOOOE+00
O.OOCOE+OO
O.OOOOE+OO
0.16E9E>04
0.1S07E+04
^OV8238E+03
' 6/5031E+04
O.OOOOE+00
O.OOOOE+OO
O.OOOOE+00
0.OOOOE+OO
O.OOOOE+00
C.3149C+OS
O.181BE+03
0.3688E+04
0.4362E+04
0.3998E+04
0.9071E+03
0.1242S+05
0.1147E+03
0.
0.1104E+03
0.2398E+03
O.1434E+03
O.1390E+03
0.1378t+03
0.&12SE+O4
O.1209E+04
0.1498E+03
0.4267E+04
0.9947E+04
O.7418E+O4
"O. 1723E+03
0.2333E+03
0.141SE+OS
0.2233E*04
0.138SE+04
O.SS4CE+03
0.3649E+03
0.3640E+03
O.934.1E+03
0.i061E+04
0.120SE+04
0.9O32E+03
0.1373E+04
0.19A8E+04
0.1103E+O4
O.4287E+O1
0.4030E-01
0. 1634E+00
0.2800E+00
0.1679E-O3
0.1S81E+02
0.3747E+01
0.2599E+04
0.2954E+04
0.2S87E+04
0.1843E+Q4
0.1674E+04
0.1947E+04
0.2013E+O4
0.24&1E+04
0.2708E+04
0.3370E+O4
. 0.3056E-O3
0.2930E+pt
- p.&835E+g2
*-0. 15S9E+b2
O.4433E+OO'
O.A997E+01
0.7833E-01
^S^*00
" 0.9790E-
O.3166E+O1J
0.4922E+01
O.6363E+02
0.
O.
0.3223E+03
0.33S2E+03
0.3693E+03
O.S7&OE+03
O.S190E+03
0.
O.
O.
0.
0.3982E+03
0.6233E+O3
0.6222E+03
0.6433E-03
0.6234E+03
.0.6390E+03
O.63&&E+O3
.4378E+03
.6331E+03
0.62O4E+O3
0.6517E+03
0.6461E+00
0.6991E+OO
0.6820E+00
0.7437E+00
0.77&4E+00
0.713SE+OO
0.8069E+00
0.77S3E+00
O.7424E+00
0.7332E+OO
0.1B77E+04
O.B164E+O3
0.6320E+03
0.3531E+03
C.522&E+O3
O.3OO1E+03
O.SO21E+03
0.7349E+03
0. 1391E->O4
0.1760E+O4
rO. 1730E+O4
0.6839E+00.
O.6S39E+OO
0.6364£^00
0.&863E+00
6. 72OOE+00
0.11B3E+04
O.1OS4E+04
O.8377E+O3
O.4492E+03
0.346SE+03
0.8048E-
O.7312E+00
0.7378E+OO
0.7007E+00
0.72O6E+00'
-O/7OA8E+OO
O.3822E+O3
•O.A369E+03
0.11OOE+O4
0.1B12E+O4
TOMS63E+04
O. 1223E+OS
0.1340E+OS
0.3I96E+O4
0.1279E+03
0.19&9E+03
0.7480E+04
0.4789E+03
0.2464E+03
0.1420E+03
0.3937Et01
0.18SSE+04
0.230tE+pp
0.2646E+O3
0. 1740E+03
0.2113E+O5
0.1643E+03
0.9235E+04
0.4001E+04,
O.2213E+03
0.2396E+03
O.3903E+O4
0.1681E+03
O.19&9E+OS
0.1243E+03
-0.9570E+O4
-0.1O27E+03
O.4221E+04
0.9928E«-04
0.1147E+OS
0.10&OE+03
0.1112E+05
-0.4927E+04
-O.
-0.
-O. U78E+03
-0.1182E+05
-O.10&3E+OS
-0.1OO9E+O3
0.4135E+04
O.9597E+04
O.11O9E+03
^.0. i054E+03
".O.1136E+OS
-O.4874E+04
-0.1130E+03
-O.1O95E+OS
-OillO3E+05
0.19O8E+03
-0.6S47E+03
O.S957E*-04
0.7750E+04
-0.9139E+03
-C. 1OSSE-^O5
-0.1S46E+OS
-O.2701E+O4
-O.1374E+04
O.1993E+04
-O.200SE+04
-0.1848E+OS
-O.8142E+04
O.3863E+O4
O.31B3E+04
0.7446E+04
0.9932E+04.
O.144SE+04
-O.43B3E+04
-O.SO39E+04
-0.4112E+02
-0.2198E+O4
E 0.263eE-HiS 0.2732E-0& 0.4223E*O5 O.1306E+O4 0.1423E+O3 0.173BE+O2 O.2311E-rOS O.3aQ2E+C6 -0.4716S+OS -0.1949E-M3S
ute: Units eq/month.
Input Is Atmospheric compartment to Lake,
Input 2: Snow compartment to Lake; "•''" '.
Input 3: Soil campartmant to Lake,
Input. 4: Unaaturated zona comportment to Lake,
Input 3> Overland *low to Lake,
Input 6: Groundwatitr compartment to Lake,
Out It Lake compartment bo Qroundnater,
Input 7; Dry depoaition to the Lakit,
Out 2: Outflow from tha Lake,
Reaction: Internal production in the Laker compartment,
Storigo; Changa In Ldlcu utcracjd.
-------
TABLE 7. MSB Evaluation for the
Calibration Period.
16
Variable Units
Discharge m3/s
Cum. Discharge m/yr
Chloride ueq/L
Alkalinity jieq/L
Sulfate veq/L
MSB
RMSE
730
730
77
77
77
0.00225
0.00133
3.t7
319.9
129.75
0.0175
0.036U
1.86
17.88
11.1
-------
TABLE 8. MSB Evaluation for the
Verification Period.
17
Variable
Discharge
Cura. Discharge
Chloride
Alkalinity
Sulfate
Units
m3/5
ra/yr
yeq/L
yeq/L
yeq/L
MSB
RMSE
365
365
45
45
^5
0.00468
0.0144
3.57
216.25
111.13
0.0684
0.1202
1.89
14.70
10.54
-------
18
TABLE 9. Sensitivity Analysis of Hydrological
Parameters for Outflow
fMSENEW "
» % MSE change - l Sg= ^ * 10°
MSEopt
PERCENT MSE OF OUTFLOW CHANGE*
Parameter x + lOJtx x -
BETA -0.466. -6.335
KAPPA 2.236 -1.180
KPAN2 6.584 -t.863
KPAN3 0.435 -0.497
KPAN5 0.745 -0.839
KLAT3 -2.050 -1.739
KLAT4 -0.186 0.124
KPERC3 0.683 -0.745
KPERC4 0.124 -0.186
D1 0.093 -0.124
FRAX 0.217 0.062
ALF1 0.000 0.000
ALF2 -0.093 0.062
-------
19
TABLE 10. Sensitivity Analysis of Chemical Parameters
for Alkalinity,
PERCENT MSB OF ALKALINITY CHANGE*
PARAMETER x + 10*x x - 10$x
CF 0.310 J0.207
RE3 -0.035 0.000
KHM -0.070 0.070
KH5 2.178 -1.522
KH6 0.000 0.000
KOM 0.000 0.000
K05 1.660 -0.830
K06 0.000 0.000
KP3 0.207 -0.207
KP4 0.000 0.000
K -1.660 2.141
KP6 0.000 0.000
MSB - MSE .
r new opt)
* JMSE Change - k rrr= ' * 100
MSEopt
-------
Acid
Depositior
Input
Allcalinit
Concentr.
(ANC)
10 .
100 • 1000
A. Input
10000
B. Not Sensitive
(no response)
Concentr
(AflC)
C. Moderately
Sensitive
D. Extremely
Sensitive
10 100 1000 10000
Tine in years ——->
Figure 2. Direct and delayed responces of surface waters to a step-
function increase in acid deposition (showing watershed
sensitivity classes).
-------
Ability to Retain
——————-*- Low
8 High
sl !
*S
> <=
^. o '
°- ~ 1
3.;
° v t
*" ^» — '
— .2 « i
£ c S 1
Low
J Capacity "Protected" /
| (no response /
| over centuriesl /
\ /
I/
/
\
' • X
!"""
i
i
i
i
-
t Quick Response
J (rate limited)
1
1
1 l
Slower Response
(may become rate
limited [quick
response] in
years or decades)
FIGURE 1. A qualitative presentation of the effect of two
major soil properties—the ability to retain sulfate and
the ability to supply base cations—on the rate at which
streams and lakes respond to changes in acid deposition.
(NRC Panel on Lake Acidification, 1984)
-------
20
Figure 3. Precipitation data for the calibration period.
Figure 4. Evaporation data for the calibration period.
Figure 5. Temperature data for the calibration period.
Figure 6. Precipitation data for the verification period.
Figure 7. Evaporation data for the verification period.
Figure 8. Temperature data for the verification period.
-------
LAKE WOODS
PERIOD : 9/78 - 8/80
100 200
300 400 800
DAYS
000 700
•00
-------
LAKE WOODS
PERIOD : 9/78 - 8/80
100
700 100
-------
LAKE WOODS
PERIOD : 9/78 - 8/80
CO
111
111
1C
a
Ul
a
*
u
a
D
OC
Ul
a
ui
•10
.20
-30
700
•00
-------
LAKE WOODS
PERIOD : 9/80 - 8/81
350 400
-------
LAKE WOODS
PERIOD : 9/80 - 8/81
300 400
-------
LAKE WOODS
PERIOD : 9/80 - 8/81
400
-------
21
Figure 9. Three years data of wet and dry sulfate loading.
Figure 10. Three years data of wet and dry acidity loading.
-------
LAKE WOODS
TIME SERIES OF SULFATE DEPOSITION INPUT
SIMULATION PERIOD: 9/1/78 - 8/31/81
Legend
• WBT DEPOSITION
O DRV DEPOSITION
100
200
100
400
§00 000
DAYS
TOO
•00
900 1000 1100
-------
LAKE WOODS
TIME SERIES OF ACID DEPOSITION
SIMULATION PERIOD: 9/1/78 - 8/31/81
Legend
*»T OBPO8ITIOH
DEPOSITION
100
200
300
400
SOO 600
DAYS
TOO
800
000
1000 1100
-------
22
Figure 11. Calibration results of cumulative outflow.
Figure 12. Calibration results of daily outflow.
Figure 13. Results of daily outflow error.
Figure 1H. Calibration results of lake chloride.
Figure 15. Calibration results of lake sulfate.
Figure 16. Calibration results of lake alkalinity.
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/78 - 8/80
2.S
o
u.
til
1.5
I <
o
0.5
100
200
300 400 800
DAYS
000
700
Legend
D FIELD DATA
•00
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/78 - 8/80
40
o
u.
30
25
20
15
10
100
200
aoo
400
DAYS
BOO
Legend
• fJMCAAT.IOH.OATA
D FIELD DATA
000
700
•00
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/78 - 8/80
40
S
Q
20
10
g .
flC
ff
III
>. -10
<
a
-20
-so
-40
100
200
900
400
DAYS
800
600
700
SOO
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD CONCENTRATIONS
PERIOD : 9/78 - 8/80
100
Legend
• IMULATIOH BATA
HELD DATA
700
100
-------
LAKE WOODS
COMPARISON OF SIMULATION AND FIELD SULFATE
CALIBRATION PERIOD: 9/1/78 - 8/31/80
300
Legend
• «IMUtATIQN OATA
Q MELD DATA
700
•00
-------
LAKE WOODS
COMPARISON OF SIMULATION AND FIELD ALKALINITY
CALIBRATION PERIOD: 9/1/78 - 8/31/80
200
•100
TOO
Legend
PAT*
O FIELD DATA
• 00
-------
23
Figure 17. Verification results of cumulative outflow.
Figure 18. Verification results of daily outflow.
Figure 19. Verification results of daily outflow error.
Figure 20. Verification results of lake chloride.
Figure 21. Verification results of lake sulfate.
Figure 22. Verification results of lake alkalinity.
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/80 - 8/81
2.5
o
1.S
01
1
2 1
0.5
80 100 110 200 250 300
DAYS
350
Legend
Q FIELD DATA
400
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/80 - 8/81
Legend
PAT*
380
400
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/80 - 8/81
30-
20-
10-
•20-
•90-
H,Atr-«
W "
so
100 150
200
DAYS
280
900 950
400
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD CONCENTRATIONS
PERIOD : 9/80 - 8/81
100
to
3
*
z
o
tc
fr-
ill
o
z
o
o
III
Q
-------
LAKE WOODS
COMPARISON OF SIMULATION AND FIELD SULFATE
VERIFICATION PERIOD; 9/1/80 - 8/31/81
300
Legend
»IIIUt»TIOM DATA
FIELD DATA
400
-------
LAKE WOODS
COMPARISON OF SIMULATION AND FIELD ALKALINITY
VERIFICATION PERIOD: 9/1/80 - 8/31/81
200
•100
Legend
PAT*
O FIELD OAT*
380
400
-------
Figure 23. Lake sulfate projection under various loading scenarios.
Figure 2U. Lake alkalinity projection under various loading scenarios.
-------
LAKE WOODS, NEW YORK STATE
PROJECTION OF LAKE SULFATE UNDER VARIOUS LOADING SCENARIOS
400
Legend
Mi$EMT LOAOIMQ
LOAOINO
1870
iaao
1990
2000
2010
2020
2010
-------
LAKE WOODS, NEW YORK STATE
PROJECTION OF LAKE ALKALINITY UNDER VARIOUS LOADING SCENARIOS
200
Legend
• POUBIB LOADING
D >«e»eMT LOADIHQ
-200-
• HAL* LOADINQ
1978 19tO 1«a5 1990 199S 2000 200S 2010 2018 2020 202S 2030
-------
Technical Report No. CEE-ARRG-86.0M
Modeling Short and Long Term Impacts
of Acid Precipitation Using the
Enhanced Trickle-Down Model:
Lake Panther Case Study
by:
Nikolaos P. Nlkolaldls
Jerald L. Schnoor
Konatantlne P. Georgakakoa
Civil and Environmental Engineering
The University of Iowa
Iowa City, Iowa 52242
First Draft: December 1986
-------
MODELING SHORT AMD LONG TERM IMPACTS OF ACID PRECIPITATION USING THE ENHANCED
TRICKLE-DOWN MODEL: LAKE PANTHER CASE STUDY
INTRODUCTION
The Enhanced Trickle-Down (ETD) model (Nikolaidis et al., !986a) is being
used to assess the impacts of acid precipitation of Lake Panther's
watershed. Lake Panther is a 1.24 km2 forested watershed located in the
Adirondack Park region of New York State. Lake Panther is considered neutral,
with a typical outlet pH of 7. The lake receives an annual precipitation of
1.16 m and the mean annual air temperature is 5° C. The vegetation of the
watershed is dominated by Sugar Maple, Beech, Yellow Birch and Red Spruce.
The watershed is underlain by hornblende granitic gneiss bedrock. The depth
to bedrock is 24.5 m. The basin is covered by thick glacial till. The soils
are predominantly spodosols that have been developed on the till and average
less than 1 m in depth.
Lake Panther is located 10 km south of Lake Woods (Nikolaidis et al.,
I986b). Both watersheds receive the same quality loading, and on the average
both have similar vegetation, soils and bedrock type (with the exception of
the depth of the till - Panther has 10 times deeper till than Woods), but they
have different responses. This was the topic of research that was undertaken
by the Integrated Lake-Watershed Acidification Study (ILWAS). The ILWAS
project investigated why three lakes (Woods and Panther Lakes were included)
that receive similar input of acidity respond differently in neutralizing this
acidity. The thesis of the ILWAS project as it was presented by Peters et
-------
al., 1985, was that water moves through the acidic upper soil horizons rather
than the lower ones and that's why it is more acidic.
The scope of this research is to calibrate, verify, and perform a
sensitivity analysis on the parameters and examine the long term response to
various acid loadings of Lake Panther, using the ETD model.
MODEL CALIBRATION
PROCEDURE
The calibration of ETO for Lake Panther was obtained by decoupling the
hydrologic, sulfate and alkalinity submodels, similar to the calibration of
Lake Woods (Nikolaidis et al., 1986b). However, in this case it was not
necessary to use the standardized optimization package, IDESIGN (Arora et al.,
1985), since adequate initial estimates for evaporation and snowmelt had
already been established after the calibration of Lake Woods. Keeping the
same guideline for establishing the optimum value of discharge, by trial and
error the lateral and vertical percolation flow parameters were adjusted.
These adjustments helped in capturing the seasonal variability as well as the
peaks and valleys of the daily discharge. In order to obtain closure of the
cumulative flow during the calibration period, the evaporation and snowmelt
parameters were adjusted as well. Continuous refinements of the hydrologic
parameters were applied until agreement between the simulation and field data
was achieved. Trial and error adjustments to sulfate and alkalinity
parameters were made starting with the initial values of Lake Woods in a
similar manner.
-------
IHPUT DATA
The time series data of surface precipitation, evaporation and
temperature for the calibration and verification periods are presented in
Figures 1, 2 and 3. Precipitation and temperature data were collected at the
LMI station of the ILWAS project. Evaporation measurements were obtained by
using van Bavel's combination method (Nikolaidis, I986c).
Figures 4 and 5 present the wet and dry daily sulfate and acidity loading
time series. An analysis of wet and dry deposition is presented in Table 7.
Approximately 37% of the total acidity is dry deposition and 47$ of the total
sulfate loading is dry sulfate. The total acidity loading is 1205 eg/ha/yr
and the total sulfate loading is 1313 eg/ha/yr.
RESULTS
Table 2 presents a partial list of the hydrologlc and chemical watershed
descriptors that were input to the model. The optimum values of the
calibrated parameters are presented in Table 3* Comparison between lake
discharge, chloride, sulfate and alkalinity simulations and field data are
presented in Figures 6 through 11. Figures 12 through 15 contain the percent
saturation, alkalinity, pH and sulfate plots of the terrestrial compartments.
Table 4 contains a complete hydrologic budget for Lake Panther. Direct
precipitation accounts for 1H.5 > of the total inflow to the lake. The
majority of the water is coming through the unsaturated zone (40.5JO.
Finally, 23$ is due to snowmelt and 22$ is from the soil compartment.
An alkalinity budget for Lake Panther is presented in Table 5. The
majority of the acidity input to the lake is coming through soil and
-------
unsaturated zone, with 23 and 12? respectively. Wet deposition contributes
135K and dry 9.5 % of the total acidity input. According to the simulation
results, 58$ of the neutralization of the total acidity input to the watershed
occurs in the terrestrial compartments and the rest is due to the in-lake
processes (weathering and sulfate reduction).
A sulfate budget is shown in Table 6. The majority of sulfate input to
the lake is coming through the soil and unsaturated zone with 25 and 48.2$
respectively. Wet and dry deposition contribute 9$ each. The net sulfate
reduction in the lake sediments is 9$ of the total sulfate deposition. About
25$ of the total sulfate deposition is sorbed or stored in the terrestrial
compartments.
Table 7 presents the MSB evaluation for the calibration period for
discharge, cumulative discharge, chloride, alkalinity and sulfate.
-------
VERIFICATION
The verification simulations of Lake Panther were performed by using one
more year of input field data time series not included in the calibration
period. The simulation results between lake outflow, chloride, sulfate and
alkalinity for the verification period (9/80-8/81) and field data are
presented in Figures 12 through 17. The results are comparable with the
calibration ones. Table 8 presents the MSE evaluation for the verification
period.
SENSITIVITY ANALYSIS
The sensitivity of lake outflow to each of the hyrdologic parameters was
checked by varying the parameters 10% above and 10J below their optimum
values. The percent change of the mean square error (MSE) of the outflow is
presented in Table 9. The most sensitive parameters were: the snow parameter
BETA, and the soil lateral and vertical permeability correction coefficients.
The sensitivity of outflow alkalinity to each of the chemical parameters
was similarly checked. Table 10 contains the percent change in MSE. The most
sensitive parameters were: the lake compartment weathering constants KH5 and
K05, the lake sulfate reduction reaction rate K, and the unsaturated zone
ligand attack rate, KOI.
LONG TERM SIMULATION
In order to perform long term simulations, the existing 3 year record was
input repeatedly for 17 times, which extended the simulation period to 51
-------
years. Three simulation runs were made: one for present loading, one for
half loading and one for double loading. In order to establish the half and
double loadings, both wet and dry sulfate and alkalinity inputs were halved or
doubled respectively. Figure 18 and 19 show the projection of lake alkalinity
and sulfate for the next 50 years under three different loading scenarios.
The prediction indicates that Lake Panther is a direct response system and
that it would only take a few years for the lake to respond to a decrease in
acid deposition loading. The lake currently has reached steady state with a
mean alkalinity of about 170 ueq/L and a mean sulfate of HlO yeq/L. If the
loading were to be halved, the mean alkalinity would be 210 yeq/L and the mean
sulfate concentration 75 ueq/L. On the other hand, if the loading were to be
doubled, the mean alkalinity would be 60 ueq/L and the mean sulfate
concentration 310 ueq/L.
DISCUSSION
An issue that has to be addressed is the importance of the in-lake
processes with respect to the neutralization of acidic deposition. One could
raise the question why the selection of reducing sulfate in the lake sediments
instead of adsorpting it in the terrestrial compartments was made, or, by the
same token, why such an amount of alkalinity had to be neutralized in the lake
instead of in the soil and unsaturated zone.
In order to answer these questions, the calibration exercise of the
terrestrial compartments has to be validated. Figures 20 through 23 present
the simulation results of percent saturation, alkalinity, pH and sulfate of
the terrestrial compartments. Peters et al., 1985, evaluated the sulfate and
alkalinity concentration for summer and winter periods using tension lysimeter
-------
for the soil compartment. They found that sulfate concentrations range
between 1*12 to 180 yeq/L, (the low values occurred during the winter periods)
and alkalinity from -77 to -90 peq/L. The simulation results give a mean
sulfate concentration of 130 yeq/L and a mean alkalinity of -60 peg/L.
Comparing the simulated and observed values, one could notice that the
simulated mean sulfate concentration is less than the observed, which
indicates that adequate sulfate is being adsorbed in the calibration
exercise. Similarly, simulated mean alkalinity is higher than the measured
values, which indicates that sufficient weathering has been applied in the
calibration of soil chemistry.
Since the calibration of the terrestrial compartments is in accordance
with the limited field data, it is obvious that the validity of the importance
of the in-lake processes can be Justified.
-------
REFERENCES
Arora, J. S., Thanedar, P. B. and Tseng, C. H. (1985). Users manual for
program IDESIGN, version 3.4 for PRIME computers. Optimal Design Laboratory,
College of Engineering, University of Iowa, Technical Report No. ODL 85.10.
Cronan, C. S. (1985). Biochemical Influence of vegetation and soils in the
ILWAS watersheds. Water, Air and Soil Pollution, Nol. 26, No. 4.
Nikolaidis, N. P., Rajaram, H., Schnoor, J. L. and Georgakakos, K. P.
(1986). Enhanced Trickle-Down model description. Civil and Environmental
Engineering, University of Iowa, Technical Report No. CEE-ARRG-86.01.
Nikolaidis, N. P., Georgakakos, K. P., and Schnoor, J. L., (1986b). Modeling
short and long term impacts of acid precipitation using the Enhanced Trickle-
Down model: Lake Woods case study. Civil and Environmental Engineering,
University of Iowa, Technical Report No. CEE-ARRG-86.03.
Nikolaidis, N. P. (1986c). Estimation of dally potential evaporation. Civil
and Environmental Engineering, University of Iowa, Technical Report No. CEE-
ARRG-86.02.
Peters, N. E., and Murdoch, P. S. (1985). Hydrologic comparison of an acidic-
lake basin with a neutral-lake basin in the west-central Adirondack Mountains,
New York. Water, Air and Soil Pollution, Vol. 26, No. 4.
-------
TABLE 1. LAKE PANTHER
ANALYSIS OF WET AND DRY DEPOSITION
ACIDITY (eg/ha-yr) SULFATE (eq/ha-yr)
Period: Wet Dry Wet Dry
9/78-8/79 807.4 446.0 791.4 666.3
9/79-8/80 702.6 476.3 - 697.6 648.7
9/80-8/81 746.0 437.7
Average 752.0 453.3
Total average Acidity » 1205.3 eq/ha »yr
Dry Deposition - 37.6? total Deposition
Total average Sulfate - 1343.3 eg/ha »yr
Dry Deposition - 47.0$ total Deposition
Sulfate Loading (Average)
Wet - 36.1 kg/ha/yr
Dry - 30.3 kg/ha/yr
Total - 66.4 kg/ha/yr
On the average, 8.25 metric tons of sulfate are deposited on Lake Panther
watershed (or 2.7 metric tons of sulfur)
-------
10.
TABLE 2. LIST OF WATERSHED DESCRIPTORS USED FOR MODEL CALIBRATION,
GENERAL WATERSHED CHARACTERISTICS:
AQUATIC AREA •> 0.1800E+06 SQUARE MET2RS
TERRESTRIAL AREA - 0.1060E+07 SQUARE METERS
CHARACTERISTIC DISTANCE - 0.65003+03 METERS
DEPTH TO BEDROCK - 24.5000 METERS
PARTIAL PRESSURE OF ATM C02 - 0.0003 ATMOSPHERES
SURFACE WATER PC02 IS 1.50 TIMES SATURATED PC02
SOIL COMPARTMENT CHARACTERISTICS:
POROSITY - 0.2700
DEPTH OF SOIL LAYER - 0.6300 METERS
SUM OF BASES - 113.1000 EQUIVALENTS/KILOGRAM
SOIL DENSITY - 1283.0000 KILOGRAMS/CU. METER
UNSATURATED ZONE COMPARTMENT CHARACTERISTICS:
POROSITY - 0.2000
TRANSPIRATION COEFFICIENT - 0.0010 INCHES/DAY
BARE-GROUND FROST COEFFICIENT - 0.1000
REDUCTION IN FROST COEFFICIENT - 0.0800
DAILY THAW RATE - 0.1200 DEGREES C
INITIAL FROST INDEX - 0.0000 DEGREES C
LIMITING FROST INDEX - -3.0000 DEGREES C
THAW COEFFICIENT - 0.2000
BULK DENSITY - 15^9.0000 KILOGRAMS/CU.METER
SURFACE WATER BODIES CHARACTERISTICS:
STREAM BED ELEVATION AT OUTFLOW - 23.5530 METERS
GROUNDWATER COMPARTMENT CHARACTERISTICS:
POROSITY - 0.2000
-------
11
TABLE 3. List of optimum values of the calibrated parameters.
a) Hydrologic Parameters
-Snow:
BETA = 0.8095
KAPPA - 1.1423
KPAN2 - 0.1766
-Evaporation:
KPAN3 - 1.5171
KPAN5 - 1.5381
-Lateral and Vertical Flows:
KLAT3 - 1.5281
KL.AT4 - 4.9791
KPERC3 - 4.3117E - 3
KPERC4 - 4.1741E - 3
b) Alkalinity Parameters
RE3 - 8.1 E-8 m3/eg/day
KH4 - 2.0E-2 raeq/nr/day
KH5 - 6.5E-1 meq/mVday
KH6 - l.OE-2 meq/raVday
K04 - 1.1E-3 meq/or/day
K05 - 3.9E-1 meq/mVday
K06 - 1.1E-3 raeq/m2/day
c)Sulfate Parameters
CF - 2.0247
KP3 - 4.3E-5 eq/kg/eq/m^
KP1 - 8.5E-6 eq/kg/eq/m3
K - 2.033E-3 a/day
KP6 « 4.7E-7 eq/bg/eg/m3
-Groundwater
D1 - 0.5963
FRAX - 0.4804
ALF1 » 6.2375E - 4
ALF2 • 3.4045E - 2
-------
TABLE 4. MONTHLY HYDROLQGIC BUDGET OF LAKE PANTHER
M-<^£ ^ *-••• *',••* •\'-'IJ' . '
INPUT 1 "
0
O
O
0
O
0
0
O
O
.*.
'/
0
0
O
0
0
0
O
0
)
O
0
Q
0
0
. 1906E-01
.164QE-01
. 6342E-02
.4129E-02
.•6638E-02
. 2434E-02
.ISISE-OI
.I23SE-M
. 1490E-01
* 1 Tt* t?_fc.".*5
• Vl Jr^t-^JjJ
. 1043H -01
.2633E-01
.2109E-C1
.134.*:E-01
. 1 3372-01
. 4645E-02
. 4335E-02
.OOOOE-OC
. 1043E-O1
. 1346E-O1
. 70O&E-O2
. 1744E-01
.2157E-M
.973SE-0-
**•*•*&•** i4'<&i
"INPUT 2 "**
""6TooooE-oo~
O.OOOOE-OO
6. 1346E-03
0.2312E-01
0. 3344E-02
0.223rE-?2
o.3c.i!?z-::
0. 199SE+00
G.2315E-C1
*•• r-.^i*-i*e-«..^'"i
*.- ^ t- J**VSl V J
O.OOCOE-00
O.OOCOEK>0
O.OOOCE+OO
O.C-OOOS-00
0. 1&02E-G1
0.2006E-01
0.2283E-01
0. 4742E-O2
O.67&OE-01
0.2071E-01
0. OOOOE-00
O.OOOOE-OO
O.OOOOi-OO
O.OOOOE-OO
S.'.*^ •'• ' :-
INPUT 3
0.
0.
0.
0.
0.
0.
C.
0.
0.
•
0.
0.
0.
A
•.'.
0.
0.
0.
0.
0.
0.
0.
0.
0.
c.
2336E-01
2149E-OI
1092E-01
8611E-02
9723E-02
4147E-02
2^73H»Oi
2773E < 1
2193E-01
C 7i^»*?— ^*^
C / C iSfc. *•— ,
H43E-01
S330E-01
34O1E-01
1914E-01
1953H-C*
t294E-Ot
992BE-02
1 129E-02
1531E-01
2367E-01
3761 £-02
2746E-01
23722-51
1497E-01
-• ' • ' ••-'*•••*&
INPUT 4
0. 2789E-01
0.2318E-01
"b."i&73E-01
0. 1347E-01
0.917BE-02
O.B193E-C2
0.2421E-C1
0.2922E-01
0.2T41E-0:
*" ' T * 5™ *.»^ *
'-• - -. J A 7^, "V *
0. S629K— 12
0.3327E-C1
O.B902E-01
O.A012E-01
0.4S6AE-01
0. 4A38E-01
0.3301E-01
O. 1442E-O1
O.163BE-01
0.3B91E-01
0.2EOOE-01
0.4B70E-C1
0.3774E-01
O.314SE-O1
-,'?«"*«•" •>*::.".•-•
~jr "
INPUT 3
0.2141E-03
.O.3043E-03
O. 1055E-0&
0. 1339E-07
0.9840E-07
0. 1604E-08
C.2M1E-05
0. ii20E-OS
0.902SE-C4
.-. fT;* « | "t^_. ,C
C.4790E-07
0. 1S6AE-04
0.7231E-OA
0. 1031E-OA
0.2341S-C3
0.6222E-07
O.7O36E-06
0.i319E-l&
0.3722E-03
O.7&97E-OA
0.2013E-0&
O.S275E-04
0.23C1E-04
0. 1327E-06
INPUT A
0.2314E-03
..0.2233E-03.
O.2033E-03
O.2087E-03
0. 1992E-03
O. 1739E-07
0. 1B90E-03
C.1753E-03
0.1S09S-03
! '^O*?'?— .1*1 T
' - i^*Tr» -t—V-A
C.1223E-03
0. 1240S-O3
0. 1172E-03
O. 1270E-03
0.1273E-03
O. 1319E-03
0.1341E-03
O.1199E-03
0. I354E-03
0. 1236E-03
0. 1273E-03
0.1222E-03
•I.12S4E-03
0. 1272S-03
-•• •-*e->w,s,j
OUT 1
O.iiolE-03
0..609SE-05
0.5S03E-03
0.3&46E-03
O.S390E-03
0. 4706E-03
0.3113E-03
0.4749E-03
O. 4O83E-03
OlfCt-' t 2T~.t^«t
* C w « 1 Z, t»'2
0.3314E-OS
0.3410E-0S
0.3172E-05
O.T433E-03
0. 3444E-05
0.3369E-03
O.3&2BE-05
0.3244E-03
0.3&&3E-03
O.339BE-03
0.3444E-05
0.3307E-0S
O. 3474E-03
O.344:E~OS
***&m&&&
OUT 2
O.2301E-01
0.2327E-O1 .
0.15&4E-01
0. OOOOE-00
O.OOOOE+00
0-.0000E-00
O.OOOOE-OO
0. 0000S-0O
0.2£37E-0i
0"f*j»rt^ff— A 1
* <&£**• .*SL -,*i
0.2663E-01
0.210SE-01
0. 1963E-01
0.1534E-C1
0.8245E-02
O.OOOOE-OO
O.OOOOE-OO
O.OOOOE-OO
0.0000E-00
7T'0.i'2A7E-Ot
0. 2963E-01
0.2984E-01
0.271BE-01
C-.2232E-01
H mmm*' ^»* <
#*Hr •:
'our 3
0.4406E-01
•O.4003E-01
0.2220E-01
0.4914E-01
0.2697E-01
0.232SE-01
O.S3SOE-01
0. 2A90E+00
0.9fc3«E-:i
O^^.^™U \ •
*^.V*^.'I» J*
C.3635E-02
O.3998E-O1
0.1307E+00
0.60-US-O1
0.8334E-01
O.894&E-01
0.7429E-01
0.273SE-01
O.BB21E-O1
O.1179E-OO
O.2819E-01
O.5373E-01
0.47&3H-C1
0-i384E-Oi
STORAGE
0.3B89E-02
0.6603E-04
-0.3S03E-02
0.3Q87E-03
0.410SE-O2
-0.A3S4E-02
O. lOlSE-'M
0.26iSE-;C
-O. l?rtrE-Oi
— <'• ^A 1 rtC ."» •
•*• — "Hvt— Unit: .« af water per month
(over the entire w*t3rai-»ecj-124 ha)
Input :. .',t-.c*.:: ..• ic comp-irtmen * tc Lake,
I.iput 2: CiTOH ^O'-ipartment to La :•:«?,
Input 5s Soil compartment to Lake,
Input 4t Un*flturaLed zone ccmp*rt!.Mint to Lake,
Input Si Ovurljnd flow to Lake,
Input 6: Grcur>dw.itcr compartment to L
Cut 1: L»i;if eo/r.par t.r.e:ii. to Si
Out 2: Evaporation from the Lake,
Out 3: Out-flow -from the Lake,
2 :_r;iyj: Change in LJKQ storage.
-------
i i
1
JO O O 6 O O C- O :•<> O O O O O O O O O O O O O '->
IUI Ul U UJ Ul UJ Ul U LJ HI bl Ul 111 Ul Ul Ul UJ UJ U Ul Ul Ul Ul III
Id* -0 Ol O
10> N O -< 0) N M -J I' CJ "• «• 01 * CD CD CS - « -0 in 1'J ni III
I -0 "0 M 'O M 0* —« JJ •""• -0 ID 0* N -0 H O tf> f •< O M S CD r» N
d d'd d d d o •"* •:> d d o d o d o o o d o o d d o
o o o o o o c- •-*<:• o y o o o o o o o o o o o o o
tut W Ul Ul W Ul i'l ''J !il UJ UJ Ul UJ III (tt W UJ W W W Ul UJ 111 Ul
I|K ^ «o PI *0 1^ C" I'1* T *** CH N Is 0 ^MOWQJ^4lh^N**
N 0* 03 ITJ M *0 tu '•'> ' • t i W W n £* 0* 41 VI tD Tl- f I') M M '•"
IS ft i* -« «4 »4 ..« f I M ri PJ N N "< *« »4 •* «"i »•« (S< M PI Cl M
ooooo'-oooooooooooo
•4 0 tfl^jtt«t"*(t-n)(nooooo9o9o9°9O9oo
+ f?+ .+. + * + + * f. + .+. + A + .+. + .+. + .+. + .*. * .*.
»• M O t-. -< M O» *a
-4 N W O «" •* P *4
UoooooOf/ooooo. oooooooooooo
lo'Sbo
iuii
0000000000000000000
,. 4.4.4.^4.
U UI Ul Ul III Ul |t| III W III 111 UJ III
nt o> at »«•««« ii t*. ~* n -<
n n
o> N n -o t-i ••> -n
o s -* w -« ri o m
n
n
n
r>
UJ
t*
UJ
r.
-o
r* ** •» o »< in -<
S w M n w « rt N
Ul Ul Ul Ul
(D -0 O 01
0> 0) O 0*
« fl M CM
oocooooo
o o o o o °^'9 0990000000
!??u?????ff??f?????°'??f?f?
I^OQOo^OOO'^OOOOOOOOOOoOOO
1H W W 'O Or -* M ^ t-< N M N JM O H ^ QJ W C *• -* •':•
Jl "J C* I .
N O tt) N -'1
to r i ri to -
88
S§
oo
ui iti iu ui uj ui w ui u u; uj UJ
§O«» S»OH^OOO'>
O vl -fl (>• If' ^ Vt O O O O
So i-) ^) o N •) t
O O O O (j
(i U/ Ul til li'l
•OO
OOOOOOOOOOOO
I I I I I I
».'•
-?-
r-»
Ki^
W-4
1? N
.M»fK)WOM«tK>'
?OOOOOOOO'_
+ 4- t 4- + * .+. + t
UlUIUUIUlUllUUUJUJuiUJ
£?
i ' : i i
i o O
I I
O O
I I I
O O O O O O O O
I I ( I I I I
18
1-4
18
jo
!o
1
10
1C
Stir
I?
I?
IUJ
g
s
a
n
n
\i
^
!?
IUI
10
I*
ItJ
IN
8
i=:
19
I?
In
>9
!u'j
IS
!fl
Id
o
41
(I
je
J 4J
o g
41 .-«i
zm
01*
*> z *
4* C «
» ^ n I
o
0
I
0
5
*
.£«
*• 0 O
L. M A« U
a*i 4> a
6 C C C
o * « o
u a e ii
•f •"
- U ». L. TI
r. -< « ^ a
v v. a. a *«
§» 6 z a
£ 0 0 L.
6 a If U 3
a a s -- *
u E 0 -« «
i! c a c
Ui
•• C <9
O L
M a
3 •)
Jt E t 3)
« */ (U ^T
_i i. 44
• -a a
x u r c £. o
O » O
••« t s ^«
L. «{
a -i
-
«»- n j- jj e
*• i. -. o -< c
-
5 -< r4 n «*
•• 3 3 i 15
f) o. tx a a
«; i; c c e
no «4 c 4. o X>«.MX:
o eo .« i. *" u
tQ ^J 3 M
•• ^^J «• Q c **
n m N o
-------
Ul
U
* in t ••< O •* S CD <• O-
i «* o p-1 • in w -fl N ^ • CD iO ** 0- •* rt *f "fl ^
idddddddddddddodddddddodo
i ^
IO O O O O O O O 6 O O O O O O O O O O O O O O O
I'jlUIUlUIUIIlJUIIlJUJUIUJUJUJUlUJUJUIUJUJUJUJUJUllll
I^I^ONCOMO-1 •OWOjNNC'IO-^l'-OOCa.iMO-O
IfoNOCIfrOOOMC-inNO'MOO'Ot^O'OCD-OlvM
t-* f o on i
l«r -o «•«-*<
id o do d o <
o o o o o o
(l I I I t I I
ro
oooooooo
IUI 111 U Ul
inRoK
100 m m o 4 6
in n no
UlUlUlUIUlUIUlUlUJUJ
0- —< »-< O O CD N CD *t -Q
to o o o o o o o ooooooooooooooo
I4 + + + 4- + + + + + + + + + + + + +- + + + + +- -
||||UlUllUUlUlU]lllUlUlUjlllUJUlUjUJlilUJ|iJUJUjUJUlUJ
11 IO O O O OO O O. O O O O O O O O O O O O.O O O O
toooooooooooooooooooooo
toooooooooooooooooooooo
144.444444444444. .(. 444 •(• + + +
o o
" • "• ^r «4 o M o
-< o- N M M « H ^J — n in -> cb N «j CD rl -o n o- -o n rt n
d d d d d d d d dddddddddddddddd
^.
_SSWKSy.H{SKgg!S!SKS«Na
!&5?S528£S?S£§2SS8K2B
INMNNM-*'*'''1''1*-***''**'^"'*1''**"*
^^
»»
U1UI Ul
CD M -U
O |v O
NriN
i o o o o o o o o o o o o o o o o o o o o o o
-.fq-o-nS*i (b 01 r> o- ^ 11)
o o o
goiS
•0 M K) -0
o nnw
M co ca -<
IOOOOOOOOOOOOOOOOOOOOOOOO
ooooooooooooooooO
.+. * f. + A + + + .+ + A > A + A + A
H I
|^g**»»i35*?»
iiL uiui
IC-I »j M
(COM h-
If'I -0 0>
1-0 t x
UJlillllUlUlli)UIUJUIlijllJUlUJlillllllJUJ
n I . 03 fl CO N O -• - M N r< -fl J3 M JJ bl
8nSS!?o:PrH~1fl-!010 " "
-• m f •! to - i >
n>MnNOni*.tvO-«*
• WN• -0 f.
5K8R
o> in n n
OOoOOOOOoOoOOOOOoOOOOOOO
?o ^ •
00-
lOOOOOOOOOOOOOOOOO
lUlUlUlUlUIHIUItilUIUJUIUIUlUlUlUlUl
io«-«wtvNnO"roooo~ — ~
I- . IO O 4| -4J -O (0 f f t-t O O O O
Q. IOOM-*(-|-«»«NHO.
M id oddddddd1'^ dddododddd
OOO
+ + +
m ui
0000
o o o o •> o o o o o o o o ii o o o 6 o o o o o o
ibl Ul Ul Ul III Ul HI f O
-< tl N -* K O N f» O •*) * 0* ID - ~* n »* 0" to •< K ••• 'i o- N co D n» M •« o M 01 •« •< M •<
d -o d o d o d o o d o d d «•>•'• J o o o d d d o
IB!
I'J
10
I VI
181
l-t
If
In
10
!«• * » *• * *onn » « f in in « n n * n n «r * » n \-o
loooooooooooooooooooooooo 10
II.• lit •.! ••! t*l t.l •.• 1*1 Ikl III I.I III I.I III 111 III I.I III III III lit III III tfl Mil
!"•
100 o ooo o.p o oo o o o o o o o o o.o o o o 10
is
ii
^^ !KJ
•4 *-• IN
Id
u
i?
In
'2
II
f«
M
i-
s.
ei
•M ^
C «
fJ
O
o
je
«
_1
Q
*J
a
a
u
lj-'i-
*• o o u
IS
u e
•ut
£ -4 IB
*fc*
O C 0
6 a u
V W
SO X
e o
JJ c
« «en
TT^ 'T
5-
5
."3
**.-
s3
4J g, o
gg"
1^1
">(!*•
0
JC •
5fe
i<
01-
Ji S *
rj A
MCI
IS a -
41 o j: in
io'"S
a.-" n J
o *• ** a
U C C £
« u *>
^ *>'?i G
u \. - O
« 1 J \.
S a -^ •••
•0 6 -x _
•a
•a
K •
•»4 3t
T
. c It
So
4* *»
u w
3«
£-5
M C
° O
«n 3 o io jc v.
4 Q
t tr> -o
C <•
»<£
U
.«
C •*
a u
*< u*
j-**j«'*j*i«j 3 U^
u a a. a a a. a K u. ••-• -n o
tr»i»»-*-*-/--i — -iii.*_i
333333
•
Q M
-------
12
TABLE 7. MSB Evaluation for the
Calibration Period.
Variable Units n MSB RMSE
Discharge m3/s 730 0.0012 0.031*
Cum. Discharge m/yr 730 0.003 0.056
Chloride jieq/L .102 26.3 5.13
Alkalinity jieq/L 102 6828.0 82.6
Sulfate yeq/L 102 128 11.3
-------
fABLE 8. MSB Evaluation for the
Aerification Period.
Variable Units n "SE RMSE „
Discharge m3/s 365 0.001H 0.038
Cun. Discharge m/yr 365 0.0026 0.052
Chloride |ieq/L 52 5.85 2.J2
Alkalinity yeq/L 52 4619.2 70.0
Sulfate weq/L 52 136.2 11.7
13
-------
TABLE 9. Sensitivity Analysis of Hydrological
Parameters for Outflow
PERCENT MSE OF OUTFLOW CHANGE*
x - 10* x
-1.7
11.3
0.0
0.3
0.0
-0.5
-0.4
-7.1
-7.5
0.0
0.0
0.0
0.0
100
PARAMETER
BETA
KAPPA
KPAN2
KPAN3
XPAN5
KLAT3
KLATU
KPERC3
KPERC4
D1
FRAX
ALF1
ALF2
MSE
x + 10JS x '
-0.8
13.5
-0.2
-O.U
-0.1
-10.2
-10.1
-0.3
-0.1
0.0
0.0
0.0
0.0
" ""opt) .
__ A.
-------
15
TABLE 10. Sensitivity Analysis of Chemical
Parameters for Alkalinity.
PERCENT MSB OF ALKALINITY CHANGE*
x - 10* x
0.9
-0.2
0.0
-1.6
0.0
12.8
-4.9
0.0
0.2
0.1
1.0
0.0
100
PARAMETER
CF
RE3
KH4
KH5
KH6
K04
K05
K06
KP3
KP4
K
KP6
MSE
*
-------
16
Figure 1. Precipitation data.
Figure 2. Evaporation data.
Figure 3. Temperature data.
Figure 4. Wet and dry sulfate loading.
Figure 5. Wet and dry acidity loading.
-------
LAKE PANTHER
PERIOD : 9/78 - 8/81
100 200
aoo
400 BOO 000 700
DAYS
800 800 1000 1100
-------
LAKE PANTHER
PERIOD : 9/78 - 8/81
100 200 300 400 100 800 700 800 800 1000 1100
-------
LAKE PANTHER
PERIOD : 9/78 - 8/81
-so
100
200 100
400
800 000
DAYS
700 100 800 1000 1100
-------
LAKE PANTHER
TIME SERIES OF SULFATE DEPOSITION
SIMULATION PERIOD: 9/1/78 - 8/31/81
Legend
• WET DEPOSITION
D DRY DEPOSITION
100
200
SOO
400
800 eoo
DAYS
700
too
800 1000 1100
-------
LAKE PANTHER
TIME SERIES OF ACID DEPOSITION
SIMULATION PERIOD: 9/1/78 - 8/31/81
•o
a
•C
"•»,
er
o
E
H
Q
O
U.
O
z
o
<0
o
D.
ttl
O
Legend
ID WET DEPOSITION
Q DRY DEPOSITION
100
200
• 00
400
soo eoo
DAYS
700
800
900 1000 1100
-------
17
Figure 6. Calibration results of cumulative outflow.
Figure 7. Calibration results of daily outflow.
Figure 8. Calibration results of daily outflow error.
Figure 9. Calibration results of lake chloride.
•Figure 10. Calibration results of lake sulfate.
Figure 11. Calibration results of lake alkalinity.
-------
LAKE PANTHER
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/78 - 8/80
2.5
UJ
1.5
a
o
0.5
Legend
• SJMULAT.ION.OATA
Q FIELD DATA
100
200
300
400
DAYS
soo
600
700
•00
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/78 - 8/80
100
Legend
D f tetp OAT*
700
100
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/78 - 8/80
40
20
10
cc
O »
cc
cc
111
>: -10
-20
•30
-40
100 200 300
400
DAYS
500 BOO 700 SOO
-------
LAKE PANTHER
COMPARISON OF SIMULATION AND FIELD CHLORIDE
CALIBRATION PERIOD: 9/1/78 - 8/31/80
200
Legend
OAT»
Q FIELD DATA
100
-------
LAKE PANTHER
COMPARISON OF SIMULATION AND FIELD SULFATE
CALIBRATION PERIOD: 9/1/78 - 8/31/80
300
Legend
|B SIMULATION DATA
O FIELD DATA
700
• 00
-------
LAKE PANTHER
COMPARISON OF SIMULATION AND FIELD ALKALINITY
CALIBRATION PERIOD: 9/1/78 - 8/31/80
300
-100
700
tOO
Legend
SIMULATION DATA
FIELD DATA
-------
18
Figure 12. Verification results of cumulative outflow.
Figure 13. Verification results of daily outflow.
Figure 11. Verification results of daily outflow error.
Figure 15. Verification results of lake chloride.
Figure 16. Verification results of lake sulfate.
Figure 17. Verification reauls of lake alkalinity.
-------
LAKE PANTHER
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/80 - 8/81
2.5
in
1.5
S
| '
o
0.5
Legend
D PI CUP DATA
80 100 150
200
DAYS
2SO 300 350 400
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/80 - 8/81
Legend
PIELO DATA
350
400
-------
LAKE WOODS
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
PERIOD : 9/80 - 8/81
400
-------
LAKE PANTHER
COMPARISON OF SIMULATION AND FIELD CHLORIDE
CALIBRATION PERIOD: 9/1/80 - 8/31/81
200
ISO-
50-
O
Legend
D tlliULATION DATA
Q HELD DATA
100
150
200
DAYS
280
300
850
400
-------
LAKE PANTHER
COMPARISON OF SIMULATION AND FIELD SULFATE
VERIFICATION PERIOD: 9/1/80 - 8/31/81
300
D
100
180
200
DAYS
250
300
Legend
DATA
FIELD DATA
8*0
400
-------
LAKE PANTHER
COMPARISON OF SIMULATION AND FIELD ALKALINITY
VERIFICATION PERIOD: 9/1/80 - 8/31/81
300
•100
350
Legend
• IIMUIATIOH DATA
D HELD DATA
400
-------
19
Figure 18. Lake sulfate projection under various loading scenarios.
Figure 19. Lake alkalinity projection under various loading scenarios.
-------
LAKE PANTHER, NEW YORK STATE
PROJECTION OF LAKE SULFATE UNDER VARIOUS LOADING SCENARIOS
400
350
800-
Z
2 250
Z 200
O
Z
O
O 160
-I 1001
D
(0
50-
Legend
LOAOINQ
P«e»BNT LOADINQ
HALF
1070
1910
1990
2000
2010
2020
2030
-------
LAKE PANTHER, NEW YORK STATE
PROJECTION OF LAKE ALKALINITY UNDER VARIOUS LOADING SCENARIOS
400
5- 300
9
z
o
S
oc
200-
1U
u
z
o
O 100
Z
J
<
3 o
-100
Legend
• DOUBLE LOADING
Q PRESENTLO*DINO
• MAIP LOADINO
1B70
1880
1000
2000
2010
2020
2030
-------
20
Figure 20. Percent moisture saturation of terrestrial compartments.
Figure 21, Alkalinity concentration of terrestrial compartments.
Figure 22. PH of terrestrial compartments.
Figure 23. Sulfate concentration of terrestrial compartments.
-------
LAKE PANTHER
CALIBRATION PERIOD: 9/1/78 - 8/31/80
100
Legend
• son
D UNSATURATED
• QROUNDWATER
100
700
•00
-------
LAKE PANTHER
CALIBRATION PERIOD: 9/1/78 - 8/31/80
300
Legend
SOIL
Q UNSATURATED
• QROUNDWATER
•00
-------
LAKE PANTHER
CALIBRATION PERIOD: 9/1/78 - 8/31/80
7.5
Legend
SOIt
O UNSATURATED
• QROUNOWATER
100
700 100
-------
LAKE PANTHER
CALIBRATION PERIOD: 9/1/78 - 8/31/80
350
Legend
• SOIL
D UNSATURATED
• QBOUNDWATER
700
•00
-------
Technical Report No. CEE-ARRG-86.05
Modeling Short and Long Term Impacts
of Acid Precipitation Using the
Enhanced Trickle-Down Model:
Clear Pond Case Study
by:
Nikolaos P. Nikolaidia
Jerald L. Schnoor
{Constant! ne P. Georgakakos
Civil and Environmental Engineering
The University of Iowa
Iowa City. Iowa 52212
First Draft: December 1986
-------
MODELING SHORT AMD LONG TERM IMPACTS OF ACID PRECIPITATION USING THE ENHANCED
TRICKLE-DOWN MODEL: CLEAR POND CASE STUDY
INTRODUCTION
The Enhanced Trickle-Down (ETD) model (Nikolaidis et al., 1986a) was used
to calibrate a two year record of Clear Pond watershed. Clear Pond is 5.21
km2 forested watershed located in the Adirondack Park region of New York
State. The pond is considered neutral, with a mean ANC of 100±19 yeg/L. It
receives on the average of 1 m of annual precipitation. The vegetation of the
watershed is dominated by Northern white Cedar, Paper and Yellow Birch, Red
Spruce, Sugar maple, Beech, and Balsam Fir. The watershed is underlain by
metanorthosite and anorthositic gneiss bedrock. The depth to bedrock is
6.3 m. The predominant soil series is Becket fine sandy loam. Clear Pond is
located about 60 miles northeast of Lake Woods and Panther.
MODEL CALIBRATION
PROCEDURE
The calibration of ETD for Lake Panther was obtained by decoupling the
hydrologic, sulfate and alkalinity submodels, similar to the calibration of
Lake Panther (Nikolaidia et al., 1986b). Keeping the same guideline for
establishing the optimum value of discharge, by trial and error the lateral
and vertical percolation flow parameters were adjusted. These adjustments
helped in capturing the seasonal variability as well as the peaks and valleys
of the daily discharge. In order to obtain closure of the cumulative flow
-------
during the calibration period, the evaporation and snowmelt parameters were
adjusted as well. Continuous refinements of the hydrologic parameters were
applied until aggrement between the simulation and field data was achieved.
Trial and error adjustments to sulfate and alkalinity parameters were made
starting with the initial values of Lake Panther in a similar manner.
IHPOT DATA
The time series data of surface precipitation, evaporation and
temperature for the calibration and verification periods are presented in
Figures 1, 2 and 3. Precipitation and temperature data were collected as part
of the RILWAS project. Evaporation measurements were obtained by using
van Bavel's combination method (Nikolaidis, I986c).
Figures 4 and 5 present the wet and dry daily sulfate and acidity loading
time series. An analysis of wet and dry deposition is presented in Table 1.
Approximately 29.5$ of the total acidity is dry deposition and 40? of the
total sulfate loading is dry sulfate. The total acidity loading is 849
eg/ha/yr and the total sulfate loading is 790.2 eg/ha/yr.
RESULTS
Table 2 presents a partial list of the hydrologic and chemical watershed
descriptors that were input to the model. The optimum values of the
calibrated parameters are presented in Table 3. Comparison between lake
discharge and field data are presented in figures 6, 7 and 8. Figure 6
contains the cumulative discharge of simulation and field data only for the
days of measured discharge. The missing data days were eliminated from
-------
the simulation results In order to construct this plot. Figure 8 shows the
daily error between simulation and field data.
Figures 9 through 11 present the chloride, sulfate and alkalinity
simulations and field data comparison. In order to obtain these simulations
the sulfate loading was increased 1.5 times, and the alkalinity and chloride
loading was increased 1.3 times the loading of the field data presented in
table 1 and Figures U and 5. The rationale of increasing the measured loading
of Clear Pond is partially presented in Figure 12. Figure 12 presents a
simulation of sulfate under actual measured data. In this simulation, there
is a definite trend of sulfate decline in the lake. When a calibration of the
terrestrial compartments was attempted, it was found that the sorption
partitioning coefficients had to be at least 2 orders of magnitude less than
the optimum coefficients of Woods and Panther lakes and still there was a
definite trend of decline. This prompted the fact that a definite increase in
loading is necessary.
Table 4 contains a complete hydrologic budget for Lake Panther. Direct
precipitation accounts for 11.2$ of the total inflow to the lake. The
majority of the water is coming through the unsaturated zone (36.0$).
Finally, 26.1$ is due to snowmelt and 26.7$ is from the soil compartment.
An alkalinity budget for Lake Panther is presented in Table 5. The
majority of the acidity input to the lake is coining through soil and
unsaturated zone, with 41.6 and 22.6$ respectively. Wet deposition
contributes 13.8$ and dry 7.8$ of the total acidity input.
A sulfate budget is shown in Table 6. The majority of sulfate input to
the lake is coming through the soil and unsaturated zone with MO and 46.6$
respectively. Wet and dry deposition contribute 5$ each. The net sulfate
reduction if the lake sediments Is 13.6$ of the total sulfate deposition.
-------
Table 7 presents the MSB evaluation for the calibration period for
discharge, chloride, alkalinity and sulfate.
-------
REFERENCES
Nikolaidis, N. P., Rajaram, H., Schnoor, J. L. and Georgakakos, K. P.
(1986a). Enhanced Trickle-Down model description. Civil and Environmental
Engineering, University of Iowa, Technical Report No. CEE-ARRG-86.01.
Nikblaidis, N. P., Georgakakos, K. P., and Schnoor, J. L. (I986b). Modeling
Short and long term impacts of acid precipitation using the Enhanced Trickle-
Down model: Lake Panther case study. Civil and Environmental Engineering,
University of Iowa, Technical Report No. CEE-ARRG-86.04.
Nikolaidis, N. P. (1986c). Estimation of daily potential evaporation. Civil
and Environmental Engineering, University of Iowa, Technical Report No, CEE-
ARRG-86.02.
-------
TABLE 1. CLEAR POND
ANALYSIS OF WET AND DRY DEPOSITION
Period
8/82-7/83
7/83-8/Sit
Average
ACIDITY (eg/ha
Wet
507.0
690.0
SULFATE (eg/ha
Wet
117.2
503.1
175.3
Total average Acidity - 849 eg/ha »yr
Dry deposition - 29.5$ of total deposition
Total average Sulfate - 790.2 eg/ha *yr
Dry deposition - 40.0$ of total deposition
Sulfate Loading (Average)
Wet - 22.82 kg/ha/yr
Dry - 15.12 kg/ha/yr
Total - 37.9^ kg/ha/yr
On the average, 19.1 metric tons of sulfate are deposited on Clear Pond
watershed (or 6.5 metric tons of sulfur).
-------
TABLE 2. LIST OF WATERSHED DESCRIPTORS USED FOR MODEL CALIBRATION.
GENERAL WATERSHED CHARACTERISTICS:
AQUATIC AREA -
TERRESTRIAL AREA -
CHARACTERISTIC DISTANCE -
DEPTH TO BEDROCK -
PARTIAL PRESSURE OF ATM C02 -
0.7300E+06
0.41I80E+07
0.1100E+4 METERS
6.3000 METERS
0.0003 ATMOSPHERES
SQUARE METERS
SQUARE METERS
SURFACE WATER PC02 IS-1.50 TIMES SATURATED PC02
SOIL COMPARTMENT CHARACTERISTICS:
POROSITY -
DEPTH OF SOIL LAYER -
SUM OF BASES -
SOIL DENSITY -
0.2700
0.5500 METERS
33.8000 EQUIVALENTS/KILOGRAM
1140.0000 KILOGRAMS/CU. METER
UNSATURATED ZONE COMPARTMENT CHARACTERISTICS:
POROSITY - 0.2000
TRANSPIRATION COEFFICIENT - 0.0010
BARE-GROUND FROST COEFFICIENT « 0.1000
REDUCTION IN FROST COEFFICIENT - 0.0800
DAILY THAW RATE - 0.1200
INITIAL FROST INDEX - 0.0000
LIMITING FROST INDEX - -3.0000
THAW COEFFICIENT - 0.2000
BULK DENSITY - 1590.0000
INCHES/DAY
DEGREES C
DEGREES C
DEGREES C
KILOGRAMS/CU.METER
SURFACE WATER BODIES CHARACTERISTICS:
STREAM BED ELEVATION AT OUTFLOW - 18.0000 METERS
GROUNDWATER COMPARTMENT CHARACTERISTICS:
POROSITY - 0.2000
-------
TABLE 3. List of optimum values of the calibrated parameters.
a) Hydrologic Parameters
-Snow:
BETA - 0.8095
KAPPA - 1.1M23
KPAM2 - 0.1766
-Evaporation
KPAH3 - 1.3171
KPAN5 - 1.0381
-Lateral and Vertical Flows:
KLAT3 - 54.781
KLATM - 55.791
KPERC3 - 2.3117E-2
KPERCM - 2.1741E-2
b)
RE 3
KH4
KH5
KH6
KOU
K05
K06
Alkalinity Parameters
5.5E-7 m3/eq/day
9.0E-2 meg/my day
3.1E-1 raeg/nr/day
2.0E-3 meg/mVday
1.1 E-2 raeg/nr/day
7.0E-2 meg/nr/day
0.1 E-3 meg/mz/day
c) Sulfate Parameters
CF - 2.02M7
KP3 - 5.0E-J« eg/kg/eg/m^
KP4 - 4.0E-H eg/kg/eg/m3
K - 1.033E-3 a/day
KP6 - 9.0E-6 eg/kg/eg/m3
-Groundwater:
D1 - 0.6963
FRAX - 0.680M
ALF1 - 6.2375E-4
ALF2 - 3.40U5E-2
-------
TABLE 4.' MONTHLY HYDROLQSIC BUDGET OF CLEAR POND
INPUT 1
INPUT 2
INPUT 3
INPUT 4
INPUT S
INPUT 6
OUT t
OUT 2
OUT 3
STORAGE
;.-44'*SE-02
0.11656-01
0.69076-02
O. 69696-02
0.93616-02
O. 13406-02
0.11186-02
0.27336-02
;. 323-f E-0^
.-.:4?se-.u
, . I606£-0i
0.36666-02
C. 12316-01
:! , ^1 7^6-01
0.80756-02
•" vrPfKiBE— 01*
O.OOOCEn-OO
0. OOOOE+OO
O. OOOOE+OO
O. OOOOE+OO
0.2994E-02
0. 19646-O1
0.5243E-03
0. 1627E-01
C.S69SS-M
-• . ;'73'>E-02
J.310BE-01
O.OCOCE-rtfO
C. OOOCE+00
C.OOCOEH50
0. OOOOE+OO
'VsO . OOOOE*OO i
0. l?94EiOll. 6-..14.07E-61
0.11026-O1
0. OOCOE+00
0.5S35E-02
0.74966-03
0. 14816-01
?. 17726-01
;.•'.' 3: 6-02
'Oi 2930ErO3^
" 0.4040E-03
0.42OBE-01
O. 1321E-O1
0.21B9E+OO
0. OOOOE+OO
5. OOCOE+00
o.2ees£-o2
0.2471E-01
0.1435E-01
0.16626-01
0.208O6-01
0.6232E-02
0.3291E-02
0.73416-02
0.11193S— 01
0.22126-0.
C. 409«.£— Ci
0.2244E-01
0.1214S-01
e.rSCOE-Ol
0.16266-01
;LOj,'23B6E-01 .
^0/i 348^6-0 1^
2SS5164E-01,,
0. 1248E-01
0. 1422E-01
0. 9240E-02
0.
0.
0.
0.
0.
0.
0.
0.
,%
J.
0.
3.
o.
0.
7O;.^E-J2
3402E-01
2056E-01
2330E-01
2812E-01
1924E-01
110SE-01
10596-01
it^c is-o;.
l-J^2E-«ii
4o5'i£-01
3900S-C1
1161S-01
0.2932E-01
0.
$°-
".••»•
O.
0.
0.
2C13E-Oi
i547jiErOl
3477E— 01
.1741E-O1
2132E-01
0.3=3aE-01 0.2873E-01
0.411 IE— M
:..i- «=.o;
o.
0.
5634E-01
,32oiS-Ci
0. =37 IS- 56
0.7742E-03
O.68296-O6
O.SOS6E-OS
0.10O2E-04
0.7200E-09
0.2373E-OB
O.2103E-C6
C. 140l5-'5c
0.3.34 56- i->
0.1058E--03
0.67646-03
O.S370E-04
0, 143?E-03
0. 23632^04
• O. 1494E-O4
0.2036E-03
O. 1348E-O2
0. OOOOE+OO
0. 1737E-03
0.22S4E-08
O. 17826-03
0. 1643E-O3
0.2E6.E-OS
O.1613E-04
0. 1SS8E-04
0.16136-04
0. 15636-04
0.16146-04
0. 1613E-C4
0.1436E-04
C. liT'Si-iV-,
•>. i2dic-.i4
3. i637E-04
0. 13646-O4
0. 1601E-04
0. 1617E-O4
0. 13386-04
0. 1616E-04
0.1371E-04
0. 16236-04.
O. 16156-04
0. 1467E-04
0.1620E-04
O. 1604E-O4
O. 16376-04
0.123:^-04
0. tC7cE-v6
O. 67736-06
0. 6344E-06
0.4775E-06
O. 6S64E-O6
0.6773E-06
0. 67756-06
0.61136-06
0. 335GSI-.26
0. -5373£- ;*;
0.iS73E— C6
0.6S676-04
0.6726E-06
0.679iE-'>Ji
O. &545E-06
0.67806-06
O. 66006-06
O. 681 96-06
O.6783E-O6
0.6160E-06
0.68O26-06
0. 6736E-06
O.6877E-06
C.522S£-W
"~or24rs"i-D2~"
O. 177^E-01
0. 13SOE-01
0. 1O91E-01
0.8889E-O2
0.2447E-02
0. OOOOE+OO
O.OOOC.E+OO
0^ 176-5E--C7
C.SlAfL'CO
0. 107SE-01
O. 19435-01
0. 2352E-01
O. 1284E-O1
0. 1530E-01
0. U22E-01
:0/3398E-O2
. O. OOOOE+OO
O. OOOOE+OO
0. OOOOE+OO
0. OOOOE+OO
O. OOOOE+OO
0. 1S24E-01
ij. 174C-E-O1
C. 1B9SE-32
0.4773E-01
O.3049E-01
O. 3907E- Lako compartmwnt to
.'ut 2: EViipor jticn rron tha Lake,
-at Cs Outflow frc.ii h^* Like,
Star .11; a: Chang J in L.«k« storage.
-------
UJ
a
<£
e
a
5
H
§
it
10
1 »
IUI
10
ICO
in in
o o
+ f
UJ UI
-an
n «• «*• n n
o o o o o
. + . + .
n
u iu ui ui
H r» M -«
•0 -0 03 -0
id
I
lin
10
I*
ItU
1C-
vi in m ui n
;»o p o o
l>l I'l Ul 1:1 UJ Ul 1
m co -• N in CD
_
tfl -«
O O O
O O
I
'.-• O oo
I
OOO
II
n in
0 o
+ *
UIUl
1 M
CD
?« in « in :n n m y n n ti I i I I I I -I. I I I I
j UtOOOOOOOOOOOoOOOOOoOO
j lo O O O O O O O O O O O O O .-.) O O O O O O
J fl+f-f + *-4-+-|4-T+'4-tf4-T4- + +-4.
' liu ui ui ui ui til ui ui 1*1 f J iti ui ui'." ui ui ui ui tj ui ui
- i «u 0* ^3 *4 ^^ m m ^^ t *^ fii fs ^^ »•• M) *Q Q* 0^ Q« c^i ^^
^4 I CD O ^f ^0 ^s ^^ f*« CD fs f^f C^ C^ C4 ^3 «i 0^ tO) 4J 4^ ^B CD
f O O
Id 111
12???
IUI 111 u Ul
10 NO 03
~.rt rt -
t??^
U IU Ul U)
•H «0 «* CT*
M w n -•
it * -0 (H
«-<«•«
•ri ?l n •« N M ^ •« N W O O .} M H -«
';. O O O O O ,> O O O O P O{O O O
its Ml
HI HI |;J (•) III Ul i(| Ul
-i M J) .« o- f-1 4 '
n " o •' w K r
in XI 41 t 0-
idoo o
j I I I I
ib o o o
IS
It
UIUI
SB
It V O> II
10- r» |s O
jtj n 5 i
0000
|88|:
SSiSS
n r; PI M
H * K -
idddddddd
it • i i i i
Ill UJ UJ UJ UJ Ul nj 111 Ul Ul lil Ul Ul Ul Ul Ul
i" '-• fi ct .< rs w CM K •« M o» o» ffJ r« -<
10 r-i o N o o or> -« n -o -«t -o -• -« —
(O •' i'« -O H N H «0 K N l> N
OOOOoOoOoOOOOOOOoOOO
^88§f
I Ul UJ Ul Ul Ul
1 III ^
•»
.
UlUIUIIUlilUtluUIUJUIUlUJbit/J
n -< n f-i -o -« o (> H co o- « u) q
"• 9 <" N HJ co n n o O' «• n n- N
ca * -« n
too o o
ipppo
to d d d
o o * « w
?????
Ill UJ Ul Ul III
O O t -O 111
o o -o in tr
o o w r j o
O O w 0' »i
Ul Ul Ul I
5 K m o o
n t m o o
N t c- ':> o
-« M ri '•• '•>
io o o o
jiti w ui i
1C- O O M
IWJ * 0> CD
| .4 t •» (N|
l.~. o O O
oo
I I
8W
c/
O O
UJ lil UJ Ul
•Q-Q a~ n
-«r -«
N t p•
».' IN . '
o o 6 o
000000,:
UIUI
M t
o -o •
ill -n i
•* 111 o
O O O O O
III Ul 111 UJ IU
O *^ 03 t O
ra -o ^ n o
o o
ijj C^ 03 '•'J CJ
n -4 M *< o
"• o c- o p p o o o p o
t M t V *
o o o o o
U UJ UJ Ul Ul
-( O ~< CU —
is O K) * *
o* rs M o w
** n ^ "f" ^*
I -'I
10
( I
IUI
in
IM
10
' IK
IC2I
H
U
O CMO M ,« OJ ft- O N •« ID -O «fW N K
!o
UJ W
)-QOKMO>«K
« M M fl N « « (M H -H
V
I ill
llf
1ST
u»
If
i?
I.-:
K4
1
lid
•R
Is
0
!?
ID
M
\i
i-<
11(1
IV
IV-
I <
I -'I
f-
1 1
in
\l
1-1
I *
IO
•I
0
Jt
a a
-u j; J»
C fl «
W_l _l
•woo
V..M *•
t*< *•
c c
ugg
4J *•
• U L L.
j: -« ig <•
c^&y
o /: a c
e a u "
v w
.- o i -•
M e a -»
At c O
••
T
a -f n -i
«= v
n *> -v
u •-• c -c
v n ••
L. 'i -<
f,i At 4^ v
4' V. -,» '1
<6 •' (A I.
s a o •«•
u t; a
r n o 2
J -• n t»
a •«
». w >•»-
Cl — t JJ
•en 3
•* Mj •« ri
-0 K
•• ••
*« •-• 4J r-(
•.
i_ •»
0 f!
-i O
41 J'
XI *
O A
i. fl
a i
n. :\
r. •••
k-4 t «
u
•»
J. "*
-. '•*
.*' Si
«I V.
•J '.t
-------
ti
UI
o
M
N
5
JE
a.
to o o o o o o o o
IW Ul Ul Ul Ul W M UJ 1.J
• ••!•) n N >> * o •-•• !'*
IM !') q- N f| -«O -'I T
i-ij -. co QJ *N net -«i
n-> M ID f> n •*-•*• n
i--j n i') O O C- O O
+ + +:*••• +
UJ III Ul 01 C>l III Id (if III W
c- HI n o- f i n tn N ifi o
-n o- 1,*- ••• M o» ui «t K; N 0- O 0> f n
>i _l .H .1 _l o o o o o o o o o
I I I I
O O O O v-
IO O O O O O O O
!••[ Ul Ul Ul Ul 111 III 111 lil
iv «• to n o t--) oti-* •*•
t -0 W N N 01 (D (O II* i:
IV - 0* M JJ 11 H M, N
:> v> nn
o o o o
+
I
i «n w n w « w n wi
1 O O O O O O O O <
iii HI III Ul Ul
•o *-n o (•> o»
•f ffi «V1 ^J {h*
N 6 ft- o O
> -• W -i W N
ui uj
I"
, i N
f -I
K.OOC
O O O O
ui ui ui ui ui u ui w w
t »t r«- K» o» "««r «•« P- n n oa •* t* -o o o- - m n •
-j o t N w o> •> ^ M o- o m N N * « n n o- a-
•"""""""•^RasffaBB*
, .. . N * « n « »; o- ra
»Cj* 0» IX (Q i^% ^} 00 «Q (^| 0* •<}
ojn^Min-oiotn^oaM
i;
iKiwvvttiiHMfo^fl-Tfl'^'r'Tionioi'jn^'^T
loooooooooooooooooooooooo
IUJ Ul It] Ul Ul Ul til Ul til Ul til Ul Ul Ul lil Ul Ul Ul til Ul Ul Ul Ul W
if * * o o is •* r» m tj CQ ui -o CD •* -o M ri •* n ** N o 01 n
ICO * * N CD f -" N m /> «•« f>« n !•) n pi »«-a n • +: 44 + 4+ H- + + *••»•+• + + + +* + + +*
t|ilUI|i|blultiJuiUIUlllJ|flUI|UlillJJUIlUUJbJUIUII'llilUl
10^ (0 ft W *^ *^ ~n ^ *w- ^ *•"* n*rth'^""h'^"***'* »••*•*—*K.j»/%
10* W N W IO *O
l-fl *.* » * *
id o'd.o do
I ' '•'
15?$5?l
IbJ Ul lit Ul Ul • •
!-< <> in -o •HP
dododododddddddodo
i bi o- n n -»-o
I Cl O ID *•• 0s •*
l«i (D N tB FN CO n) K CB S
idddddd
??????????????????
«»!8KS8l!!8S«!l!|!!1!j!ll!a{!!
K>-.a)M-Otf-MOM-)ininNfDpi
•k b A «k ^^ A.^ ^ _ ^A. ^^ .^&. L «• T^ KJk A . ^ A^ •
- >0
•4-gl
11^aRB88S8aES«B8S
dddddddddddddddddo
10 oo o o o
i' r?
iMO'^N*OW*
O^^90%OQ'*0*4tl}QIQ9'Q*lJOO*6oM||')r4l>fc*t
»-i «i 0* M «•• « N M r> W » |-> •? 01 »* O N VI M W •<
dododododododododo
«rfn
ooooooooooooo
&A&ii*-+-•*•>--*••*•-*•*•
a- n 6 .5 S » w 5r ih N ii
-0 n- "i N N r» '-••-• M M I'' M «* •< »* K) M
O
W Ul (il (if It/ Lil Ul UJ til lil
- • « n o r- o r ^ - •
ID -O I- Jl •< |
OOOOOOOOOOOOOOOOOOOOOOOO
«r « w n n *
+AAAAA
U'
M I
OOoOOOOOOOoOoOOO O.O
Ul tal Ul Ul IU UJ Ul Ul lil Ul ui Ul Ul Ul i Ul itl Ul
oi
r% a •< * M «r o* N »•• w p) in vi o~ M •* * o
•rc^cotn-otnujo 11 -f ts o-1>> o »< N CK <>
V -0 (h -< H ••< o* N • • W N •* N 01 fcl •* -' «H
jo ooooo
IO O OO O •*•
i?°???v
ooooo
oo
ooo
li| '.ll Ul Ul ti| til Ul Ul III Ul Ut Ul IU Ul Ul UJ UJ Ul
^. M O W r» O O O <^ O OB -« OT -• th 0) O O
«^J ^r » -o ri o o o .-» o « 01 « w o •« o o
v- -o n -* r< <• o o o o ^ o- « o M -o o o
~< <*• -a -« ,0 o o o •. ;. o o o o o o o o o o
- lil Ul til Ul Ul Ul t'l Ul Ul Id Ul Ul bl Ul Ul bl
!•>, Q* «f n r+11 v tt' t t-'i fa <> o o> •< M !•> i -
^ " "~ " " «Z fy^ y i^ ft ,^^ >..
_ - ---oooNo^oin
M «T » H Cl o 6 o 6 o o o •"> o 6 6 6 6 6 o cS 6
to
Iw
i
ui uJ bJ ui i itl
o at -a p in
KJJl>f1 III
III
id
I i
fe
liillci Ul bl
!jr £ J- ^ 1 r>
iH 0* N •* O
iri oo> n M. ..
!»« M •« N W W N «* .(I 11 .< I
tdodododonodo6o6o6o666ooo jo
L
it
I?
id
o
o
!*
!^
8
i
l-n
••i
• I
-O • i
Ct «••
^n
Iii
4J .-<
3v
5o
a v
J5
SI
"J
" 5
n
:,l
O
^ •
'fi
f- ''I
in
a-J
E
O •>•
"Ji
i *:
: -• ft IT •*• IO
*J JJ JJ
3 ::• -j
a a. *t
c u -.
i. 5)
«^s
«. -'
wi
J 3 -J
rx a. if'
C c c :i
»-« I"C t-( I.I
"m c •
^i?
"J .K ..« a
•« A» i>
:• J u «
•.i n
S -! 3
_ jis r» .i
Cl •" >.. -fl
— _ .1-1
— 1 .1 C
j) ; i •-«
o •<• c
a i. "'
a s a ^
T> n w +"
** c: •»
:»«. — r
L -J '-I
rj J ..
•• o - -
?•- i.* *J
.. ^ -n
.•j n i.- 5
J •-* M
a--§ / 4
1-1 r' ,! 1.1
-------
TABLE 7. MSB Evaluation for the
Calibration Period.
Variable
Discharge
Chloride
Alkalinity
Sulfate
Units
m3/s
ueg/L
ueg/L
ueg/L
n
HH3
2U
21
21
MSB
0.026
5.61
829U.O
79.7
RMSE
0.16
2.37
18.6
8.9
-------
10
Figure 1. Precipitation data.
Figure 2. Evaporation data.
Figure 3. Temperature data.
Figure *». Wet and dry sulfate loading.
Figure 5. Wet and dry acidity loading.
-------
CLEAR POND
PERIOD : 7/82 - 7/84
800
-------
CLEAR POND
PERIOD : 7/82 - 7/84
50
40
30
z
o
1
ui
DC
0.
20
10
100
200
300
400
DAYS
soo
600
700
800
-------
CLEAR POND
PERIOD : 7/82 - 7/84
CO
ill
u
DC
0
UJ
D
•>
111
DC
IU
a
S
ui
H
-6
.10
-15
-20
I I
100
1 ' '
__^_^^J________—^^^^^_^^^^^^^^^^^MJ
1 1 1 I ' ' 1 1 I
200 300 400 BOO 600 700 800
DAYS
-------
CLEAR POND
TIME SERIES OF ACID DEPOSITION
SIMULATION PERIOD: 7/27/82 - 7/24/84
Legend
D WET
D DRY
100
700
800
-------
CLEAR POND
TIME SERIES OF SULFATE DEPOSITION
SIMULATION PERIOD: 7/27/82 - 7/24/84
m
•o
40
35
30
o
. 25
Ul
(0
It.
O
z
o
E
o
Ou
U
O
20
15
10
100
200
300
400
DAYS
500
Legend
m WET
D DRY
600
700
800
-------
11
Figure 6. Calibration results of cumulative outflow.
Figure 7. Calibration results of daily outflow.
Figure 8. Calibration results of dally outflow error.
Figure 9. Calibration results of lake chlorine.
Figure 10. Calibration results of lake sulfate.
Figure 11. Calibration results of lake alkalinity.
Figure 12. Sulfate simulation without loading increase.
-------
CLEAR POND
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
CALIBRATION PERIOD : 7/27/82 - 7/24/84
O
u
I
D
2
D
O
Legend
D FIELD DATA
100
150
200 250
DAYS
aoo
350
400
450
-------
CLEAR POND
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
CALIBRATION PERIOD : 7/27/82 - 7/24/84
40
35
30
25
3* 20
3
IL
S is
10
5
i>,.i^\ i ^
Vv
100
200
300
400
DAYS
soo
Legend
FIELD DATA
600
700
800
-------
CLEAR POND
COMPARISON OF SIMULATED VERSUS FIELD OUTFLOWS
CALIBRATION PERIOD : 7/27/82 - 7/24/84
.40
50
100
1SO
200 2SO
DAYS
soo
3SO
400
4SO
-------
CLEAR POND
COMPARISON OF SIMULATION AND FIELD CHLORIDE
CALIBRATION PERIOD: 7/27/82 - 7/24/84
200
150-
Ill
u
z
o
u
III
S
E
o
X
o
50
D
a DLJ UDD LJDU u DDDT3—ann D- nn
100
200
300
400
DAYS
Legend
SIMULATION DATA
FIELD DATA
SOO
600
700
800
-------
CLEAR POND
COMPARISON OF SIMULATION AND FIELD ALKALINITY
CALIBRATION PERIOD: 7/27/82 - 7/24/84
z
O
Ul
O
z
O
O
300
250
200
150
100
.50
-100
Legend
0 SIMULATION DATA
O FIELD DATA
700 800
-------
CLEAR POND
COMPARISON OF SIMULATION AND FIELD SULFATE
CALIBRATION PERIOD: 7/27/82 -7/24/84
900
260
200
U
O
z
O
O
150
100
100
200
300
400
DAYS
Legend
B SIMULATION DATA
D FIELD DATA
BOO
600
700
800
-------
CLEAR POND
COMPARISON OF SIMULATION AND FIELD SULFATE
CALIBRATION PERIOD: 7/27/82 -7/24/84
300
D
D
n
100
200
300
400
DAYS
SOO
600
700
Legend
• ACTUAL DATA
D FIELD DATA
800
-------
APPENDIX A.1-2
integrated Lake-Watershed Acidification Study (ILWAS)
-------
Direct/Delayed Response Project
ILWAS MODEL CALIBRATION AND SENSITIVITY ANALYSIS
FOR WOODS LAKE, PANTHER LAKE AND CLEAR POND
Ronald K. Munson
Steven A. Gherinl
Margaret M. Lang
Robert M. Newton
Smith College/Tetra Tech, Inc.
3746 Mt. Diablo Boulevard, Suite 300
Lafayette, California 94549
(415) 283-3771
-------
INTRODUCTION
The ultimate goal of the Direct/Delayed Response Project is to provide
Congress with a scientifically-based estimate of the degree to which surface
waters in the northeast and southeast regions (as defined by U.S. EPA) will
be affected by acidic deposition over the next fifty years. This estimate
will be based, in part, on the results obtained from dynamic computer
simulation models. The models will be applied to several lake- and stream-
watershed systems in both regions. Classification rules will then be
developed based on the model simulations and used to predict surface water
behavior for the entire regions. Clearly, the application of these models
to many watersheds dictates that the data record for calibration will, in
most cases, be very short. Therefore, before the models can be applied
regionally, their ability to simulate the behavior of single lakes or
streams having relatively long data records must be demonstrated. In
addition, the sensitivity of the models to various model parameters must
also be demonstrated.
The ability of the Integrated Lake-Watershed Acidification Study (ILWAS)
model to simulate northeast lakes with relatively long data records has been
demonstrated with the calibration of Woods Lake, Panther Lake, and Clear
Pond. These calibrations were performed using data generated as a part of
the field components of ILWAS and RILWAS (the regionalization of the ILWAS
project). The results of the calibrations will be presented below as will a
short description of the ILWAS model, a discussion of the physical and
chemical characteristics of the three watersheds, and the results of the
model sensitivity analyses.
ILWAS MODEL OVERVIEW
ILWAS Model Description
The ILWAS model conceptualization, development, application, and theory have
been described elsewhere (Chen, Gherini, and Goldstein, 1979; Goldstein,
Gherini, and Chen, 1984; Gherini, Mok, Hudson, Davis, Chen, and Goldstein,
-------
1985; Davis, Whipple, Gherinl, Chen, Goldstein, Chan, and Munson, 1986).
What follows below is a brief summary of the model.
The ILkAS model was developeo to simulate the chemical response of surface
waters in forested ecosystems to changes in deposition acidity. To do this
the model simulates the major physical ana biogeochemical processes
occurring in the tributary watersheds as well as in the surface waters.
This is achieved in the model by routing incident precipitation through the
forest canopy, soil strata, streams and lakes (See Figure 1). Concurrently
the model simulates the major processes which add acid or base to the
throughflowing water.
The ILWAS project demonstrated that for drainage basin systems, surface
water quality is largely determined by where precipitation flows en route to
becoming surface water (Goldstein, Gherini, and Chen, 1984; Peters, 1985;
Schofield, 1985). To calculate the distribution of water between flow paths
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 relationships for surface
waters. Snow and ice formation and melting are simulated by the model using
multiple-term heat budgets.
In simulating the acid-base chemistry of the water, the ILWAS model
calculates the concentrations of 16 chemical constituents (See Table 1) in
throughfall, surface runoff, soil solution, and in surface waters. This is
achieved using mass-balance techniques and both kinetic and equilibrium
formulations to represent the major processes which add acid or base to the
water (Table 2). Concentrations of chemical species which are not readily
mass balanced are computed from the concentrations of other constituents.
For example, the model calculates H*-ion concentration from alkalinity,
total inorganic carbon, total monomeric aluminum, and total organic acid
analogue.
The model is applied by dividing a lake or stream-watershed system into
catchments, each of which discharge water directly to either a stream or
lake. Each catchment is divided vertically into compartments with
-------
deposition
wet I I dry
Transpiration
Evaporation
UTTER LAYER
ORGANIC LAYER
4 Sublimation
SNOW PACK
^^iK^:^m ^m&£^^
y^^:^^ ^ J^^V^J1^: W/
////!
"INORGANIC" -'-
Of MINERAL
Stream
Flow
Interlayer Advection
and Dispersion
LAKE SEDIMENTS
T
Figure 1. Lake-Watershed System Showing Hydrologic Setting
-------
r ">
Table 1
Solution-Phase Chemical Constituents
Cations Anions
Solutes CaT S04_
tracked ^
by mass MgT NO3T
balance
KT CIT
NH4j | F-
Solutes H+ AI(OH)*
which
can be Al3* HCQj
calculated
from those AI(OH)2* COf"
above
AI(OH£ R13*
AIF2+ HR12'
AlFj H2R1'
AI{SO/ R2'
AIR22* AIF4
Ai(R2)j AIF|-
AIF3-
A I/ CO ^"
r\l\&\JA)2
Neutral
Species
Analytical Gases
Totals HnoutOnlv)
H4Si04 Alk(ANC)
Cp2(aq)
AIF3
AIR1
AI(R2^
AI(OH|
H3R1
HR2
Org Acid Lig. 1
(triprotfc, R1)
Org Acid Lig. 2
(monoprotic, R2)
Ah-
CT (TIC)
FT
Note: Subscripted T"
total analytical solution
concentration
(SOX)
(NOX)
indicates
phase
^
-------
Table 2
Chemical and Physical Processes Simulated
by the ILWAS Model
Surface Water Processes
Gas Transfer
Mixing (Advection & Dispersion)
Heat Exchange
Ice Formation & Melting
Algal Nutrient Uptake
Nitrification
Reductive Loss of Strong Acid Anions
Solution Phase Equilibration
Canopy Processes
Dry Deposition
Foliar Exudation
Nitrification
Solution Phase Equilibration
Washoff
Snowoack Processes
Accumulation
Sublimation
Leaching
Nitrification
Soil Processes
Heat Transfer
Biomass Loop
Litter Accumulation
Litter Decay
Organic Acid Decay
Nitrification
Nutrient Uptake
Root Respiration
Abiotic Processes
Mineral Weathering
Competitive Cation Exchange (Al, Ca, Mg, K, Na, NH4, H)
Anion Adsorption (SO4, PO4, organic acid ligand)
CO2 Exchange
AI(OH)3(am) Dissolution-Precipitation
Solid-Liquid-Gas Phase Equilibration
-------
homogeneous characteristics: forest canopy, vegetation, and separate soil
layers. Streams are divided Into longitudinal segments and lakes are
divided vertically into well-mixed layers.
Input to the model includes time invariant parameters which characterize
each compartment (e.g., for a soil layer -- bulk density, organic and
mineral composition, cation exchange capacity, etc.) and the initial
solution and solid phase concentrations (Initial conditions). Time variant
model Input consists of meteorological data (e.g., dally air temperature,
and precipitation amounts) and chemical data (monthly dry deposition and
precipitation solute concentrations). As output the model calculates the
?+ OA x X ^
aqueous concentrations of the base cations (Ca , Mg , K , Na , NH.),
2
strong acid anions (S0|~, N03, Cl ), alkalinity, silicic acid, organic acid
analogue, total monomeric aluminum (AU), organically complexed monomeric
aluminum, and pH - in throughfall and in soil and surface waters.
BASIN CHARACTERISTICS
Woods Lake, Panther Lake and Clear Pond are all located in the Adirondack
Par* region of New York State (Figure 2). Woods and Panther Lakes are both
in the west central region of the Adirondack Park and are located less than
20 miles from each other. Clear Pond is located near the eastern boundary
of the park and is slightly farther north than hoods Lake.
All three sites are forested watersheds with largely deciouous cover. The
2
basins range from 1.2 to 5.7 km in surface area and lie at approximately
the same elevation. Clear Pond has 3 to 4 times the relief of the other
basins and also has a wioer variety of till depths with thin till (less than
%
1 m) near the top of the watershed and thick till (as deep as 55 mj near the
lake (average aepth = 6.5 m). Panther Lake soils are more uniformly oeep
(average depth = 24.5 m) while Moods Lake soils are thin (average depth -
2.3 m;. The physical characteristics of all three basins are summarized in
Table 3.
-------
Adirondack
Park
N.Y. state border
N
20 km
Figure 2. The ODRP Intensively Studied Lakes In the Adirondack Park of
New York
-------
r
CHARACTERISTICS OF
Basin area (km2)
Lake surface elevation3 (m)
Relief6 (m)
Forest cover (%}
Coniferous (%)
Deciduous (%)
Mean till depth (m)
Lake
Area (km2)
Mean depth (m)
Maximum depth (m)
Lake Outlet
Alkalinity (peq/i)
pH
aRelative to mean sea level
Difference between highest and
Table 3
PANTHER, WOODS,
Panther
1.2
557
170
98
3
95
24.5
0.18
3.5
7
•35 to 240
4.5-7.2
lowest elevation
AND CLEAR LAKES
Woods
2.1
606
122
95
5
90
2.3
0.26
4.0
12
-60 to 30
4.4-5.9
in the watershed
Clear
5.7
580
518
95
0
95
6.5
0.74
6.0
24.4
80 - 170
6.1-7.4
^A
8
-------
Although Woods and Panther Lakes each receive about the same amount of
precipitation of nearly identical quality, the alkalinities (ANC) of the
lake outlet waters are significantly different. The alkalinity at woods
Lake is about -10 Meq/l while at Panther Lake the average alkalinity is
nearly 150 neq/i. Clear Pona receives less total precipitation than Wooas
and Panther and the acidity of the precipitation is slightly lower as well.
The lake outlet alkalinity at Clear Pond averages 100 ueq/A. The chemical
characteristics of both the wet ana dry deposition at all three lakes is
summarized in Table 4.
CALIBRATION
Application of the ILWAS model begins with calibration. Basin data are usea
to quantitatively characterize the system to be simulated. Initial
conditions (e.g., lake stage, soil and surface water quality) are
established as a simulation starting point. The model is then run using
actual meteorological and air quality oata as input. The model output, the
quantity and quality of water at various points in the system, is made to
coincide with observed values by adjustment of calibration parameters (e.g.,
evapotranspiration coefficients). The simulated results are typically
compared to the observed data using graphical procedures. Simultaneous
calibration against observed data for several points within a watershed is
recommended if data are available. For example, throughfall quantity,
snowpack depth, and flows at various points in any streams, should be
calibrated together with lake discharge. Since the ILUAS model simulates
many processes, calibration of these processes must follow a logical order
to proceed with minimal effort. The general rules for calibration are: I)
calibrate the system's hydrologic behavior before calibrating the chemical
behavior; 2) calibrate in the same order as water flows through the basin;
and 3) calibrate on an annual basis first, then seasonally, and finally
calibrate to the Instantaneous (daily) behavior.
-------
Table 4
MEAN ANNUAL WET DEPOSITION LOADING
(equivalents/hectare-year)
Constituent
Sulfate
Nitrate
Chloride
Ammonium
Calcium
Magnesium
Sodium
Potassium
Clear8
470
250
46
130
65
23
38
14
Panther
670
370
52
200
150
40
49
22
Moods'
560
320
44
180
110
36
44
17
MEAN ANNUAL DRY DEPOSITION LOADING
{equivalents/hectare-year)
Constituent
Sulfate
Nitrate
Chloride
Ammonium
Calcium
Magnesium
Sodium
Potassium
Clear'
83
39
15
37
37
11
14
12
Panther
120
67
15
36
89
18
17
12
Woods*
100
50
15
29
62
16
17
12
Sampling period was for a two year period from August 1982
through July 1984,
Sampling period was from March 1978 through December 1981.
10
-------
Hydro logic Cali brat Ion
The first step in the hydrologic calibration is the adjustment of the
evapotranspiration parameters so that the predicted cumulative lake outflow
matches the observed data. Observed lake outflow is checked against the
observed rainfall to ensure that all major storm events cause either
increases in lake outflow or at least retard recessions. The seasonal
patterns in the predicted cumulative outflow are controlled by the length of
the snow cover period and by seasonal variations in evapotranspiration. The
snow formation temperature and the snowmelt rate coefficients, along with
the sublimation rate, are adjusted to match observed snowpack depth and lake
outflow during the snowmelt period. A seasonal calibration parameter is
used to adjust the variation in potential evapotranspiration. Canopy
interception storage, monthly leaf area indices, and root distribution
determine the allocation of evapotranspiration to the different compartments
of the system. Depending on the depth of the water table, the distribution
of potential evapotranspiration will influence the actual amount of water
lost (i.e., a large distribution factor for the upper soil layer together
with a deep water table will give a low total evapotranspiration rate
despite a high potential).
The peak flow and recession curve characteristics are influenced by the
snowpack, soil field capacity, soil-specific yield, and soil hydraulic
conductivity. The fine adjustment of the instantaneous lake discharge rate
can be performed in conjunction with the chemical calibration. However, the
ratio of base flow to the sum of shallow and surface flow should be
reasonably estimated using flow separation techniques before proceeding with
chemical calibration.
Chemical Calibration
Lake chemical characteristics are primarily determined by the intial lake-
water quality, in-lake processes, soil solution quality and the routing of
water through the oifferent soil layers. The concentrations of several soil
solution constituents is fairly constant with time because of equilibration
of the aqueous phase with the solid phase by cation exchange and anion
11
-------
adsorption reactions. Hence, a reasonable calibration of lake outflow
chemistry can be obtained by establishing the initial solute concentrations
in the soil and lake waters, the soil cation exchange selectivity
coefficients and anion adsorption coefficients, and the in-lake rate process
coefficients. The initial solute levels in the soil solution must be such
that a volumetric flow-weighted average of the soil solution concentrations,
plus direct precipitation onto the lake surface, will give the average lake
concentrations for species which oo not undergo reaction in the lake (e.g.,
Cl~,). The volumetric flows used in the above averaging are the soil layer
lateral flows obtained from the hydrologic calibration. The cation exchange
selectivity coefficients and anion absorption coefficients are adjusted so
that the initial soil solution concentrations are at equilibrium with the
adsorbed concentrations. The soil nitrification rates, lake nitrification
rates, and monthly algal production rates are aojusted to calibrate the
ammonium and nitrate concentrations in the lake outflow.
Observed lysimeter and ground water chemical data can be used to check the
hydrologic calibration. The concentrations of solutes which do not
equilibrate with the solid phase may drift quickly away from their initial
levels if the rate processes in the soil are not calibrated properly. The
aluminum dissolution rate is adjusted to maintain a fairly constant level of
aluminum in the applicable soil layers from year to year (aluminum
concentration may fluctuate seasonally with the nitrate cycle). The
nitrification rate is adjusted to prevent buildup of nitrate in deep soil
layers. The observed silicic add concentrations are used in setting the
mineral weathering rates.
If observed data are available, the throughfall chemistry can be calibrated
by adjusting the gas and particulate deposition velocities, canopy
nitrification rates, and foliar exudation rates. .This also provides a means
of quantifying total atmospheric deposition (wet and dry). The snowpack ion
levels are calibrated by adjusting a solute leaching coefficient.
The dynamic behavior of the lake outflow water quality is influenced by the
lake.thermal profilet the fraction of solute retained in the lake ice, and
12
-------
the fraction of flow that comes in as surface runoff or ground water
seepage.
The calibrated and observed cumulative lake outflow, instantaneous lake
outflow, pH, and concentrations of ANC, NH., AU, Ca, hg, K, Na, SU^, N(X,
Cl, and Silicic Acio are shown in Figures 3 through 16 for Panther Lake,
Figures 17 through 29 for woods Lake and Figures 30 through 42 for Clear
Pona. The figures indicate good agreement between simulated and observed
dynamic concentrations. The calculated mean square errors between the
simulated ana observed concentrations are presented in Tables 5, 6, and 7.
Discussion of Results
The results indicate that the primary reason for the difference in the
acidity of Wooos and Panther lake waters is the depth of till in the
watersheds. Panther basin has thick till and thus a larger reservoir of
exchangeable bases and weatherable minerals to neutralize acid deposition.
Sensitivity analysis indicates that if the depth of till at Panther Lake
were reduced, it too would be more acidic (see next section). For both
Woods and Panther Lakes, the majority of the base supplied to throughflowing
waters comes from cation exchange (70%).
The simulation of Clear Pond led to some interesting comparisons of sources
of alkalinity. For example, although the average till thickness at Clear
Pond is only 6.5 m, the thick till in the immediate vicinity of the lake
provides significant buffering capacity. Another factor in the alkalinity
supply at Clear is the presence of anorthosite minerals. These are similar
to the minerals at Woods and Panther but have a higher ratio of calcium to
sodium. The calcic minerals at Clear Pond weather faster than the sodic
minerals at woods ano Panther and thus provide a much larger fraction of the
base supply. Mineral weathering contributes is over 50 percent of the base
supply at Clear compared to less than 30 percent at Woods and Panther.
13
-------
107fl
• SinULRlEO
1B79
1980
1981
DflTE
PflNTHER OUTLET
Figure 3. Simulated and Observed Cumulative Outflow for Panther Lake
-------
SOMOJFnflflJJflSOMDJFrtflMJJflSOWDJFMflMJJfl
1978 i879 18B0 1981
DftTE
- SIHULRIED
PRNTHER OUTLET
Figure 4. Simulated and Observed Instantaneous Outflow for Panther Lake
-------
Lft
3- i
U5
• i i f
SON.DJFMRMJJfiSONDJFNRMJJRSOKDJFHRrlJJfi
1976 1979 19S0 1961
- CflLCULRTED
O OBSERVED DRTfi
PRNTHER OUTLET - PH
Figure 5. Simulated and Observed pH for Panther Lake
16
-------
SCNDJFMfiMJJfiSOMDJFMfiMJJflSOMOJFKRHJJR
197S 1979 I860 19B1
CRLDULftTED
OBSERVED DflTfi
PRNTHER OUTLET - RLK
Figure 6. Simulated and Observed Alkalinity for Panther Lake
17
-------
i
R
a
if>
t\j
UJ
R J
B-
SDNDJFNflMJJRSDNDJFMflrTJJflSOMDJFMflMJJ
197S 1979 I960 19S1
CfiLCULflTEO
OBSERVED DSTfl
PflNTHER OUTLET - NH4
7. Simulated and Observed Ammonium Concentration for Panther Lake
18
-------
S3
ID
B-l
cu
SOMDJFMflKJJflSOMDJFMflMJJRSOfcJCJFnRKJJR
1876 J979 J9B0 1861
CRLCULRTED
OBSERVED ORTfl
PflNTHER OUTLET - RLT
Figure 8 Simulated and Observed Total Monomeric Aluminum Concentration for
3 Panther Lake
19
-------
SONDJFMflMJJRSONDJFMRMJJflSOMDJFMflMJJR
197S 1973 J9B0 LB81
- OTLCULflTEO
D OBSERVED DflTfi
PRNTHER OUTLET - CR
Figure 9. Simulated and Observed Calcium Concentration for Panther Lake
20
-------
UJ
!M
LU g
c\i
s>
us
SDWDJFMflMJJRSOWDJFMRMJJflSOKDJFMRMJJR
197B 1979 19B0 i9Si
CflLCULflTED
OBSERVED DflTfi
PRNTHER OUTLET - MG
Figure 10. Simulated and Observed Magnesium Concentration for Panther Lake
21
-------
E3
ta
151
U> _J
8-1
SOMDJFMRMJJflSOMDJFMR MJJRSONDJFMflMJJR
1976 1979 I960 1961
- CflLCULRTED
O OBSERVED DfiTfi
PflNTHER OUTLET - K
Figure 11. Simulated and Observed Potassium Concentration for Panther Lake
22
-------
Si
in
fU
UJ J
CE
Si
LT5
SQ 0 N D J^M RKJJflSONDJFMflfUJflSDNDJFKRKJJR
CflLCULflTED
OBSERVED DRTfl
PflNTHER OUTLET - Nfl
12. Simulated and Observed Sodium Concentration for Panther Lake
23
-------
K)
SDKDJFMRMJJflSDMDJFMflMJJflSOMCJFMRMJJR
1976 1979 13B0 1961
- CRLCULflTED
P OBSERVED DnTfl
PRNTHER OUTLET - 804
Figure 13. Simulated and Observed Sulfate Concentration for Panther Lake
24
-------
LU
ro
o
SOWDJFKflNJJflSONDJFMRKJJRSOMOJFKflMJJR
197S J979 J960 1961
CRLCULRTED
OBSERVED DflTn
PflNTHER OUTLET - N03
Figure 14. Simulated and Observed Nitrate Concentration for Panther Lake
-------
Si
BE
Si
Si
S2
L3 S -J
B-
D °D
SONDJFM-fl^JJflSOMDJFMflMJJflSO M DJFMRKJJR
1378 1979 19B0 1861
- CflLCULflTED
D 055£R!/ED DBTn
PRNTHER OUTLET - CL
Figure 15. Simulated and Observed Chloride Concentration for Panther Lake
-------
Si
l-l
O
LU
IX
& —
C\i
>—i S -
SOWDJFNRKJJRSOWDJFKflMJJRSCKCJFMflKJJR
197B J979 19S0 1961
- CRLCULflTED
O OBSERVED DflTfl
PnNTHER OUTLET - SI04
Figure 16. Simulated and Observed Silicic Acid Concentration for Panther
Lake
27
-------
ro
CD
en •
cc
|n|f'ri|T'lT ...... "IT "p""if" iw
IT
60NOJFMflMJJfl60NDJFHflrtJJfl60NOJFnflMJJfl
- SinULflTEO
O OBSERVED
DflTE
WOODS OUTLET
Figure 17. Simulated and Observed Instantaneous Outflow for Woods Lake
-------
Si
(A
g_
CA
Q_ ^
S3
»
Ss
§.
*
a-
8.
rt SDWDJFMflMJJRSOWDJFMflMJJflSONDJFMflMJJR
197B 1979 1980 1981
DRTE
- CRLCULRTED
o
Figure 18. Simulated and Observed pH for Woods Lake
29
-------
in
N.
ru
in
N.
LU
in
CM
o
SDWDJFMflMJJRSONDJFMRMJJRSDNDJFMRMJJR
197B 1979 19B0 1981
HOT-
L?I i I 1—
- CRLCULflTED
o
Figure 19. Simulated and Observed Alkalinity for Woods Lake
30
-------
SONDJFMflMJJRSDNDJFflflMJJRSOMDJFnRMJJR
137B 1979 I960 iSBi
- CRLCULflTED
D OBSERVED DflTfl
WOODS OUTLET - NH4
iFigure 20. Simulated and Observed Ammonium concentration for Woods Lake
31
-------
KJ
SO
3-
B-
LU
0: 1
§
SONOJFMflMJJRSONOJFMflMJJflSOMOJFMflMJJfl
1978 1979 J9B0 19B1
DRTE
- CflLCULflTED
O .^B,.™
Figure 21, Simulated and Observed Total Monomeric Aluminum Concentration
for Woods Lake
32
-------
Si
Si
=r
Si
IB
Si
LJ S
1—I—I—I—I—I—I—I—I—I—I—I—I 1 I—r~l—I—'—'—'—'—'—'—'—'—'—I—r~t—'—'' ' r~
SDWDJFMflMJJRSOWDJFMflMJJflSDWDJFMflMJJR
1978 1979 1980 1981
r» r^Tfr
Dn i h
CRLCULflTED
OBSERVED DflTfl
WOODS OUTLET - Cfl
Figure 22. Simulated and Observed Calcium Concentration for Woods Lake
33
-------
SONDJFMflMJJflSDWDJFMRMJJRSDNDJFMflMJJfl
1978 1979 I960 1931
o
CflLCULflTED
Figure 23. Simulated and Observed Magnesium Concentration for Woods Lake
34
-------
ED
S5
sr
Si
(B
f«5
s
C\J
1/5
SDNDJFMRMJJRSOWDJFMRMJJflSONDJFMRMJJfl
197B 1879 I960 1981
D-Tr
- CRLGULflTED
O ™00™WOO[)S _
Figure 24. Simulated and Observed Potassium Concentration for Woods Lake
35
-------
ED
K>
=r
is;
SONDJFMflMJJflSDNDJFMflMJJflSONDJFMflMJJR
187B J979 • I960 I9B1
CR"
- CflLCULflTED
o
Figure 25. Simulated and Observed Sodium Concentration for Woods Lake
36
-------
D
SDMOJFMRMJJRSDNDJFMflMJJRSONDJFMRMJJfi
1978 1979 19B0 1961
DRTE
- CRLCULflTED
o
Figure 26. Simulated and Observed Sulfate Concentration for Moods Lake
37
-------
SI
U5
CO
§
6 0 N D J F M fl H J J fl S 0 N D J F M R M J J fl S 0 N D J F M fl M J J R
197B 1979 J9B0 19B1
CflLCULflTED
C6SERVED OflTfl
DRT
WOODS OUTLET - N03
igure 27. simulated and Observed Nitrate Concentration for Woods Lake
38
-------
K;
&
3-
If)
CJ
, -
C_J U>
U5
I I I
SONQJFMRMJJRSONDJFMRKJJflSONDJFMRMJJfi
1976 J979 I960 19Bi
- CRLCULRTED
0 OBSERVED DRTfl
WOODS OUTLET - CL
Figure 28. Simulated and Observed Chloride Concentration for Woods Lake
39
-------
t\j
ss
U5
O
LU
O g —
12-
B-
o
SONDJFMflMJJflSDWDJFMflMJJflSONDJFMRMJJR
19 76 1979 J9B0 1961
DRTE
- CRLCULflTED
Figure 29. Simulated and Observed Silicic Add Concentration for Woods Lake
40
-------
Ld
g
LiJ
O
cn
H-
cn
o
CL
ct:
CE
LU
_l
CJ
_,I ,J I _JI I I I ._l
h- e-cu et aa-ei o$-ei atrei IC-BI wei arai «a at
So
•o
o
a.
(O
3
C7J
NIBN
CU1
41
-------
CO
§_
tfi
X g_
Q_ *
B J
*
CO
ftSDNDJFMflMJJflSDNDJFHflHJJ
1982 1983 19B4
- SIMULRTED
OBSERVED
DRTE
Figure 31. Simulated and Observed pH for Clear Pond
42
-------
sa
CM
IS)
K.
LU
!ZJ
*:
_i
a:
RSDWDJFMflMJJRSDNDJFMRMJJ
1982 19B3 J9B4
DRTE
- SIMULRTED
resERVED CLERR POND OUTLET [CD
Figure 32. Simulated and Observed Alkalinity for Clear Pond
43
-------
RBDNDJFttflMJJflSONDJFMflriJJ
1982 J9B3 19S4
DRTE
- SIMULRTED
BBSBWED CLERR POND OUTLET CCL)
Figure 33. Simulated and Observed Ammonium Concentration for Clear Pond
44
-------
S3
in
CU
IS
CM
LU
in
ru
g-
DO
RSONDJFMRMJJflSONDJFMfiMJJ
1982 1983 18B4
DRTE
- SIMULRTEO
OBSER.EO
Figure 34. Simulated and Observed Total Monomeric Aluminum Concentration
for Clear Pond
45
-------
cu
S3
H
IB
ID
is
cu
IT-
U_
C_3
flSONDJFIIflMJJflSDWDJFMFlMJJ
1982 19B3 19B4
DRTE
- SmULflTED
0 OBSERVED
(Figure 35. Simulated and Observed Calcium Concentration for Clear Pond
46
-------
i
CU
8-
ID
i , f
flSOWDJFMflMJJflSDWDJFMRMJJ
1982 19B3 19S4
DRTE
- SIMULflTED
resERVED CLEflR POND OUTLET CCL)
Figure 36. Simulated and Observed Magnesium Concentration for Clear Pond
47
-------
CVJ
1/5
LD
8 J
s>
QQ
Q pOD Q-"uDQ
U
RSOWDJFMRMJJRSDWDJFMRMJJ
1982 1983 1984
- SIMULRTED
D OBSERVED
D.RTE
CLEflR POND OUTLET COL)
Figure 37. Simulated and Observed Potassium Concentration for Clear Pond
48
-------
cu
ni
rr
LL-
o o
-------
RSDNDJFMRMJJRSDNDJFMRMJJ
1982 19B3 J9B4
DRTE
- SIMULATED
SERVED
Figure 39. Simulated and Observed Sulfate Concentration for Clear Pond
50
-------
ru
8 J
CM
B-J
CM
X
O
LU
K2 J
O 2
in _
N. '
Q O
o o Q--QI
RSOMDJFMflflJJRSONDJFMflrtJJ
1382 1983 1984
DfiTE
- SIMULRTED
"SBWED CLERR POND OUTLET [CD
Figure 40. Simulated and Observed Nitrate Concentration for Clear Pond
-------
a
U)
B-
ru
§ sJ
-J g-l
LJJ -
tf> _
B-
us —
XL
^W^e-o Q D a D B 'on °Q
Q -D
1 1 1 1 1 1 1 r
RSDNDJFMFIMJJfiSDNDJFMFlKJJ
1982 19B3 1984
- SIMULflTED
D OBSERVED
DRTE
CLERR POND OUTLET [CD
Figure 41. Simulated and Observed Chloride Concentration for Clear Pond
52
-------
8
ru
cu
CU
§
CU
U)
is.
a
O
LU
Da
a
U5 _
B-
RSDNDJFMflflJJflSDNDJFMfiMJJ
1982 1983 1984
- SIMULRTED
O C6SERVED
CLERR POND OUTLET [CD
Figure 42. Simulated and Observed Silicic Acid Concentration for Clear Pond
53
-------
Table 5
Panther Lake Outlet
Parameter
Discharge
Cl
NO,
sof
Ca2*
Mg2+
Na*
K+
NH*
4
pH
Alk
AIT
* MSE = S
Number of
Data Points, n
547
250
247
250
250
250
250
250
222
159
205
65
(xoBS-xs|M)2
n
Mean Square Standard Estimate
Error (MSE) * of Error (SEE)"
0.0002 ( m3/sf
2
20.7 (jieq/l)
279
312
1350
76.7
66.3
4.0
8.3
.28 (S.U.)2
2220 (i-ieq/l)
i
10.5 ((imol/l)
/
-SEE-, /»<,,
V
0.013 rrP/sec
4.6 jieq/l
16.8
17.7
36.9
8.8
8.2
2.0
2.9
.53 S.U.
47.3 neq/l
3.3 (nmo!/l)
3S-XsJ2
n-2
J
54
-------
Table 6
Woods Lake Outlet
i
Parameter Number of
Data Points, n
Discharge 1 095
Cl 263
NO3 261
SO*' 263
Ca2* 263
M^+ 263
Na* 262
K+ 262
NH* 263
pH 187
Alk 205
AlT 136
* MSE = I(XOBS-XSjMf
n
\
Mean Square Standard Estimate
Error (MSE)* of Error (SEE)"
0.0065 ( rrP /sf
11.1 (|ieq/l)
129
301
277
44.7
42.0
10.5
16.8
0.10(S.U)22
983 (p^q/0
1 2.3 Oimol/l)2
** SEE«t / KX0
V
0.08 m3/sec
3.3 jieq/l
11.4
17.4
16.7
6.7
6.5
3.3
4.1
0.32 S.U.
3.5 ^imol/I
>8s-Xs,M)2
n-2
55
-------
Table 7
Clear Pond Outlet
Parameter
Discharge
Cl
N03
so;
Mg2*
Na+
K+
NH*
PH
Alk
AIT
* MSE =
Number of Mean Square Standard Estimate
Data Points, n Error (MSE)* of Error (SEE)
102
23
23
23
23
23
23
23
11
23
23
21
HXOBS-xsj2
n
0.001 1 ( m? /s)
2.1 (neq/lf
43.3
114.0
558.0
25.4
39.9
1.0
4.5
.05(S.U)22
320 (M-e
-------
APPENDIX A.1-3
Model of Acidification of Groundwater in Catchments (MAGIC)
-------
Model of
Acidification of
Groundwater
In
Catchments
r&J^S&g^S?:
s:•:•••>'•: :":•>!
••:••-*-.: - :••-;-;
• *fc ••"-•«> --»••-**-"•.•.%,•*
:••:•:-:•;•"<•:.:*;•:•:*:•!..*.::•
•• * w. *"* *".*--, .v- »".vXv"
«s?
i%'XA»>K"» «t-"-"..v.v.".".*.'.-1v.-I-l-;'it-
."*~.~.v*
stl
W?s
«l
SSWK-K"
Llii
«s
S>5«5J:?ffij-i:ft.J...-l
i^^>-
PM^
IK
;-;-fla
^^^f^
I'-viv
*-"~^"»
•^^
-* •
.•*•«?•
^>
*..•.":<£?
,&3*
£&&
'ffK-e.
»z^<^<:i
Plg^rrrrr:.-^ Base ca«,on^
. "^r—-*.H
SA^V^V:-! ••-.. ^
^j'.J'-.-;
s5f/X<,<
*VJS
^^!wra^^^t^v^5^
B. J. Cosby R. F. Wright G. M. Hornberger J. N. Galloway
-------
PROTOCOL FOR CALIBRATION PHASE OF
DDRP MODELING
B.J. Cosby, G.M. Hornberger, P.F. Ryan
and D.M. Wolock
Department of Environmental Sciences
University of Virginia
Introduction
Four tasks must be completed in the calibration phase of our DDRP
modeling project. 1) First, all inputs must be specified. Also, all "fixed"
parameters must be set and an (objective) procedure for selecting these values
must be specified. By "fixed" parameters we mean those that are not adjusted
in calibration. 2) Second, the remaining ('adjustable') parameters are
selected so that a loss function, in this case an unweighted sum of squared
errors (SSE), is minimized. 3) Third, an analysis of how well the adjustable
parameters have been estimated must be accomplished. This is, the sensitivity
of the SSE to changes in the optimal values of the adjustable parameters must
be determined. 4) Fourth, with the adjustable parameters fixed at their
optimal values, a conventional, univariate sensitivity analysis for the
'fixed' parameters must be done. Again this will be a sensitivity with
respect to the SSE.
The first task is relatively straightforward. Inputs will, in general,
be specified using measured data. If data are unavailable, suitable
extrapolations will be specified by objective procedures. In a similar vein,
some model parameters will be fixed (i.e., have values specified) using
observations in conjunction with well-defined objective procedures. These
parameters will not be adjusted in the calibration.
-------
2
The second task requires the minimization of the SSE between model
predictions and observed data. In general terms the model can be written.
_Y - _f (_x, JB ) (1)
where _Y is a vector of 'outputs', ^x is a vector of 'inputs' (and fixed
parameters), e is a vector of adjustable parameters of order k, and _f is a
vector function relating the other quantities. We then seek to minimize
* < e. > - "S mS eua2 <_G )
a-1 u-1 (2)
where eua presents the error in the u"1 component of the vector at the atl1
t
temporal observation:
eua - - Yua 0»<>del>]. <3>
The order of the output vector is m and the number of observations is n.
There are any number of techniques for determining a value of _© that
minimizes * ( J& ). (For example, see Bard, 1974). We propose to use the
Rosenbrock (1960) method to optimize, although we may experiment with other
methods (e.g., the Harquardt (1963) algorithm).
Once a parameter vector _e has been chosen such that the objective
function is (approximately) a minimum for these values, questions about the
reliability and precision of our estimates must be answered. This is the
third task. Essentially we would like to know something of the sensitivity of
the objective function to parameter changes, of the parameter estimation error
covariance matrix, and of the goodness-of-fit of the model.
Useful approximations for nonlinear models with normally distributed
errors using maximum likelihood estimation are given below.
-------
1. The « - indifference region is defined by the set of values
of _© for which
* ( _© ) - ** S e ,
where ** is the estimated minimum value of the objective function.
This region is defined approximately by
(3_e ) H* (3 _6 ) < 2 €, (4)
where 3 JB - _e - j&* and where H* is an estimate of the Hessian matrix
with elements H a« ( © ) - 32 */3©a 3e» (a and & are indices of the k x k
matrix). This equation defines ellipsoids about the estimated minimum in
which the objective function does not vary greatly.
2. The covariance matrix of the estimates is defined by
V - E (6 J3* S J3 *T )
where d _© is the shift in the parameter estimates that would
be caused by an error in the measured variables of 6^. (That is,
8 J& is the estimation error in parameters that arises from
measurement error in the variables.) For a wide class of maximum
likelihood estimates with normally distributed errors,
V wH*"1.
Generally speaking, the V computed from any given data sample an
only be regarded as a rough estimate, correct to within an order of
magnitude. Approximate confidence ellipsoids can be computed by
S T H*5 0 < c (5)
-------
4
where c must be determined based on the sampling distribution. Thus
confidence regions coincide with e-indifference regions (compare
equations 4 and 5). For a given confidence level, we choose c such
that
Prob [ S e_T V -1 *_e * cj - a.
When the sampling distribution is normal, unbiased and with known
covariance V J9 , then the quantity S JB*V ~ SjB is
distributed as y* with 1 degree of freedom.
3. Two linear approximations for confidence limits are given in Conway
et. al., 1970.
The fourth task is to determine the sensitivity of the SSE to changes in
the fixed parameters in at least a crude fashion. Even though these
parameters will be specified using an objective procedure and thus may be
considered part of the model structure, they cannot be considered to be known
with certainty and we think that it is prudent to determine how small changes
in these parameters affect the SSSE. We cannot estimate the error covariances
of these parameters and cannot comment on how well they have been estimated,
but we believe that a simple sensitivity calculation is worthwhile
nevertheless. In this task, we will individually perturb the fixed parameters
a prescribed amount (in both the positive and negative directions) and record
the change in SSE over the base case.
The DDRP modeling at the University of Virginia involves two models, the
hydrological model TOPMODEL and the chemical flux model MAGIC. In the
following sections each model is described and its inputs, fixed parameters,
-------
adjustable parameters, and outputs are specified.
Summary
The four tasks in the calibration phase of our DDRP modeling were
outlined in the introduction. Each task will be performed on each model. The
results will include the sensitivity of the SSE to changes in each fixed and
adjustable parameter.
References
Bard, Y. 1974. Nonlinear Parameter Estimation. Academic Press, N.Y., 2341
PP.
Beven, K.J. and M.J. Kirkby. 1979. A physically based variable contributing
area model of basin hydrology. Hydro. Sci. Bull.. 24; 24:43-69.
Cosby, B.J., R.F. Wright, G.M. Hornberger and J.N. Galloway. 1985a. Modeling
the effects of acid deposition: assessment of a lumped parameter model of
soil water and streamwater chemistry. Wat. Resour. Res.. 21: 51-63.
Cosby, B.J., R.F. Wright, G.M. Hornberger and J.N. Galloway. 1985b. Modeling
the effects of acid deposition: estimation of long-term water quality
responses in a small forested catchment. Wat. Resour. Res.. 21. 1591-
1601.
Cosby, B.J., G.M. Hornberger, J.N. Galloway, and R.F. Wright. 1985c.
Freshwater acidification from atmospheric deposition of sulfuric acid: a
quantitative model. Environ. Sci. Tec.. 19: 1144-1149.
Cosby, B.J., G.M. Hornberger, R.F. Wright, E.B. Rastetter and J.N. Galloway.
1986a. Estimating catchment water quality response to acid deposition
using mathematical models of soil ion exchange processes. Geoderma (in
press).
Cosby, B.J., P.G. Uhitehead and R. Neale. 1986b. A preliminary model of
long-term changes in stream acidity in southwest Scotland. J. Hvdrol..
(in press).
Chow, V.T. 1964. Handbook of Applied Hydrology. McGraw-Hill, New York.
Conway, G.R., N.R. Glass and J.C. Wilcox. 1970. Fitting nonlinear models to
biological data by Marquardt's algorithm. Ecology 51: 503-507.
-------
Corps of Engineers, U.S. Army, North Pacific Division. 1956. Summary Report
of the Snow Investigation, Snow Hydrology.
Hamon, W.R. 1961. Estimating potential evapotranspiration. J. Hvdraul. Dlv.
Am. Soe. Civ. Eng. 87:107-118.
Hornberger, G.M. , K.J. Seven, B.J. Cosby, and D.E. Sappington. 1985.
Shenandoah watershed study: Calibration of a topography-based, variable
contributing area hydrological model to a small forested catchment. WRR
21: 1841-1850.
Marquardt, D.tf. 1963. An algorithm for least-squares estimation of nonlinear
parameters. J. Soc. Indust. Appl. Math.. 11: 431-441.
Neal, C., P.G. Whitehead, R. Neale and B.J. Cosby. 1986. The effects of
acidic deposition and conifer afforestation on stream acidity in the
British uplands. J. Hvdrol.. (in press).
Rosenbrock, H.H. 1960. An automatic method for finding the greatest or least
value of a function. Comput. J. 3: 175-184.
Whitehead, P., Hornberger, G., and R. Black. 1979, Effects of parameter
uncertainty in a flow routing model. Hydrol. Sci.. Bull.. 24: 445-464.
Wright, R.F., B.J. Cosby, G.M. Hornberger and J.N. Galloway. 1986.
Interpretation of paleolimnological reconstructions using the MAGIC model
of soil and water acidification. J. Wat. Air Soil Pollut.. (in press).
-------
SECTION A: SUMMARY (BOTTOM LINE, FOLKS)
Table A.I Values of MSE, RMSE and RMSE expressed as a percentage of the
observed mean value of the DDRP variables for Woods Lake
(calibration and corroboration periods)
Table A.2 Values of MSE, RMSE and RMSE expressed as a percentage of the
observed mean value of the DDRP variables for Panther Lake
(calibration and corroboration periods)
Table A.3 Values of MSE, RMSE and RMSE as expressed a percentage of the
observed mean value of the DDRP variables for Clear Pond
(entire period of record)
Figure A.I Output of calibrated model: long-term hindcast for Panther Lake
Figure A.3 Output of calibrated model: long-term hindcast for Clear Pond
OTHER ASSORTED PRELIMINARIES:
Figure A.4 Conceptual basis of the hydrological model (TOPMODEL)
Figure A.5 Conceptual flux routing in the chemical flux model (MAGIC)
Figure A.6 Conceptual linking of TOPMODEL state variables to flow routing
parameters in MAGIC
-------
TABLE A.I
MSE SUMMARY FOR UOOOS LAKE
MSE RMSE XMEAN
Discharge n3/s .0029 .05
Cun Disc. m/yr
Chloride meq/rn3 22.8 4.8
Sulfate meq/m3 358.2 19.2
Calcium neq/m3 43.7
Magnesium meq/m3
G.0
Sodium meq/nS 17.6
Potassium meq/n3 3.6
Alun(tot) mol/m3 41.4
neq/m3 49.5
6.6
2.4
4.2
1.9
6.4
7.0
50. S
14.7
Alk(0rn) meq/n3 617.0 24.8 -183.7
9.1
13.1
21.5
29.7
61.7
36.5
.0046
19.2
173.0
266.2
47. t
5.9
8.9
2.8
46.2
64.B
.07
4.4
13.2
16.3
6.9
2.4
3.0
1.7
6.8
8.0
-
48.4
10.7
-429.0
9.6
13.1
14.5
26.6
102.0
52.6
I September 1978 ~ 31 May 1980
(2) 1 June 1980 — 31 August 1981
-------
TABLE A.2
MSE SUMMARY FOR PANTHER LAKE
(based on OORP variables of interest)
MSE - Mean Squared Deviation, RMSE * Square Root of MSE
XMEAN = 100.*RMSE/ 1 September 1978 — 31 May 1980
<2) I June 1980 ~ 31 August 1981
-------
TABLE A.3
USE SUMMARY FOR CLEAR POND
(based on DORP variables of interest)
M5E = Mean Squared Ceviation, RM5E « Square root of M5E
XMEAN - 100.»RHSE/(Mean of observed data/
Period of Record (1 >
VARIABLE UNITS MSE R«SE XHEAN
Discharge m3/s .0219 .15
Cum Disc. n/yr
Chloride neq/n3 21.7 4.7
Sulfate neq/m3 90.2 9.5
Alk(gran) neq/n3 347.4 13.6
Calcium meq/m3 444.0 21.1
Magnesium neq/m3 22.0 4.7
Sodium meq/m3 25.0 5.0
Potassium meq/m3 .5 .7
Alum(tot > nol/m3 1.4 1.2
i+ neq/«3 .0 .2
-
63.5
7.5
17.9
12.7
14.7
12.8
18.4
514.3
28.6
(1) 27 July 1982 — 24 July 1934
-------
FIGURE A.I
§
13*1
OODS LAKE (O
ALK
WOODS L^vKE (OB)
VOLUME WEIGHTED ANNUAL AVERAGE
1981
-------
FIGURE A.2
PANTHER LAKE -:'C5)
VOLUME WEIGHT5D AMMUAL a-. ERASE
SBC
SAA
ALK
1961
1981
TEAR
PANTHER LAKE (OB)
WDOMTEC AIHNUAL
3.
•4-
1O -
Al
H+
tot
1881
1001 1921
YEAR
1981
1981
-------
FIGURE A.3
CLEAR POND OB
8"
I
i
SBC
CLEAR POND (QB)
VOLUMC VNDGMTCD ANNUM, AVERAGE
30 H
20 H
10 H
1S&4
192«
YEAR
Al
tot
1984
-------
FIGURE A.4
evapo transpiration
precipitation
contributing
area
In (a/tan B)
-------
FIGURE A. 5
Fl
i
F2
A+B SOIL
F3
C SOIL
SURFACE WATER
-------
FIGURE A.6
A+B
QA+B
Q
Q
-------
SECTION B:
OPTIMIZATION PROTOCOL FOR TOPMODEL (HYDROLOGICAL MODEL)
Description of optimization protocol
Table B.I Values and sources of fixed parameter values
Table B.2 Ranges and optimal values of adjustable parameters (optimized
using both the calibration period and the entire period of
record for Woods and Panther)
Table B.3 Results of Hessian analysis on optimized parameters (Woods
Lake, calibration period)
Table B.4 Results of Hessian analysis on optimized parameters (Panther
Lake, calibration period)
Table B.5 Results of Hessian analysis on optimized parameters (Clear
Pond, entire period of record)
Table B.6 Results of Hessian analysis on optimized parameters (Woods
Lake, entire period of record)
Table B.7 Results of Hessian analysis on optimized parameter (Panther
Lake, entire period of record)
-------
10
Hydrological Model
The hydrological model, TOPMODEL (Beven and Kirkby, 1979), will be used
to determine routing parameters for use in the chemistry model. TOPMODEL is a
topography-based variable contributing area catchment model. The model has an
upper and a lower storage zone, with precipitation input and
evapotranspiration loss occurring only within the upper store. Flow paths
within the model include overland flow directly into the stream, drainage from
the upper zone to the lower zone, and baseflow from the lower zone into the
stream. Precipitation can also bypass the upper zone and flow directly into
the lower store.
A detailed description of TOPMODEL is given in Hornberger et al. (1985).
They also present the results of a sensitivity analysis which suggest a
reduced TOPMODEL structure that captures the critical elements of catchment
hydrological behavior. Based on their findings, a simplified version of the
model is used for the DDRP. The inputs and parameters listed below correspond
to those in Hornberger et. al. (1985).
A simple snow accumulation and melt model was added to the "front end" of
TOPMODEL to be used for model applications where winter snowpack are
significant. The optional TOPMODEL parameters are determined during the warm
months when possible errors from a snow model need not be considered. The
optimal TOPMODEL parameters are then taken as fixed and the optimal snow model
parameters are determined using the entire period of record.
Since TOPMODEL simulates the hydrology of only the terrestrial portion of
watersheds, it was necessary to include a simple reservoir routing routine on
the "back-end" of TOFMODEL for those catchments with lakes. This routine
utilizes hypsographic information to delay water entering the lake before it
-------
11
flows out through the outlet.
Inputs:
PPT - Measured precipitation [nun/day]. Snow accumulation and melt
calculated using empirical formulas given in Chow (1964), and in Corps of
Engineers (1956). The cutoff temperature for snow accumulation and the
temperature-induced snowmelt parameter were optimized whereas the rain-induced
snowmelt parameter was fixed to the value given in Chow (1964).
PET - Potential evapotranspiration [mm/day]. Daily values calculated
from mean temperature and daylength data using the equation of Hamon (1961).
TOPOGRAPHICAL INFORMATION. Includes A/TANB distribution, total blue line
stream length and total catchment area. The constants A and TANB are the
upslope area and the local slope respectively. All the information is
derived from topographic maps.
DISCH - Measured flow [mm/day].
Fixed Parameters:
SUBV - Kinematic streamflow velocity [km/day]. Estimated from typical
values given in the literature.
RIP - Riparian area [fraction]. Calculated as:
RIP - WID * BLUE / CATCHAREA
where WID is the average stream width, BLUE is the total
catchment blue line stream length and CATCHAREA is the total
catchment area.
ARIAK - Lake area [fraction].
UO - Saturated hydraulic conductivity [mm/day]. Estimated from available
data or literature sources.
-------
12
SZQ * Maximum baseflow rate [mm/day]. Calculated as:
SZQ - KMAX * DEPTH * ( 2 * BLUE + PERIMETER)
TERRAREA where KMAX is the maximum saturated hydraulic conductivity,
DEPTH is the average till depth, BLUE is the blue line stream
length, PERIMETER is the lake perimeter and TERRAREA is the
terrestrial area of the catchment.
SRMAX - Maximum storage in upper layer [mm]. Calculated as SRMAX -
ROOTDEPTH * FIELDCAP where ROOTDEPTH is the rooting zone depth and
FIELDCAP is the volumetric moisture proportion at field capacity.
Adjustable parameters:
PMAC - Macropore flow parameter [fraction]. Optimum value determined
from the range of 0 - 0.75.
SZM - Baseflow recession parameter [mm]. Optimum value determined from
the range 4-180.
Outputs:
QTOT - Total streamflow per day [mm/day].
QOF - Total overland flow per day [mm/day].
SMDEF - Average soil moisture deficit [mm].
-------
TABLE B.I
FIXED PARAMETER VALUES
Parameter
Woods
Panther
Clear Pond
RIP
ARLAK
SRMAX (mm )
SUBV12
U0' '* (mm/day)
SZQ
-------
TABLE B.I (CONT.)
FIXED PARAMETER vALUES (continued)
Parameter
total blue3
126 2
lake area ' ' (km >
rooting depth (m>
average till depth '
1 0
max permeability (mm/day)
4 1 0
field capacity '
lake perimeter
-------
TABLE B.2
ADJUSTABLE PARAMETERS
Ranae for Dot irnizat ion
Parameter Minimum
SZM (nn) 4.0
PMAC 0.0
TCUT (°F> 10.0
SNOPROP 0.0
Optimized Values
Period of
Catchment Dot imizat ian SZM(mm)
Woods Calibration1 7.43
Woods All Data2 10. 64
Panther Calibration 11.22
Panther All data2 14.82
Clear Pond All data3 12.19
Maximum
180.0
0.7S
50.0
0.5
PMAC TCUT( °F )
0.36 Z6.Z4
0.51 25.81
0.49 26.89
0.53 2S.46
0.30 32.62
SNOPROP
0.03
0.03
0.06
0.04
0.04
4/14/78 to 5/31/80
4/14/78 to 12/19/81
8/1/82 to 7/31/84
-------
TABLE B.3
HESSIAN ANALYSIS
UiQQDS - CALIBRATION PERIOH
TQPMODEL PARAMETERS
PARAMETER
SZM
PMAC
ESTIMATED
OPTIMUM
7.426
.359
STO OEU
ESTIMATE
2.742
.095
% STD ERR
ESTIMATE
36.32
26.32
CORRELATION MATRIX OF PARAMETER ESTIMATES
SZM PMAC
SZM
PMAC
100.0
50.7
50.7
100.0
EPSILON INDIFFERENCE REGION IS: ,19073E-01
(BASED ON HESSIAN MATRIX WITH 2 PARAMETERS)
SNOWPARAMETERS
PARAMETER
TCUT
SNOPROP
ESTIMATED
OPTIMUM
26.244
.030
STD DEV
ESTIMATE
1 .405
.005
X STD ERR
ESTIMATE
5.36
16.60
CORRELATION MATRIX OF PARAMETER ESTIMATES
TCUT SNOPROP
TCUT
SNQPROP
100.0
-11.0
-1 1 .0
100.0
EPSILON INDIFFERENCE REGION IS: .51813
(BASED ON HESSIAN MATRIX WITH 2 PARAMETERS)
-------
TABLE B.4
HESSIAN ANALYSIS
PANTHER - CALIBRATION PERIOD
TQPMODEL PARAMETERS
PARAMETER
SZM
PMAC
ESTIMATED'
OPTIMUM
11.223
.492
STO OEY
ESTIMATE
19.625
.457
X STO ERR
ESTIMATE
174.87
92.79
CORRELATION MATRIX OF PARAMETER ESTIMATES
TCUT SNOPROP
TCUT 100.0 37.8
SNOPROP 37.8 100.0
EPSILON INDIFFERENCE REGION IS: 7.317J
(BASED ON HESSIAN MATRIX WITH 2 PARAMETERS)
-------
TABLE B-5
HESSIAN ANALYSIS
CLEAR - ALL DATA-
TQPMQDEL PARAMETERS
ESTIMATED STD DEV X STO ERR
PARAMETER OPTIMUM ESTIMATE ESTIMATE
SZM 12.193 9.402 77.11
PMAC .304 .393 123.15
CORRELATION MATRIX OF PARAMETER ESTIMATES
TCUT SNOPROP
TCUT 100.0 -1.2
SNOPROP -1.2 100.0
EPSILON INDIFFERENCE REGION IS: 2.4680
-------
TABLE B.6
HESSIAN ANALYSIS
WOODS - SJULJOfllfl-
,4
TQPHQDEL PARAMETERS
ESTIMATED STQ DEO X STO ERR
PARAMETER OPTIMUM ESTIMATE ESTIMATE
SZM 10.645 5.02! 47.17
PMAC .505 .064 12.75
CORRELATION MATRIX OF PARAMETER ESTIMATES
SZM PMAC
SZM 100.0 -38.2
PMAC -38.2 100.0
EPSILON INDIFFERENCE REGION IS: .11522
(BASED ON HESSIAN MATRIX WITH 2 PARAMETERS)
SNQU PARAMETERS
PARAMETER
TCUT
SNOPROP
ESTIMATED
OPTIMUM
25.810
.030
STD DEU
ESTIMATE
1.426
.006
X STO E
ESTIMA
5.52
19.75
CORRELATION MATRIX (R2«100) OF PARAMETER ESTIMATES
TCUT SNOPROP
TCUT 100.0 1.4
SNOPROP 1.4 100.0
EPSILON INDIFFERENCE REGION IS: .43858
(BASED ON HESSIAN MATRIX WITH 2 PARAMETERS)
-------
TABLE B.7
HESSIftN ANALYSIS
PANTHER - ALL DATA
TOPMODEL PARAMETERS
PARAMETER
SZM
PMAC
ESTIMATED
OPTIMUM
14.818
.533
STO DEV
ESTIMATE
**********
*»••*•**•**«
X STO ERR
ESTIMATE
**••»**»*•
»•**»»•»*•»
CORRELATION MATRIX OF PARAMETER ESTIMATES
TCUT SNOPROP
TCUT
SNOPROP
100.0
3.7
3.7
100.0
EPSILON INDIFFERENCE RESIGN IS: 1.0230
(BASED ON HESSIAN MATRIX WITH 2 PARAMETERS)
-------
13
SECTION C: EVALUATION OF MSB FOR THE HYDROLOGICAL MODEL
Table C.I MSE of cumulative flow
Table C.2 MSE of daily flow (in units of m3/s)
Table C.3 MSE of daily flow and efficiency (in units of mm/day)
Table C.4 MSE of average monthly discharge and efficiency (in units
of mm/day)
In each table, entries are made for:
a) Snowmodel vs. TOPMODEL contributions to MSE
b) Calibration vs. corroboration period MSE values
(for Woods and Panther)
c) MSE values for the models calibrated to all available data
(for Woods and Panther)
-------
TABLE C.I
MSE OF CUHULnTIUE FLQU
Watershed
Woods
Woods
Panther
Panther
Clear Pond
Period of
OntiMiration
Cal ibrat ion
nil data2
Cal ibrat ion
nil data2
All data3
Number of
— Tine ateos
1287
1287
1287
1287
372
MSE Um«]2>
5986.
11975.
103788.
9872S.
356E4.
USE (It
2.69 x
5.38 x
i.S2 x
1 . 44 x
9. 57 x
-------
TABLE C.2
WATERSHED
Woods
Woods
Woods
Woods
Woods
Woods
Panther
Panther
Panther
Panther
Panther
Panther
Clear Pond
Clear Pond
PERIOD
Calibration
Calibration
Corroborat ion
Corroboration
All data3
All data
Calibration
Calibration
Corroborat ion
Corroboration
All Data3
All Data
Ail Data*
AH Data
MODEL
TOPMQOEL3
SNOW5
TOPMODEL
SNOU
TOPMODEL
SNOU
TOPMODEL
SNOU
TOPMOOEL
SNOU
TOPMOOEL
SNOW
TOPMOOEL
SNOU
MSE <
69. x
288.
188.
455.
101 .
253.
7. x
124 x
214.
249.
7, x
45. x
'20 x
2190.
[«3/sJZ>
x I0~S
X 10^g
X 10
X 10 _
X 10
I0~s
x 103?
X 10
'•Is
10 "
'•"*«
X 10
4/14/78 to 5/31/80
6/1/80 to 12/19/81
4/14/78 to 12/19/81
8/1/82 to 7/31/84
Optimizing TOPMODEL parameters from June to October.
Optimizing SNOU parameters during entire year.
-------
TABLE C.3
USE (Cmm/day] ) AND EFFICIENCY OF DftlLY FLQU
Watershed
Woods
Woods
Uoods
Woods
Woods
Wooda
Panther
Panther
Panther
Panther
Panther
Panther
Clear Pond
Clear Pond
Period
Calibrat ion
Calibration
Corroborat ion_
Corroborat ion
All data^
All data
Calibration
Calibration
Corroborat ion
Corroboration
All data!?
All data
All data*
All data
Model
TOPMDOEL5
SNOW
TOPMODEL
SNOW
TOPMOOEL
SNOW
TOPMODEL
SNOW
TOPMODEL
SNOU
TOPMODEL
SNOU
TOPMODEL
SNOU
2
MSEU nn/dav] )
1 .15
4.78
3.12
7.56
1.67
4.19
0.38
6.31
3il2
7,56
0.38
2.Z9
0.33
6.02
Number of
time steps
296
721
306
S67
602
1288
295
720
306
567
601
1287
186
372
£ff
0.54
0.01
0.50
0.20
0.53
0.39
0.63
-0.75
-6.73
-5.38
0.70
0.21
0.76
0.38
4/14/78 to 5/31/80
6/01/80 to 12/19/81
4/14/78 to 12/119/81
8/01/82 to 7/31/84
Optimizing TOPMODEL parameters form June to October.
Optimizing SNOU parameters during entire year.
-------
TABLE C.4
HSE (Inm/davl ) AND EFFICIENCY OF MONTHLY-AVERAGED FLQU
MSE(Cnm/dav]
Number of
tidesteps Eff
Woods
Uoods
Woods
Woods
Uoods
Uoods
Panther
Panther
Panther
Panther
Panther
Panther
Clear Pond
Clear Pond
Calibration
Calibration
Corroborat ion_
Corroboration
All data^
All data
Calibration
Calibration
Corroboration
Corroboration
All data?
All data
All dataj
All data
TOPMQOEL5
SNOW
TOPMODEL
SNOW
TOPMOOEL
SNOU
TOPMODEL
SNOU
TOPMODEL
SNOU
TOPMODEL
SNOU
TOPMOOEL
SNOU
0.18
0.30
0.31
.1 .32
0.30
0.82
0.17
0.76
0*.25
0.72
0.16
0.63
0.22
3.08
9
23
10
18
19
41
9
23
10
18
19
4-1
6
ft
0.78
0.86
0.7S
0.43
0.76
0.64
0.69
0.69
0.56
0.24
0.72
0.64
0.70
0.48
4/14/78 to S/31/80
6/01/80 to 12/19/81
4/14/78 to 12/119/81
8/01/82 to 7/31/84
Optimizing TOPMOOEL parameters form June to October.
Optimizing SNOU parameters during entire year.
-------
15
Chemical flux model.
The chemical flux model is MAGIC (Cosby et al., 1985 a,b,c). Details of
MAGIC and examples of its use have been given elsewhere (Cosby et al., 1986b,
Wright et al., 1986, Neal et al., 1986), including a sensitivity analysis
(Cosby et al., 1986a). MAGIC has been modified for this work in the following
ways: the model now contains two soil layers; atmospheric inputs can be
measured values when available (rather than inferred annual means);
atmospheric inputs can bypass the upper soil layer (macropore flow);
atmospheric inputs can be accumulated and released from a snowpack. Several
outputs from TOPMODEL are used to set routing parameters and to control the
snow accumulation and melt in MAGIC. The basic chemical reactions modeled in
MAGIC are unchanged. The inputs and parameters listed below correspond to
those in Cosby et al, , (1985b) with the exception of the routing parameters
(identified with a *) which were introduced for this two layer version of
MAGIC.
Inputs:
Q-M measured streamflow [mm/day].
(xx - Ca, Mg, Na, K, SO., Cl, NO3, F) - Measured atmospheric
P
deposition of ions [meq/m /day]. The rate at which the measured
deposition is added to the soil is controlled in the case of snow
accumulation and melt by the outputs of TOPMODEL.
(xx - Ca, Mg, Na, K) - Weathering inputs of base cations
*\
[meq/m*/yr]. Measurements of these rates are not available. These
inputs will be treated as adjustable parameters that must be
optimized (see below).
-------
16
W „ ( xx - SO,, Cl, NO,, F) - Net uptake rates of anions [raeq/m^/yr].
These net uptake rates simulate biological utilization of the
anions. The annual rates of these net uptakes will be calculated
from measured net fluxes of the anions.
PCO- - Partial pressure of CO, in the two soils [atm]. These partial
pressures are linearly scaled to temperature (a measured input).
The scaling factor (S^) is an adjustable parameter that is
optimized (see below).
T - Temperature of the two soil layers and the surface water [deg. C].
Taken from measurements or literature.
Fixed parameters:
CEC - Cation exchange capacities for the two soil layers [meg/kg]. Taken
from appropriately weighted measurements.
C - Sulfate adsorption parameters (C — half-saturation constant
•>
[meq/m ]), for the two soil layers. Taken from appropriately
weighted measurements.
D - Average depth of the' two soil layers [m]. Taken from appropriately
weighted measurements.
BD - Bulk density of the two soils [kg/m3]. Taken from appropriately
weighted measurements or the literature.
P - Porosity of the two soils [fraction]. Taken from measurements or
literature.
- Equilibrium constant for Al(OH)3 solubility in the stream.
Calculated from observed pH-Al3+ relationships.
-------
17
K_» - Lumped equilibrium constant for the solubility of A1(OH>2 in each
soil layer.
OF* - Fraction of deposition (or snow melt) that bypasses the soils and
enters surface waters directly [fraction]. Derived from the
calibration of TOPMODEL.
PMAC* - Fraction of deposition (or snowmelt) that bypasses the upper soil
layer and enters the lower soil layer directly [fraction]. Derived
from the calibration of TOPMODEL.
IF* - Fraction of water leaving the upper soil layer and flowing directly
to the surface waters [fraction}. Derived from the calibration of
TOPMODEL.
K.-... - Thermodynamic equilibrium constants for the yy aqueous chemical
reactions that occur in the soil and surface waters (see Cosby et
al., 1985b). Derived from the literature.
Adjustable parameters:
St* - Scaling parameter that relates PCO_ in the two soil layers and the
surface waters to the temperature [atm/deg].
W^ (xx - Ca, Mg, Na, K) - Mineral weathering inputs of base cations for
the soil layers [meq/m2/yr].
SA1*XX ^xx " Ca» **g, Na- K) " Selectivity coefficients for exchange of
Al and base cations on the two soil types.
Emx - Maximum sulfate adsorption capacity for each soil (meq/kg).
Outputs:
The output variables of MAGIC are summarized in Cosby et al. (1985b).
-------
TABLE D.I
UOQDS - CALIBRATION PERIOD (UNITS OF USE ARE (M3/S)2 )
GRADIENTS
PARAMETER DEL |