The Hydrogeologic Framework of the Roswell Groundwater Basin, Chaves, Eddy, Lincoln, and Otero Counties, New Mexico

The hydrogeologic framework of the Roswell groundwater basin, Chaves, Eddy, Lincoln, and Otero Counties, New Mexico

Item Type Dissertation-Reproduction (electronic); text

Authors Havenor, Kay Charles, 1931-

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Download date 23/09/2021 15:36:08

Link to Item http://hdl.handle.net/10150/191196 THE HYDROGEOLOGIC FRAMEWORK OF THE ROSWELL

GROUNDWATER BASIN, CHAVES, EDDY,

LINCOLN, AND 01 BRO COUNTIES, NEW MEXICO

Kay Charles Havenor

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF GEOSCEENCES

In Partial Fulfillment of the Requirements For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

1996 2

THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE

As members of the Final Examination Committee, we certify that we have read the dissertation prepared by Kay Charles Havenor entitled The hydrogeological framework of the Roswell groundwater basin,

Chaves, Eddy, Lincoln, and Otero Counties, New Mexico

and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of Doctor of Philosophy

April 1, 1996 Date

aittX4/1" April 1, 1996 Date

April 1, 1996 Date

April 1, 1996 Date April 1, 1996 Charles Glass Date

Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requir t.

April 1, 1996

Dissertation Director Date Spencer FL Titley 3

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of the requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgement of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgement the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED: 4

ACKNOWLEDGEMENTS

My wife, Deborah, was primarily responsible for motivating and implementing my return to the University of Arizona to pursue this research. Without her encouragement and support it would not have been feasible. The Department of Geoscience provided a Teaching Fellowship and arranged for a BP Graduate Scholarship to make it possible.

Kevin Horstman, a fellow graduate student, was of immeasurable help with both data acquisition and advise on image processing. This project would have been exceed- ingly difficult without his generous assistance. The Bureau of Land Management, Roswell, New Mexico, assisted in obtaining publications and map data. The United States Geological Survey, Tucson, Arizona, obtained the digital TM image data, and Dr. Douglas McAda of the United States Geological Survey, Albuquerque, New Mexico office provided the hydrochemical data.

The New Mexico Bureau of Mines and Mineral Resources, Director Dr. Charles Chapin, and Dr. John Hawley, my mentor, provided publications and encouragement. Eastern New Mexico University, Roswell Campus, and the Southeastern New Mexico Economic Development Council provided office space and a telephone during the research for and preparation of this dissertation. Mr. Jack Ahlen, Roswell, New Mexico, a fellow geoscientist, provided field and subsurface information, encouragement, and friendship as he always has for many years.

My professors and advisors and the Geoscience staff at the University of Arizona, particularly Dr. Spencer R. Titley and Dr. Jay Quade, provided insight and solutions to both scientific and procedural problems. Each extended their kindnesses well beyond the requirements of their positions. Special thanks are due to Dr. Judith Totman Parrish for her assistance in editing this manuscript. 5

TABLE OF CONTENTS

LIST OF ILLUSTRATIONS

LIST OF TABLES 17

ABSTRACT 21

INTRODUCTION 23 Purposes and scope 25 Methodology 30 Previous Investigations 32 Topography 33 Physiography 34 Drainage 37

STRATIGRAPHY 40 Introduction 40 Late Pennsylvanian and Wolfcampian Stage 40 Leonardian—Guadalupian Stages 44 Yeso Formation 46 Glorieta Sandstone 48 San Andres Formation 50 Artesia Group 55 Grayburg Formation 58 Queen Formation 64 Seven Rivers Formation 65 Yates Formation 66 Tansill Formation 67 Ochoan Stage 68 6

TABLE OF CONTENTS—Continued

Tertiary Formations 69 Quaternary Alluvial Deposits 71 Aquifer Stratigraphy 75 Permian Aquifers 75 Quaternary Aquifer 81

REGIONAL TECTONICS 84 Introduction 84 Early and Middle Paleozoic Tectonics 89 Late Paleozoic Tectonics 90 Marathon Uplift 92 Summary of Late Paleozoic Tectonics 95 Mesozoic Tectonics 97 Cenozoic Tectonics 97 Wrench Tectonics 103 Capitan Lineament 107 Cenozoic Tectonics Summary 109

HYDROCHEMISTRY AND HYDROGEOCHEMISTRY 112 Introduction 112 Methodology 115 Stage One: Data Accumulation 115 Stage Two: Data Analysis Using Wateq 116 Stage Three: Hydrochemical Plots 117 Ternary diagrams 118 Piper diagrams 119 Fingerprint diagrams 119 Stability diagrams 120 7

TABLE OF CONTENTS—Continued

Hydrochemical Plots 121 San Andres Formation Hydrochetnical Signatures 123 Artesia Group Undifferentiated Hydrochemical Signatures 132 Quaternary Alluvium Hydrochemical Signatures 144 Santa Rosa Sandstone Fingerprint Signature 147 Stage Four: Interpretation of the Data 149 Hydrogeo chemistry 155 Introduction 155 Reconstruction of minerals contributing ions 158 General methodology of mineral reconstruction 158 San Andres Formation Signature 162 Artesia Group Signature 169 Quaternary Alluvial Aquifer Signature 178 Summary of Normative Mineral Reconstructions 181 Mineral Stability Diagrams 188

Ca02-Si02 Diagrams 196

CONCLUSIONS 199 The Hydrogeologic Framework 199 Recommendations for Future Work 203

REFERENCES CITED 205

APPENDIX A 218 Well and Spring Analysis Data 218 Analysis Quantities 218 Geologic Formation Codes 218 Well Numbering System 219 8

TABLE OF CONTENTS—Continued

Map Scale 220 Wells List 221 Springs List 225

APPENDIX B 228 Hydrochemical Plots 228

APPENDIX C 260 Normative Mineral Tables 260 9

LIST OF ILLUSTRATIONS

Figure 1.1. Location and general features from satellite imagery 26 Figure 1.2. Location map and outline of the groundwater basin 27 Figure 1.3. Highly generalized topography of study area 34 Figure 1.4. Diagrammatic section across Roswell groundwater basin 36 Figure 2.1. Regional paleogeographic setting of the Guadalupian Permian 41 Figure 2.2. Chronostratigraphic diagram Late Leonardian-Middle Guadalupian 45 Figure 2.3. Diagrammatic lower San Andres depositional environment 51 Figure 2.4. Surface and sub-Quaternary distribution of Artesia Group 54 Figure 2.5. Distribution of Artesia Group formations along the Pecos River 57 Figure 2.6. Location of major dextral strike-slip faults and respective structural blocks 62 Figure 2.7. Landsat TM image of City of Roswell, New Mexico 63 Figure 2.8. Generalized distribution of Quaternary deposits and western boundary of the shallow Quaternary unconfined aquifer 73 Figure 2.9. Generalized western margin of the confined artesian carbonate aquifer 74 Figure 3.1. Location of Capitan and Sierra Blanca Mountains 85 Figure 3.2. Regional paleogeographic setting of the Guadalupian Permian 86 Figure 3.3. Geologic map of Pajarito Mountain area 86 Figure 3.4. Bouguer gravity map of the Sacramento Mountains 87 Figure 3.5. Map of the Rio Grande Rift 91 Figure 3.6. Map legend for Figure 3.5 91 Figure 3.7. Salt Basin graben and structures of Roswell groundwater basin 93 Figure 3.8. Paleogeographic reconstructions 94 Figure 3.9. Location of major dextral strike-slip faults and respective structural blocks 98 Figure 3.10. The Tinnie-Dunken fold trend 100 Figure 3.11. K-M fault shown on Landsat TM image 102 1 0

LIST OF ILLUSTRATIONS—Continued

Figure 3.12. Thickness change of erosion-beveled San Andres Formation due to juxtaposition across strike-slip fault 104 Figure 3.13. Trend of the Capitan Lineament in New Mexico 107 Figure 4.1. Location of T12S-R24E samples 123 Figure 4.2. Ternary plot of 5 samples T12S-R24E 124 Figure 4.3. Piper diagram of 5 samples from T12S-R24E 127 Figure 4.4. Fingerprint diagram of 5 samples from T12S-R24E 128 Figure 4.5. Fingerprint diagram of 2 samples from T11S-R23E 129 Figure 4.6. Fingerprint overlay of Fig. 4.4 and 4.5 129

Figure 4.7. Location of T11S -R23E samples 129

Figure 4.8. Fingerprint diagram of 1 spring sample from Section 28, T11S-R14E 130

Figure 4.9. Fingerprint diagram of 1 spring sample from Section 27, T10S-R16E 130

Figure 4.10. Fingerprint diagram of 1 spring sample from Section 18, T14S-R14E 130 Figure 4.11. Fingerprint diagram overlay of spring waters 132 Figure 4.12. Location of T10S-R25E samples 133 Figure 4.13. Ternary plot of 6 samples, T10S-R25E 135 Figure 4.14. Piper diagram of 6 samples from T10S-R25E 135 Figure 4.15. Fingerprint diagram of 6 samples from T10S-R25E 137 Figure 4.16. Location of TIOS-R24E samples 137 Figure 4.17. Fingerprint diagram of 6 samples from T10S-R24E 138 Figure 4.18. Location of T9S-R24E samples 139 Figure 4.19. Fingerprint diagram of 9 samples from T9S-R24E 139 Figure 4.20. Location of S/2 T8S-R24E samples 140 Figure 4.21. Fingerprint diagram of 8 samples from S/2 T8S-R24E 140 Figure 4.22. Location of N/2 T8S-R24E samples 141 Figure 4.23. Fingerprint diagram of 7 samples from N/2 T8S-R24E 142

Figure 4.24. Ternary plot of 7 samples, N/2 T8S-R24E 143

Figure 4.25. Fingerprint diagram of 1 sample from Section 31, T6S-R24E 143 11

LIST OF ILLUSTRATIONS—Continued

Figure 4.26. Ternary plot of 4 samples, T11S-R25E 145 Figure 4.27. Fingerprint diagram of 4 samples from T11S-R25E 146 Figure 4.28. Location of T 1 0S-R25E samples 146 Figure 4.29. Fingerprint diagram of 3 samples from T10S-R25E 146 Figure 4.30. Location of Section 22, T1S-R27E spring sample 147 Figure 4.31. Piper diagram of 1 spring sample from Section 22, T1S-R27E 148 Figure 4.32. Fingerprint diagram of 1 spring sample from Section 22, T1S-R27E 149 Figure 4.33. San Andres carbonate aquifer Fingerprint diagram signature 149 Figure 4.34. Artesia Group aquifer Fingerprint diagram signature 150 Figure 4.35. Quaternary alluvial aquifer Fingerprint diagram signature 150 Figure 4.36. Ternary plot of three aquifer host water sample sets illustrating anion grouping by source 150 Figure 4.37. Ternary plot of three aquifer host water sample sets illustrating cation grouping by source 151 Figure 4.38. Distrubution of San Andres Formation and Artesia Group Fingerprint diagram water signatures 153 Figure 4.39. Normative mineral chart Section 27.21333, T12S-R24E 163 Figure 4.40. Normative mineral chart Section 22.41333, T12S-R24E 164 Figure 4.41. Normative mineral chart Section 22.41333, T12S-R24E 165 Figure 4.42. Normative mineral chart Section 22.23111, T12S-R24E 166 Figure 4.43. Normative mineral chart Section 15.43111, T12S-R24E 168 Figure 4.44. Normative mineral chart Section 31.34, T10S-R24E 169 Figure 4.45. Normative mineral chart Section 33.423, T10S-R25E 170 Figure 4.46. Normative mineral chart Section 32.424333, T10S-R25E 172 Figure 4.47. Normative mineral chart Section 29.443333, Tl 0S-R25E 174 Figure 4.48. Normative mineral chart Section 14.312, T10S-R25E 175

Figure 4.49. Normative mineral chart Section 5.3, T10S -R25E 177 Figure 4.50. Normative minerai chart Section 36.144, T11S-R25E 178 12

LIST OF ILLUSTRATIONS—Continued

Figure 4.51. Normative mineral chart Section 36.14234, T11S-R25E 179 Figure 4.52. Normative mineral chart Section 36.14234A, Tl 1S-R25E 180 Figure 4.53. Correlation using normative minerals charts 181 Figure 4.54. Areas of Na+ excess in the Roswell groundwater basin 189 Figure 4.55. Mineral stability plot for samples from T12S-R24E 196 Figure 4.56. Mineral stability diagram for samples for T10S-R25E 197 Figure 5.1. Well numbering system diagram 219 Figure 6.1. Location of T12S-R24E samples 229 Figure 6.2. Ternary plot of 5 samples T12S-R24E 229 Figure 6.3. Piper diagram of 5 samples from T12S-R24E 229 Figure 6.4. Fingerprint diagram of 5 samples from T12S-R24E 229 Figure 6.5. Location of T11S-R23E samples 230 Figure 6.6. Ternary plot of 2 samples, T11S-R23E 230 Figure 6.7. Piper diagram of 2 samples from T11S-R23E 230 Figure 6.8. Fingerprint diagram of 2 samples from TlIS-R23E 230 Figure 6.9. Location of T11 S -R22E samples 231 Figure 6.10. Ternary plot of 3 samples, T11S-R22E 231 Figure 6.11. Piper diagram of 3 samples from T11S-R22E 231 Figure 6.12. Fingerprint diagram of 3 samples from Tl 1S-R22E 231 Figure 6.13. Location of T10S-R25E samples 232 Figure 6.14. Ternary plot of 6 samples, T10S-R25E 232 Figure 6.15. Piper diagram of 6 samples from T10S-R25E 232 Figure 6.16. Fingerprint diagram of 6 samples from T10S-R25E 232 Figure 6.17. Location of T10S-R24E samples 233 Figure 6.18. Ternary plot of 6 samples, T10S-R24E 233 Figure 6.19. Piper diagram of 6 samples from T10S-R24E 233 Figure 6.20. Fingerprint diagram of 6 samples from T10S-R24E 233 Figure 6.21. Location of T10S-R23E samples 234 13

LIST OF ILLUSTRATIONS—Continued

Figure 6.22. Ternary plot of 4 samples T10S-R23E 234 Figure 6.23. Piper diagram of 2 nearby wells, section 34, TIOS-R23E 234 Figure 6.24. Fingerprint diagram of 2 nearby wells, section 34, T10S-R23E 234 Figure 6.25. Location of T9S-R24E samples 235 Figure 6.26. Ternary plot of 9 samples, T9S-R24E 235 Figure 6.27. Piper diagram of 9 samples from T9S-R24E 235 Figure 6.28. Fingerprint diagram of 9 samples from T9S-R24E 235 Figure 6.29. Location of T9S-R23E samples 236 Figure 6.30. Ternary plot of 3 samples, T9S-R23E 236 Figure 6.31. Piper diagram of 3 samples from T9S-R23E 236 Figure 6.32. Fingerprint diagram of 3 samples, T9S-R23E 236 Figure 6.33. Location of Section 31, T8S-R25E sample 237 Figure 6.34. Ternary plot of 1 sample, T8S-R25E 237 Figure 6.35. Piper diagram of sample from Section 31, T8S-R25E 237 Figure 6.36. Fingerprint diagram of sample from Section 31, T8S-R25E 237 Figure 6.37. Location of S/2 T8S-R24E samples 238 Figure 6.38. Ternary plot of 8 samples, S/2 T8S-R24E 238 Figure 6.39. Piper diagram of 8 samples from 5/2 T8S-R24E 238 Figure 6.40. Fingerprint diagram of 8 samples from S/2 T8S-R24E 238 Figure 6.41. Location of N/2 T8S-R24E samples 239 Figure 6.42. Ternary plot of 7 samples, N/2 T8S-R24E 239 Figure 6.43. Piper diagram of 7 samples from N/2 T8S-R24E 239 Figure 6.44. Fingerprint diagram of 7 samples from N/2 T8S-R24E 239 Figure 6.45. Location of T8S-R23E samples 240 Figure 6.46. Ternary plot of 2 samples, T8S-R23E 240 Figure 6.47. Piper diagram of 2 samples from T8S-R23E 240

Figure 6.48. Fingerprint diagram of 2 samples from T8S - R23E 240 Figure 6.49. Location of T7S-R26E samples 241 14

LIST OF ILLUSTRATIONS—Continued

Figure 6.50. Ternary plot of 3 samples, T7S-R26E 241 Figure 6.51. Piper diagram of 3 samples from T7S-R26E 241 Figure 6.52. Fingerprint diagram of 3 samples from T7S-R26E 241 Figure 6.53. Location of Section 32, T7S-R25E sample 242 Figure 6.54. Ternary plot of 1 sample, Section 32, T7S-R25E 242 Figure 6.55. Piper diagram of 1 sample from Section 32, T7S-R25E 242 Figure 6.56. Fingerprint diagram of 1 sample from Section 32, T7S-R25E 242 Figure 6.57. Location of Section 19, T7S-R24E sample 243 Figure 6.58. Ternary plot of 1 sample, Section 19, T7S-R24E 243 Figure 6.59. Piper diagram of 1 sample from Section 19, T7S-R24E 243 Figure 6.60. Fingerprint diagram of 1 sample from Section 19, T7S-R24E 243 Figure 6.61. Location of Section 23, T7S-R23E sample 244 Figure 6.62. Ternary plot of 1 sample, Section 23, T7S-R23E 244 Figure 6.63. Piper diagram of 1 sample from Section 23, T7S-R23E 244 Figure 6.64. Fingerprint diagram of sample from Section 23, T7S-R23E 244 Figure 6.65. Location of Section 26, T7S-R22E sample 245 Figure 6.66. Ternary plot of 1 sample, Section 26, T7S-R22E 245 Figure 6.67. Piper diagram of 1 sample from Section 26, T7S-R22E 245 Figure 6.68. Fingerprint diagram of 1 sample from Section 26, T7SR22E 245 Figure 6.69. Location of Section 31, T6S-R24E sample 246 Figure 6.70. Ternary plot of 1 sample, Section 31, T6S-R24E 246 Figure 6.71. Piper diagram of 1 sample from Section 31, T6S-R24E 246 Figure 6.72. Fingerprint diagram of 1 sample from Section 31, T6S-R24E 246 Figure 6.73. Location of Section 31, T5S-R2OE sample 247 Figure 6.74. Ternary plot of 1 sample, Section 31, T5S-R2OE 247 Figure 6.75. Piper diagram of 1 sample from Section 31, T5S-R2OE 247

Figure 6.76. Fingerprint diagram of 1 sample from Section 31, T5S -R2OE 247 Figure 6.77. Location of Section 13, Ti S-R24E sample 248 15

LIST OF ILLUSTRATIONS—Continued

Figure 6.78. Ternary plot of 1 sample, Section 13, T I S-R24E 248 Figure 6.79. Piper diagram of 1 sample from Section 13, T1S-R24E 248 Figure 6.80. Fingerprint diagram of 1 sample from Section 13, T1S-R24E 248 Figure 6.81. Location of Section 12, T1S-R23E sample 249 Figure 6.82. Ternary plot of 1 sample, Section 12, T1S-R23E 249 Figure 6.83. Piper diagram of 1 sample from Section 12, T1S-R23E 249 Figure 6.84. Fingerprint diagram of 1 sample from Section 12, T1S-R23E 249 Figure 6.85. Location of Section 18, T14S-R14E spring sample 250 Figure 6.86. Ternary plot of 1 spring sample, Section 18, T14S-R14E 250 Figure 6.87. Piper diagram of 1 spring sample from Section 18, T14S-R14E 250 Figure 6.88. Fingerprint diagram of 1 spring sample from Section 18, T14S-R14E 250 Figure 6.89. Location of Secton 28, T12S-R14E spring sample 251 Figure 6.90. Ternary plot of 1 spring sample, Section 28, T12S-R14E 251 Figure 6.91. Piper diagram of 1 spring sample from Section 28, T12S-R14E 251 Figure 6.92. Fingerprint diagram of 1 spring sample from Section 28, T12S-R14E 251 Figure 6.93. Location of Section 3, T12S-R13E sprimg sample 252 Figure 6.94. Ternary plot of 1 spring sample, Section 3, T12S-R13E 252 Figure 6.95. Piper diagram of 1 spring sample from Section 3, T12S-R13E 252 Figure 6.96. Fingerprint diagram of 1 spring sample from Section 3, T12S-R13E 252 Figure 6.97. Location of Section 28, T11S-R14E spring sample 253 Figure 6.98. Ternary plot of 1 spring sample, Section 28, T11SR14E 253 Figure 6.99. Piper diagram of 1 spring sample from Section 28, T11S-R14E 253 Figure 6.100. Fingerprint diagram of 1 spring sample, Section 28, T11S-R14E 253 Figure 6.101. Location of Section 27, T10S-R16E spring sample 254 Figure 6.102. Ternary plot of 1 spring sample, Section 27, T10S-R16E 254 Figure 6.103. Piper diagram of 1 spring sample from Section 27, T10S-R16E 254

Figure 6.104. Fingerprint diagram of 1 spring sample, Section 27, T10S -R16E 254 Figure 6.105. Location of Section 5, T11S-R25E spring sample 255 16

LIST OF ILLUSTRATIONS—Continued

Figure 6.106. Ternary plot of 1 spring sample, Section 5, T11S-R25E 255 Figure 6.107. Piper diagram of 1 spring sample from Section 5, T11S-R25E 255 Figure 6.108. Fingerprint diagram of 1 spring sample from Section 5, T11S-R25E . 255 Figure 6.109. Location of Section 3, T11S-R26E spring sample 256 Figure 6.110 Ternary plot of 1 spring sample, Section 2, T11S-R26E 256 Figure 6.111. Piper diagram of 1 spring sample from Section 2, T11S-R26E 256 Figure 6.112. Fingerprint diagram of 1 spring sample from Section 2, T11S-R26E . 256 Figure 6.113. Location of Section 22, T1S-R27E spring sample 257 Figure 6.114. Ternary plot of 1 spring sample, Section 22, T1S-R27E 257 Figure 6.115. Piper diagram of 1 spring sample from Section 22, T1S-R27E 257 Figure 6.116. Fingerprint diagram of 1 spring sample from Section 22, T1S-R27E . 257 Figure 6.117. Location of T11S-R25E samples 258 Figure 6.118. Ternary plot of 4 samples, T11S-R25E 258 Figure 6.119. Piper diagram of 4 samples from T11S-R25E 258 Figure 6.120. Fingerprint diagram of 4 samples from T11S-R25E 258 Figure 6.121. Location of T10S-R25E Quaternary Alluvium well samples 259 Figure 6.122. Ternary plot of 3 samples, T10S-R25E 259 Figure 6.123. Piper diagram of 3 samples from T10S-R25E 259 Figure 6.124. Fingerprint diagram of 3 samples from T10S-R25E 259 17

LIST OF TABLES

Table 2.1. Generalized Upper Permian stratigraphic correlation chart 56 Table 2.1. Precambrian rocks in various oil-gas test wells 88 Table 4.1. Cation concentrations in T12S-R24E samples 124 Table 4.2. Anion concentrations in T12S-R24E samples 125 Table 4.3. Cation concentrations in TIOS-R25E samples 134 Table 4.4. Anion concentrations in TIOS-R25E samples 134 Table 4.5. Cation concentrations in Section 36, T11S-R25E samples 145 Table 4.6. Anion concentrations in Section 36, T11S-R25E samples 145 Table 4.7. Cation composition of T1S-R27E sample 148 Table 4.8. Anion composition of T1S-R27E sample 148 Table 4.9. Mineral reconstruction for Section 27.21333, T12S-R24E water sample 162 Table 4.10. Weight % of minerals in Section 27.21333, T12S-R24E 162 Table 4.11. Weight % of minerals from Section 22.41333, T12S-R24E waters 164 Table 4.12. Weight % of minerals from Section 22.41333, T12S-R24E waters 165 Table 4.13. Weight % of minerals from Section 22.23111, T12S-R24E waters 166 Table 4.14. Weight % of minerals from Section 15.43111, T12S-R24E waters 167 Table 4.15. Weight % of mineral from Section 31.34, T10S-R25E sample 169 Table 4.16. Weight % of minerals from Section 33.423, T10S-R25E 170 Table 4.17. Weight % of minerals from Section 32.424333, T10S-R25E 171 Table 4.18. Weight % of minerals from Section 29.443333, T10S-R25E 174 Table 4.19. Weight % of minerals from Section 14.312, T10S-R25E 175 Table 4.20. Weight % of minerals from Section 5.3, T10S-R25E 177 Table 4.21. Weight % of minerals from Section 36.144, T11S-R25E waters 178 Table 4.22. Weight % minerals from Section 36.14234, T11S-R25E 179 Table 4.23. Weight % of minerals from Section 36.14234A, T11S-R25E 180 Table 4.24. Mineral reconstruction weight % calculation data 183 Table 4.25. Mineral reconstruction remaining dissolved ions in gmol/L 185 Table 4.26. Average composition of water in streams and in the oceans 191 18

LIST OF TABLES—Continued

Table 4.27. Chemical composition of meteoric precipitation 192 Table 5.1 Geologic formation code for water analyses 219 Table 7.1. Weight % of minerals in Section 27.21333, T12S-R24E 261 Table 7.2. Weight % of minerals from Section 22.41333, T12S-R24E 261 Table 7.3. Weight % of minerals from Section 22.41333, T12S-R24E 261 Table 7.4. Weight % of minerals from Section 22.23111, T12S-R24E 261 Table 7.5. Weight % of minerals from Section 15.43111, T12S-R24E 261 Table 7.6. Weight % of minerals from Section 8.232222, T11S-R23E 261 Table 7.7. Weight % of minerals from Section 25.331142, T11S-R22E 262 Table 7.8. Weight % of minerals from Section 22.111, T11S-R22E 262 Table 7.9. Weight % of minerals from Section 2.1, T11S-R22E 262 Table 7.10. Weight % of minerals from Section 31.34, T10S-R25E 262 Table 7.11. Weight % of minerals from Section 33.423, T10S-R25E 262 Table 7.12. Weight % of minerals from Section 32.42333, T10S-R25E 262 Table 7.13. Weight % of minerals from Section 29.443333, T10S-R25E 263 Table 7.14. Weight % of minerals from Section 14.312, T10S-R25E 263 Table 7.15. Weight % of minerals from Section 5.3, T I 0S-R25E 263 Table 7.16. Weight % of minerals from Section 35.443, T10S-R24E 263 Table 7.17. Weight % of minerals from Section 24.314, T10S-R24E 263 Table 7.18. Weight % of minerals from Section 24.333, T10S-R24E 263 Table 7.19. Weight % of minerals from Section 15.43, TIOS-R24E 264 Table 7.20. Weight % of minerals from Section 15.431, T10S-R24E 264 Table 7.21. Weight % of minerals from Section 8.333, T10S-R24E 264 Table 7.22. Weight % of minerals form Section 34.432, T10S-R23E 264 Table 7.23. Weight % of minerals form Section 34.432, T10S-R23E 264 Table 7.24. Weight % of minerals form Section 34.432A, T10S-R23E 264 Table 7.25. Weight % of minerals form Section 34.432A, T10S-R23E 265 Table 7.26. Weight % of minerals form Section 32.233324, T9S-R24E 265 19

LIST OF TABLES—Continued

Table 7.27. Weight % of minerals form Section 20.413, T9S-R24E 265 Table 7.28. Weight % of minerals form Section 14.121, T9S-R24E 265 Table 7.29. Weight % of minerals form Section 11.3, T9S-R24E 265 Table 7.30. Weight % of minerals form Section 11.133, T9S-R24E 265 Table 7.31. Weight % of minerals form Section 11.141, T9S-R24E 266 Table 7.32. Weight % of minerals form Section 2.42312, T9S-R24E 266 Table 7.33. Weight °A of minerals form Section 5.314, T9S-R24E 266 Table 7.34. Weight % of minerals form Section 5.134, T9S-R24E 266 Table 7.35. Weight % of minerals from Section 35.13, T9S-R23E 266 Table 7.36. Weight % of minerals from Section 36.133111, T9S-R23E 266 Table 7.37. Weight % of minerals from Section 20.14, T9S-R23E 267 Table 7.38. Weight % of minerals from Section 31.131, T8S-R25E 267 Table 7.39. Weight % of minerals from Section 32.411, T8S-R24E 267 Table 7.40. Weight % of minerals from Section 31.211, T8S-R24E 267 Table 7.41. Weight % of minerals from Section 27.433431, T8S-R24E 267 Table 7.42. Weight % of minerals from Section 28.413321, T8S-R24E 267 Table 7.43. Weight % of minerals from Section 29.414, T8S-R24E 268 Table 7.44. Weight % of minerals from Section 20.233113, T8S-R24E 268 Table 7.45. Weight % of minerals from Section 22.142113, T8S-R24E 268 Table 7.46. Weight % of minerals from Section 18.233, T8S-R24E 268 Table 7.47. Weight % of minerals from Section 17.143, T8S-R24E 268 Table 7.48. Weight % of minerals from Section 18.241, T8S-R24E 268 Table 7.49. Weight % of minerals from Section 15.111, T8S-R24E 269 Table 7.50. Weight % of minerals from Section 9.311, T8S-R24E 269 Table 7.51. Weight % of minerals from Section 8.413, T8S-R24E 269 Table 7.52. Weight % of minerals from Section 7.423, T8S-R24E 269 Table 7.53. Weight % of minerals from Section 5.143, T8S-R24E 269 Table 7.54. Weight % of minerals from Section 33, T8S-R23E 269 20

LIST OF TABLES—Continued

Table 7.55. Weight % of minerals from Section 1.322, T8S-R23E 270 Table 7.56. Weight % of minerals from Section 30.431, T7S-R26E 270 Table 7.57. Weight % of minerals from Section 19.243, T7S-R26E 270 Table 7.58. Weight % of minerals from Section 6.242, T7S-R26E 270 Table 7.59. Weight % of minerals from Section 32.432, T7S-R25E 270 Table 7.60. Weight % of minerals from Section 19.1, T7S-R24E 270 Table 7.61. Weight % of minerals from Section 23.243, T7S-R23E 271 Table 7.62. Weight % of minerals from Section 26.311, T7S-R22E 271 Table 7.63. Weight % of minerals from Section 31.43343, T6S-R24E 271 Table 7.64. Weight % of minerals from Section 31.34, T5S-R2OE 271 Table 7.65. Weight % of minerals from Section 13, T1S-R24E 271 Table 7.66. Weight % of minerals from Section 12.22133, T1S-R23E 271 Table 7.67. Weight % of minerals from Section 18.213, T14S-R14E 272 Table 7.68. Weight % of minerals from Section 28.432, T12S-R14E 272 Table 7.69. Weight % of minerals from Section 3.121, T12S-R13E 272 Table 7.70. Weight % of minerals from Section 28.321, T11S-R14E 272 Table 7.71. Weight % of minerals from Section 27, T10S-R16E 272 Table 7.72. Weight % of minerals from Section 5.4, T11S-R25E 272 Table 7.73. Weight % of minerals from Section 2.442, T11S-R26E spring sample . 273 Table 7.74. Weight % of minerals from Section 22.333, T1S-R27E spring sample . 273 Table 7.75. Weight % of minerals from Section 36.144, T11S-R25E 273 Table 7.76. Weight % of minerals from Section 36.14234, T11S-R25E 273 Table 7.77. Weight % of minerals from Section 36.14234A, T11S-R25E 273 Table 7.78. Weight % of minerals from Section 36.213, T11S-R25E 273 Table 7.79. Weight % of minerals from Section 32.42, T10S-R25E 274 Table 7.80. Weight % of minerals from Section 32.23, T10S-R25E 274 Table 7.81. Weight % of minerals from Section 29.44, T10S-R25E 274 21

ABSTRACT

Aquifers of the Roswell groundwater basin are unconfined and confined types in

Permian San Andres Formation and Artesia Group carbonates and evaporites, and the shallow unconfined Quaternary sedimentary and alluvial aquifer. The carbonate-evaporite aquifers were developed from solution by meteoric water, groundwater, the Pecos

River, and its tributaries.

The structural geology of the region includes Cenozoic folding and wrench faulting. Regional dextral strike-slip faults, <30 Ma to as young as 0.5 Ma, dominate the hydrogeologic framework of the groundwater basin. The faults created major lithologic and structural boundaries for the groundwater systems developed between them. The

Roswell groundwater "basin" is actually a series of en echelon structural blocks with aquifers developed in erosion-beveled, fault-displaced Permian carbonates and evaporites partly covered by Quaternary sedimentary rocks and alluvium. The confined portions of the carbonate aquifers are in the San Andres Formation, the Artesia Group, or a solutional-karstic melange of the two. The Permian aquifers developed within each structural block exhibit different hydrochemical and hydrologic properties.

The rock groups produce distinctive bulk element water chemistry signatures which are readily visible on ternary plots, Piper diagrams, and Fingerprint diagrams. San

Andres Formation waters have high HCO3-, intermediate SO 42 , and low Cl- that demon- strate a preponderance of carbonates with some evaporites. Waters hosted by the Artesia 22

Group are characterized by low HCO 3-, high SO 42-, and high Cl- that reflect evaporites with some carbonates. Quaternary alluvial aquifer waters show low Ca2+, low HCO3-, with moderately high SO 42- and Cr.

Normative mineral reconstructions identify the lithologic combinations through which the waters flowed to acquire their present chemical characteristics. Plotted as

charts the normative mineral reconstructions can be correlated as are electric well logs.

Mineral stability diagrams support exchange by sodium liberation and calcium replace-

ment in Na-smectite marine clays for altering the ce - Na groundwater chemistry.

Hydrochemical plots provide a robust means of identifying aquifer sources of groundwater and delineating their structural and stratigraphic boundaries. The work

should be expanded to include more water analyses from each group, and as a means to identify unknows, such as the sources of water to the Pecos River. 23

INTRODUCTION

The hydrogeologic framework of a groundwater basin strongly controls the quality and quantity of its water as well as the water's direction and rates of movement.

The hydrogeologic framework is the lithology, stratigraphy, and structure of the area containing the aquifers. Those components control virtually every hydrochemical and hydrogeochemical characteristic developed by the water after it falls as precipitation.

Carbonate aquifers are an important source of groundwater. It is general knowledge that water moving through carbonate aquifers becomes hard-water, but there has been little study of the geological controls on the water flow path and movement in such aquifers. Carbonate aquifers often develop excellent porosity and permeability, characteristics controlled by lithology, structure, geological history, and precipitation distribution. Unfortunately, the high porosity and permeability found in carbonates not only provides good groundwater resources, it also exposes these important aquifers to a greatly increased potential for natural and anthropogenic pollution.

Conventional hydrologic evaluation of areas dominated by carbonate lithologies does not usually provide important geological information necessary for protection of the 24 groundwater from antiropogenic contamination. Conventional hydrologic data alone do

not commonly provide all the data required for effective remediation. Conventional

hydrology, by necessity, makes the assumption that aquifer lithologies are homogeneous

and isotropic— especially carbonate aquifers. Although the conventional assumptions

are often considered acceptable for groundwater-use management, unfortunately they are

not realistic, for example, in determining:

a) changes in hydrochemistry controlled by geologic factors;

b) the delineation of preferential flow directions controlled by structural

deformation, its consequential structural boundaries and stratigraphic

variations;

c) regional estimates of groundwater velocities;

d) discrimination of groundwater recharge source areas.

Integrating the geology of the aquifers and of the different recharge areas with the

hydrological characteristics of the individual aquifers is necessary to obtain a more

thorough understanding of some or all of the above issues. I frequently point out:

Water falls as precipitation upon the rocks of the surface; what is not evaporated or transpired runs across the soil and the surface rocks; it is absorbed into the pores of the rocks; it travels through the rocks and it is produced from the rocks. The rocks control the quantity, quality, location and movement of the water.

The hydrogeologic framework describes the structural and stratigraphic nature of the

rocks and how the rocks exert their control upon the waters that flow across and within them. 25

Purposes and scope

The purposes of this investigation are to define the Roswell groundwater basin in a geological context based upon its rocks, geologic history, boundaries and physical limits. In order to do that, consideration was given to the following aspects:

1) the regional tectonics and structure of the Roswell groundwater basin;

2) the regional stratigraphy of the aquifers of the Roswell groundwater basin;

3) the use of hydrochemical or hydrogeochemical techniques to discriminate

among aquifers, storage areas, and recharge areas of the groundwater basin;

4) the use of hydrochemisty and hydrogeochemistry as geological tools to help

delineate structural and/or stratigraphic boundaries in major portions of a

groundwater basin.

5) an examination of the potential for hydrogeochemical discrimination of water-

rock reactions and the normative reconstruction of the minerals dissolved.

The greater Roswell basin area is shown in Figure 1.1. The study area covers approximately 42,000 km2. Two principal aquifers, the shallow alluvial aquifer and the confined artesian aquifer, cover 11,088 km2, slightly more than one-quarter of the entire study area. The size of the recharge area is of the greater Roswell groundwater basin is about five times that of the underground area of the aquifers. While the areal scope of the project was very large, the objectives were believed to be regionally discernable and determinable. 26

Figure 1.1, Location and general features from satellite imagery of the greater Roswell groundwater basin area in parts of Chaves, Eddy, De Baca, Guadalupe, Lincoln, and Otero counties, New Mexico. Modified from McAda and Morrison, 1993. 27

o 106 105 104

Saner I 1-1 bal. r T3I4 Fji L j TN TO RRANCE I_ 111 Ito T114

T1S 11111111111601111- Ira T2S II 111011111=11 ill 111 lil 139 34 • ini Mifil iliga:1491A1P Inlet. T 4S ii LI COLN two II T5S III u rill 11111111171M111 El 168 n 110W"'-' • I .AINIIIIIImr.A i MI T7S 1111 Il LIIIMIIII T8S NMI magaira CHAVES T9S NI „RNLI 111 Ill TI OS iiiimilEaker = ..2.111110111111• 711S Inggropplaii....,.." Nu T128 worm a • li Ian 7138 RIMINI URlf"iUU 1143 mg• ii."6, R. i1IiRR 33 •I E.FA , . Eitlillei* . 1111111111111PKIIPP 1168 Artesia Tmgm1 UNIR ,,,,„.,,as- 0 1- 178 R il oilis IUJr EDDY 7199 el. MEIN aillirlITC T19S OTERO 1UUU!IN=UU T203 T21S 114-1010P....!: 1229 I 1 1111 40 Kilometers Outline of Roswell zs IIll Groundwater basin and New Mexico recharge area Area Map

Figure 1.2. Location map and outline of the groundwater basin and recharge area. 28

An initial survey of the literature would suggest that a substantial amount of

geological work on the area of the Roswell groundwater basin has been previously

completed. There is still a serious deficiency in the integration of that geologic data into

an understanding of the effects of the surface-mapped structures and stratigraphy upon

the groundwaters, their movement, recharge, quantity, and their resulting hydrochemical

and hydrogeochemical characteristics. This investigation is addressed to studying those

interconnected relationships.

One realm in which the previously accepted hydrogeologic framework of the

Roswell groundwater basin is almost totally lacking meaningful information is in the

tectonic implications of, and the consequences of, the major strike-slip faults that cross the region. Not only is there a serious lack of geological data from the subsurface, there

have been no attempts to relate the chemistry of the groundwaters to the rocks through which they flow. The geological information that can be gleaned from the hydro-

chemistry and hydrogeochemistry of the groundwater is substantial. It will be demon-

strated in this work that the chemistry of the groundwaters can be used to relate the

stratigraphy of the aquifers to the geologic structural and stratigraphic framework of the groundwater basin and by that to a large portion of southeastern New Mexico.

Any attempt to relate the strike-slip faults of the Roswell groundwater basin to the aquifer system must deal with the problem that much information is seriously lacking.

The missing information regards the regional tectonics—the causes and timing—that might have caused the structures proposed in this work as complicating the normal juxtaposition of stratigraphic units of the aquifers under investigation. For this reason it 29 is believed important to devote part of this work to building the case for the causes of the major tectonic elements within and next to the Roswell groundwater basin.

The section on Regional Tectonics presents new arguments as to the age and nature of some major structural features. The hydrochemical and hydrogeochemical studies of the groundwaters then build upon the structural-stratigraphic presentations to show that the former can be used as subsurface geological tools for correlation and interpretation. In those ways, this work significantly affects, adds to, and sometimes challenges the previous work—including my own—as well as accepted concepts of the hydrogeologic framework of the groundwater basin. This work also will impact some ideas as to the age and style of the geological development of much of the greater

Roswell groundwater basin area and parts of southeastern New Mexico.

This work will show that the water of the Roswell groundwater basin is in large part controlled as to location, movement, and chemical nature by the presence of major strike-slip faults and the stratigraphic units placed in juxtaposition as a result of their Late

Tertiary to Quaternary movements. The faults involved include the K-M, Y-0, Six Mile

Hill, and Border Hill faults. The formations dominately involved in the distribution of the groundwater include the Permian Yeso Formation, San Andres Formation, Artesia

Group, and the Quaternary sedimentary and alluvial deposits. The locations of these major strike-slip faults are shown on Figure 2.4, page 54. It will be shown that the

Roswell groundwater basin is an en eschelon series of structural blocks, that each has individual hydrogeologic properties. 30

It will also be shown in this study that the hydrochemical and hydrogeochemical techniques employed are useful in identifying the formation in which the aquifer has developed. Those same techniques will be shown to be able to significantly contribute to the description of the hydrogeologic framework of the groundwater basin, the lithologies of the aquifers, how the structure controls the direction of recharge and flow, and the chemisty of the water at a give location.

Methodology

The large size of the study area led to the use of Landsat 5 Thermic MapperTm

(TM) digital imagery to help examine the hydrogeology of the project area. The Roswell image covers approximately 80 percent of the area shown in Figure 1.1 and 1.2.

Considering the area's expanse, the density of control points on the surface is low.

Chemical analyses, for example, average less than one per four hundred square kilometers. Geophysical well logs are about one order of magnitude fewer, less than approximately one per four thousand square kilometers. Additionally, the data are not uniformly distributed, but are locally concentrated.

The literature, as later cited, provided much information on the areal distribution of geological formations of the region. The long history of petroleum and natural gas exploration and development in the Permian basin of southeast New Mexico and West

Texas has generated a good understanding of the regional distribution of the rocks of

Paleozoic age. The literature also supplied graphical map information on the distribution of geological structures and fault zones. 31

Considerable personal knowledge about the geology of the region has been gained by me from living in Roswell, New Mexico, and working as a geoscientist in the

Permian basin and adjoining areas. The personal knowledge is the result of field work, subsurface and surface studies of both local and regional extent, and the frequent interchange of information with fellow geoscientists. The combination of hands-on geological work in the area for many years, plus professional contact, literature, and field work, often makes it difficult to cite accurately the original source from where certain knowledge was derived. I apologize for any slighting of fellow geoscientists by failure to reference or credit, but I accept full responsibility for all the statements.

Where possible, subsurface geophysical wells logs were utilized to understand subsurface lithologies and the distribution of porosity and permeability better. Structural data were modified by TM imagery. Aerial photographs were used in some areas for closer examination of features observed on the Landsat imagery. Hydrochemical and hydrogeochemical data were examined in an attempt to understand groundwater movement and quality distribution and to show the presence of lithologie and structural boundaries or controls—some of which can be seen on the satellite imagery.

Unpublished hydrologic data and maps were used to confirm the hydrogeological interpretations. Field examination and verification were made of many features and conditions. Obviously, the size of the area precluded field examination of all but the most significant features. 32

Previous Investigations

The previous geological work covering the Roswell groundwater basin will be

cited and their findings will be integrated into the appropriate following sections. The

most significant geological contributions to the groundwater aspects of the area are

discussed below.

The most extensive, and probably the most significant, published work on the

geology of the Roswell groundwater basin was by Fiedler and Nye (1933). Their

understanding of the groundwater basin, derived from the information available then, was

significant. Their work forms the foundation of all subsequent geological and

hydrological studies in the area.

As for present usefulness, the second most important published geological study

of the larger area containing the regional Roswell groundwater basin is the mapping by

Kelley (1971). Kelley recognized several major northeast trending structural zones that

transect the Roswell groundwater basin and its recharge area. These faults have been

(and are still often) called "buckles" or "zones" that Kelley suggested were possibly

strike-slip in nature.

Geologic reports covering subsurface information on the shallow and the artesian

aquifers were presented by Havenor (1968) and Kinney et al. (1968). The latter was a

study of the general stratigraphy of the groundwater basin compiled by the Roswell

Geological Society. Havenor's (1968) work was based upon a groundwater exploration program conducted for the City of Roswell, New Mexico. 33

Summers (1972) compiled and presented a comprehensive review of the published literature of the Pecos River Basin of New Mexico. While Summers' study covered an area much greater than the Roswell groundwater basin and recharge area, much of his emphasis was on the Roswell groundwater basin.

Maddox (1969) and Welder (1983) made excellent hydrological data available.

These two studies provide good background data for observing changes in water levels and general hydrologic features throughout the groundwater basin. The works of

Maddox and Welder are not readily available.

McAda and Morrison (1993) provided the most recent bibliography covering geological and hydrological information from the greater Roswell basin. Their publication is very useful in that it also provides information on springs, transrnissivity tests, weather records, water usage, and other valuable data.

Topography

The study area is on the western margin of the Permian basin of West Texas and southeast New Mexico. Topography varies from 1004 meters above mean sea level

(MSL) at the Carlsbad Municipal Airport at the southern end of Carlsbad, New Mexico, to 3659 meters MSL at Sierra Blanca on the westernmost edge of the study area.

Figure 1.3 shows the very general topography of the areas having the potential of supplying recharge waters to the Roswell groundwater basin to the east. South and east of the Capitan Mountains the sedimentary beds slope eastward at a low angle. The terrain is moderately rugged. The carbonate beds dissected by streams generally form 34 steep-walled to stair-stepped arroyos and canyons. Arroyo and canyon widths vary from a few meters to several kilometers. Seldom do the heights of the larger valley walls exceed 150 meters. Surfaces north of the Capitan Mountains are much less dissected than those to the south and directly east of the range.

The highest terrain is found in the central-western portion of the study area, and the lowest along the southward draining Pecos River. The Capitan Mountains are a major exception to the generally smooth east to southeasterly tilt and greatly influence the hydrogeology of the Roswell groundwater basin. The east-west trending mountain range is 32 km long, about 8 km wide, and reaches a maximum elevation of 3109 meters.

Immediately east of the Pecos River elevations are roughly 25 to 40 meters higher than the valley area. Farther east toward the Llano Estacado (the Mescalero

Ridge, Figure 1.1), the elevations are as much as 125 meters above the Pecos River. 34 6

Western margin' Physiography Llano Estacado

Fenneman (1931) j gsa described the present study area as being in the Great Plains Prov- ince, where it forms the Pecos 104° — Figure 1.3. Highly generalized topography of Section of the province. He de- study area capable of supplying recharge waters to Roswell groundwater basin. Modified from scribed the Pecos section as a Summers (1972). 35 trough lying between the High Plains of Texas and the Basin and Range province of western New Mexico. Of significance, Fenneman (1931, p. 47) said,

Its boundaries are marked almost throughout by steep slopes 500 to 1,500 feet high, but within these clearly marked limits the topography varies from flat plains to rocky canyon lands. The unity of the section consists in this, that the whole has been eroded below the once continuous level of the High Plains and of the Raton Section. The Llano Estacado, surfaced with wash from the mountains on the west, must have been continuous across what is now the Pecos section.

Fiedler and Nye (1933) recognized the Roswell groundwater basin area as having five erosional topographic surfaces. Fiedler and Nye (1933) named the topographically highest two as the Sacramento Plain and the Diamond A Plain. Both surfaces were erosional. The remaining three surfaces described by Fiedler and Nye (1933), in topographically descending order, are the Blackdom, Orchard Park, and Lakewood terraces. They identified the latter three as distinct alluvial fill constructional terraces.

The relationships of these surfaces are best shown by Fiedler and Nye's (1933) diagrammatic section (Figure 1.4 of this manuscript).

The Sacramento Plain is an erosional surface exposing the Permian bedrock carbonates and older horizons of the Sacramento Mountains on the western side of the study area (see Figure 1.1, page 25). The Sacramento Plain erosional surface is generally correlated with the constructional surface of the Upper Miocene-Early Pliocene Ogallala

Formation of the Llano Estacado that forms the High Plains of eastern New Mexico and

West Texas (the "Caprock" of Fiedler and Nye, 1933, shown in Figure 1.4). 36

Le. ke...cad terrace

Fitsurie I.—Generalised diagrammatic section showing Watkins of the physlographIc features In the Roswell artesian basin and the adJoInIng region to the east

Figure 1.4. Diagrammatic section across Roswell groundwater basin. From Fiedler and Nye (1933), Figure 1, p. 11.

The Diamond A Plain is developed to the west of the Pecos River. It was described by Fiedler and Nye (1933), Morgan (1942), and Summers (1972), as a pediment surface that lies about 120 to 400 meters below the Sacramento Plain. Its correlative surface east of the Pecos River, the Mescalero Plain, is more widespread.

Morgan (1942) estimated the vast majority of erosion in the Pecos River Valley, probably about 95 percent of it, took place during the Diamond A erosion.

The Ogallala Formation is believed to have completely covered the present Pecos

River Valley area (Fiedler and Nye, 1933; Thornbury, 1965; Summers, 1972). The Pecos

River is considered a notable example of stream piracy (Thornbury, 1965) in that it was initially only a small tributary of the Rio Grande and gradually captured eastward flowing streams. It also cut-away the overlying Ogallala from the study area. Hawley

(1984) stated that the oldest relict landscapes in New Mexico are the constructional and erosional surfaces in the southern High Plains and the Pecos Valley. Hawley (1984, p. 14) said, "These late Miocene and early Pliocene surfaces have formed on alluvial and eolian deposits and caprock calcretes of the Ogallala Formation." 37

Drainage

The courses of the Pecos River its significant tributaries, the Rio Hondo, Rio

Felix, Rio Penasco, Arroyo del Macho, Salt Creek (see Figure 1.2), and other lesser

streams are reflections of the fracture and fault patterns of the underlying bedrock.

Summers (1972) recognized this important control on the drainage of the Pecos River.

From central De Baca County (Figure 1.2, page 27) the Pecos River flows generally southward to old Lake McMillan, where it turns southeastward to the Texas-

New Mexico state line. Close to Carlsbad, the buried Capitan Reef of Permian age forms a sill at the southern end of the Roswell groundwater basin (Summers, 1972). The sill

creates a temporary base level and inhibits upstream down-cutting. Downstream from the sill the softer lithologies have been eroded and the gradient steepens.

The area north of the Capitan Mountains, the Pedernal Hills drainage of

Summers (1972) contains few surface drainage arroyos and no perennial streams. In the

area of this study the Arroyo del Macho and Salt Creek are the two principal drainage arroyos. The Arroyo del Macho is the major trunk and is about 160 km long (Kelley,

1971). This portion of the area is replete with sink holes. The general and long standing definition of typical karst topography is an area of soluble rocks, such as carbonates, close to the surface, containing abundant sink holes and solution features and exhibiting few, if any, through-going streams (Montgomery and Dathe, 1994).

South of the Capitan Mountains, west of the Pecos River, and essentially east of the crest of the Sacramento Mountains, the drainage to the Pecos River is dominated by 38 the Rio Hondo, Rio Felix, and the Rio Penasco. The Rio Hondo, Felix, and Penasco

Rivers are major trunk drainages and each is about 160 km long (Kelley, 1971). The Rio

Penasco's average gradient from the Pecos River westward for a distance of approximately 72 km is only 4.7 m/km. This is a very low gradient, but characteristic for the lower reaches of the three principal streams. Summers (1972) pointed out that the influence of fractures from the underlying bedrock shows strongly in the regional drainage patterns.

The drainage area south of the Rio Penasco to the Texas-New Mexico state line has been termed the Guadalupe drainage (Summers, 1972). From the north flank of the

Guadalupe Mountains to the Rio Penasco the arroyos dominantly exhibit dendritic drainage patterns, but a definite directional control on their drainage is exerted by rock structure. Some sink holes are present as well as large areas of centripetal drainage.

Summers (1972) said the drainage in the area south of the Guadalupe Mountains is dominated by two perennial streams, the Black River and the Delaware River (Rocky

Arroyo). Both streams are major trunk tributaries to the Pecos River and are about 60 km to 80 km in length. Additional major trunks to the Pecos River include Fourmile and

Seven Rivers, both of which are about the same length as the Black and Rocky Arroyo.

Both Fourmile and Seven Rivers are south of the Penasco River and north of Carlsbad.

Many sinks, some drainage in sinks, and springs are common in this portion of the area.

On the eastern side of the Pecos River, the drainage from the Mescalero Ridge

(Figure 1.1) is poorly defined south of Fort Summers. Taiban Creek, Buffalo Creek, and

Long Arroyo are the most important. As this area is generally underlain by evaporites 39 near the surface, many arroyos are short and drain into sinks or centripetal drainage basins. Sinks are common. Along the east edge of the Pecos River, immediately east of

Roswell, Bottomless Lakes State Park is developed around saline lakes in sink holes in anhydrite and gypsum strata of Permian age. The evaporites crop out along the steep eastern bluffs that delimit the Pecos River's eastward erosional migration. 40

STRATIGRAPHY

Introduction

The surface formation exposures within the Roswell groundwater basin are predominately of Permian, Tertiary, and Quaternary age. Pre-Permian rocks are exposed within the western scarp of the Sacramento Mountains, at Bent, New Mexico, and at

Pajarito Mountain on the Mescalero Apache Reservation. Some minor exposures of

Triassic and Cretaceous beds are found along the western scarp of the Sacramento

Mountains and in the Sierra Blanca region. For the most part, the formations of Permian and Tertiary-Quaternary age are the principal hydrostratigraphic units of concern in analyzing the collection, transmission, and storage of groundwater within the greater

Roswell groundwater basin.

Late Pennsylvanian and Wolfcampian Stage

The Late Pennsylvanian and Early Permian deposition was essentially continuous in many parts of southeastern New Mexico. The depositional environments of the greater Roswell groundwater basin were dominated by exposed land masses on the west and the east (see Figure 2.1). The Pedernal uplift, and to a much lesser extent the Central 41

Basin Platform, provided sediment to form the shelf facies deposits of the Northwestern

Shelf and the fine-clastic facies in the

so skAELF partly starved-basin facies of the incipi- 2 x 1401ANIA*4e51 z 1- ent Delaware Basin (Wilson and Jordon, C.uadaiwpo la,,, MDLAND (-n1') . . CI EASTERN 16 LAWARE tl, SHELF 1988). Only thin black shale was depos- 0, BASI , 0 1:33 bASI 117 4k „ /,./ Z ited in this time within the developing .Apache „„. ,);.c 04, 4:,, SUE FiELD 4, ee HAWS- / ' SOUTHERN Delaware Basin. Along the northern • ' A c.....• . SHELF

r B/A IN/ ' woRgT"Iti wtowLE 8 1-; 9 A edge of the Delaware Basin and the edge 1.4 HEx,co • 100 16D Ion Modified from king 1948 of the Central Basin Platform, carbonate Figure 2.1. Regional paleogeographic setting of the Guadalupian Permian. deposition was profuse. Carbonate bank Modified from Sonnenfeld, 1991. deposition and algal mound deposits

strike parallel to the developing hinge line of the Delaware Basin. Pennsylvanian rocks

are largely absent under the northern one-half of the Roswell groundwater basin. Meyer

(1966) showed that Pennsylvanian rocks generally overlap previously eroded older

Paleozoic sedimentary rocks and Precambrian age igneous and metamorphic rocks. The

Pennsylvanian sedimentary rocks are in angular unconformity with those of underlying

pre-Pennsylvanian age. The next major sequential overlap was by Wolfcampian deposits.

In the northeastern part of the greater groundwater basin, the rocks of Pennsylvanian age thicken southeastward from a feather-edge beneath the city of Roswell, to more than 600 meters in thickness about 80 kin to the south-southeast beneath the city of Artesia

(Meyer, 1966). 42

Wolfcampian age rocks in the region are represented by the Hueco Formation to the south and west of the crest of the Sacramento Mountains, the Bursum Formation in the scarp area of the Sacramento Mountains, and the Abo Formation over most of the

Pecos Slope and the Northwest Shelf. These units are, in part, regional facies of each other. Overall the carbonates of the Hueco Formation are dominant in the south and east, whereas the shales, sandy shales, and arkoses are preponderant to the north and west.

Wolfcampian rocks range in thickness from absent along the crest of the Pedernal uplift and Pajarito Mountain on the Mescalero Apache Reservation, southeast of

Mescalero, to about 300 meters in thickness in a north-south trending area found 40 kilometers east of Pajarito Mountain (Meyer, 1966). Kelley (1971) speculated that the north trending pre-Permian Pedernal high is without Wolfcamp age deposits. Meyer

(1966) depicted the Lower Wolfcampian rocks as present over the Pedernal area, thickening into the area 40 km east of Pajarito Mountain, then again thinning eastward toward Roswell to a maximum thickness of approximately 75 meters. The Wolfcampian rocks again thicken eastward and southeastward from Roswell.

On the west, in the Orogrande Basin located southwest of the western scarp of the Sacramento Mountains, strata of Wolfcamp age thicken to more than 1000 meters. In the Orogrande Basin the Wolfcampian of the Hueco Formation are limestones, shales, and some conglomerates (Wilson and Jordon, 1988). These formations reflect a significant and important pulse of orogenic activity in the middle Wolfcamp in the Orogrande

Basin (Armin, 1987). Evidence of this activity is also evidenced in wildcat oil wells drilled south of Pajarito Mountain (J. Cleo Thompson wells, unpubl. data provided by 43

J. Ahlen, 1994). The isopachous and lithofacies maps presented by Meyer (1966) illustrate well the general distribution of Wolfcampian beds over southeastern New

Mexico.

The lower part of the Wolfcampian rock sequence to the northwest of the greater

Roswell groundwater basin area consists of nodular gray limestones, arkose, and a dominance of reddish-purple and green shale. These rocks are overlain by dark red shales, red quartzose sandstone, and arkose. As the units approach the Pedemal uplift, the nodular limestones and dark reddish-purple zones gradually change to red shale, arkose and conglomerate with only minor amounts of limestone included. These units are generally correlated with the Abo Formation of the middle to upper Wolfcamp.

Wildcat oil and gas wells drilled along the southeastern side of the buried

Pedernal uplift and to the north of the end of the Sacramento Mountains (Guadalupe fault of Kelley, 1971) encountered substantial variations in the thickness and lithologies of the

Abo Formation (J. Ahlen, pers. comm., 1994). Samples and geophysical logs show the area was covered by Wolfcampian Abo Formation which overlies Paleozoic strata. Three wells were in Section 6, Township 16S, Range 17 East, Section 28, Township 16S,

Range 17 East, and Section 3, Township 17S, Range 17 East (see the general location map, Figure 1.2, page 27).

The Abo Formation thickens from 110.3 m in the Section 6 well, to 305.7 m less than 7.7 km away in the middle (Section 28) well. Farther southeast, approximately 4.8 km (Section 3 well), the Abo Formation is 310.9 m thick, virtually the same as in the middle well. In each of these wells the Abo Formation contained many conglomeratic 44 zones with abundant igneous and metamorphic rocks (J. L. Ahlen, wellsite geologist, pers. comm., 1994). These wells prove that uplift of the Pedernal was contributing substantial coarse-grained and conglomeratic material during Abo deposition.

Except for Precambrian igneous rocks exposed at Pajarito Mountain and where depositionally thinned or absent along the Pedernal uplift, middle to upper Wolfcampian

Abo Formation or older Wolfcampian rocks blanket the pre-Permian beds in most of the greater Roswell groundwater basin. Although the Wolfcampian rocks are not significant in the hydrostratigraphic system of the Roswell groundwater basin, they lend to understanding the structural-stratigraphic framework of the Roswell groundwater basin area.

Wolfcamp strata demark the extent of the mostly buried Pedernal uplift. The Pedernal uplift supplied significant amounts of sediment toward both the east and the west plus serving as a major land barrier to the spread of Permian seas in this area. These features will be discussed further in the section on tectonics.

Leonardian—Guadalupian Stages

Rocks of the Yeso and the San Andres Formations of the Pecos Slope are

Leonardian in age. These formations are hydrogeologically important in the Roswell groundwater basin because they form the artesian aquifers and are the zones of recharge transmission from west to east. The Yeso Formation is entirely of Leonardian age, whereas the San Andres Formation is, for the most part in the greater Roswell groundwater basin, is Guadalupian in age. In the San Andres Mountains, approximately 80 km west of the western scarp of the Sacramento Mountains, the San Andres Formation is 45

SHELF MARGIN BASIN

: 3 2 ALGERITA LAST CHANCE WESTERN ESCARPMENT CANYON ESCARPMENT AALA — SB 411,2, IAANZANITA GOAT SEEP 10 QUEEN FM. \ \ ..AX ,F..,I!E.. ..). SOUTH W

Z A' OW' me, se, Z. 07 T Z < HIATUS 2il 1 GRAYBURG FIA Epos,0,,, 0 0 ,7,' oz 5 9 ,z Jr zeiozwmr ,,,_ MI OGLE GRAYBURG 30 V 40:

-••n ,„, • A' 5 Ir

Nc- ...„„, czi i i 4 1 ? 4,0 „,,, 9.-ri., in' o -...... LI VICTORIO PEAK YESO BONE SPRING FM o FM.

Figure 2.2. Chronostratigraphic diagram Late Leonardian-Middle Guadaluptan age stratigraphic units in the western Guadalupe Mountains. (From Sonnenfeld, 1991; Sarge and Lehmann, 1986) entirely Leonardian in age, but only the lower portion of the formation is present.

The stratigraphy of the Permian in the study area, especially the Leonardian and the Guadalupian Stages, is complex and highly dependent upon the depositional environments being considered. The correlation chart, Figure 2.2, shows the currently accepted general terminology for the respective formations of the middle part of the

Permian. Stratigraphic nomenclature in the Roswell groundwater basin follows that of the Northwestern Shelf, which is predominately back-reef facies. Post-Wolfcampian sedimentation in southeastern New Mexico and West Texas was strongly dominated by the presence of reef development around the margin of the Delaware Basin (see

Figure 2.1). The Capitan Reef complex, although not directly involved in this study, is significant to the southern drainage of the Roswell groundwater basin. The hydro- 46 geologic framework, as dictated by regional stratigraphic and lithologic variations, has been greatly influenced by the massive Capitan Reef Showing the stratigraphic nomenclature from the reef complex to the back-reef is therefore important. Figure 2.2 denotes the general formational relationships for the Leonardian and the Guadalupian from the shelf through the reef and into the Delaware Basin in southeastern New Mexico.

Yeso Formation

The Yeso Formation is important to the hydrogeologic framework of the Roswell groundwater basin for two reasons. First, it forms the aquitard beneath the artesian San

Andres aquifer, and second, the upper sandstone members of the formation provide conduits for the movement of water from the recharge areas.

The Yeso Formation is regionally composed of east-northeast trending lithofacies that are generally consistent with the Northwestern Shelf depositional environment caused by the Capitan Reef (see Figure 2.1). In front of the reef the equivalent Yeso

Formation stratigraphic unit is recognized as the Bone Spring Formation. The Bone

Spring Formation is dominantly black, dense, massive, fetid limestones, and very dark to black siltstones and thin shales. The Bone Springs grades into the Victorio Peak

Formation that Lloyd (1949) considered as the fossil barrier reef core of the Capitan

Reef Toward the northwest the Leonardian backreef facies is composed predominately of dolomite and anhydrite. North of Artesia the dolomite and anhydrite facies of the

Yeso Formation grade laterally into limestone with increasing amounts of red and yellow shales. Slightly south of Roswell the east-northeast trending lithofacies begin to pickup 47 stringers of halite interspersed with the limestones, dolomites, anhydrites, and reddish to yellow shales. Continuing northward the Leonardian Yeso Formation becomes progressively more anhydritic with increasing halite and shale, and decreasing carbonates. Some anhydritic sandstones and clastic anhydrites in the east-northeast trending lithofacies examined in borehole cuttings of wells located south and east of

Roswell (Havenor, 1968) are well-sorted. This is conimonly the case with the sandstone stratigraphically positioned between the top of the Yeso Formation and the younger San

Andres Formation in areas near and east of the Pecos River. Those subsurface sandstones are commonly correlated with the Glorieta Sandstone.

In the area west of the Chaves-Lincoln County line, halite in the Yeso Formation is absent. At many locations west of the Border Hill fault where Yeso outcrops are commonly found, especially where porous beds, water-seeps, or springs occur, surface coatings or incrustations of gypsum and salt resulting from evaporation are regularly observed. These evaporative crusts are commonly seen in road cuts. Halite zones are absent from well cuttings in these areas. Therefore, the absence of halite in the Yeso

Formation here is due to a groundwater dissolution.

Close to the Sacramento Mountains escarpment the Yeso Formation varies in thickness from approximately 1,900 to 2,250 m (Summers, 1972). The general thickening of the Yeso Formation from the outcrop areas toward the east is, in small part, a function of increasingly preserved halite sections within the formation, but it depends on depositional accumulation. The top of the Yeso Formation in the subsurface is commonly chosen below the base of the San Andres Formation carbonates or the base of 48 underlying Glorieta Sandstone. This is usually accompanied by a marked change in color from dark San Andres carbonates and shales to white or tan sandstones of the Glorieta.

The Yeso Formation is most commonly yellowish in color.

Various workers have placed the Leonardian-Guadalupian boundary above, in, and below the uppermost sandstone separating the Yeso Formation from the overlying

San Andres Formation. Lloyd (1949) considered it gradational between the two.

Summers (1972) said that sedimentation was continuous from Leonardian into

Guadalupian time and the formations are therefore time transgressive. Near the Capitan

Reef complex Sonnenfeld (1991) considered the Leonardian and Guadalupian deposition to be separated chronostratigraphically by some period of nondeposition because the top of the Yeso Formation represents a sequence boundary (see Figure 2.2, page 45).

Glorieta Sandstone

The Glotieta Sandstone as present in all parts of New Mexico with a thickness range of three meters in southern Eddy and Lea Counties, New Mexico, to 158 m in

Torrence County, New Mexico (Borton, 1972). While the base of the Glorieta is clearly the top of the Yeso Formation, which is certainly Leonardian, the location of the Guada- lupian lower boundary is far less certain. In the San Andres Mountains, the San Andres

Formation is Leonardian (Kottlowski et al., 1956). In the vicinity of the Capitan Reef the lowermost part of the San Andres is Leonardian. The Glorieta Sandstone is considered as Leonardian based solely upon stratigraphic position and interformational relationships above and below. Summers (1972) pointed out that Kelley (1971), Harbour (1970), and 49

Skinner (1946), along with others, consistently suggested that within the basin the

Glorieta Sandstone be Leonardian; it has a characteristic and readily recognized

lithology; and lastly, the unit is thick and dominantly sandstone to the north, but contains

limestones and other lithologies to the south. These characteristics do provide a general

ease of identification and correlation. It is historical convention that relegated the

Glorieta to the Leonardian, nothing more. Based upon my field examinations of the

Glorieta it is probably a time-transgressive.

Lang (1937) described the sandstone as coarse-grained, calcareous, white and

yellow to brownish-red in color. Within the Roswell groundwater basin this sandstone is

generally referred to as the Glorieta Sandstone. Bean (1949) applied that name to the

unit at the base of the San Andres Formation and linked it to the previous description by

Lang (1937). Mourant (1963) also described the Hondo Sandstone Member as

Leonardian in age and noted the correlation by other investigators with the Glorieta

Sandstone. Kelley (1971) referred to sandstone units in the basal San Andres Formation

as the Rio Bonito Member of the San Andres Formation. The Rio Bonito Member was

described as dominantly limestone and dolomite in the southern portion of the area

interfingering with the Glorieta member to the north.

Havenor (1968) reported that when an exploratory water well was being drilled for the City of Roswell in Eight Mile Draw, west of Roswell, running water could be heard in the open hole in the Glorieta Sandstone. Because the test well was being drilled with air, the water flow encountered upon drilling into the Glorieta Sandstone was immediately detectable. That observation was significant because the exposed San 50

Andres Formation carbonates become confined immediately east of Six Mile Hill, less than five km east of the well location. In the area of the test well the San Andres

Formation carbonates are not water saturated. The groundwater in the area of the test well moves east toward the confined aquifer through the Glorieta Sandstone. Bean

(1949, page 9) stated, "the main water body apparently extends through the Glorieta (?) sandstone member and the overlying porous carbonate rocks with little regard for their different character."

San Andres Formation

To the north of the Capitan Reef area the carbonates of the San Andres are dominantly dolomites with thin limestones, gypsum, some mudstones, and thin shales.

Overall, the San Andres Formation in the Roswell groundwater basin can be attributed to the backreef environment developed because of Capitan Reef growth. The general lithology of the San Andres Formation can be described as up to 450 m of tan to dark brown, gray to very dark gray, dense to coarsely crystalline, oolitic limestones, and dolomites. From the exposures in the northwestern part of the groundwater basin toward the east into the deeply buried San Andres Formation of the Northwestern Shelf, the light-gray carbonates contain more anhydrite, gypsum, and halite. Also, in the northwestern portion of the groundwater basin the karstic nature of the San Andres is substantial and obvious because it is at or near the surface,. Next to the groundwater basin and east of the Pecos Valley the San Andres Formation is not yet exposed to the 51

groundwater circulation that creates the karstic conditions seen on the Pecos Slope and

north of the Capitan Mountains.

Although many local areas of the San Andres Formation within the greater

Roswell groundwater basin have been examined and reported upon (Bean, 1949;

Havenor, 1968; Kelley, 1971; Kinney, 1968; Kottlowski et al., 1956; Lang, 1937;

Mourant, 1963; Summers, 1972; and others), the first comprehensive stratigraphie study

of the San Andres Formation within this study area was by Kelley (1971).

The result of Kelley's major surface mapping program was the most exhaustive

regional division of the San Andres Formation in the large area of this study. In the

exposures close to the Capitan Reef complex, other workers have also noted the general

division of the San Andres Formation into an upper, middle, and a lower zone

(Sonnenfeld, 1991; Sarge and Lehmann, 1986; see Figure 2.2, page 45 of this

manuscript). Kelley (1971) named the members of the San Andres Formation, from

oldest to youngest, the Rio

Bonito, Bonney Canyon, HYPER/is/J/1E PONDSMSAt SAYS SEINE PAN/SALINA MLANDi SA SKIA and Fourmile Members.

The following generalized descriptions of those units are predominately after

Kelley (1971).

The Rio Bonito Figure 2.3. Diagrammatic lower San Andres depositional environment in southeastern New Mexico. Member is generally thick- Modified from Sonnenfeld (1991). 52 bedded limestone and dolomite that interfinger in their lower sections with the Glorieta

Sandstone proper. The Rio Bonito Member varies in thickness from 91 m to 198 m.

The Bonney Canyon Member varies from less than 18 m to 91 m. It is generally thin-bedded, porous limestones and dolomites. The original porosity in the Bonney

Canyon Member was probably due to erosion and/or groundwater circulation during early to middle Guadalupian time. Evidence of karst solution is common in the upper part of the Bonney Canyon Member. Many exposed erosional surfaces of the Bonney

Canyon Member exhibit evidences of solutional activity.

The Fourmile Draw Member, in part equivalent to the Cherry Canyon Formation of the Delaware Mountain Group in the Delaware Basin, commonly contains thin-bedded anhydritic or gypsiferous limestones and dolomites. This upper member of the San

Andres Formation becomes more evaporitic farther backreef, as is diagrammatically shown in Figure 2.3. The Fourmile Draw member also has many karst solution features,

some demonstrating collapse breccia in the top of the formation. Kelley (1971) suggested that in the area north of the Capitan Mountains evidence is present of post-San

Andres uplift and erosion that preceded the subsequent deposition of the overlying

Grayburg Group. Borton (1972) discussed large solutional features involving the San

Andres and overlying beds in northwestern Chaves County, New Mexico. Other workers have agreed that the top of the San Andres Formation is regionally unconformable

(Gratton and LeMay, 1969; Havenor, 1968; Young, 1965; Hayes, 1964; Tait et al., 1962).

In the subsurface to the east and southeast of the Pecos River, the San Andres

Formation is variously found to have noteworthy porosity and permeability in three 53 different zones. The uppermost zone is at the top of the formation; the second general zone is approximately the middle of the formation (Slaughter, or Pi zones of oil field terminology); the lower zone is the thick carbonate (generally limestone) sequence above the Glorieta Sandstone in close proximity with and east of the Pecos River.

The uppermost zone has yielded copious amounts of hydrocarbons from Artesia eastward to Hobbs, New Mexico, and south along the Central Basin Platform (see Figure

2.1, page 41). Having examined many cores and borehole cuttings derived from the upper San Andres Formation throughout the area, it is apparent to me that the top of the formation has been subjected to post-depositional solutional activity. The upper 30 m of the San Andres Formation often has porous and permeable zones. Those zones commonly contain dolomite and anhydrite breccia-like fragments associated with clays suggesting solutional and collapse activity. The porous and permeable zones often overlie a fine-grained siliclastic sandstone called the Lovington Sandstone in petroleum exploration and development usage. The horizons below the Lovington sand are often anhydrite-rich dolomites, limestones, and dolomitic anhydrites.

Based upon the work of Kottlowski et al. (1956), Havenor (1968) pointed-out that the San Andres Formation in the San Andres Mountains of central southern New Mexico is only about one-half as thick as is found immediately east of the Pecos River. The formation thins from the Pecos River area toward the west, with the loss being in the upper one-half of the section. The San Andres Formation is gradually beveled toward the west. 54

o 106

T3M

T211

T1P1

T1S

T2S

135 34 0

145

TSS

T1 1S

T129

T13S

T168

T17S

Tits

T1 SS

1205

T21S

1229

Outline of Roswell Groundwater basin and recharge area Tertiary intrusives Stfike-slip fault

Artesia Group

Figure 2.4. Surface and sub-Quaternary distribution of Artesia Group as projected to rest upon the San Andres Formation in the eastern portion of the Roswell groundwater basin in Chaves and Eddy Counties, New Mexico. 55

The general stratigraphy of the principal aquifers of the Roswell groundwater basin is reviewed later in a subsection titled Aquifer Stratigraphy. The confined artesian aquifer in the Roswell groundwater basin is primarily developed within the San Andres

Formation. The overlying basal Grayburg Formation of the Artesia Group is often also an important part of the confined aquifer.

Figure 2.4 shows the surface exposures and projected sub-Quaternary distribution of the progressive beveled Artesia Group atop the San Andres Formation. The shaded

Artesia Group is underlain by San Andres Formation throughout the eastern portion of the Roswell groundwater basin. For many kilometers to the west of the Artesia Group's western outcrop or subcrop the San AndresFormation is either exposed at the surface or under Quaternary age deposits. The beveled formations presented in Figure 2.4 illustrate significant offsets due to fault displacement. This displacement is discussed in the section on tectonics, but is shown here to illustrate the spatial stratigraphic-structural relationship of the San Andres Formation and the overlying Artesia Group.

Artesia Group

The stratigraphic relationships of the Artesia Group to the underlying San Andres

Formation within the greater Roswell groundwater basin are shown, in part, in

Figure 2.2 on page 45, Figure 2.4, and in the following chart, Table 2.1. The Artesia

Group was defined by Tait et al. (1962) and that definition was accepted by the Roswell

Geological Society 1965-1967 (Kinney et al. 1968). The formational boundaries are 56 recognized as sequence boundaries separated by hiatuses (Sonnenfeld, 1991; Sarge and

Lehmann, 1986; see Figure 2.2, page 27 of this manuscript).

The Late Guadalupian Artesia Group was designated by Tait et al. (1962) in a reference well drilled by the Humble Oil and Refining Company as the No. 1 Federal-

Bogle. The well is in Section 30, Township 16 South, Range 30 East, Eddy County, New

Mexico and is approximately 48 km east of the Pecos River and the eastern edge of the

Roswell groundwater basin.

Upper Permian Stratigraphic Nomenclature of Southeastern New Mexico Stage Northwestern Shelf Capitan Reef Dewey Lake Fm Dewey Lake Fm Rustler Fm Ochoan Rustler Fm Salado Fm Salado Fm Castile Fm Castile Fm Tansill Fm Tansill Fm s2. Yates Fm o 6 Seven Rivers Fm Guadalupian —m Capitan Fm a)cA Queen Fm Goat Seep Fm Grayburg Fm Grayburg Fm -

Cherry Canyon Fm San Andres Fm San Andres Fm -

Victorio Peak Fm Table 2.1. Generalized Upper Permian stratigraphie correlation chart for part of Southeastern New Mexico. 57 The Artesia Group and the

overlying Ochoan Salado and

Rustler Formations are exposed at

the surface immediately east and west of the Pecos River. Kelley

(1971) mapped these exposures along the eastern margin of the groundwater basin as shown in

Figure 2.5. From Kelley's (1971) maps and from my own field observations and examinations of well cuttings in the mapped area, upper Guadalupian and Ochoan rocks are clearly exposed across the region. The formations were up-tilted on the west with the o ir >- La 2, regional uplift of the Pecos Slope 8 La ES LREEF F ACI and have been beveled by erosion. The episodic uplift of 104 the ancient Pedernal from Late Figure 2.5. Distribution of Artesia Group formations along the Pecos River. From Kelley Permian through part of Triassic (1971). Pag=Grayburg, Pacr-Queen, Pas=Seven Rivers, Pay—Yates, Pat—Tansill, Ps1=Salado, time is reflected by the Pr—Rustler. 58 distribution of Upper Permian beds beneath the Triassic redbeds close to the town of

Capitan and also east of the Pecos River.

Just beyond the midpoint of the eastern boundary of the groundwater basin, the

Triassic cuts through the Rustler and Salado Formations into the Yates Formation. This can be observed close to Samples Lake (a playa), in Sections 7 and 8, T. 7 S., R. 27 E.,

Chaves County.

Grayburg Formation

The base of the Grayburg Formation is identified in the subsurface east of the groundwater basin as an anhydritic to calcareous very fine-grained quartz sandstone resting disconformably on top of San Andres carbonates. Near Roswell, the Grayburg

Formation and the overlying Queen Formation are difficult to separate in outcrop or subsurface and are generally called Artesia Group undifferentiated. The lower sandstone of the Artesia Group is typically called the Penrose sandstone by petroleum geologists; it is a hydrocarbon-producing horizon in the area east of the Pecos River.

The Grayburg Formation is characterized in the subsurface next to the eastern margin of the Roswell groundwater basin as interbedded anhydrites, anhydritic dolomites, red shales and siltstones, with gray to reddish sandstones that are often anhydritic and/or dolomitic. Overall, from the Capitan Reef northward, the dolomite content decreases and the gypsum-anhydrite content increases. Similarly, northward from the reef, the sandstones generally become redder in color, finer-grained, and thinner. The formation becomes more evaporitic and terrestrial northward onto the broad 59

Northwestern Shelf. This gradual facies progression is readily traceable in the subsurface and in surface exposures along and west of the Pecos River.

South of the Y-0 fault, on the Y-0 block (see Figure 2.6), the San Andres

Formation and the overlying Grayburg Formation has been identified in the subsurface based on well-sample lithology and by electric-log correlations with areas to the east

(Maddox, 1969; Havenor, 1968; Kinney et al , 1968). This suggests that in some areas about the lower one-half of the Artesia Group is present beneath the alluvium.

Within the Roswell block (see Figure 2.6), between the Y-0 fault on the south and the Six Mile Hill fault on the north, the Artesia Group is much thinner than on the block south of the Y-0 fault. Along the northwestern edge of the Roswell block the

Artesia Group is recognized in surface exposures. Some exposures of Artesia Group

(possibly lower Queen and Grayburg) are present west of the city of Roswell and west of the Six Mile Hill fault.

The available electric well logs, drill samples and driller's logs suggest a much more confusing stratigraphic relationship beneath soil and alluvial cover than simply basal Grayburg Formation on top of San Andres Formation. Havenor (1968) suggested that part of the Grayburg Formation was absent because of implied Guadalupian erosion.

Erosion of the lower Artesia Group and the upper San Andres Formation with subsequent deposition of upper Grayburg and Queen Formation would have required Guadalupian uplift of the Roswell block. Uplift is not consistent with the observations discussed in this paper. 60

The presence of what was originally called (Havenor, 1968, page 11), "Queen- like red beds that form the aquitard over the artesian aquifer," are recognized here as, in part, Queen Formation—or at least Artesia Group undifferentiated. The pinkish, gypsiferous, shaly, limestone breccias overlying what can be distinguished as an irregular, disrupted San Andres limestone sequences are considered here remnants of

Queen and Grayburg redbeds, evaporites, carbonates, and solution-brecciated carbonates of the upper San Andres. These Queen-Grayburg-upper San Andres beds have been seriously affected by the solutional activity of meteoric waters and the development of karstic features.

The obvious solutional activity was clearly noted with the observation that "the red shales forming the thin aquitard are possible terra rosa-like clays formed as the residual clay deposit from solution of limestones" (Havenor, 1968, page 11). The "terra rosa-like clays" are abundant and occur naturally as red clays within the Grayburg and

Queen formations of the Artesia Group. These clays would be concentrated with the solution of the evaporites and carbonates. Surface development of a terra rosa regolith would not be the sole environment necessary. The clays would also be readily incorporated into collapsed solutional features—as were observed and described in cores cut through the aquifer (Havenor, 1968). Collapse features are visible in local road cuts.

Two large areas northwest of Roswell were noted by Borton (1972) as having anomalously thin sections of Artesia Group and San Andres Formation. Borton found

60 m - 120 m of section missing as compared with thicknesses in wells on the flanks of the anomalies. Borton (1972, page 9) also found that more than 180 m of "red and 61 yellow clay and silt of Quaternary age and some limestone of the Artesia Formation overlie the San Andres Limestone in the sink area." These two areas occupy most of

Township 5 South, Range 20 East, and a large part of Township 7 South, Range 19 East.

Figure 2.6 shows the Township and Range locations of the features in relation to the

Border Hills fault. Both of the features reported by Borton lie north of the northeast trending Border Hills fault. Figure 2.6 permits referencing the regional location of the collapse features by Township and Range plus it shows the major right-lateral strike-slip faults in the vicinity. 62

106 104

T311

72M

71N

715

T2S

73S 34 745

75S

769

77S

785

79S

7'10S

Ills

712S

7135

714S

1169

7178

7189

719S

1209

721S

7229

Tertiary intrusives

Figure 2.6. Location of major dextral strike-slip faults and respective structural blocks discussed in text. 63

Figure 2.7. Landsat TM image of City of Roswell, New Mexico. Circled area is a possible old sinkhole. North is toward the top-left center of the figure. Image acquired 07/04/1985. Band 1, 2x, histogram stretch.

Further evidence of extensive solution development in the near-surface carbonates and evaporites can be observed using Landsat TM imagery. One such feature is found in the heart of downtown Roswell and is approximately 1.5 km in diameter.

This feature is shown in Figure 2.7. The major streets next to the annotated circular area are north-south, east-west streets found 1.6 km apart. Another sinkhole-like feature is found on the southeast side of the city. The second feature, although larger than that in the midtown locality, is within an area currently being used for agriculture. This demonstrates the magnitude of the solutional activity referred to in this study.

Sinkholes are commonly developed in the gypsiferous portions of the Grayburg

Formation of the Artesia Group, although there is also karstic development within the upper San Andres carbonates in the Roswell groundwater basin (Barton, 1972; Kelley, 64

1971). Part of the thinning anomaly on the Roswell block within the San Andres

Formation can be explained as due to extensive solution and alteration.

Understanding the stratigraphy of the San Andres Formation and Artesia Group

contact requires consideration of later events effecting that contact. The solution

development is believe by me to be due to mostly to stream and meteoric water reactions

with the rocks during Quaternary time. The Hondo River is a major tributary to the

Pecos River. The Berrendo River is a lesser, but significant, tributary. Both streams

have developed floodplains upon the Roswell block. Each has provided substantial and

identifiable sources of fresh water for carbonate-evaporite solution. Stream meander

patterns of the Hondo and Berrendo Rivers visible on satellite imagery and on the ground

further support the conclusion of extensive solutional capabilities. The mapping of soil

distributions on the Roswell block (J. W. Hawley, pers. corn. 1994, unpubl. maps)

confirms the presence of major surface drainage during the Quaternary on the Roswell block. The Six Mile Hill block and theY-0 block do not show evidences of having had

as much available surface water as did the Roswell block.

Queen Formation

The Queen Formation of the Artesia Group is composed of sandstone and anhydrite with interbedded red shales and occasional dolomites. The Queen Formation becomes more evaporitic north of the Capitan Reef Sandstones of the Queen are characteristically red, very fine-grained to fine-grained, with anhydrite to halite cement,

and frequently contain medium - grained, very well - rounded and frosted quartz grains. 65

Red sands predominate in the formation. Within the Roswell groundwater basin the

Queen Formation is difficult to distinguish from the underlying Grayburg Formation.

Where geophysical well logs are available and supported with good samples, the Queen can be identified. Unfortunately, logs and good samples do not commonly coexist.

The subcrops of the Queen Formation are to the west of and parallel to the Pecos

River. They are generally overlain by alluvium and thus are physically in contact with portions of the shallow (alluvial) aquifer (see Figure 1.4, page 36 of this manuscript).

The sandstones of the Queen Formation have varying porosities and permeabilities and can act as low-yield aquifers within the Roswell groundwater basin (Kinney et al., 1968;

Hantush, 1955). The sandstones of the Queen Formation dip regionally east- southeastward.

Seven Rivers Formation

The Seven River Formation, as is observed with the Queen, Grayburg, and San

Andres formations, transitions from dolomites in the vicinity of the Capitan Reef to progressively more gypsum, anhydrite, and halite northward onto the Northwestern Shelf

Near the Capitan Reef, namely east of Carlsbad, around the margin of the Delaware

Basin (see Figure 2.1, page 41), some dolomites of the Seven Rivers Formation are highly porous and permeable. Although the Seven Rivers Formation contains more anhydrite and gypsum north from the reef, it still contains very porous dolomites and leached evaporites in the subsurface—especially east of Artesia. 66

Usually, in the back reef area, the Seven Rivers Formation can be depicted as very porous and permeable with gradually decreasing porosity and an increasing percentage of evaporites, very fine-grained sandstones, and siltstones northward. The time of Seven Rivers Formation deposition was much like that depicted for the lower San

Andres Formation depositional period diagrammatically shown in Figure 2.3, on page 51.

It was an evaporite dominated environment that subsequently was greatly affected by the activity of groundwater and surface water.

The distribution of the Seven Rivers Formation subcrop within the Roswell groundwater basin is shown in Figure 2.4, on page 54. The Seven Rivers Formation is

completely covered by Quaternary alluvium within the Pecos River Valley. Like the underlying Queen Formation, part of the Seven Rivers is in physical contact with the

shallow aquifer.

Yates Formation

Immediately east of old Lake McMillan reservoir the thickness of the Yates

Formation has been measured at 123 m by Kelley (1971). In that area the Yates

Formation is mostly white to occasionally reddish gypsum, thin white dolomite in the lower portion of the formation, with hematite-stained siltstone at the base of the unit.

The upper portion of the formation has brown to gray siltstone and very fine-grained sand stone. The upper sandstones probably correlate with the wide-spread subsurface Yates

sandstone mapped across much of the Permian Basin. 67

The Yates Formation is a valuable structural subsurface mapping horizon throughout the Northwestern Shelf, the Central Basin Platform and West Texas. Some

Yates Formation surface exposures in the near-reef shelf areas were mapped extensively by Hayes (1959), Kelley (1971) and others. The Yates Formation crops out east of the

Pecos River (see Figure 2.5) from approximately 40 km north of Roswell to the northern

end of the Brantley dam and reservoir located north of Carlsbad (see Figure 2.6).

Tansill Formation

The Tansill Formation is the uppermost unit in the Artesia Group (see Table 2.1,

page 56). The Tansill Formation is a thick-bedded dolomite that transitions into and

overlaps the Capitan Reef carbonates south of the Roswell groundwater basin. North from the reef the Tansill Formation thins rapidly and contains more gypsum and

anhydrite. The Tansill Formation thins toward the Roswell groundwater basin and

disappears within a short distance northeast of Hagerman, north of highway NM 289, in

Township 13 South, Range 27 East ( reference Figure 2.6). North of this thin zone of

Tansill Formation outcrops of the Triassic strata rest directly upon the Yates Formation.

Erosion by the Pecos River and its western tributaries has removed the Tansill Formation

from most of the area. 68

Ochoan Stage

The Ochoan adjacent to the Roswell groundwater basin is composed of the Salado and the Rustler Formations. The Salado Formation is predominately evaporites. It becomes difficult to distinguish from the underlying Tansill Formation. The evaporites, mainly anhydrite, gypsum, and halite, often attain an orange-red color from mixing clays due to ablation of salt near the surface (Kelley, 1971).

The Rustler Formation overlaps the Salado Formation. The Rustler Formation to the south of the Roswell groundwater basin contains more dolomite than does the Salado

Formation, but it is still very evaporitic in nature. Sandstones and siltstones are present in the Rustler Formation south of the study area.

The Salado and Rustler Formations are present east of the Pecos River as is shown in Figure 2.4, ( page 54). Neither of the formations is recognized within the groundwater basin proper. The lower Ochoan Castile Formation is not recognized until much further south near Carlsbad, New Mexico. The Salado Formation, although no longer present in the Roswell groundwater basin, affected the groundwater by virtue of its high concentrations of halite in the formation.

The Rustler Formation, like the Salado, has been progressively stripped from the uplifted Pecos Slope by down-dip erosion of the Pecos River and its tributaries. The

Rustler Formation in the area is dominated by gypsum, red siltstones, red sandstones, and dolomites. Remnants of the Salado Formation are occasionally observed disrupting

Rustler outcrops because of solution or flow of the underlying salts of the Salado (Kelley,

1971). 69

Tertiary Formations

The rocks of Tertiary age within the Roswell groundwater basin are present in the

Sierra Blanca Mountain volcanics and the detrital sedimentary rocks derived from them.

The Cub Mountain Formation (Kelley, 1971) predates and is distributed mainly around the Sierra Blanca Mountains. The Cub Mountain Formation is not considered hydrologically significant to this study.

Tertiary alluvial materials may be present throughout the Roswell groundwater basin, but their ages have been determined by lithologic association and not by paleontological or radiometric methods. Except in the area where the Cub Mountain

Formation is recognized, the alluvial materials are usually classified as Quaternary in age

(Kelley, 1971). However, the Late Miocene to Pliocene Ogallala Formation is believed to have originally covered much of this portion of New Mexico and remnants may be exposed at high elevations in the southern and western parts of the Roswell groundwater basin (Kelley, 1971).

Knowledge of the distribution of Ogallala Formation west of the Pecos River would be helpful in the problem of dating some Tertiary and Quaternary age events in the

Roswell groundwater basin area. To my knowledge there has not been positive identification of Ogallala remnants along the Pecos River in the study area.

A. L. Boucher (pers. comm., 1992) reported gravel remanents south of Dunkin, New

Mexico, that correlate with Ogallala deposits. During field examinations with Mr.

Boucher in February 1994, we found many Quaternary alluvial deposits south of Dunkin, but those examined were not Ogallala Formation. We may not have visited the critical 70 sites. His verbal description of the calcrete-like indurated conglomerate deposits, based upon many notable years of field experience in the region leaves open the possibility of finding and examining potential Ogallala remnants.

Calcrete occurrences in the Ogallala along the western edge of the Llano

Estacado, to the east of the Mescalero Plain, are very common. Machette (1985) reported early Pleistocene calcic soils (Stage V) in alluvium that included reworked clasts of older calcic soils (Stage VI) of Miocene Ogallala Formation. Hawley (1986) also reported the presence of zones of secondary-carbonate accumulation in the Ogallala Formation. The highly developed calcretes in the Pecos River Valley area of the Roswell groundwater basin are among the most highly indurated K horizons (Stage VI) found in the arid southwestern United States. Birkeland et al. (1991), using carbonate morphology have estimated the ages of the Roswell-Carlsbad Stage VI calcretes to be as old as late

Miocene. The highly predictable pattern of stages of carbonate morphology development with time provides a powerful tool for the correlation of these deposits (Birkeland et al.,

1991).

Hawley (1993) summarized recent work on the late Miocene-early Pliocene

Ogallala Formation, latest Tertiary-Quaternary Gatufia Formation, and the late Pliocene-

Pleistocene-Recent Pecos Valley fills. Overall Hawley (1993) concludes that considerable detail is lacking on the subsurface nature of the Ogallala-Gaturia and the alluvial-fill in the Pecos Valley. Hawley (1993, page 267) said: 71

An important (and certainly the most controversial) conclusion in this report is that, for at least the past 10-12 Ma, an ancestral "lower" Pecos fluvial system followed a course through southeastern New Mexico and western Texas that is very close to the present Roswell-Carlsbad- Pecos Valley trend. Deposits of this system (e.g., Ogallala, Gaturia and "quartzose conglomerate") have progressively and differentially subsided as dissolution of evaporites of Leonardian to Ochoan age (Yeso to Rustler Formations) occurred in various parts of the region. . . . Rates of dissolution in the "lower" Pecos Valley of New Mexico are significantly slower than those of models that limit most of valley and solution- subsidence depression formation to post-Ogallala time (past 4 Ma).

Quaternary Alluvial Deposits

The alluvial deposits of the Roswell groundwater basin are generally mapped as

Quaternary in age. The alluvium is derived from the western Pecos Slope, Capitan

Mountains, Sierra Blanca Mountains, and from sediment transported by the Pecos River.

Tributaries east of the Pecos River transport materials from Triassic, Tertiary, and

Quaternary age deposits. The tributaries west of the Pecos River derive their sediment

from Permian sedimentary strata, Tertiary igneous intrusives and extrusives, and older

alluvial deposits. The gravelly to fine-grained beds of late Cenozoic age that compose

the older valley fill have been left essentially unstudied beneath surface exposures. The

deposits are heterogeneous and lithologically complex.

The surface distribution of the Quaternary deposits was mapped by Kelley (1971).

Kelley's maps augmented by Landsat imagery and field observations form the basis for

observing that even the lithified strata of Quaternary age display a spatial pattern that is 72 influenced—if not directly controlled by—the northeast trending strike-slip faults that cross the Roswell groundwater basin. Figure 2.8 is presented to illustrate that spatial relationship. Discussion of the faults is deferred to the next chapter, Regional Tectonics. 73

100 105

T311

75S

70S

175

T8S

TSS

TiOS

712S

T13S

Western boundary of shallow Quaternary unconfined aquifer I 5 '' ' 1 - • •

....M9,11Mr,4inMr:aM2Mniqrz.M.rt;MikInriAZ- 40 Kilometers 4 &Mine of Roswell 25 Mlles Groundwater basin and recharge area Tertiary intrusives Stri ke-slip fault Quaternary undifferentiated

Figure 2.8. Generalized distribution of Quaternary deposits and western boundary of the shallow Quaternary unconfined aquifer. Primary sources: Kelley (1971) and Welder (1983). 74

106

T3M

T2N

F -_ T1M r glfeto c.ree , 1- 1S F DEBACA : ..", T2S V 4 1 g .... 4 t---1- a 1

.1e&

T5S

TSS E ?p,xtoVo T7S Capitan Mtns.

Pilamogordti

s(xidfie Kin

„. OTERO

Tertiary intruslves Strike-slip fault

Approximate western edge of confined aquifer

Figure 2.9. Generalized western margin of the confined artesian carbonate aquifer within the Roswell groundwater basin and recharge area. Modified from Kinney et. al (1968), Maddox (1969), and Welder (1983). 75

Aquifer Stratigraphy

The Roswell groundwater basin is composed of three principal aquifers in which storage occurs. The confined carbonate aquifer includes the San Andres Formation and/or the Grayburg Formation. The unconfined carbonate aquifer is also composed of the San Andres Formation and/or the Grayburg Formation. The unconfined shallow aquifer is in Quaternary alluvium and sedimentary units. Figure 2.8 shows the distribution of the undifferentiated Quaternary deposits. Figure 2.9 illustrates the western extent of the confined carbonate aquifer (Kinney et. al., 1968; Maddox, 1969; Welder,

1983).

Permian Aquifers

The formations that become aquifers of the Roswell groundwater basin are of

Permian Leonardian, Guadalupian (see Table 2.1, page 56), and late Tertiary-Quaternary age. In the topographically higher western parts of the Roswell groundwater basin the older formations act as the aquifers. The younger formations compose the aquifers in the eastern portion. The upper Yeso Formation acts both a local aquifer and as a transmission zone for the movement of groundwater toward the east and ultimately the

Pecos River.

No single stratigraphie formation, unit, or horizon is solely responsible for the movement of groundwater from the recharge areas to the aquifer reservoir areas within the Roswell groundwater basin. The geologic, hydrologic, and geomorphic conditions 76 prevailing across the large recharge area result in a complex delivery system. The western edge of the confined aquifer has been delineated and is shown in Figure 2.9. The western edge of the unconfined Quaternary aquifer is shown in Figure 2.8.

In many areas of the western Pecos Slope where Yeso Formation is exposed, the sandstones of the formation serve as the groundwater transmitting horizons. A stratum acting as a groundwater conveyor is often not a good aquifer unless local structural or stratigraphie conditions allow for accumulation of the moving water. The dips of the

Permian strata on the Pecos Slope are very gentle; they are less than five degrees to the east-southeast (except in areas of structural deformation). The Glorieta Sandstone is a principal conduit for recharge water to move from the mountainous areas of the west toward the east and the reservoir portions of the aquifers of the Roswell groundwater basin.

Where exposed, the Yeso Formation sandstones absorb meteoric waters and surface runoff. The zones of enhanced porosity and permeability in the strata exposed at or very near the surface may cross depositional bed boundaries—especially in the sandstones interbedded with thin carbonates or evaporites. As the eastward dip of the erosional surface is less than the dip of the formations, the solution-enhanced porosities and permeabilities migrate stratigraphically upward and by that cross bed boundaries of progressively younger strata.

As the recharge water is transmitted toward the eastern portion of the Roswell groundwater basin, it progressively moves stratigraphically into younger strata where enhanced porosity and permeability are present. Carbonates of the San Andres 77

Formation thicken eastward from the western crest of the Sacramento Mountains owing to less erosional stripping. The fractures, original and secondary porosity, and solution features of the gradually thickening San Andres Formation carbonates expand the thickness of the vadose zone from the west toward the east. This is due to the general absence of an effective underlying confining horizon within the carbonate section of the formation and the presence of good transmission strata beneath.

In the recharge areas to the west of the Six Mile Hill fault zone (refer to

Figure 2.9) the groundwaters percolate downward through the San Andres Formation carbonates until they encounter the Glorieta Sandstone underlain by the less porous and permeable Yeso Formation. With the upper Yeso Formation acting as a basal aquiclude, the Glorieta Sandstone becomes the primary conduit for transmission of the eastward moving groundwater. Locally the distribution of porosity, solution features, and fractures in the San Andres Formation will deter or temporarily detour the downward percolation of groundwater into the Glorieta Sandstone. Retarding the downward movement of the groundwater locally creates perched water table conditions. These local conditions create springs and provide some domestic wells with water producible from the San

Andres Formation carbonates. Throughout the area of transmission, the moving groundwater in the Glorieta Sandstone causes most wells constructed in the formation to be low- capacity producers.

Transmission of groundwater eastward in the Glorieta Sandstone (also see discussion on page 49), and sometimes within zones of enhanced porosity and permeability within the San Andres Formation toward the Pecos River Valley, requires a 78

significant upward stratigraphic shift as the water crosses the Six Mile Hill fault zone onto the Roswell block. On the Roswell block the confined artesian carbonate aquifer is

developed within and below the solutional melange in the uppermost San Andres

Formation and lowermost Artesia Group undifferentiated. The transfer of groundwater from the lower San Andres Formation and Glorieta Sandstone into the aquifer on the

east-side of the Six Mile Hill fault zone requires structural -stratigraphie juxtapositions as

accomplished by strike-slip faulting. The changes accompanying the switching of

aquifer units are reflected in transmissivity, hydrochemistry, and in the nature of flow path development.

In the Y-0 block, the artesian aquifer zone is within the lower San Andres

Formation. In that portion of the Roswell groundwater basin, the greater thickness of the

San Andres Formation shows that it has not been as deeply beveled. The solutional activity of the fresh groundwater has been concentrated within the lower-porosity horizon so that the artesian aquifer is developed at a much greater drilled depth below the surface and a much lower stratigraphic horizon than seen on the Roswell block.

The upper aquiclude (aquitard of some workers) on the Roswell block includes the shales, carbonates, and evaporites of the undifferentiated Artesia Group (Havenor,

1968; Kinney et al., 1968; Maddox, 1969; Welder, 1983). While the upper aquiclude has been catagorized by Hantush (1955) as a leaky aquitard, that conclusion is questioned here, by Havenor (1968), and Barroll (1993). The underlying aquiclude of the carbonate- 79 evaporite aquifers is comprised of dense, low-porosity carbonates and evaporites of the

San Andres Formation.

South of the KIM fault zone, near the Pecos River, problems of water quality and sediment content described to me by the Artesia Water Department suggest the Glorieta

Sandstone and the Yeso Formation might be the aquifer horizons for their municipal water well field. Insufficient well and sample analysis data was available to investigate the stratigraphic relationship of that problem area in this study.

The lithology and juxtaposition of the Seven Rivers Formation to the Quaternary alluvial shallow aquifer in the Pecos River Valley suggest that the Seven Rivers

Formation be an important part of the shallow aquifer system. Kinney et al. (1968) suggest that the hydraulic parameters of the Seven Rivers Formation would closely approximate those of the shallow alluvial aquifer.

In the reef margin the Seven Rivers Formation forms the top of the carbonate sequence composing the Capitan Reef complex. Reef pinnacles of Seven Rivers

Formation dolomites are hydrocarbon-productive because the very localized reef buildups form traps into which hydrocarbons accumulate. Beneath the reef pinnacles the water within the Capitan Reef moves preferentially around the basin, the Central Basin

Platform, and ultimately toward the Sheffield Channel in West Texas. Evidence for the movement of water through the reef is that salinities in the reef increase rapidly from near meteoric salinities on the west, near the surface, to oil field brine water salinities eastward from Carlsbad where the reef plunges into the subsurface. Potientiometric gradients in the Capitan Reef aquifer also prove that movement (Hiss, 1971). 80

Along the southern margin of the Roswell groundwater basin, close to Carlsbad, the Seven Rivers Formation is exposed at the surface over a considerable area. The newly completed Brantley dam and its reservoir are developed on the Pecos River in the thin alluvial deposits and the immediately underlying Seven Rivers Formation. The significant regional porosity of the Seven Rivers dolomites and the regional flow of groundwater from the surface into the Capitan Reef (the Capitan aquifer) has long been recognized (Hiss, 1971).

From the Pecos River and the surface exposures of the Capitan Reef, the groundwater flow in the Capitan aquifer is regionally east to the Central Basin Platform, then south into Ward and Winkler County, Texas, via the reef. Measurements of the salinity of subsurface waters show a progressive increase in chloride concentration along a line from the Capitan Reef south of Carlsbad, New Mexico, along the Pecos River channel area, across the reef, and eastward a total distance of approximately 40 km. The salinities range from less than 100 ppm to more than 20,000 ppm (Forest Miller, 1980, pers. comm.). The Capitan Reef acts as a natural outlet or drainage for water moving into the Seven Rivers Formation, from direct contact with the Pecos River, or contact with the unconfined Quaternary aquifer above.

The southern exposures of Yates Formation at the Brantley Dam reservoir and in the area southeast of the Brantley dam both expose high porosity, permeable strata for the transmission of water southward. It is more than reasonable to suggest that many thousands of acre-feet of water are annually lost to beneficial use by down-stream water users due to its seepage into the subsurface and subsequent movement southeast. While 81 the seepage is recharge to the regional Capitan aquifer system, its quality rapidly worsens

by mixing with oil field brine waters within the formation.

Quaternary Aquifer

The Roswell groundwater basin's shallow aquifer is principally developed in

sediments and rocks of Quaternary age. By local convention, the Quaternary aquifer

would include any sediments or rock units composed predominately of quartz sands and

conglomerates and would be called the Quaternary "alluvial" aquifer. This local usage

would then include upper Tertiary deposits such as the Ogallala Formation, the Gatufla

Formation, or the calcretes of Hawley (1993) and Machette (1985)—if they were found

present. In the southeastern portion of the Roswell groundwater basin the Quaternary

alluvial aquifer rests unconformably upon strata of the middle and upper Artesia Group.

Where the shallow alluvial aquifer is in hydrologic communication with the porous and

permeable Yates and Seven Rivers Formations, it is often difficult to closely delineate the

contact between the Permian and the Quaternary.

The Quaternary alluvial aquifer has at least two major divisions. Any subdivision

of the shallow aquifer should define the aquifer as in the unconsolidated, very young

sands and gravel of the Pecos River channel, or in the older, consolidated, floodplain

strata. The youngest sediment is very loose sand, silt, and gravel that fill the channel area of the Pecos River. The older alluvium is the result of floodplain deposition by the

Penasco, Felix, Hondo, Spring, Berrendo, Salt, and Arroyo del Macho tributaries to the

Pecos River and would correspond to the Gaturia Formation of Hawley (1993). 82

The floodplain deposits reflect that the courses of the Pecos River tributary streams were meandering to braided. The extent of the alluvial deposits show that the early Quaternary stream-flow volumes were much greater than those recorded in the past century. The patterns of deposition can be readily seen on satellite imagery. They are also readily apparent in the distribution of surface soils in the northern portion of the

Roswell groundwater basin (J. W. Hawley, unpubl. maps, pers. comm. 1994).

The rich textural and lithologic variation in floodplain deposits—vertically and especially laterally—will strongly affect both recharge and water migration. Major channels such as developed by the Hondo River will have more continuous lateral porosity and permeability than will the lenticular overbank deposits. The substantial variations in lithologies of the Quaternary deposits bear heavily upon the ability of the shallow aquifer to interact with the Pecos River's gain or loss of water during flow across these different facies of the shallow aquifer system.

Although the shallow aquifer has been considered an isotropic and homogeneous unit it is far from that. Great variability is apparent in the composition, texture, and structure of the deposits that comprise the shallow "alluvial" aquifer. Great variability should therefore be expected, and is observed, in the hydrologic properties from different portions of that aquifer (Kinney et al., 1968; Maddox, 1969; Welder, 1983).

The classification of the aquiclude separating the artesian aquifer and the shallow

"alluvial" aquifer by Hantush (1955) as a leaky aquitard was probably the basis for the general acceptance of the idea that an important amount of recharge to the shallow aquifer—and thereby to the Pecos River—has derived as upward leakage. My 83 questioning the leakage from the artesian aquifer stems from the composition of the

Artesia Group with an abundance of impermeable red shales and the lack of thermal changes in the temperature survey study by Barroll (1993). This clearly raises the question of if, where, or how the waters of the artesian aquifers in fact recharge the alluvial aquifer and thereby contribute to the flow of the Pecos River. 84

REGIONAL TECTONICS

Introduction

An understanding of the local and regional tectonic history of southeastern New

Mexico is necessary to appreciate the interaction of structure, stratigraphy, porosity and

permeability distributions, the origin, distribution, and controls of varying hydrologic

conditions in the Roswell groundwater basin.

In the introductory chapter it was stated that one of the purposes of this

dissertation is to show that structures generated during different tectonic episodes from the Precambrian to the present have influenced Cenozoic groundwater processes and

chemistry in the Roswell groundwater basin. The area containing and surrounding the

Roswell groundwater basin was involved in an unknown number of Precambrian and three episodes of Phanerozoic deformation: post-Meramec Mississippian to Late

Permian, Late Cretaceous to early Tertiary Laramide, and middle Tertiary through

Quaternary. Each episode left a unique signature on the region.

The purpose of this section is to review previous work on the tectonics of the

region and propose some new interpretations of the tectonics. The new interpretations

are justified partly on the basis of the detailed chemical work in the latter part of this 85

a 10S 105

T3N

T211

T1 Ii

TTS

T2S

T5S

TSS

T7S

40 KlIwneters Outline of Roswell New Mexico o 25 MI1,25 Groundwater basin and recharge area

Area Map

Figure 3.1. Location of Capitan and Sierra Blanca Mountains in the study area. 86

study and will be discussed further in the illASkAE° 11 section on hydrochemistry and hydrogeo- IAOR -04e9.1 Z '— Guadalupe Was , MIDLAND . Oft, (.:. '. P chemistry. In the following discussion, EASTERN SHELF LAWARE BASIN 0, ST".11r I 4/64 0 tO / / important elements of the structural bASI 111 7 50 a AZ le 1.- // l/ Z 4). 0. F.h.. ,9 , S / environment of the Roswell groundwater HEPFIELO CMANHEL / .#°' - - di' 14‘. SOUTHERN 4» c...... _ SHELF 4C31 8/A I N M AR PkT173 04 basin include the buried Pedernal uplift MOEALE AS AA,cf CO I 1 and the Tertiary tectonic activity •Ii O SO 6O xrn ModTed Rom King 1948 Figure manifested by the Capitan and Sierra 3.2. Regional paleogeographic setting of the Guadalupian Permian. Blanca Mountains (Figure 3.1). Modified from Sonnenfeld, 1991.

The regional paleogeographic

4, s \ RI RICE \ NI elements of southeastern New Mexico R,, \s, N t n • 91R\1 and western Texas discussed \\ N in the text \ \ N,.. N W V ' N \ ' \ n \.\\ :'• N '.‘, are shown again in Figure 3.2. ' N.' \ Ck 'N'‘,.,,, ' Pi

Par .s\ \ N. Precambrian Structures PO N P.' Mid 0 14 t n The Precambrian tectonic history Rob -IT • • • A% Lein I:‘ .s.:- It,. T MS \,..,\,...x.\,\ -4i41t4 Nrprii. `111 TI35 . I MILE , of the region is limited by the lack of \\ Par 1 kikit basement rock exposures. Two areas of

Precambrian exposure are known. One is Figure 3.3. Geologic map of Pajarito Mountain area (From Bowsher, 1991, a small inlier of melasyenite, syenite, New Mexico Geological Society). gbr - gabbro; PC - Precambrian syenites; Psb - quartz syenite, alkali granite, and Bonney Canyon Member of San Andres Fm; Psr - Rio Bonito Member of San pegmatite at Pajafito Mountain on the Andres Fm; Py - Yeso Fm. 87

Mescalero Apache Reservation (Kelley,

9. 1971; Moore et al., 1988; Bowsher, 10 _____/------.,,,,,,—,---•^"

. • Cad,. , 01 -20 1991; Foord and Moore, 1991) is shown -mow 33 3cr

Had. innie in Figure 3.3). Pajarito Mountain is an

\ 20

7N. e"--p.jardo erosional remnant of the Pedernal uplift -1

tient \ (Kelley, 1971) onlapped by Leonardian \ V raderota -9 n oo. age Abo Formation. The second, the .7 ( _20 • 00,..k,al 0 _ K.

a „,IV,1 \:.... 0 Bent dome, is a small area near Bent, ni do -1°".•-,-dOunken

- 'a • gacrarrlyn o about 8 km west of Mescalero. The cp - 0 exposure is a few hundred square meters • Pinon 0

c--- -._ f? (Lindgren et al., 1910; Bachman, 1954, 07 z -i) m ,o( MU 2 -.....\ / ' 6 '''''' ib ;r). icr Figure 3.4. Bouguer gravity map of the 1960; Foster, 1959; Moore et al., 1988; Sacramento Mountains. From Bowsher, 1991, New Mexico Geological Society. and Foord and Moore, 1991). The Bent

dome is a core of diorite and granite overlain by Cambro-Ordovician sandstone and

limestone. Foord and Moore (1991)

concluded that the Bent dome is a Precambrian core uplifted as early as Cambrian to as

late as Tertiary.

Precambrian rocks are known from the subsurface and are summarized in

Table 3.1 (Bowsher, 1991). In addition, basement rock has been investigated using

geophysical methods. Bowsher (1991) proposed that the high intensity magnetic and gravity anomalies concentrated east and south of Ruidoso may be the result of intrusions

of layered ultramafics into syenites and gabbros (Table 3.1) dated ranging from 1.1 to 88

Well Name Location Lithology Comments Yates Petroleum 8 km SW Lincoln 36 m gabbro dated at 1284 Ma Corp #1 Munoz Sec 10 T10S-R15E Yates Petroleum Sec 15 TI8S-R15E 30m diabase and Precambrian by Corp 41 Dog gabbro with stratigraphic Canyon abundant antigorite position Yates Petroleum Sec 18 TI8S-R15E 3.65 m diabase Precambrian by Corp 41 One Tree approx 1/2 antigorite stratigraphic position Yates Petroleum Sec 29 T18S-R16E 84 m diabase with Precambrian by Corp 42 One Tree magnetite and stratigraphic abundant antigorite position Gulf Oil Co 41U Sec 10 T18S-R16E 23 m marble and Precambrian by talc, 93 m stratigraphic metarhyolite and position talc Lubbock Machine Sec 3 T16S-R16E 15 m pink granite Precambrian by & Tool 41 or granite gneiss stratigraphic Anderson position National Sec 21 T11S-R18E bottomed in dark Possibly Precam- Exploration 41 brown "granite" brian mafic igneous Picacho Souther Production Sec 5 T17S-R21E 28 m gabbro, Precambrian by Co. 41 Cloudcroft diabase, monzonite stratigraphie position Table 2.1. Precambrian rocks in various oil-gas test wells reported by Bowsher (1991).

1.284 Ga. A similar feature, the Central Basin Platform, has been confirmed by the drilling of about 5 km of mafic intrusive rocks dated at 1.1 Ga (Keller et al., 1989). The pronounced negative gravity anomaly might be a Proterozoic basin up to 8 km deep

(Roberts et al., 1991) and is shown in Figure 3.4. 89

Evidence of middle Proterozoic extension in the Sacramento Mountains and the

Central Basin Platform of southeastern New Mexico (Figure 3.2) is present as mafic igneous intrusive bodies delineated by gravity and magnetic anomalies (Adams and

Keller, 1994). This extension may have caused the features described by Bowsher (1991)

and Roberts et al. (1994).

The consensus in the literature is that Precambrian structures influenced and

possibly controlled the locations of Phanerozoic structures. Although, as seen from the

descriptions above, little evidence exists to illustrate exactly where or to what extent.

Early and Middle Paleozoic Tectonics

In the New Mexico portion of the Permian basin, and particularly on the

Northwestern Shelf (Figure 3.2), there is a high degree of consistency to the folding style

of the lower to middle Paleozoic strata. The folds usually incorporate Cambro-

Ordovician through Mississippian rocks that are fault-bound on the western side of the

structures and are usually down-thrown toward the west. The antiformal axes commonly

strike north to north-northwest.

Within the study area, Paleozoic deposition was affected by the Pedernal uplift, of which Bend dome and Pajarito Mountain are probably a part. The Abo Formation unconformably overlies the early Paleozoic units directly next to the Bent dome and thickens away from it to overlie late Paleozoic formations (Pray, 1959; Foord and Moore,

1991; Baurer and Lozinsky, 1991). The Ordovician El Paso Limestone and the underlying Cambrian rocks appear to have been truncated by erosion, suggesting that 90 uplift occurred after the Ordovician (Baurer and Lozinsky, 1991). A well-developed

Permian regolith on the Precambrian rocks of Pajarito Mountain is overlain by

Leonardian Yeso Formation (Kelley, 1971).

The Pedernal positive area (Willis, 1929), also referred to as the Pedernal uplift

(Kelley, 1971; Meyer, 1966; and others), is known to have been a major contributor of elastic material for deposits of Pennsylvanian and Permian age and to have generally dominated sedimentation patterns. By implication, the uplift formed sometime before the

Pennsylvanian and possibly much earlier as is suggested by the previous discussion of the Bent dome. However, the origin, timing, and even areal extent of the Pedernal uplift is a source of general speculation. Adams and Keller (1994) attributed this positive feature to reactivation of Proterozoic structures during the Paleozoic. Other workers are less specific, regarding the uplift as simply pre-Permian (Meyer, 1966; Kelley, 1971;

Summers, 1972; Bowsher, 1991; and many others).

Late Paleozoic Tectonics

In contrast to the Precambrian through Middle Paleozoic, evidence for tectonic activity during the Late Paleozoic is abundant, partly because of subsurface data generated during exploration and development of this oil-rich region. This tectonic activity can probably ultimately be traced to the effects of the collision of Gondwana and

Laurasia. Major tectonic uplifts are recorded in the rocks of Late Mississippian (Chester) through Middle Pennsylvanian age and possibly later. These uplifts include the

Northwestern Shelf, Pedernal uplift, and a northwestward protrusion of the Marathon

Uplift (Kelley, 1971; Pearson,

LEGEND 1978; Chapin and Cather, 1981;

Basin,

Bowsher, 1991; Roberts et al., O Green River lips • Denver lips • Echo Pon lips

1991; Adams and Keller, 1994). Gereilleren *rest bell les s?

#. 141 ts b,e.f Evidence for these uplifts cornes 00 V low, principally from stratigraphic

CO -LOR ADOi `,--3 (,..}cc information. For example, [ -'no i . , !!} \ c e regional isopach maps of the j \l,,, ) )py .... „,,, , Mississippian rocks (Meyer, pLATEAu'i -., = ; : e , r L 1966; Havenor, unpubl. maps)

show that they pinch out around

the Northwestern Shelf.

174 Continental deposits grade into Figure 3.5. Map of the Rio Grande Rift and shelf and backreef facies and related tectonic and depositional features. From Chapin and Cather (1981, Fig. 1). See legend on Figure 3.6.

Map of the Colorado Plateau-Rocky Mountain area showing Eocene uplifts, basins, and selected structural features. Dotted line indicates approximate boundary of the Colorado Plateau in Eocene time. The over thrust belt (013), and the Mogollon Highland (MH) are two thrust-related uplifts that border the area to the west and south, respectively. Monoclinal uplifts of the Colorado Plateau are the White River (,VhR), San Rafael (SR), Uncompahgre (tin), San Juan (Si), Circle Cliffs (CC), Monument (Mon), Kaibab (K), Echo Cliffs (EC), Defiance (D), Zuni (Z), Lucero (Lue), Apache (A), and Morenci (Mor) uplifts. Basement-cored uplifts of the Laramide Rocky Mountains arc the 13cartooth (Bt), Bighorn (Bh), Black Hills (BPI), Owl Creek (OC), Wind River (WR), Granite (6), Laramie (L), Itartville (H), Uinta (U), Sierra Madre (SM), Park (P), Medicine Bow (MB), Front Range (FR), Sawatch (Saw), San Luis (SL), Sangre de Cristo (SC), Wet (W), Nacimiento (N), Sandia (San), Sierra (S), and Tularosa (T) uplifts. Green River-type basins are the Bighorn (Bhb), Powder River (PRb), Wind River (Wrb), greater Green River (ORb), Shirley (Sb), Hanna (Hb), Laramie (Lb), Uinta (Ub), Piceanec (Pb), San Juan (Sib), and Baca (13b) basins. Denver-type basins are the Denver (Db), Raton (Rb), and Sierra Blanca (SI3b) gasins. Echo Park type basins are the North Park-Middle Park (NPb), South Park (SPb), Echo Park (EPb), Iluerfatto Park (HPb), San Luis (SLb), Galisteo-El Rito (Gb), Carthage-La Joya (CJb), and Cutter sag-Love Ranch (CSb) basins. Figure 3.6. Map legend for Figure 3.5. From Chapin and Cather (1981). 92 finally into fine-grained basin deposits. The general paleogeography and major elements

(Fig. 3.2) are well known

Controversy centers around controls on the geometry of the uplifts and their effects on other tectonic features and events (Chapin and Cather, 1981; Bowsher, 1991; Roberts et al., 1991; Adams and Keller, 1994). The most important of these are effects on the geometry of the Rio Grande Rift, discussed in the section on Cenozoic tectonics. The zone within which the Rio Grande Rift now occurs (Figures 3.5 and 3.6) extended along the western side of the greater Roswell groundwater basin and into the Salt Basin (Figure

3.7), and predates the rifting as a tectonic feature (Chapin and Cather, 1981; Bowsher,

1991; Roberts et al., 1991; Adams and Keller, 1994).

Marathon Uplift

Pearson (1988) suggested the presence of a southward continuation of the Rio Grande

Rift system that he designated as the Trans-Pecos rift, a structural zone that continues for about 480 km southward. The Trans-Pecos rift extends from the vicinity of the

Sacramento Mountains, through the Salt basin, Valentine basin, Marfa basin and the

Tornillo basin, all four being in West Texas, to the Rio Grande River, and southward at least 48 km into Mexico. Pearson's Trans-Pecos rift, which is at least a lineament, would provide a feasible physical connection from the location of the Marathon orogeny to the present study area. This is an idea with considerable merit.

Pearson (1988) observed that the northwestward protrusion of the Marathon thrust in the Marathon uplift of West Texas may be the result of Tertiary right-lateral movement, 93 rather than as a Permian (Wolfcampian) salient into the Marfa Basin between the Diablo platform buttresses, as he had earlier proposed (Pearson, 1978). While there has certainly been right-lateral Tertiary wrench displacement along the Rio Grande Rift, as will be discussed later, the data presented by

Adams and Keller (1994), Bowsher

(1991), Roberts et al. (1991), Chapin and

Cather (1981), and Kelley (1971), clearly

suggest the origination of the entire zone

was much older than Tertiary (see

Figures 3.5, 3 6, and 3.7). Pearson's

generalized relationship of the location NEW MEXICO " TEXAS -

of the Salt basin graben of Otero County,

New Mexico, and Hudspeth County,

Texas, to some larger present-day

structures of the greater Roswell ground water 0 80 basin study area is shown in Figure KM 3.7. This model fits well with the Figure 3.7. Salt Basin graben and detailed tectonic map presented by structures of Roswell groundwater basin. From Pearson (1988). Kelley (1971). The northeastern edge of

Pearson's (1988) Salt basin would correspond to Kelley's (1971) Guadalupe fault.

Two episodes of tectonic activity occurred during the Marathon orogeny (King

(1937, 1978, 1980) that are consistent with the timing of major folding and uplifts in 94

A EARLY MISSISSIPPIAN B LATE MISSISSIPPIAN 0

_-, NMI Ti .-1 Ail-- - K NMI TX I O K AR - - i L\

‘410"

• - . ..

GONDWANA „.„-----111- V-- GONDWANA Figure 3.8. Paleogeographic reconstructions. Subduction dipping southward occurs south of microcontinents. Vertical dotted line area = microcontinents; shaded area = bathymetric lows; black arrows = bottom-current pathways in Ouachita basins; white arrows = direction of plate movement. From Nobel (1993). southeastern New Mexico. The presence of detrital materials derived from a possible island arc to the south of the Marathon basin is supported by Gleason et al. (1995).

Using paleogeographic and paleoceanographic information, Noble (1993) reconstructed the margin as shown in Figure 3.8. The closure of the seaway resulted in north-verging thrusts and a series of landlocked basins that filled with elastics and were progressively buried by carbonates and evaporites during the Permian.

Given this connection, it is here suggested that the zone now containing the Rio

Grande Rift, Roswell groundwater basin, and Salt Basin may have been the boundary area of a salient along the Marathon-Ouachita collision zone. That boundary would parallel Pearson's (1988) Trans-Pecos rift shown in Figure 3.7 and would be consistent with pre-Tertiary wrench faulting along the same zone (Beck and Chapin, 1991). 95

Summary of Late Paleozoic Tectonics

The observations of King (1937, 1978, 1980) that two episodes of tectonic activity occurred in the Marathon orogeny are also consistent with the timing of major folding and uplifts in southeastern New Mexico. The linking of structural deformation and uplift of Precambrian basement rocks of the study area to the regional deformation generated by the collision of North America and Gondwana during the Paleozoic is largely circumstantial, but it is not beyond the known evidence, nor is it unreasonable. In fact, the hypothesis offers potential solutions to various local and regional structural, tectonic, and stratigraphie events of the Phanerozoic. The Late Paleozoic deformational timing and subsequent regional Permian sedimentary onlap seem ideally explained by a causal relationship such as the Marathon orogeny.

Given the data presented by Bowsher (1991), Roberts et al. (1991), and Adams and

Keller (1994), concerning Precambrian rock distribution and crustal thicknesses, it seems

reasonable to consider adding a Rio Grande salient to Figure 3.8 to coincide with

Pearson's Trans-Pecos lineament and the left-lateral wrench system predecessor in the general location of the present-day Rio Grande Rift. Rifting, subduction and transform fault displacement probably have played far greater roles in the development of the Late

Proterozoic and Paleozoic paleogeography of southeastern New Mexico than have previously been allowed. There is good circumstantial evidence to postulate that even the earlier—as well as later—major regional tectonic features of southeastern New

Mexico and West Texas were related to crustal scale anomalies. These features may have been microcontinental pieces broken off the rifted and parting masses of Laurentia 96 and/or Gondwana during the Late Proterozoic and Early Paleozoic as suggested by Nobel

(1993) and Lowe (1985).

Beck and Chapin (1991) pointed to ancestral Rocky Mountain deformation in the

Late Paleozoic as having been a north-trending, divergent, left-lateral wrench zone extending through much of central and southern New Mexico. They further cited data in the Sacramento Mountains that suggests a similar period of north-striking normal faults that exhibit dominantly dip-slip movement with a definite left-slip component. These faults are characterized as ancestral Rocky Mountain in age.

The location of the left-lateral wrench proposed by Beck and Chapin (1991) is closely related to wrench faulting in the pre-Rio Grande Rift zone of Late Tertiary to Quaternary age (albeit the latter is right-lateral in its movement). Beck and Chapin (1991) recognized the reversal of the direction of wrench movement between the ancestral

Rocky Mountain and the pre-Rio Grande Rift periods. They also pointed out that

Laramide and Cenozoic structural overprinting has made reconstruction difficult.

However, repositioning of the earlier ancestral Rocky Mountain age movement by

Tertiary right-lateral wrench movement during pre- to early Rio Grande rifting would essentially remove the seemingly anomalous nature of left-stepping breaks in the otherwise north trending Late Paleozoic right-lateral wrench zone.

The Late Paleozoic tectonic effect on the development of the study area's tectonic framework as demonstrated in this study establishes that regional zones of non-incidental wrench-type faulting—as would be expected with major tectonic plate convergence—did occur in the area. As will be shown later, Tertiary and Quaternary movement along these 97

same zones generated strike-slip faults that created fault blocks in which the aquifers

were developed. The major Late Paleozoic structural activity generated by Ouachita-

Marathon tectonic movement is recognized here as creating the locations within this area

for wrenching.

Mesozoic Tectonics

In Triassic time southeastern New Mexico, along with most of the North American

craton, was undergoing general emergence. Within the study area rocks of Upper

Triassic age are mostly confined to locations parallel to and east of the Pecos River.

They are generally absent from the Pecos River to the crest of the Sacramento Mountains.

Within the study area, from the Pecos River to the crest of the Sacramento Mountains

on the west, Jurassic rocks are almost nonexistent. Kelley (1971) concludes that the area

had been reduced to a great peneplain by the Late Triassic and that Cretaceous Dakota

beds gradually overstep the progressively older Permian beds until they overlie the San

Andres Formation.

Cenozoic Tectonics

Considerable Tertiary tectonic activity has occurred in the Roswell groundwater

basin. This is manifested by the presence of the Capitan Mountains and Sierra Blanca

Mountains (see Figure 3.1, or 3.9). The Capitan Mountains is a major east-west intrusive

stock of Late Oligocene age (26.2 ± 1.2 Ma, Allen, 1988). Mafic dikes strike eastward from the Capitan Mountains area (Elston, 1991; Kelley, 1971). The Sierra Blanca 98

40KHOMNeM ClutlineafRosyvell 25 M IIem Groundwater basin and recharge area Tertiary intrusives Strike-slip fault

Figure 3.9. Location of major dextral strike-slip faults and respective structural blocks discussed in text. 99

Mountains is an igneous complex of Eocene to Oligocene (37.3 to 25.8 Ma, Cather,

1 99 1 ) felsic intrusions, extrusive rocks, and dikes. Fault block uplift on the west of the study area has created the major Sacrament dip slope to the east. The Tularosa Basin to the west of the study area further confirms the extensive tectonic and plutonic activity within the region.

The structural histories of some parts of the greater Roswell groundwater basin area are reasonably well constrained. Features such as the Sierra Blanca and Capitan intrusive complexes are closely dated (Segerstrom and Ryberg, 1974; Thompson, 1972; Moore et al., 1985; Allen and Foord, 1991). Other features such as the Six Mile Hill fault, Border

Hills fault, and Y-0 fault (Figure 3.9), are less constrained. The upper Permian carbonates of the Pecos Valley are overlain by Quaternary sedimentary rocks and alluvium.

Kelley (1971) described the faults of the Pecos country as (p. 44) "buckles." He observed that the surface expressions of the Six Mile Hill, Border Hills and Y-0 zones were both faults and folds with considerable evidence that the zones have strike-slip movement. He further noted that the sharply upturned beds range from a few tens of feet to approximately 1200 m wide. The Six Mile Hill zone was thought by Kelley (1971) to be as much as six km wide in part of its anticlinal expression. Kelley (1971) reported the vertical throw along these zones to be virtually from zero to about 150 m.

Kelley (1971, page 44) said the evidence for wrenching action was present as "(1) steeply plunging drag folds and left-branching spur faults along the axis or fault, and (2) echelon diagonal folds of some length between the buckles." Fiedler and Nye (1933) 100 reported that the Border Hills zone appeared to have thrust fault characteristics and even reverses itself at several places.

Most of the formations exposed in the western portion of the Roswell groundwater basin are dipping easterly at low angles. The dips are commonly between one-half and five degrees. In many areas of this large region gentle dips are disrupted by structural features that are also generally of low relief. These structural features include anticlines,

synclines, domes, and numerous igneous intrusions. Kelley (1971) mapped and

discussed these features in considerable detail.

The Dunken-Tinnie trend, shown in Figure 3.10, disrupts the gentle eastward dip of the sedimentary strata. This trend is discussed by Kelley (1971), Summers (1972),

Bowsher (1991), and Yuras (1991). Each attaches different significance to the trend.

Kelley (1971) considered the Tinnie-Dunken fold trend and the "buckles" of the area at

length.

The Lincoln folds may be coeval with the Tinnie folds, but the asymmetry and fold vergence show a direction of tectonic transport. Here, the westward vergence would

suggest a westward direction of movement (Davis, 1975; Yuras, 1991). concluded that the Lincoln folds were of late Laramide (Eocene?) age. The Tinnie folds have a relation to the timing of strike-slip movement and deformation. This is pointed out in that the

Tinnie folds terminate against the Capitan stock, but are younger than the stock (Kelley,

1971). The C apitan pluton was emplaced at + 26.5 Ma (Allen, 1988).

Many investigators have discussed these important fault zones of the greater Roswell groundwater basin. Some of these include Nye (1927, 1928), Bean (1949), Mourant 101

(1963), Havenor (1968), Maddox (1968), T8S Kinney, et al. (1968) and Kelley (1971). n It appears that of those only Kelley (1971)

began to appreciate the overall

significance of the complexities and

interrelationships of the normal and thrust

faulting, folds, reversals, splays, and other

features.

My early conclusions (Havenor, 1968)

considered the faults to be normal. Those

assumptions were strongly based upon

differential structural positions in the

subsurface around the City of Roswell.

While the apparent throw on the Y-0

fault and the Six Mile Hill fault may have

been locally correct, the regional

implications were wrong. Upon

extending my subsurface discovery of the

subsequently named K-M fault, I began to

realize the interrelationships of the struc-

Figure 3.10 , The Tinnie-Dunken fold tural features and their inferences. After trend. From Bowsher, (1991), New Mexico Geological Society. many discussions with Vincent Kelley 102

(pers. comm.,

1969- 1970) and

George Maddox

(pers. comm. 1968-

1969) I began to appreciate that substantial strike- slip movement along structural FA' the ‘4:- 41:, Figure 3.11. K-M fault shown on Landsati1 ,1 iBand 5, "buckles" within 07/04/85 image date. the Roswell groundwater basin had occurred—in addition to some local vertical uplift components such as along the Roswell block.

The realization of the extent of strike-slip faulting was strengthened by the surface and subsurface projection of the K-M fault (Figure 3.11). A major offset of the course of the Pecos River falls directly on the K-M fault's surface projection and suggests a right- lateral shift of ten km. No strike-slip stratigraphic offsets in pre-Tertiary strata have been mapped on the surface along the K-M zone. However, Kelley (1971, plate 5) mapped down to the south structural throw of about 60 m. It is apparent from the satellite imagery that a distinct lineation exists in the groundwater basin approximately parallel to the K-M fault zone. The K-M fault shows a geomorphic expression in the deviation in the Pecos River and the drainage development in Long Arroyo, located to 103 the northeast of the Pecos River. Additional support for the location of the fault is provided by the anomalous distribution of irrigation on the south of the K-M fault—which is here believed to be controlled by the K-M fault acting as a groundwater boundary in the alluvial and artesian aquifers.

Wrench Tectonics

Basic wrench tectonics were discussed by Wilcox et al. (1973). They describe common relationships of en echelon folds and conjugate strike-slip faults where low- angle faults, synthetic strike-slip faults, have the same sense of displacement as the wrench fault. The high-angle faults, antithetic strike-slip faults, have a displacement that is opposite to that of the wrench fault.

The above descriptions typify the structural features found in and beside the major faults within the greater Roswell groundwater basin (see Kelley, 1971, Plate 5; Figures

3.9 and 3.10 of this manuscript). Wilcox et al.'s (1973) description of wrench faults and their commonly associated features closely fit the general distribution of structural features of the Sacramento Mountains, Guadalupe Mountains, and greater Roswell groundwater basin areas. Comparison of the strikes of the Tinnie-Duncan fold belt with the strikes of the Border Hills, Six Mile Hill, and Y-0 faults show an almost ideal angular relationship as described by Wilcox et al. (1973). The wrench faults and compressional folds resulting from movement along the faults have developed very close to the right-lateral wrenching shown in Figure 3.10, on page 100. 104

The conclusion that strike-slip faulting has been primarily responsible for the

development of the Tinnie-Duncan fold belt and the major strike-slip faults on the eastern

side of the Pecos Slope leads to the question of when they were formed. The evidence of

structural offset of stream drainage and groundwater distribution strongly argues for—at

the least—late Tertiary to Quaternary movement along the K-M, Y-0, Six Mile Hill and

Border Hills faults (the major right-lateral strike-slip faults are shown on Figure 3.9).

The northeast to southwest strike-slip faults divide the Roswell groundwater basin

into a series of en echelon fault blocks

responsible for the displacement of SE

stratigraphic units and places beveled units Orchard Park block containing or in which the aquifers develop

Roswell block in juxtaposition as shown in Figure 3.12.

The right-lateral movement of the Roswell

block in relation to the Y-0 block (Orchard Figure 3.12. Thickness change of erosion-beveled San Andres Formation Park) on the south, along the northeast due to juxtaposition across strike-slip fault. Diagrammatic. Not to scale. trending Y-0 fault would serve to place a

thinner, erosion-beveled wedge of the lower San Andres Formation of the Roswell block

in juxtaposition with a thicker wedge on the Orchard Park (Y-0 block) side of the fault.

However, it should again be emphasized that wrench faulting is replete with complex movements and folding, including vertical uplift.

Havenor (1968) suggested that the area between the Six Mile Hill and Y -0 fault zones (termed the "Roswell block") had significant erosion and uplifting as compared 105 with the San Andres Formation in the subsurface found immediately south of the Y-0 fault. That early conclusion was based primarily upon the apparent differential thickness of the San Andres Formation between the Six Mile Hill block, the Roswell block, and the

Y-0 block (Orchard Park) immediately south of the Y-0 fault. In retrospect it seems logical to visualize the thickness of the San Andres Formation across the Roswell block as a function of right-lateral displacement along the Y-0 and Six Mile Hill strike-slip faults plus subsequent dissolution and/or erosion rather than just vertical faulting and subsequent erosion. That conclusion also better fits the sequence of regional tilting of the

Pecos Slope, down-slope erosional migration of the Pecos River, and strike-slip faulting of the Pecos Slope.

Kelley (1971) suggested that the Tinnie folds were younger than the Capitan Moun- tains emplacement. This would place the age of the folds at post ±26.5 Ma (Allen, 1988).

Kelley's (1971) detailed maps of the region show a few very small areas where the Qua- ternary alluvium (Qa) and Quaternary caliche soils (Qs) are shown to be involved in the folding and faulting rather than covering them.

As to the regional considerations of the causes of the strike-slip faults in the Roswell groundwater basin, Kluth and Coney (1981), Kluth (1986), and Beck and Chapin (1991) have discussed the Paleozoic ancestral Rocky Mountains deformation involving, in part, southern New Mexico. Beck and Chapin (1991, page 189) concluded that "the north- striking normal faults within the Sacramento Mountains may represent reactivation of basement structures." Their studies led to the proposal of an ancestral Rocky Mountain left-lateral wrench zone through central New Mexico. Movements attributable to that 106 late Paleozoic wrench movement would not account for the right-lateral sense of movement seen in the major faults of the Sacramento Mountains dip slope and the

associated right-lateral folds of the Tinnie-Dunken fold belt. Those observations then

lead to the deduction that the right-lateral fault-fold system was then at least post-

ancestral Rocky Mountain deformation.

Chapin and Cather (1981), and Chapin (1983), discussed the evidence for the general

development of the Rio Grande Rift since the mid-Tertiary (see Figure 3.5, and Figure

3.6, on page 91) and the evidence for a mid-Tertiary right-lateral wrench fault system that

deformed Eocene depositional basins developed in the present Rio Grande Rift. The

Sacramento Mountains fault scarp forms the eastern boundary of the Rio Grande Rift

zone and would be the approximate east side of the right-lateral mid-Tertiary wrench

system. Continuing tectonic activity along this eastern margin of the rift zone is

evidenced by the occurrence of Pleistocene to Recent fault scarps in alluvial deposits

near Alamogordo and Tularosa (Machett, 1987).

The discussion by Chapin and Cather (1981) concerning the Rio Grande Rift and the

earlier Tertiary development of a right-lateral wrench system is significant for

understanding the development and age of the major faults and fold belts of the Roswell groundwater basin. The major right-lateral strike-slip faults of the Pecos Slope are

attributed here to be synthetic strike-slip or wrench faults generated from movements of the right-lateral wrench faults described by Chapin and Cather (1981).

The ultimate cause of this wrench movement, according to Chapin and Cather (1981), can be related to the Farallon plate reaching its lowest angle in latest Pliocene to earliest 107

Eocene. The viscous coupling with the Colorado Plateau lithosphere caused a rotation in the axis of compression to about N 45° E. The first-order shears became roughly parallel to the trend of the southern Rocky Mountains in New Mexico and the Colorado Plateau began to decouple along right-lateral wrench faults.

Capitan Lineament

The east-west Capitan lineament (Allen and McLemore, 1991; Cather, 1991; Chapin and Cather, 1981; Kelley and Thompson, 1964) is a well-recognized lineation of major

length and tectonic significance.

Chapin and Cather (1981) referred

Colorado to the displacement of the Capitan Plateau lineament that crosses the Rio Pecos Slope Grande Rift as evidence for the Capitan Lineament right-lateral wrench movement • Capitan Socorro Pluton along the margins of the younger Basin CLI

'Ruldoso Roswell Rio Grande Rift. L- The lineament (9 Range extends from the Matador-Electra o • Carlsbad arch of West Texas westward

through Datil, New Mexico near

the Arizona-New Mexico state Figure 3.13. Trend of the Capitan Lineament in New Mexico. Diagrammatic representation after line, into Arizona. The offset of Allen and McLemore, 1991, New Mexico Geological Society. the Capitan lineament is shown by 108

Allen and McLemore (1991) in their analysis of the distribution of post-Laramide Rocky

Mountain alkalic igneous rocks (see Figure 3.13). Kelley and Thompson (1964) report

left-lateral movement along the Capitan lineament of 15 to 30 km based upon offset of

structural contours on top of the Permian strata on either side of the lineament. The left-

lateral displacement along the Capitan lineament would fit the mode and angle for

antithetic movement due to north-south right-lateral wrench fault displacement. The

suggested north-south displacement for the Tertiary right-lateral wrench fault is in the

magnitude of 100 km (Chapin and Cather, 1981).

The Capitan Mountains with the Transverse Range of southern California and the

Unita Mountains of northeastern Utah are the three major east-west ranges in the United

States. The Western Transverse Range of southern California has been described by

Molnar and Gipson (1994) as a piece of crust rotated clockwise during the past 15 Ma,

resulting in its east-west orientation. Bohannon and Parsons (1995) discuss at length the

extensional tectonics of the Basin and Range Province as related to the North American

continent overriding the East Pacific spreading center. Their descriptions of California's

Basin and Range tectonics offers interesting similarities for considering causes of the

east-west orientation of the Capitan Mountains.

Kelley (1971) mapped a right-lateral strike slip fault, a splay from his Bonito fault, that cuts through and offsets the Capitan Mountain intrusive. That splay, by virtue of cross-cutting relationships, would then set the age of that right-lateral fault at younger than 26.5 Ma (see page 105). The regional relationship of the faulting and folding of the area suggests they too are younger than the implacement of the Capitan Mountains. 109

To the east of the Capitan Mountains, but still west plus immediately east of the

Pecos River, the terraced Quaternary alluvial deposits and the "caliche soil" deposits have been cut by the Six Mile Hill and the Border Hills faults (Kelley, 1971, Plate 2, also see fault location map, Figure 3.9 of this manuscript). Older calcified Quaternary alluvial deposits of the area underlie the Quaternary terrace deposits and have been estimated by

Birkeland et al. (1991), and Machette (1985) to be from 0.5 Ma to as much as 5 Ma.

Hawley (1993) suggests the calcretes may be as old as late Miocene (±12 Ma). It is possible to suggest—albeit on sparce data—that the right-lateral strike-slip movement in the Roswell groundwater basin could be as recent as 0.5 Ma. With the cooling age of the

Capitan intrusive at ±26.5 Ma, it is quite possible that Late Tertiary wrench faulting and strike-slip movement continued into Pliocene, Pleistocene or possibly even younger time.

Cenozoic Tectonics Summary

Laramide (?) through Early Tertiary displacement along the right-lateral pre-Rio

Grande Rift wrench was followed by Eocene through Oligocene volcanic activity and deep-seated emplacement of the Capitan pluton. The decrease in the angle of subduction of the Farallon plate (Coney, 1987) resulted in compressive stresses that caused decoupling of the Colorado Plateau and its right-lateral wrench movement toward the north (Chapin and Cather, 1981). Movement along the eastern side of the Rio Grande

Rift, right-lateral wrench fault resulted in right-lateral strike-slip or wrench faults in the

Pecos Slope. Accompanying the Pecos Slope fault development was the generation of a right-lateral fold belt and the Tinnie-Duncan anticlinorium. 110

Based upon the cutting and truncation of Quaternary alluvial deposits younger than calcified alluvium estimated to be 0.5 Ma to as much as 5 Ma (Machette, 1985), the right-lateral Border Hills and Six Mile Hill faults could be as young as Late Pliocene to early Pleistocene. Conversely, they are not older than the ±12 Ma late Miocene upper age for the calcretes discussed by Hawley (1993). The K-M fault offset of the Pecos

River stream bed near Artesia suggests that a minimum right-lateral displacement of

6 km occurred since the stream established its position on the north side of the fault relative to the present south side. Even concurring with Hawley's (1993) inclination that the Pecos River's course has remained close to its present location during the past 10-12

Ma, does not diminish the probability that lateral stream migration down the stratigraphic and structural slope has occurred since Late Miocene. The strike-slip faulting, differential uplifting of the wrenched blocks, the progressive erosional and solutional alteration of the wrench faulted blocks, the geomorphic drainage, and land-form expressions all argue a need for more time than 0.5 Ma. On-the-other- hand, the Y-0 fault offset of the Pecos River at Roswell, the K-M fault offset of the Pecos River and

Long Arroyo near Artesia, and similar offsets along Salt Creek, Rio Hondo, Rio Felix, and Rio Penasco all argue strongly for the more youthful that ±12 Ma age for the faulting. This conclusion then suggests that the strike-slip faulting is younger than any previously proposed. These combined events are then genetically related to on-going, recurrent activity in the Rio Grande Rift.

The Cenozoic tectonics caused the Roswell groundwater basin to be sliced into a sequence of northeast trending structural blocks. This allowed the seemingly erratic 111 carbonate aquifer development by Tertiary-Quaternary solution development. The significant movement resulting this faulting, especially as suggested by displacement of the Pecos River along the K-M fault, imposed the structural-stratigraphic controls observed on recharge, groundwater flow, and the hydrochemistry of water in the artesian and alluvial aquifers of the Roswell groundwater basin.

To understand the hydrology of the aquifers of the Roswell groundwater basin it is vital that the geological structure of the region be fully incorporated. Wrench faulting has modified the distribution of the Permian strata and the development of the aquifers.

The hydrology of the Roswell ground-water basin has a very strong temporal and spatial relationship to the tectonics of the region. 112

HYDROCHEMISTRY AND HYDROGEOCHEMISTRY

Introduction

Two objectives of this study are to evaluate hydrochemical and hydrogeochemical techniques for 1) discriminating between the lithologies of the aquifers, storage areas, and recharge areas of the groundwater basin and 2) using these techniques as geological tools to help delineate structural and/or stratigraphic boundaries in major portions of a groundwater basin. With an adequate density and distribution of water analyses, the hydrochemical makeup of a large carbonate-evaporite dominated basin should be discernable because of the water-rock reactions that have occurred. As the geological history of an area is revealed from the layers of its rocks, so should the history of its waters be revealed from its hydrochemistry.

The hydrochemical and hydrogeochemical applications employed in this study were considered and selected for three basic purposes. Consideration was first given to their potential as regional methodologies to study carbonate or carbonate-evaporite groundwater basins. The second consideration was to develop methods that could use data that were already available or could be obtained without incurring substantial costs. 113

The third consideration regarded future studies of the interaction of the Roswell groundwater basin and its relationships with the Pecos River.

Hydrochemical and, to a lesser degree, hydrogeochemical and isotopic analyses have been used for both the derivation and support of geological data in many areas. Only in recent years have they been used as a tool for studying groundwater flow and mixing.

Winograd and Thordarson (1975) successfully used hydrochemistry to help them in the

Ash Meadow basin to (page Cl) "determine the direction of ground-water movement in the major aquifers and aquitards" in the Amargosa Desert of the Great Basin of

California and Nevada. Their study is of interest owing to its involvement with both carbonates and complex geological structure, as well as their effective early hydrogeologi cal use of hydrochemical data.

Unfortunately, Winograd and Thordarson (1975) were probably more constrained in their analytical applications because U.S. Geological Survey computerized tools such as

Wateq (Ball and Nordstrom, 1991; Nordstrom and Munoz, 1994) and NetPath (Plummer et al., 1994) were as yet undeveloped. These tools, and others, provide the potential power for basic geological utilization of hydrochemical data in the development of the hydrogeological framework of a groundwater basin such as the Roswell basin. Other significant studies of hydrochemical, hydrogeochemical, and isotopic applications are discussed by many other investigators including, but not limited to, Mazor (1991), Faure

(1991, 1986), and Phillips et al. (1989).

The Roswell groundwater basin is not different from any other developed groundwater source in that anthropogenic effects have been profound. Use has 114 historically affected supplies. In the Roswell groundwater basin State management was imposed to control over drafting. As a result, water levels have been observed to effectively but gradually (in a nongeological time perspective) recover. Saline encroachment in parts of the basin has fluctuated. The flow of the Pecos River across the basin has not significantly changed in response to elevated water levels in the basin during the past quarter-century.

Water table elevation maps and potentiometric surface maps for the artesian aquifer of the Roswell groundwater basin were prepared by Welder (1983) for the Office of the

New Mexico State Engineer for the years 1926, 1950, 1969, and 1975, and by Maddox

(1969). In the use of these maps over a period of years I have been greatly concerned by the observation that the aquifers were shown to be connected, essentially homogeneous

(save major areas of differing transmissivity). The disturbing condition is that it can be easily shown that some down-gradient areas of the artesian aquifer contained lower concentrations of chlorides and sulfates than did areas immediately up-gradient. If the water quality, expressed as mg/L chloride, was an order of magnitude higher in the up- gradient well as compared with the down-gradient location, how then does the subsurface artesian aquifer water—as depicted by conventional hydrology— move down-gradient through the confined (and by that separated) aquifer and improve in quality?

Questions of the above nature are common and should logically be answerable by use of hydrochemical, hydrogeochemical and isotopic (when available) analyses. The solution revolves around the basic idea that I believe can be expressed as: 115

Water falls as precipitation upon the rocks; what is not evaporated or transpired runs across the soil and the surface rocks; it is absorbed into the pores of the rocks; it travels through the pores of the rocks, and it is produced from the rocks. The rocks control the water's quantity, quality, location, and movement.

With the availability of Wateq or similar programs, handling a large number of water analyses is possible. Wateq evaluates the usability of a water analysis in terms of ionic balance, speciation and mineralogical phase, and arrives at potential rock compatibilities.

Wateq4f is the version used in this study.

Methodology

In order to make hydrochemical and hydrogeochemical interpretations to identify water-rock reaction signatures and aquifer lithology indicators the study was divided into four stages. Stage one: Accumulate the data. Stage two: Analyze the data using Wateq and conventional chemical computations. Stage three: Graphically plot the data using

Ternary diagrams, Piper plots, Fingerprint diagrams, mineral phase stability diagrams, and other graphical displays from the reconstruction of dissolved minerals. Stage four: interpret the data.

Stage One: Data Accumulation

Published sources of water analysis data for the Roswell groundwater basin are very limited. The principal source of water analyses has been Dr. Douglas McAda of the

United States Geological Survey (USGS), Albuquerque, New Mexico. This data set 116 includes analyses on 1,884 water samples from wells and 104 spring samples. The

USGS format for reporting the data includes about eighty items—the great majority of which are absent and not reported. The lack of reporting is due to the ions or elements not being among those surveyed or their being below detection limits. The reported data on each sample ranged from a minimum of chloride concentration to containing the more usable conventional calcium, magnesium, sodium, potassium, bicarbonate, sulfate, fluoride, nitrate and silica concentrations.

The USGS water analyses have been accumulated over a period of more than fifty years. Most of the samples were collected before 1970. Samples are most concentrated within the irrigation areas of the Pecos Valley, but representative samples are present for widespread portions of the Roswell groundwater basin and its recharge area.

Water analyses from the USGS were tabulated into a spreadsheet. Only 233 of the the wells contained sufficient ion analyses data to be considered for use in this study. A database was compiled containing the 233 analyses and is presented in condensed form in

Appendix A, beginning on page 218. The well and spring analysis data considered most adequate and of best possible use in this study was only 82 of the wells in Appendix A.

Stage Two: Data Analysis Using Wateq and Conventional Chemical Computation

Wateq was used to obtain speciation, evaluation of ion balances and to prepare for the calculation of normative mineral compositions. The wells were processed individually, but grouped to Wateq output files by Township and Range. Wateq enabled the calculation of data for the CaO and silica stability plots by providing the necessary ion 117 activity product (IAP) derived from the analyses. Mineral reconstructions were prepared on a spreadsheet from the normative mineral composition calculations.

Stage Three: Hydrochemical Plots

The suite of hydrochemical plots is comprised of ternary diagrams, Piper plots,

Fingerprint diagrams, and mineral phase stability diagrams. All the plots are included in

Appendices B, C, and D, including those used within the text.

The ion concentrations reported by the USGS were generally as milligrams per liter

(mg/L) or micrograms per liter (,ug/L) for some ions such as iron. These were converted to milliequivalents per liter (meq/L) for plotting.

Water analyses are denoted by the USGS as to the formation from which the waters were produced (where known). Only the samples identified as code 110 for the shallow

Quaternary alluvial aquifer, code 300 for the Artesia Group, and code 313 for the Artesia

Group-San Andres Formation-Glorieta Sandstone-Yeso Formation undifferentiated were used in this study. The ternary plots, Piper diagrams and Fingerprint diagrams are designated as springs, alluvial aquifer, or carbonate aquifer, based upon the USGS designation. On all of the following figures the use of the phrase "Reported as carbonate wells" or "as carbonate aquifer" shows the wells were reported as constructed in the

Artesia Group-San Andres Formation-Glorieta Sandstone-Yeso Formation undifferentiated or the San Andres Formation—even though all locations are not confined aquifers, or for that matter not all composed only of carbonate. 118

Ternary diagrams

Three ternary plots, a Piper diagram and a Fingerprint diagram were prepared for each Township. All samples plotted that originated within a Township were plotted on the diagram. This plotting scheme was selected because of the size of the area and the distribution of water sample locations. Thirty-one Townships are referenced for a total of

82 plotted samples.

The hydrochemical plots for all 82 samples are attached to this study as Appendix B

(see page 228). For each Township a plat shows the location of the wells or springs. A combination ternary plot of cations and anions (as in Figure 4.2 on page 124) is presented for each Township. A Piper diagram (as in Figure 4.3 on page 127) comprises the third figure of each set. A Fingerprint diagram as the fourth figure completes each set.

Appendix B contains the complete set of four figures for each set in all thirty-one

Townships examined.

Three ternary plots were made using the analyses of well and spring waters for each of the 82 samples used. Plots were made for a cation ternary plot, an anion ternary plot and a combination ternary plot on which both cations and anions were shown. The cation

vertices are Na + 1(±, Mg' and Ce. The anion vertices are Cl - , HCO3 - and SO42 . Plots were made for all samples used from a Township and Range. The points plot as percentages of the normalized cations or anion present in the water sample. Only the combination ternary plots are presented in the text and Appendix B. 119

Piper diagrams

Piper diagrams reproduce the individual cation and anion ternary diagrams using the

same vertices. The central diamond of the Piper diagram portrays the combined plot of

cations and anion differently than seen in the combination ternary plot. The points on the

Piper diagram plot are also plotted as a percentage of the normalized cations and anions

present in the sample analysis.

Eitigerpriitt diagrams

The use of Fingerprint diagrams in hydrochemical investigations is discussed by

Mazor (1991). The Fingerprint diagrams prepared in this study use seven ions, when

available, and are accumulated in terms of meq/L for plotting. Only a few analyses were

plotted that lacked either sodium or silica values.

On each of the Fingerprint diagrams the plots follow the convention of the X axis representing the ions being graphed and the Y axis representing the concentration of each ion in milliequivalents per liter (meq/L) logarithmically. The ions are identified as 1 = sodium plus potassium, 2 = magnesium, 3 = calcium, numbers 4, 5, and 6 are left blank for spacing, 7 = silica, 8 = bicarbonate, 9 = sulfate, and 10 = chlorine. Sodium and potassium are, as discussed below, combined because only a few analyses reported the ions separately.

Silica is reported in the original data as dissolved silica in mg/L. The activity of amorphous silica as Si02 under normal surface conditions (25°C) is [1-14 SiO4] = 10-2.74

(Faure, 1991). This is silicic acid with an effective ionic charge of zero. Therefore, 120

silicic acid in any amount from none (not reported) or greater would plot as a zero if converted to milliequivalents. However, by making the justifiable assumption that

dissolved silica in groundwater is in its various states of dissociation, we can consider

silica on the Fingerprint diagram.

The states of H4 SiO4 dissociation and their equilibrium constants (Faure, 1991) are as

follows:

H4S104 H3SiO4 + H + K1 = 10 -9-71

H3S104 H2SiO 42 + H K2 - HY "

H2S7042 HSiO 43 + H K3 = 10986

K4 = HSiO 43- + H+ iO30

From the equations, the vast majority of dissolved silica will be as H3 SiO4 - . The second

equilibrium constant is so small that little silica will be in it or later stages of dissociation.

For that reason the use of a minus one (-1) valence for H 3 SiO4- (formula mass = 95.12

g/mol) and a resulting milli-equivalents per liter conversion (0.01051 times the mg/L

dissolved silica reported) is obtained for use on the Fingerprint diagrams.

Stability diagrams

The data for the CaO and Si0 2 silica stability diagrams were derived from the Wateq

output for each water sample. The diagram uses the log of the IAP's for calcium (Ca')

divided by hydrogen (Er), and silica as silicic acid (H4 SiO4). These diagrams are presented and discussed in a separate section (see page 188). 121

Hydrochemical Plots

The reader is referred to Figure 3.9 on page 98 for the regional positions of the

Township and Range plats that are shown in the following presentation of hydrochemical plots. As a verbal reference for location, the Pecos River marks the eastern practical boundary of the Roswell groundwater basin and flows from the north toward the south in and between Ranges 25 East and 26 East. The northernmost water samples discussed are from Township 1 South (Tl S) and the most southerly is from Township 14 South (T14S).

The water samples are distributed from Range 13 East (R13E) to Range 27 East (R27E).

This represents a maximum north-south and east-west distance of approximately 135 km.

The hydrochemical plots and comparisons of data are limited by the distribution of water sample analyses. Although many more sample analyses were available most were not identified as to the producing aquifer or formation or they lacked determination of most of the basic ions used in this study.

Throughout this study the aquifer conditions of temperature, pressure, and CO2 content are considered constant. The USGS database reported a few pumped well water temperatures, no pressures at producing levels were reported, and no CO2 values were reported on the well analyses used in this study. The data used here is typical of many aquifer areas in that many "ordinary" analyses are available, but lack analyses on many of the "would like to have" ions or details on the producing environment. A challenge then confronts the investigator as to usable methods to obtain meaningful information concerning the aquifers and the groundwater. 122

The plots in this study show that the chemical characteristics of waters from the basin vary according to the rocks through which they flow. In studying these plots we can derive important chemical information to distinguish the lithologies that are dominating the water-rock reaction. The data may then be used to characterize the aquifer, or

"fingerprint" the formation constituting the aquifer. The spatial changes in the chemistry

(at or near surface conditions) are controlled by the rocks through which the water moves

and yield important indications of both the local and regional flow paths of the water.

Waters in the same flow path will systematically change chemical composition from

one point to another as they flow through an aquifer at quasi-constant temperature and

pressure. The chlorine ion concentration in the water, for instance, will not decrease with continued movement along the same flow path, but will increase or remain unchanged.

Therefore, if waters of two samples are compared and have dramatically different ion contents, care must be given to conclusions that they are from the sanie flow path. Some

sequences of water samples denote a common path of flow whereas others simply could

not be in the same flow path given the existing hydrochemical composition and hydraulic gradient.

Not all the water samples used in this study were obtained from the confined portion

of the carbonate aquifer (Artesia Group-San Andres Formation-Glorieta Sandstone-Yeso

Formation undifferentiated). Large portions of the groundwater basin are unconfined.

Some samples used were obtained from springs. 123

San Andres Formation Hydrochemical Signatures

The first set of plots shown are from T I2S-R24E. This set is derived from the Y-0 block, south of the Y-0 fault and approximately 15 km west of the Pecos River. The

wells from which these samples were collected are found in an irregular Township as

shown in Figure 4.1. This Township is also

shown regionally in Figure 3.9, on page 98. R.24E 4 3 2 1 In Figure 4.1 (and other similar plats) the

line connecting the wells is only to

9 10 11 12 emphasize their location. Well location

16 1 5 M 13 numbering is described in Appendix A,

Figure 5.1, on page 219.

2'1 r'il" ilb 23 24 The well water samples in this Town- / o

28 27 25 25 ship were chosen for several reasons. First,

the shallow alluvial aquifer is not present

33 31. 35 35 beyond the eastern edge of this Township. Figure 4.1, Location of T12S-R24E Andres samples. Section 22 contains 3 Second, the top of the San locations. One section is 1.6 km wide. Formation in these wells is approximately

100 m (determined from well samples)

beneath the suiface; the top of the confined aquifer is an additional 30 m deeper, and the

bottoms of the completed depth of the wells range from 213 m to 381 m beneath the

surface. These conditions are more than sufficient to assure the water samples were

obtained solely from the carbonate artesian aquifer. Additionally, all of the waters were 124

Cations Ca2+ Me Na++ K+ Ca2+ Mg2+ Na++ K+

Twn - Rge - Sec mg/L mg/L mg/L meq/L meq/L meq/L 12S.24E.27.21333 120 36 26 6.0 3.0 1.1 125.24E.22.41333 150 39 17 7.5 3.2 0.7 12S.24E.22.41333 140 32 14 7.0 2.6 0.6 125.24E.22.23111 140 38 32 7.0 3.1 1.4 12S.24E.15.43111 130 38 34 6.5 3.1 1.5 Table 4.1. Cation concentrations in T12S-R24E samples.

sampled within a period of 67 days. The data obtained from these analyses are considered here the reference standard for known San Andres Formation aquifer waters.

This is especially important for the correlation of Fingerprint diagrams.

Table 4.1 shows the mg/L and meq/L makeup of the cation portion of the samples.

Correlation of the data with the ternary plot reveals that the tight cluster of three cation

points is from the three

Ternary Plot westernmost wells of

SO 2- Mg2+ 4 the group (Section 27,

west side of Section

22, Section 15) and

the two remaining

samples are from the aim - - / GB 3 wells in the eastern Ha++ IV-50 Ga2+ portion of Section 22.

Figure 4.2. Ternary plot of 5 samples T12S-R24E. The character of Percentage of ion concentration in meq/L. San Andres Formation. these waters is first 125 portrayed in the combination ternary plot shown in Figure 4.2. Cations are plotted in the lower right of the triangle, and exhibit a noticeable lineation. Five samples are shown.

Three of the samples cluster closely together on the left side of the cation group. The cluster is composed of the samples taken from Section 15, the westernmost location in

Section 22, and the well in Section 27. The two remaining well samples are from the eastern side of Section 22.

In the anion portion of the combination ternary plot (Figure 4.2) there appear to be only four points, but the upper of the farthest right group is two analyses. The double point is from the two sample locations in Section 22.41333 and the single adjacent point is from Section 27.21333. The anion points show a directional change that is approximately parallel to the cation trend.

tic03- s042- cr fic03- s042- cr

Twn - Rge - Sec mg/L mg,/L mg/L meq/L meq/L meq/L 12S.24E.27.21333 230 270 30 3.8 5.6 0.8 12S.24E.22.41333 250 270 30 4.1 5.6 0.8 12S.24E.22.41333 170 270 53 2.8 5.6 1.5 12S.24E.22.23111 240 280 30 3.9 5.8 0.8 12S.24E.15.43111 220 290 50 3.6 6.0 1.4 Table 4.2. Anion concentrations in T12S-R24E samples.

Table 4.2 is presented to show both the mg/L and meq/L relationship of the ternary plot that shows the normalized percentage concentrations in meq/L

The ternary plot shows the water sample chemistry exhibits a directional change. The change is caused by the water-rock reaction as the water moves through the aquifer. The cation plot in Figure 4.2 shows a clustering of the Section 27, Section 22.23111 and 126

Section 15 samples that signifies an increased percentage of Na + K and Mg2+ ions compared with Ca24 ions and contrasted with the Section 22.41333 wells located farther- down the structural dip and down the hydraulic gradient.

Visualizing absolute variations in concentrations between locations is usually difficult because the ternary diagram shows only the percentage of concentration. This is of concern in hydrogeologic evaluations because of the importance of the direction of water movement. Therefore, comparing the quantitative ionic change between sample locations is also necessary.

Going from southwest to the northeast of the Township, from Section 27 to Section

22.23111 to Section 15, the cation and anion concentrations in mg/L change in the following order:

Na' + K+ 26 -*32 -+34 HCO3- 230 4 240 4 220

Mg + 36 -438 4 38 SO42" 270 4 280 4 290

Ca2+ 120 4 140 .4 130 a- 30 -+30 4 50.

These changes are, for the most part orderly, consistent and suggesting groundwater movement essentially parallel to the strike of the formation and the aquifer. However, it does not preclude the argument that in fact the recharging waters are moving down the hydraulic gradient from northwest to southeast and the water-rock reaction is effecting only slight change because of slight lithologic variations. The down-gradient direction is toward the southeast and the Section 22.41333 wells.

The two Section 22.41333 samples have the following mg/L concentrations in:

Na+ +1C+ 14 and 17 HCO3- 170 and 250 127

me- 32 and 39 SO42" 270 and 270

Ca' 140 and 150 Cl" 53 and 30.

Note that in these samples Mg', Naf + K+, and HCO3 - ions are showing some values less than the up-dip, up-gradient samples is significant.

The ternary plot is informative, but

Pi per Di agr am the Piper diagram has the added ability of

showing the mutual relationships of the

so:* cations and anions from a group of

samples. The Piper diagram for the

rro; T12S-R24E waters is presented in

Figure 4.3. Piper diagram of 5 samples Figure 4.3. While the anion and cation from T12S-R24E. Reported as carbonate aquifer wells. San Andres Formation. presentations in the Piper diagram are repetitive of the ternary plot, the combination diamond of the Piper diagram is important to understanding the mineralogical background of the water-rock reaction. The Piper diagram in Figure 4.3 shows more readily that the lineation observed in the ternary plot is a shift from bicarbonate toward sulfate as opposed to an increase in chloride ion content.

While the variation of ion concentrations in T12S-R24E has been discussed, the primary purpose has been to establish reference points with which comparisons of other water within the basin can be made and conclusions as to their aquifer rocks can be derived. The goal is to identify a signature for the various waters and derive meaningful correlations with the aquifers from which the waters are produced. 128

Fingerprint diagram for the The —o—Fingerprint Diagram water samples in T12S-R24E is presented in Figure 4.4. This figure presents the San Andres Formation carbonate signature for the Roswell 0. 1 T1'29-R24E 5 ,e1I's groundwater basin. The definition of 1rtsnian Aquifor 0.01 I i 01 2345 6 78 9 10 this predominately carbonate aquifer Na1K=1 N1g=2 Ca=3 SiO =7 H03O=8 S0=9 C1=10 2 3 is important in understanding the Figure 4.4. Fingerprint diagram of 5 samples from T12S-R24E. Reported as carbonate distribution of the "carbonate aquifer wells. San Andres Formation signature. aquifer" lithology even where that lithology is the unconfined carbonate aquifer. As will be seen in subsequent Fingerprint diagrams, the lithologic change between San Andres Formation and Artesia Group—even across the transition between the two where solutional activity has played some significant role—is strongly reflected in the waters shown in the Fingerprint diagrams.

Additionally, strong variations in the ionic composition of the aquifer waters from area to area also will be readily seen in the Fingerprint diagrams.

Waters from Tl I S-R23E, Sections 27 and 8, a distance of 8 km and 15 km northwest of the northernmost location of samples in Figure 4.4, are shown in the Fingerprint diagram, Figure 4.5. Even without the values for silica in the anion area of the

Fingerprint diagram, the signatures are very similar. To illustrate the close similarity,

Figure 4.6 is an overlay of the plots in Figure 4.4 and Figure 4.5. The samples from

T11 S-R23E are from 8 km to 27 km from the location of samples in T12S-R24E. The

129

-c- Firgerorint Diagram -o-ringerpr nt Diagram! a•,

102

13 -;

TIES-RZE ' Artesin Aquifer 0.01 1 2 3 4 5 6 7 8 9 10 2 3 4 5 6 7 8 9 10 9a4K=1 Mg=2 Ca-3 00 =7 43J S4J =9 0=10 Nail1=1 Mg=2 8a=3 4 3 4 SiO =7 PEO =9 S0=9 0=10 2 3 Figure 4.6. Fingerprint overlay of Fig. Figure 4.5. Fingerprint diagram of 2 4.4 and 4.5. Note conformity of samples from T11S-R23E. Reported as signatures from samples 8 to 27 km carbonate wells. San Andres Formation apart. San Andres Formation signatures. signatures.

Figure 4.5 samples were collected from wells in an area of unconfined carbonate aquifer

production. The most northwestern location (Section 8) in the Figure 4.5 samples is

shown in Figure 4.7. This is very close to, or upon, the Six Mile Hill fault zone. The San

Andres Formation crops out at the

R 23E surface. The Section 27 location is 6 1 Mile 'H fault n: within the Lakewood Terrace (see Figure

9 10 11 12

_.±...' 1.4, page 36) in the Rio Hondo drainage

I g 17 16 15 14 13 and floodplain. In the Section 27, T11S-

19 20 21 23 24 R23E area, the San Andres Formation is

30 29 28 • 26 23 covered with approximately 30 m of

31 32 33 34 35 alluvial material (J. Hawley, 1994, pers.

comm., unpubl. map) that is thinning Figure 4.7. Location of T11S-R23E samples. One section is 1.6 km wide. rapidly toward the northwest. 130

-o-Fingerprint Diagram 106Y3

100

0. 1

0.01 0 1 2 3 4 5 6 7 8 9 10 Na+ K =1 Mg=2 Ca=3 Ha-W=1 Mg=2 Ca=3 S 0 =7 033=8 61=10 Si 02 =7 HCM3 SO 4=9 61=10 4r9 Figure 4.8. Fingerprint diagram of 1 Figure 4.9. Fingerprint diagram of 1 spring sample from Section 28, TlIS- spring sample from Section 27, T10S- R16E.R14E. Reported as Glorieta-Hondo Reported as Glorieta-Hondo Sandstone-Yeso Formation. San Sandstone-Yeso Formation. San Andres Formation signature. Andres Formation signature.

The next two water samples are from the far-western portion of the recharge area.

They are presented to illustrate the similarities of the Fingerprint diagram signatures for

San Andres Formation waters separated by

more than 100 km. Again, only one of -o- Fingerprint Didgram

these two samples has an analysis for silica,

but even without the plot of that component

the comparison is valid. Sample waters

illustrated in the Fingerprint diagram, 0.01 Figure 4.8, are from Section 28, T11S-

R14E. This sample was taken from a Figure 4.10. Fingerprint diagram of 1 spring sample from Section 18, T14S- spring on the south side of the Village of R14E. Reported as Glorieta-San Andres-Artesia Group aquifer. San Ruidoso Downs. Yeso Formation is Andres Formation signature. 1.31 exposed in the lower areas and is overlain by carbonates of the San Andres. The ternary plot and Piper diagrams for this spring (see Figure 6.98 and Figure 6.99, Appendix B) show a stronger Mg' and SO,' ion influence that is also reflected in the Fingerprint diagram (Figure 4.8 and Figure 6.100 of Appendix B). Still, the overall pattern of the signature is strongly shape-correlative with the San Andres Formation waters shown in

Figure 4.6. This spring is more than 100 km due west of the Figure 4.8 sample location.

Spring water sample analysis of water from Section 27, T10S-R16E is illustrated by a

Fingerprint diagram in Figure 4.9. This spring is in the Rio Hondo stream valley and is associated with outcrops of Yeso Formation that are overlain by San Andres Formation carbonates. The chemistry of this spring shows similar SO42" concentrations, but slightly higher Mg' relative to Ca', and reduced Na + K.- as a per-centage of total cations present (see ternary plot and Piper diagram, Figure 6.102 and 6.103 in Appendix B) as compared to the more westerly signature in Figure 4.8.

The spring sample represented by the Fingerprint diagram in Figure 4.10 is from

Section 18, T14S-R14E, in Elk Canyon on the Mescalero Apache Indian Reservation.

Elk Canyon spring is about 125 km west of the Pecos River and 25 km south of Ruidoso,

New Mexico. The spring is at an elevation of 2,311 m. Elk Canyon is cut into Permian

Yeso Formation (Kelley, 1971, Plate 3) overlain by San Andres Formation. The canyon drains to the southeast to join the Penasco River in the northwest corner of T16S-R16E.

The signature from the Elk Spring sample (Figure 4.10), particularly the anion portion, correlates well with the Fingerprint diagrams in Figures 4.8 and 4.9 dispite the order of magnitude difference in ion concentrations. The decrease in Mg' concentration 132

--,-Fingerprint Diagram is attributed by me to a generally lower 141'3 percentage of dolomite content of this 100

to area within the San Andres Formation To illustrate the correlativity of the

0.1 Fingerprint diagrams more emphatically, OvLae.F of Fiyu, es 4. ao0 4.10

0.01 A I 2 3 4 5 IS 7 8 10 an overlay of the three spring water NaiK=I Mg=2 Ca=3 SiO =7 HCC SU =9 C1=10 4 sample analyses onto one Fingerprint Figure 4.11. Fingerprint diagram overlay of spring waters from Fig. 4.8, 4.9, 4.10. diagram is presented as Figure 4.11. The

Township location plots, ternary plots,

Piper diagrams and duplicate Fingerprint diagrams for each of these three spring water samples in Figure 4.8, 4.9 and 4.10 are Figure 6.97 to 6.100, Figure 6.101 to 6.104, and

Figure 6.85 to 6.88, respectively, in the attached Appendix B.

Artesia Group Undifferentiated Hydrochetnical Signatures

The interval discussed in this portion of the hydrochemical evaluation includes the

Artesia Group undifferentiated (see Table 2.1, page 56) and the solutional breccia and karstic complex that involves both the Artesia Group and the underlying upper San

Andres Formation. Based upon the examination of many well samples and upon cores from the aquifer (Havenor, 1968; Kinney et al., 1968; Maddox, 1969) it is believed here that in those many cases where this condition occurs, the hydrochemical character of the

Artesia Group is most pronounced. The lithologies of the two formations differ in that 133 the Artesia Group is dominantly anhydrite, fine-grained siliclastics, red bed clays and dense, anhydritic dolomite, with halite and other associated evaporitic minerals.

Where the Artesia Group and the San Andres Formation contact area are exposed to the surface and near-surface action of meteoric waters and the waters within the aquifer, the dissolution often results in a lithologic melange along the highly susceptible contact zone. The inclusion of red clay from the Artesia Group often marks the solutional interval, along with "spongy" carbonates, anhydrite and carbonate-anhydrite breccia, sulfur lined vugs, and occasionally scattered inclusions of silica sand or sandstone.

Where the contact area of the San Andres Formation and the overlying Artesia Group is affected by solution, the waters from the interval express the characteristics of the

Artesia Group. Although the San Andres Formation carbonates continue to contribute

Ca', Me and HCO 3-, the contact of the water with anhydrite, halite, and other evaporite minerals markedly changes its hydrochemical composition—and by that its R 25E

6 5 4 3 2 1 signature. Because of the increase in the o

11 12 water-rock reaction with both anhydrite

18 17 16 15'.›14 and halite a change in Na + K, SO42-, 10 // 19 20 21 22 23 zZ4/ and Cl- concentrations becomes more // // // 30 29 28 27 /27 25 noticeable. // Y-0 fault zor /, :1fo 3 34',// 35 36 The group of wells in T10S-R25E is / / presented in Figure 4.12 and is selected Figure 4.12. Location of T10S-R25E as representative of the water samples samples. One section is 1.6 km wide. 134 r Cations Ca2+ Na++K± Ca Mg2+ NatEK+ Twn - Rge - Sec mg/L mg/L meq/L meq/L meq/L 10S.25E.31.34 POTTE 370 I 2700 18.5 9.9 0.8 10S.25E.33.423 ROSW 640 130 1700 31.9 10.7 0.6 10S.25E.32.424333 0 1000 230 9059 49.9 18.9 0.9 10S.25E.29.443333 P 470 130 2600 23.5 10.7 0.8 10S.25E.14.312 PVAC 1400 190 10000 69.9 15.6 0.8 10S.25E.05.3 BTR LK 390 80 760 19.5 6.6 0.6 Table 4.3. Cation concentrations in T10S-R25E samples.

Anions HCO3 -+CO 32- SO4 2- a HCO 3 +CO3 2- SO42- Cl- Twn - Rge - Sec mg/L mg/L mg,/L meq/L mcq/L meq/L 10S.25E.31.34 POTTE 3.4 20.4 121.3 0.02 0.14 0.84 10S.25E.33.423 ROSW 3.0 37.5 76.2 0.03 0.32 0.65 10S.25E.32.424333 0 3.1 64.5 394.9 0.01 0.14 0.85 116.25E.29.443333 P 3.3 31.2 110.0 0.02 0.22 0.76 10S.25E.14.312 PVAC 3.0 75.0 451.4 0.01 0.14 0.85 10S.25E.05.3 BTR LK 3.1 25.0 31.0 0.05 0.42 0.52 Table 4.4. Anion concentrations in T10S-R25E samples. obtained from the Artesia Group to San Andres Formation interval subjected to the solutional activity discussed above. The well in Section 14 is found 1.5 km east of the

Pecos River, and the Section 33 well is1.5 km west of the river. The subsurface projection of the Y-0 fault zone is shown on the location plat (Figure 4.12). These wells were drilled through the deeper areas of the Quaternary alluvium occupying the valley and containing the Pecos River stream bed, the subject wells are all artesian; although they are close to the river, an aquiclude or at least an efficient aquitard is present and is separating the alluvium from the artesian aquifer.

Following the previous method of presentation of data to establish the nature of the waters from the San Andres Formation above, the waters from the aquifer developed across the San Andres-Artesia Group contact can be considered in the same manner. 135

Ternary Plrit This discussion will be

so, 2- 41 / \ followed by the

r:-atk-- , r12• Anions presentation of other

Fingerprint diagrams

33 within the Roswell 29 32 • groundwater basin and 14 a cr 4 4 \ /44 1, • its recharge areas m3±-1-K+ 3141- i3 5 50 29 , Sample location Section number showing signatures

Figure 4.13. Ternary plot of 6 samples, T10S-R25E. Ion similar to T10S-R25E, percent plot meq/L. Artesia Group. yet different from those seen for the San Andres Formation.

Table 4.3 shows the cation concentrations in both mg/L and meq/L for easier comparison of their equivalencies. The anion concentrations are presented in Table 4.4.

The distribution of cations and anion in the samples of T10S-R25E is shown in

Figure 4.13, the combination Piper Diagram cation—anion ternary plot. This plot has been annotated with the section numbers to help identify the areal change in the distribution of ions. The Na and Cl" are highest in Sections 14 and 32; the Ca',

HCO3-, SO42" and Mg' are highest in Figure 4.14. Piper diagram of 6 samples from T10S-R25E. Reported as carbonate Section 5. These relationships are shown aquifer. Artesia Group. 136 more clearly in the Piper diagram, Figure 4.14. The diamond of the Piper diagram illustrates the linear progression of percentage composition changing from SO,' dominated in the northwest of the Township to having a Na and Cl" domination in the southwest of the Township.

The higher concentrations of Na and Cl" around Section 14 and the southwest corner of the Township in Sections 31 and 32 are bordered to the southeast by the projection of the Y-0 fault (see Figure 4.12). Much lower chloride concentrations (Havenor, 1968;

Welder, 1983) are present south of the Y-0 fault. Data from the area north of the Y-0 fault have historically recorded significant fluctuations of Ct ion concentration. The southwest corner of this Township has produced water with chloride concentrations of more than 7000 mg/L. The concentrations have been observed to reduce with increased potentiometric levels in the artesian aquifer and reach high levels when the head is reduced by over-pumping in the basin (Havenor, 1968). Southeast of the fault the artesian water has chloride concentrations one to two orders of magnitude lower.

In this Township the southeast flank of the saline encroachment lobe parallels the north side of the Y-0 fault projection from the southwest. Welder (1983) showed that in

1979 the chloride concentrations ranged from 7000 mg,/L in the SW/4 SW/4 of Section

32, T10S-R25E, to 50 mg/L 6 km directly south. The potentiometric gradient in T10S-

R25E is toward the southeast. Given the hydrochemical data in this Township it is difficult to visualize how water moves toward the southeast. Without a significant geological boundary close to the subsurface projection of the Y-0 fault, the chemistry of the waters would preclude a flow path to the southeast. 137

The waters in T10S-R25E are on the Roswell block between the Y-0 fault and the Six Mile Hill fault to the northwest. The signature of these waters is considerably different from TIOS-R256 6 wells the waters seen in earlier figures from Artesl an Aquifer

T12S-R24E. The Fingerprint Na-FK-L Mg-2 Ca=3 S02 =7 KO =8 SO =9 C1=10 2 4 diagram for wells in this Township is Figure 4.15. Fingerprint diagram of 6 samples from T10S-R25E. Reported as shown in Figure 4.15. The carbonate aquifer. Artesia Group.

Fingerprint diagrams plot a distinctively different picture of the artesian water to the south on the Y-0 block (see

Figure 4.4) compared with this evaporite influenced groundwater on the north side of the

Y-0 fault.

The next set of samples is also on the R 24E

6 3 4 3 ,2i I Roswell block, between the Y-0 fault

7 0 9 10 12 and the Six Mile Hill fault, in T10S-

18 17 16 15 14 13

R24E. These analyses represent six 10

19 20 21 22 -NNILI,...N.. 24 samples that give coverage for most of

30 29 28 27 26 25 the Township. The distribution of these

31 32 33 34 35 36 samples within the Township is shown in v

Figure 4.16. The well in Section 24 is Figure 4.16. Location of T10S-R24E 7 km west of the Pecos River. samples. One section is 1.6 km wide.

138

Fingerprint diagram for these The —o— F ngerprint Diagram ia samples is shown in Figure 4.17. The 100 — ternary plot and Piper diagram are in to Appendix B as Figure 6.18 and 6.19 -41

along with copies of the location plat ‘e' e.1 TIO-Ric4E 6 wells and the Fingerprint diagram pre Arteian-- Aquifer • 0.01 i 1 2 3 4 5 6 7 8 9 10 sented here. The ternary plot flaiK=1 Mg=2 Ca=3 SiO =7 HCO=0 SO =9 C1=10 2 3 4 illustrates a lineation similar to that Figure 4.17. Fingerprint diagram of 6 samples from T10S-R24E. Reported as seen in T I 0S-R25E, Figure 4.13 carbonate wells. Artesia Group signatures.

above. These samples show a greater

spread toward the HCO3" and Ca', with lower relative percentages of Na + lc- and cr.

Like the previous Fingerprint diagram (Figure 4.15) this one exhibits the signature of a

strong evaporite influence on the water-rock reaction.

The HCO3 - content of each of these samples is steady at 210 mg/L (200 mg/L in the

Section 32 sample). Because the Cl" concentration varies from 230 mg/L to 2400 mg/L

and the points plot as a percentage of total ions, the shift of points is large on the ternary

plot and the Piper diagram. The Fingerprint diagram plotting of points is not influenced

by the concentrations relative to other ions. The Fingerprint diagram is a logarithmic

plot and shows the relationship of the actual ionic concentrations.

Reference to Figure 2.4 (page 53) emphasizes that the Artesia Group subcrops and

outcrops are confined to a narrow (but important) belt that is present along and parallel to

the Pecos River. The limited presence of the Artesia Group west of the Pecos River, 139 combined with scattered available R 24E

6 3 2 1 analyses, restricts the breadth of o i 01 7 7

oS comparison possibilities for the 7 8 9 10 11 12 31X Mile -Till fault zone

Fingerprint diagram signatures 18 17 16 15 14 13

9 observed for the group. 19 21 22 23 24 o

Figure 4.18 shows the locations 30 28 27 26 25 of nine well water samples from T9S- o 32 33 34 35 36 R24E and their positions relative to the Six Mile Hill fault zone. Two of Figure 4.18. Location of T9S-R24E samples. the samples came from the Six Mile One section is 1.6 km wide.

Hill block, and the remaining seven samples are from the Roswell block on the south side or the Six Mile Hill fault zone. The easternmost well, Section 11, is 9 km west of the

Pecos River.

is the Fingerprint Figure 4.19 --o—Fingerpr int Diagram diagram for the water samples in 100 T9S-R24E. The cation signature is

highly correlative with that seen from

samples in the next Township south 0.1 , 1S-P24E 9 HeIls (Figure 4.17). The anion signature, Artesian Aqui er 0.01 I I 0 34 5 6789 10 while still having a strong SO 42- tia+K=I Mg=2 Ca=3 Si° =7 HCO =8 SO =9 C1=10 2 9 4 point, has a decreased Cl" Figure 4.19. Fingerprint diagram of 9 samples from T9S-R24E. Reported as concentration. The lower chloride carbonate aquifer. Artesia Group signatures. 140

level would be expected from the R24E distribution of chlorides observed in 6 5 4 3 2 I the T10S-R25E samples (Figure 7 8 9 10 11 12

4 13) and being farther removed ° \ 18 17 16 15 14 13

8 \ 22 0 toward the northwest. A similar 19 20 21 23 24 change is apparent in the Township I 29 28 • 27 26 25 south (Figure 4.17) • 31 32 33 34 35 36 The relationship of the ions in this set of samples is more frilly Figure 4.20. Location of S/2 T8S-R24E shown in the complete set of plots in samples. One section is 1.6 km wide.

Appendix B. The location plat, ternary plot, Piper diagram and the above Fingerprint diagram are Figures 6.25 - 6.28 in the appendix.

In Township T8S-R24E, the

number of samples required dividing

the Township approximately into

north-south halves for the purposes of

plotting data. All of the samples in

both halves of the Township were

collected within a few days of each

other. The sample locations for the

2 Figure 4.21. Fingerprint diagram of 8 south-half of T8S-R24E are shown in samples from S/2 T8S-R24E. Reported as carbonate aquifer. Artesia Group signatures. Figure 4.20. 141

All of the samples in this portion, except Section 31, show an Artesia Group signature

in the Fingerprint diagram, Figure 4.21. The Section 31 location is very close to the

westernmost surface exposures of Artesia Group. San Andres Formation crops out less

than 5 km to the west and for many Townships to the west.

The Section 32 well is positioned about 1/2 km south of Salt Creek. As the name

implies, this creek had, according to local comments, historically carried ephemeral flows

of higher salinity waters than do other streams in the area. The Section 22 well is 2 km

west of the Arroyo del Macho and 11 km west of the Pecos River.

The Arroyo del Macho is an ephemeral stream. The drainage area of the Arroyo del

Macho is much larger than that of the Salt Creek. The Arroyo del Macho drains from the

Jicarilla Mountains in T5S-R12E, 140 km to the west-northwest, through Section 15 of

this Township, and into the Pecos

River 11 km to the east. R24E

Except for the sample from 6 /9 4 3 2 I

Section 31, the values from the 7 8 9 10 11 12 s'o / analyses plot in close groups on the o 18 1----1-T—° 16 15 14 13 ternary and Piper diagrams (see 8 19 20 21 22 23 24 Figure 6.38 and 6.39, Appendix B). 30 29 28 27 26 25 The groups are showing 31 32 33 34 35 36 approximately equal meq/L concentrations of SO42- and Cl-, with Figure 4.22. Location of N/2 T8S-R24E low HCO3- levels. samples. One section is 1.6 km wide. 142

Samples from the north half of 10 the Township are as shown in Figure too 4.22. All of the samples from this F to Township are situated north of the

Six Mile Hill fault. 0.1 NaT8S-4424E 7 wells The well in Section 18, akin to Artesian Aquifer 0.01 0 9 1.0 the well in Section 31 (Figure 4.20), Na+K=1. Mg=2 Ca=3 910 =7 F02=8 S0=9 C1=10 2 4 is found where the San Andres Figure 4.23. Fingerprint diagram of 7 samples from N/2 T8S-R24E. Reported as Formation is close to the surface—if carbonate aquifer. Artesia Group signatures. not directly beneath the thin

Quaternary alluvial cover in the Arroyo del Macho.

Figure 4.23 is the Fingerprint diagram for the samples collected from the locations in

Figure 4.22. The one deviation from the Artesia Group cation signature is the sample from Section 18. The complete set of hydrochemical plots are shown in Figure 6.41 through Figure 6.44 in Appendix B.

The steady vertical variation in the cation distribution of Figure 4.23 is representing an orderly progression from east to west of increasing Ca' and decreasing Nat The anion lineation, while not quite as straight-line as the cations, shows a north to south change from SO,' to Cl.

One additional sample will be presented to help illustrate that the signatures for the

Artesia Group lithologies can be recognized regionally. The sample is from Section 31,

T6S-R24E. Kelley (1971) mapped the area of this well as having extensive surface 143

ToarNwy exposures of the

Seven Rivers

Clati()ns 4 Anions e Formation of the

Artesia Group. 59 This location is

8 km east of the Arroyo

HCO 3 del Macho and 21 ca2+ km west of the

Pecos River. The Figure 4.24. Ternary plot of 7 samples, N12 T8S-R24E. Ion percent plot meg/L. Annotated with sample section numbers. well is within Artesia Group. one km of the farthest northeast surface expression of the Border Hill fault zone as mapped by Kelley

(1971) and is visible on the satellite

-o-Fingerprint Diagram imagery.

The Fingerprint diagram for this 100 K well water sample is presented in

Figure 4.25. It shows a signature

consistent with other samples such as T6S-4124E Set. 31 I Well Prtesian Aquirer 0.01 Figure 4.23 presented above. 0 t 2 2 4 5 E. 7 8 9 10 NalK=1 M3=2 Ca=3 SiO =7 NCO =8 SO =9 0 =10 Additional correlations, as with 2 4 Figure 4.25. Fingerprint diagram of 1 Section 19, T7S-R24E (Figure 6.60 sample from Section 31, T6S-R24E. Reported as carbonate aquifer. Artesia Group signature. 144 of Appendix B), can be found by visual inspection of other Fingerprint diagrams in

Appendix B.

Quaternary Alluvium Hydrochemical Signatures

The one remaining aquifer that is widely recognized in the area is developed in the

Quaternary alluvium. The alluvial deposits have not been geologically detailed in the

Pecos Valley. The underlying Artesia Group is lithologically distinct and can be recognized in well cuttings—when representative samples have been properly caught and preserved. Determination of a hydrogeochemical signature for the Quaternary alluvium would allow a better understanding of the relationships of the artesian aquifer(s), the aquitard(s), the alluvium and the Pecos River.

Many well sample analyses were available in the USGS data set for which there was no producing formation recorded. While many of these wells might be deduced as having been constructed in the Quaternary alluvium based upon location and some depth information, only two sets were identified as Quaternary alluvium completions.

The first set comprises 4 closely spaced wells in Section 36, T11S-R25E. The cation and anion analyses are shown in Table 4.5 and Table 4.6 respectively. The analysis for the Section 36.213 sample differs significantly from the other three samples. Examination of the cation/anion electrical balance shows the sample is not significantly less balanced than the other samples. The logical remaining conclusion is that this producing formation was misreported and the sample is from a different aquifer than the first three.

The ternary plot for the T11S-R25E samples is given in Figure 4.26. The anions and 145

Cations c2+ ml,y2+ Na4 +K+ Ca2'- Mg2+ NatF1(- Twn - Rge - Sec mg/L mg/L mg/L meq/L meq/L meq/L 11S.25E.36.144 GW S 240 1300 2500 12.0 106.9 108.9 11S.25E.36.14234 360 1200 2100 18.0 98.7 91.6 11S.25E.36.14234A 340 1500 3200 17.0 123.4 139.4 11S.25E.36.213 PVAC 590 91 26 29.4 7.5 1.2 Table 4.5. Cation concentrations in Section 36, T11S-R25E samples.

Anions HCO; SO42" Cl- HCO3- s0 42- cr Twn - Rge - Sec mg/L mg/L mg/L meq/L meq/L meq/L 11S.25E.36.144 GW S 854 5500 3700 14.0 114.5 104.4 11S.25E.36.14234 793 5500 3000 13.0 114.5 84.6 11S.25E.36.14234A 952 6100 4900 15.6 127.0 138.2 11S.25E.36.213 PVAC 171 1700 20 2.8 35.4 0.6 Table 4.6 Anion concentrations in Section 36, T11S-R25E samples. cations in this plot are closely clustered except the Section 36.213 sample. The lone

deviation plot, Section 36.213, shows a high percentage of HCO 3- and Mg' at the

expense of Na4 and C1 - .

The Fingerprint diagram for the samples listed as Qua- Ternary Plot ternary alluvial waters is in

Figure 4.27. The three samples from Section 36 that have similar analyses all present a distinctive signature.

The signature for the fourth sample (Section 36.213) sug- Figure 4.26. Ternary plot of 4 samples, T11S- R25E. Ion percent plot meg/L. Quaternary gests that water has more of a alluvium. 146

San Andres Formation signature —o-Fingerprint Diagram 1000 (compare with Figure 4.4 on page

128).

The second set of samples reported as Quaternary alluvium is 0.1 TIIS-R25E Sec. 36 4 Wells from the south west quarter of TI OS- Alluvial Aquifer I 1 2 3 4 5 6 7 6 9 10 R25E. The locations of these wells Na4K=1 Mg =2 Ca=3 Si 02 =7 ECU-8 SO 4=9 C1=10 2 are shown in Figure 4.28 and are Figure 4.27. Fingerprint diagram of 4 samples from T11S-R25E. Reported as close to samples plotted in Figure Quaternary alluvial aquifer. Quaternary alluvival signature. 4.12, page 133.

The Fingerprint diagram for the samples in T11S-R25E is shown in Figure 4.29. The

signature of these samples is apparently

R 25 E

6 3 4 2 2 1

7 8 9 11 12

18 17 16 105 14 13

19 20 21 22 23 24

30 29 28 27 26 25

31 32 i 33 34 35 36

0 I 2 3 4 5 6 7 8 9 10 Ha+K1 Vi2 Ca=3 Si02=7 1-1:00 SO 4=9 Cl =10 Figure 4.28. Location of T10S-R25E Figure 4.29. Fingerprint diagram of 3

samples. Reported as Quaternary samples from T10S -R25E. Reported Alluvium shallow aquifer. One section is from Quaternary Alluvial aquifer. 1.6 km wide Quaternary alluvial signature. 147 unlike that of Figure 4.27. The signature of these water samples is correlative to other water samples from TIOS-R25E that bear an Artesia Group signature (compare Figure

4.27 with Figure 4.15 on page 137).

Santa Rosa Sandstone Fingerprint Signature

The Santa Rosa Sandstone is absent

from the study area. This spring water R27E

6 5 4 3 2 I sample analysis is included for reference to

8 9 10 ii 12 an outside-of-the-basin aquifer for com-

IR 17 16 15 14 13 parison with the basic signatures developed

19 20 21 22 23 24 in this study. Cibola Spring is in Section

30 29 zs 27 26 25 2, T1S-R27E, DeBaca County, New

31 32 33 34 35 36 Mexico, Figure 4.30, and is the location

where this water sample was collected. Figure 4.30. Location of Section 22, The spring produces from the Triassic T1S-R27E spring sample. One section is 1.6 km wide. Santa Rosa Sandstone. Cibola Spring is found approximately 32 km south-southeast of Fort Sumner, New Mexico, and 10.5 km southeast of the Pecos River.

The cation and anion composition of the Santa Rosa Sandstone sample is presented in Table 4.7 and Table 4.8 respectively. In addition, the dissolved silica for this sample is 28 mg/L or 0.294meq/L. The Piper diagram for this sample is in

Figure 4.31. 148

Cations Ca2÷ Me+ Na++K+ Ca2+ Mg2+ Na++K+

Twn - Rge - Sec mg/L mg/L meq/L meq/L meq/L 01S.27E.22.333 COO 49 24 67 2.4 2.0 2.9

Table 4.7. Cation composition of T1S - R27E sample.

Anions Hc03- +CO32- SO42 Cl" HCO3- +CO32- SO 42 Ct

Twn - Rge - Sec 01S.27E.22.333 CIBO 250 100 33 4.1 2.1 0.9 Table 4.8. Anion composition of T1S-R27E sample.

The ternary cation portion of the Piper diagram illustrates that sodium and calcium dominate over magnesium ions in the water. The ternary anion portion of the diagram shows that bicarbonate is dominating over both Cl- and SO42- ions. The higher concentration of silica in this sample over what is normally seen in the Roswell ground water basin is reflected in the Fingerprint diagram, Figure 4.32. The Fingerprint diagram

illustrates the sig-

Piper Ciagrern nature of a non-ma-

rine fluvial quartz

sandstone and is

distinctive corn

bonate and evapo-

rite lithologies of

+ 50 ca2. Cr Ei0 HCC0 3 the Roswell

Figure 4.31. Piper diagram of 1 spring sample from Section 22, groundwater basin. T1S-R27E. Reported from Santa Rosa Sandstone (Triassic) aquifer.

149

-o-Pingerprint Diagram -o-Fingerprint Diagram

100

‘E' 0.1 TISHR21E S€-t. TIT:riCE E Sorioo 5 welli :Santa Ro5a S20d5tdno ktuslan Aqulfer 0.01 0.01 0 L 2 4 5 6 7 8 9 10 0 2 3 4 5 6 7 8 9 10 NaiK=1 11g=2 Cal tia+K=1 M3=2 Ca=3 Si02 =7 HCO9=0 SO 4=9 61=10 SiO =7 WO =8 60-O 61.10 2 3 4 Figure 4.32. Fingerprint diagram of 1 Figure 4.33. San Andres carbonate spring sample from Section 22, T1S- aquifer Fingerprint diagram signature. R27E. Reported from Santa Rosa Copy of Fig. 4.4. See text discussion. Sandstone (Triassic) aquifer.

Stage Four: Interpretation of the Data

This portion of the study shows that the principal aquifers of the Roswell groundwa-

ter basin each carries a distinctive hydrochemical signature. Three separate signatures

have been determined, one signature each for the San Andres Formation portion of the

aquifer system, one for the Artesia Group dominated portions of the aquifers, and the

third for the shallow Quaternary alluvial aquifer. The three basic signatures observed in

the Roswell groundwater basin are again shown as Figures 4.33, 4.34, and 4.35.

The hydrochemical plots are outstanding examples of tracking the water-rock

reaction in relation to the known local geology. An excellent sample of tracking change

shows clearly in the ternary plot, Figure 4.24. The general patterns of the Fingerprint

diagrams are representative of the lithologie composition of the aquifer. By the process

of correlation and observation of the subtle changes to those patterns in the Fingerprint 150

ngerpr nt Oiaqr.I

0.1 IL

1 2 3 4 5 6 7 8 9 10 71aq.! 181=2 Car3 SiO =8 SO C1=10 23 4 5 6 7 8 lia4 =1 119=2 Ca=3 Si g =7 1-1079A3 M =9 CI =10 4 Figure 4.35. Quaternary alluvial Figure 4.34. Artesia Group aquifer aquifer Fingerprint diagram signature. Fingerprint diagram signature. Copy Modified from Fig. 4.27. See text of Fig. 4.15. See text discussion. discussion. diagrams the ever-varying lithologies and structural boundaries can be detected. Hydro chemical signatures developed in the ternary and Piper diagrams are also important and support the correlation and delineation of aquifer host rocks.

The ternary plots, like the Fingerprint diagrams, Ternary Plut

SO 2- illustrate distinctive water-rock reactions. Fig- ures 4.36 and 4.37 show the unique clustering patterns of the plots of waters

50 HCO from the three different 3 source aquifers, the San Figure 4.36. Ternary plot of three aquifer host water Andres Formation (see sample sets illustrating anion grouping by source. 151

(rnary Plot Figure 4.2), the Artesia

mg 2 , Group (see Figure 4.13),

Quaternary and the Quaternary sedi- alluvium mentary rocks and allu- 50 San Andres Fm. vium(see Figure 4.26, ex-

cluding one non-Quater-

nary sample).

Use of these signatures Figure 4.37. Ternary plot of three aquifer host water sample sets illustrating cation grouping by source. with a more thorough and uniform set of chemical analyses would enable not only better distinction between producing aquifer horizons, but would also aid in delineating the general and local groundwater flow path patterns and directions. The ability to discriminate between paths of flow based on hydrochemical plots and Fingerprint diagrams is very effective. Again, knowing which direction groundwater cannot move is often as important as seeing the path along which it does flow. Mazor's (1991, p. 83) statement that, cited again, best describes the necessity for geoscientific evaluation of flow paths is: ". . . one can never deduce flow directions from water levels alone."

In the context of this study, Mazor's statement is important. The use of water levels for construction of a water table map in any given area yields a classic portrayal of hydraulic gradient. From the water level contours it is concluded that the net flow direction is perpendicular to any given contour on the water table map. In "net flow direction" the operative word is net. "Net" is a vectorial function of the direction of 152 water flow. For many purposes, such as local site evaluation of groundwater withdrawal, the vectorial summation may be sufficient, adequate, and realistic. Also, in aquifers of high lithologic uniformity the water table map may portray both the local and regional gradient and flow direction adequately. Unfortunately, many major aquifers do not

possess the ideal qualities of being homogeneous and isotropic.

The characterization that the Fingerprint diagrams represent the waters produced

from a particular lithologie assemblage is graphically shown in Figure 4.38. On this map the plot of water concluded to be from the aquifer dominated by the Artesia Group are

plotted with an "A." Those waters dominated by the San Andres Formation are plotted

with an "S."

Data are present in sufficient amounts to confirm the thesis of this study, that the

hydrochemistry reflects the rocks through which the water has moved. The map further

displays in the northern portion of the basin (where the available data are concentrated)

that the distributions of the "A's" and the "S's" conform to the structurally controlled

stratigraphie distribution of the Artesia Group as is projected in Figure 2.4, on page ?.

It is significant here that the distribution of the samples used to prepare the Finger-

print diagramswere not usually identified in the USGS database as to well depth. The

assumption here is that the development of a water well in an artesian zone will generally be sought at the shallowest depth at which it can be satisfactorily constructed and is thereby an indirect indication of the development of major porosity and permea-bility.

The Fingerprint diagram delineates the stratigraphie unit from which the water was produced. Within the Roswell block, between the Six Mile Hill fault and the Y-0 fault, 153

1 O

Tertiary intrusives

Strike-slip fault

A Artesia Group signatures

S San Andres Formation signatures Figure 4.38. Distrubution of San Andres Formation and Artesia Group Fingerprint diagram water signatures in the Roswell groundwater basin and drainage area. See accompanying text for discussion. 154 and on the Y-0 block (Orchard Park) located south of the Y-0 fault the distribution of water types "S" and "A," (Figure 4.38) support the conclusions of this study as to the significance of the right-lateral strike-slip faults on the hydrogeologic framework of the

Roswell groundwater basin. The distribution of the water types, particularly on the

Roswell block, is considered strong confirmation of the validity of Mazor's (1991, p. 83) statement about the direction of groundwater flow that is repeatedly cited above.

In the Roswell groundwater basin the indications of groundwater flow directions determined from regional water table and potentiometric level mapping (Maddox, 1969;

Welder, 1983), in my opinion, are often misleading—or meaningfully incorrect. This conclusion is substantiated in Figure 4.38 by the consideration of the hydrochemistry of the groundwater along its flow path, as Mazor states above, and as was discussed in the introduction to this section. 155

Hydrogeochemistry

Introduction

The hydrogeochemical considerations of the Roswell groundwater basin are used in this study to mean the relationships of the groundwater to the rocks over and through which the water flows. The chemical nature of the groundwater is controlled by the chemistry of the aquifer rocks.

One aspect of studying the hydrogeology of a basin is to consider the hydrogeochemistry of the groundwater. The contributions from the lithologies over and through which the waters have passed should provide an avenue to reconstruct the minerals that could have been dissolved to produce the hydrochemistry now present in a given water sample. This information would be useful in the identification of the individual aquifer lithologies, the determination of flow paths, and the influences of structural/stratigraphic boundaries.

The reconstructed mineral assemblages must reflect a reasonable approximation of the aquifer lithologies that influenced the water-rock reaction. They should be considered in relation to the known geology of the basin and used with the hydrochemical ternary plots, Piper diagrams, and the Fingerprint diagrams presented earlier. The reconstructed mineral assemblage should be used to strengthen the interpretations—and plausibility—of using hydrochemical plots such as the Fingerprint diagram signature.

Widespread fluvial distribution of igneous detritus and silicate minerals has occurred within the Roswell groundwater basin. Tertiary igneous complexes in the area are mostly 156 felsic (Kelley, 1971; Moore, Thompson and Foord, 1991; Pertl and Cepeda, 1991). The felsic presence serves as one reasonable source for silica plus small amounts of alkali metal ions. While felsic igneous rocks, their detritus, or resulting arkoses are not the only source of silica, it must be kept in mind that groundwater obtains silica as silicic acid derived primarily from the weathering of silicate minerals. Detrital silicate minerals are also present in carbonates and shales, both of which are abundant in the region—and to which the groundwater has access. Even in areas essentially lacking igneous detritus sufficient silicate particles are available within the marine carbonates and their contained clays to supply the small amounts of silica carried by most groundwaters.

Albite is the generic-type mineral selected for this study because it is a logical source as a specific sodium silicate mineral source (Faure, 1991) and for its normally widespread occurrence in felsic igneous rocks such as those found in the greater Roswell groundwater basin. Potash is also present in felsic igneous rocks which can result in higher concentrations of K20 than Na20. Albite is used in this study to normalize silica because the amount of potassium reported in the analyses of the Roswell groundwater basin is nominal compared with sodium.

Much quartz (Si0 2) is present as sandstones in the Artesia Group, the upper horizons of the San Andres Formation, the Glorieta Sandstone, and the Yeso Formation. The weathering and water-rock reaction between groundwater and free-quartz contributes almost no silica to groundwater (Faure, 1991).

Calcium, magnesium, and bicarbonate are available to groundwater by the dissolution of limestones and dolomites of the Permian San Andres Formation and the Artesia 157

Group, and to a lesser extent from the Yeso Formation. A substantial amount of sodium and chlorine are available from halite beds and inclusions within the Artesia Group and from the more evaporitic northern facies of the San Andres Formation. Similarly, calcium and sulfate are available for solution by groundwater from the Yeso Formation, the Artesia Group—and commonly from within the San Andres Formation.

The analyses are hardly perfect; they show varying degrees of cation/anion ionic electrical imbalance. The abundance of both bicarbonate and sulfate sources allows using these two ions to balance the anion/cation electrical charges from the analyses.

In the study of some mineral reconstructions discussed below, it was found that sodium ions sometimes appear in the remaining "dissolved" fraction. The "dissolved" fraction contains ions not used in reconstructing normative minerals. If sodium is in that remaining fraction it is in excess within the system. The excesses of sodium ions in the remaining fractions are anomalous in this study and will be discussed below.

This study uses mineral stability diagrams to help evaluate the possibility of sodium- calcium replacement in Na-smectite clays as a source for those anomalous sodium concentrations remaining in the "dissolved" fraction after the reconstruction of the normative minerals. The sodium and chloride contents are generally reliable in most chemical analyses. The quantitative amounts and the ratios of sodium and chloride ions can also be useful in determining the paths of surface and subsurface movement of water. 158

Reconstruction of Minerals Contributing Ions

The ions used in this study were sodium plus potassium, calcium, magnesium, sulfate, chloride, bicarbonate, and dissolved silica as Si02 . In this study most of the analyses (see Appendix A, page 218) reported sodium and potassium as sodium plus potassium. Of the total 1,885 water samples available for this study, only 105 sample analyses reported sodium and potassium ions separately. One of those samples was exceptionally anomalous compared to the remaining 104 samples. The 104 samples average 98.64% sodium and 1.36% potassium. These values were used to estimate individual amounts of Na. + K+ from the undifferentiated sodium plus potassium values in the samples used in this part of the study.

General Methodology of Mineral Reconstruction

Each water sample analysis was converted from mg/L to micromoles/L (umol/L).

Where sodium plus potassium was reported as combined, they were apportioned based on

98.64% sodium gmol/L and 1.36% potassium gmol/L.

The first step in the reconstruction involves the correction for ion concentrations typically found in rainwater of the interior continental United States. These values

(Faure, 1991) are shown on the mineral reconstruction table in ,umol/L. The rainwater

,umol/L ion values were deducted from the sample values to obtain the net ,umol/L of each ion. This correction for rainwater is to not include ocean water salts obtained in the meteoric evaporation phase (see example in Table 4.27, page 192). 159

The next step required the balancing of the anion/cation ionic charges. The procedure for this step was to use bicarbonate to adjust the total anion charges to the cation charges.

Bicarbonate is required to combine with silica to form albite, the first mineral reconstruction. Therefore, if a bicarbonate adjustment to the anion charges left a deficit for the silica requirement, then sulfate ions were used to allow sufficient bicarbonate to remain for combination with silica. Bicarbonate is generally considered as abundantly present from the dissolution of carbonates and is not usually a limiting ion. Sulfates are often observed to be in excess after all the mineral reconstructions are completed.

The first mineral reconstructed is albite, NaAlSi 30 8 . As discussed above, albite is present in the felsic igneous rocks (in situ and detrital) of the region in association with the alkali feldspar group--of which albite is considered a member. The water-rock reaction results in the formation of kaolinite as the stable phase according to the following equation:

2NaAlSi 30 8 + 2CO2 + 3H20 Al 2 Si20 5 (OH)4 + 2Na+ + 2HCO 3" +45i0 2

The equation is balanced with the assumptions that Al is conserved and HCO3- is produced through the dissociation of carbonic acid which is ubiquitous in the carbonate environment. The result is that the dissociation of one mole of albite in carbonic acid will yield one mole of Nat, one mole of HCO3", and two moles of Si0 2 . In the Roswell groundwater basin silica is not in excess, but commonly ranges from 10 to 20 mg/L. One mg/L Si02 equals 16.64 pmol/L.

With albite removed from the water sample, halite (NaCl) is next reconstructed according to the dissociation equation for the mineral: 160

NaCl (s) + H20 -4 Na (a) + C1-(,) + H20

This is an important relationship in that both sodium and chlorine obtained from halite are in groundwater in an ionic molar ratio of 1:1. This association will be often cited.

The next mineral reconstructed is dolomite, a common mineral in the San Andres

Formation of the Roswell groundwater basin. The use of dolomite rather than limestone was due to the water-rock reaction being very similar save that dolomite produces a Mg' ion besides the two bicarbonate HCO 3- ions. Limestone dissolution is ubiquitous in the

presence of groundwaters. Limestone dissolves more readily in groundwater than does

dolomite. Because of the high dolomite content of the carbonates of the Permian Basin, its use, in my opinion, produces a more conservative reconstruction. It also follows the

example of Faure (1991). Dolomite reacts with carbonic acid to yield:

CaMg(CO 3 )2 + H20 + CO2 Ca24 + Mg" + 2HCO3-

This is an ion ratio of Ca' : Mg 2+ : HCO3- of 1:1:2.

Similarly, anhydrite as CaSO 4 dissolves in water to yield Ca+ 2 + SO4- 2 on a ratio of

1:1. In the Permian Basin anhydrite is probably the major source of sulfate ions, but it

must also be kept in mind that marine clays contain sulfate ions adsorbed onto their

surfaces. These sulfates are subject to removal as meteoric groundwater contacts the clays deposited with marine carbonates and shales (J. Quade, pers. comm., 1994).

Langbeinite (K2Mg2(SO4)3) is a common marine evaporite. As evaporites are abundant in the Roswell groundwater basin, this mineral was selected as a likely reconstruction mineral for the area. Langbeinite is produced commercially from the 161 potash mines of the Permian Delaware Basin and is a common evaporite mineral throughout the Permian Basin (Doug Heyn, Head Chemist, IMC Fertilizers, Carlsbad,

New Mexico, pers. comm., Oct., 1995). Langbeinite is dissolved from the evaporites, probably Artesia Group or possibly the Yeso Formation, in the mole ratio of

K':Mg2 :SO42- of 2:1:1.

Sodium nitrate present as Na4- and NO; is probably dissolved from nitratite, a mineral commonly associated with gypsum and caliche in the Permian Basin area. Both gypsum and caliche are abundant in the Roswell groundwater basin. NO; is also a common contaminant of groundwater owing to its use as fertilizer in agriculture and from dairy and cattle feed lots. This source of nitrate could contribute to the samples collected from some irrigated areas, but the vintage of the samples probably reduces that possibility. For conservative estimations, all the nitrate reported in the water analyses used in this study is attributed to the mineral nitratite. The NO; levels reported by the USGS and

compiled in Appendix A are low. Nitratite solution provides a molar ratio Na :NO - 1:1.

Kieserite, a hydrous magnesium sulfate, is a common evaporitic mineral and has been included in the mineral reconstructions partially to account for high magnesium excesses that occur. Kieserite is one of the principal magnesium sulfates produced from the IMC

Fertilizer mine in the Eddy County, New Mexico portion of the Permian Delaware Basin

(Doug Heyn, Head Chemist, IMC Fertilizers, Carlsbad, New Mexico, pers. comm., Oct,

1995). The dissolution of kieserite also yields a molar ratio of Mg:SO42- of 1:1. Kieserite is shown on the following tables as MgSO 4 rather than as the hydrous MgSO4 • 7 H20. 162

Na ' IC Na . -fiC Ca' r Me SO:- HCO, ' CO — CI" Silica NO, - Analysis mp'3, 26 120 36 270 230 30 15 4 p moll 1115 9 1124 2941 1481 2811 3769 846 250 65 Rain. p moll 17 5 35 4 31 2 121 0 Net p moll 1098 4 2906 1477 2780 3767 834 250 65 Electric neut4.1eql lations = 9869 4nions = 10161 ation-Anion 96 difference = 2.8 / 3 Adjusted p mo1/1, 1098 4 2906 1477 2780 3475 834 250 65 .411nte,p moll 97.3 4 2906 1477 2780 3351 834 0 65 ,VaClp moll, 139 4 2906 1477 1780 3351 0 0 65 Dolonzite,p mold__ 139 4 2069 640 2780 0 0 0 65 Anllydrite, y moll 139 4 0 640 711 0 0 0 65

/till g ., (SO. y moll 139 0 0 635 705 0 0 0 65

AraNO, , p eq/L 75 0 0 635 705 0 0 0 0

:11gSO4 , peq/L 75 0 0 0 70 0 0 0 0 Table 4.9. Mineral reconstruction for Section 27.21333, T12S-R24E water sample.

Mineral p mol/L mg/L Weight %

An example of the full calculation albite 124.8 32.7 5.34% halite 834.3 53.0 8.65% process is presented in Table 4.9, a nor- dolomite 837.6 154.5 25.19% anhydrite 2068.5 281.5 45.91% mative mineral reconstruction analysis of 1ÇMg 2 (SO4)3 6.4 1.1 0.18% NaNO3 64.5 5.5 0.89% the well water sample from the San MgSO4 635.3 76.5 12.47% dissolved 144.2 8.4 1.37% Andres Formation signature group of Totals 613.2 100.00% Table 4.10. Weight % of minerals in analyses as discussed beginning on Section 27.21333, T12S-R24E waters shown in Table 4.7. page 123

San Andres Formation Signature

Table 4.10 is a summary of the ,umol/L, mg/L and mineral weight percentage of the reconstruction showing the weight percent of minerals dissolved that will produce the waters shown in Table 4.9. This sample was discussed as to its hydrochemical makeup, along with others from the same Township, starting on page 123.

Figure 4.39 is a bar chart prepared to represent the reconstructed mineral assemblage visually. The correlation value of these charts will be discussed later in this section. 163

Normative Mineral Reconstruction

50.00

45.00 40.00

35.00

30.00

25.00

20.00

15.00 10.00

5.00 0.00 -

Section 27.21333, T12S-1224E

Figure 4.39. Normative mineral chart Section 27.21333, T12S-R24E. San Andres Formation.

The remaining dissolved component of Table 4.10 is made-up of 75 ,umol/L of Na+

and 70 ,umol/L of SO4 . In this sample accounting for the probable origin of each of the

major analysis constituents has been possible. The small remaining amounts of Na and

SO42- can readily be attributed to ce replacement of Na and extraction of SO42- from

Na-smectite clays, as is discussed later in the section on mineral stability diagrams.

Some water samples evaluated in this study exhibit the characteristic of having excess

Na (anion limited). This trait is thought here to be anomalous and requires consideration.

Table 4.11 and its bar chart in Figure 4.40, are for a sample from the same township as the sample in Table 4.10. Consistency is shown between the two samples in that the 164 major mineral constituents are dolomite Mineral 1.z mol/L me, Freight %

albite 149.8 39.3 5.7/o and anhydrite. In the Table 4.11 sample, halite 562.4 35.8 5.27% dolomite 1163.5 214.5 31.61% the remaining dissolved component totaling anhydrite 2478.0 337.3 49.69% K21vi82(SO4)3 1.6 0.3 0.04% 449.7 pmol/L is composed of 136 pmol/L NaNO3 0.0 0.0 MgSO4 300.2 36.1 5.32% dissolved 449.7 15.5 2.29% Mg', 272 ,umol/L Cl-, and 42 pmol/L Totals 678.8 100.00% NO - This water exhibits a deficit of so- Table 4.11. Weight % of minerals from Section 22.41333, T12S-R24E waters. dium (relative to chloride) and an excess of 3.3 mg/L magnesium and 9.6 mg/L chloride. This is a Mg2+ :C1- pmol/L ratio of 1:2.

That ratio is discussed below (see page 176). The deficit of sodium in absolute terms is small, but apparent. The original analysis showed only 17 mg/L of Ne + K. Approxi-

Normative Mineral Reconstruction

50.00 45.00

40.00

35.00

30.00

25.00 20.00

15.00 10.00

5.00

0.00 -s o o

Section 22.41333

Figure 4.40. Normative mineral chart Section 22.41333, T12S-R24E. San Andres Formation. 165

iu mol/L mg/L Weight % mately 91% of the original Mg2+ and 68% Mineral

albite 124.8 32.7 5.27% have been accounted for by mineral of the Cl" halite 4557 29.2 4.70% dolomite 708.5 130.7 21.05% reconstruction. anhydrite 2687.9 365.8 58.95% K2Mg2(SO4)3 0.0 0.0 The sample shown in Table 4.12 is NaNO3 0.0 0.0 1\4004 91.9 11.1 1.78% from the same well as in Table 4.11, ex- dissolved 1575.3 51.2 8.24% Totals 620.6 100.00% cept that it was taken 13 months earlier. Table 4.12. Weight % of minerals from Section 22.41333, T12S-R24E waters. Both wells were sampled during the winter Sampled 13 months prior to analysis in Table 4.9. months of January and February when irrigation is minimal. The earliest sample (Table 4.12) shows an excess of sodium, magnesium, and chloride. It also shows good correspondence with the other samples

Normative Mineral Reconstruction

60.00

50.00

40.00

30.00 n-•

20.00

10.00

0.00

Section 22.41333, TI 2S-12.24E (Earliest sample)

Figure 4.41. Normative mineral chart Section 22.41333, T12S-R24E. San Andres Formation. 166 from this Township as to the percentages Mineral A mol/L mg/L, Weight % of reconstructed minerals shown in arbite 141.5 37.1 5.37% hdrite 834.3 53.0 7.67% dolomite 1132.6 208.8 30.22% Table 4.12 and Figure 4.41. anhydrite 2263.8 308.1 44.59% K2Mg2(SO4)3 9.6 1.7 0.24% The dissolved remainder is composed NaNO3 75.8 6.4 0.93% lvigSO4 420.5 50.6 7.32% of 512 pmol/L Mg', 1024 pmol/L cr, dissolved 494.0 25.2 3.65% Totals 691.0 100.00% and 39 ,umol/L NO 3 - (12.4 mg/L, 36.3 Table 4.13. Weight % of minerals from Section 22.23111, T12S-R24E waters. mg/L, and 2.4 mg/L respectively). In both samples this well yielded very low sodium plus potassium values. The sample

analysis reported 14 mg- /L Na + All of the sodium was consumed using the silica to reconstruct albite plus forming about 4.0 x 10 moles of halite/L. This sample also

Normative Mineral Reconstruction

45.00

40.00

35.00

30.00

25.00

20.00

15.00

10.00

5.00

0.00 - o 0 o o 0 CO ;.‘g

Figure 4.42. Normative mineral chart Section 22.23111, T12S-R24E. San Andres Formation. 167 has a Me:C1- pmol/L ratio of 1:2. The significance of the Me:C1- ratio is discussed below.

Table 4.13 shows the mineral reconstruction of the fourth well sample from this

Township (see locations on Figure 4.38, page 153). The well is in section 22.23111, close to the other samples shown in Table 4.10, -4.12 above. In the Table 4.13 sample, anhydrite and carbonate play the most significant roles in providing the ions required for this water. That relationship is shown on the bar chart in Figure 4.42.

The Section 22.23111 water sample shows an excess of 304 ,umol/L Na and 190

,umol/L SO 42- . Although this is not quantitatively a large amount of Na (7.0 mg/L), it is believed here to be an important consideration in examining the water-rock reaction and potential flow paths. This sample shows no excess of either Cr or Me+.

The last sample from T12S-R24E, shown in Table 4.14 and Figure 4.43, is from section 15.43111. The mineral reconstruction shows the waters in this flow path were in contact with more halite and lesser dolomite as carbonate Mineral y mol/L mg/L Weight % than the sample taken 1.6 km albite 133.1 34.9 5.15% south in section 22 (see Table halite 1308.2 83.2 12.26% dolomite 840.6 155.0 22.86% 4.11). The remaining dissolved anhydrite 2310.7 314.5 46.37%

K2Mg2(SO4)3 10.6 1.9 0.27% component is composed of 5 NaNO3 0.0 0.0 pmol/L Mg' (1.1 mg/L), 90 MgSO4 666.7 80.2 11.83% dissolved 203.1 8.5 1.25% pmol/L Cl - (3.2 mg/L, and 68 Totals 678.1 100.00% Table 4.14. Weight % of minerals from Section pmol/L NO; (4.2 mg/L). This 15.43111, T12S-R24E waters. 168

Normative Mineral Reconstruction

Figure 4.43. Normative mineral chart Section 15.43111, T12S-R24E. San Andres Formation.

sample also has a remaining dissolved Me:Cl" ,umoUL ratio of 1:2.

Table 4.10- 4.14 and Figure 4.39 -4.43 have presented the mineral reconstruction summaries for the T12S-R24E water samples used to define the San Andres Formation

Fingerprint diagram signature in the previous section (see the discussion beginning on page 123). These tables reflect the dissolution of minerals that could produce the chemical composition of the waters analyzed in the samples used. Similar data will be presented to contrast the lithologies and support the concept of water signature definitions for the Artesia Group and the Quaternary alluvial aquifers within the Roswell groundwater basin. Reconstruction tables for all the other water samples (including those shown herein) are included in Appendix C, beginning on page 260. 169

Artesia Group Signature Mineral Atmol/L mg/L Weight %

0.0 hydrochemical signature albite 0.0 The of the halite 115792.9 7361.0 80.06% dolomite 763.5 140.8 1.53% waters moving in the Artesia Group portion attliydrite 8270.1 1125.6 12.24% IÇMg2 (804 )3 1432.3 249.6 2.71% of the carbonate aquifer is summarized in NaNg 0.0 0.0 MgSCA 468.6 56.4 0.61% the Fingerprint diagram, Figure 4.15, dissolved 8239.0 261.5 2.84% Totals 9194.8 100.00% page 137. Mineral reconstruction from the Table 4.15. Weight % of mineral from Section 31.34, T10S-R25E sample. waters in the six samples comprising the representation of the Township begins with the sample from Section 31.34, T10S-R25E and is shown in Table 4.15. The graphical chart representation of the Table 4.15 data is shown in Figure 4.44. This chart illustrates the significant lithologic differences compared with San Andres signatures in Figure 4.43.

Normative Mineral Reconstruction

Sec. 31.34, 1 . 10S-R25E

Figure 4.44. Normative mineral chart Section 31.34, T10S-R24E. Artesia Group. 170

The mineral reconstruction for the Mineral ji moll, mg/L Weight %

albite 174.7 45.8 0.62% Section water sample from 33.423, T10S- halite 72725.6 4623.2 62.16% dolomite 306.6 56.5 0.76% R25E is shown in Table 4.16. The chart anhydrite 15344.6 2088.4 28.08% K2Mg2(SO 4) 3 899.0 156.7 2.11% for this sample is displayed in Figure 4.45. NaNg 0.0 0.0 MgSO4 2463.6 296.5 3.99% These first two charts of Artesia Group dissolved 5417.2 170.5 2.29% Totals 7437.6 100.00% signature waters reflect a substantial Table 4.16. Weight % of minerals from Section 33.423, T10S-R25E. mineral reconstruction difference from those seen above for San Andres Formation water sample signatures. The principal component change between the formations waters of the San Andres Formation and the

Artesia Group is the strong shift from carbonate in the San Andres Formation to halite in

Normative Mineral Reconstruction

13 40

.4 30

I- 20

1 0

O oZ-1\- cs7;

Sec. 33.423, TIOS-R25E

Figure 4.45. Normative mineral chart Section 33.423, T10S-R25E. Artesia Group. 171 the Artesia Group. That change is also accompanied by a shift in anhydrite percentage.

The mineral weight percentage of dissolved anhydrite is higher in the carbonate aquifer

and lower in the evaporitic aquifer. Comparison of the charts for the Artesia Group

reconstructed mineral assemblages illustrates that the weight percentage of the

reconstructed minerals varies from location to location, but the minerals present are

consistent in their relative abundances.

The percentage of halite and anhydrite present will fluctuate depending upon the

makeup of the remaining solution. The carbonate will be limited because of the

deficiency of bicarbonate in the sample. The Artesia Group water samples show a

consistency of halite and anhydrite dominance, just as dolomite and anhydrite show

consistent dominance in the carbonate signatures of San Andres Formation waters. Both

the San Andres Formation signature waters and the Artesia Group signature waters are

very comparable to their respective aquifers' lithologic composition.

The next sample in this Township is from Section 32.424333. The mineral

reconstruction is presented in Table 4.17. The graphical chart for this sample is in

Figure 4.46. The relative concentrations - Mineral p mol/L nig/L Weight % of the minerals reconstructed are albite 108.2 28.4 0.10% consistent with the two previous Artesia halite 388438.5 24693.0 83.53% dolomite 0.0 0.0 anhydrite 24474.8 3331.0 11.27% Group water samples shown above. K2Mg00 4) 3 4823.2 840.5 2.84%

NaNO3 0.0 0.0 One less apparent difference between MgSq 2942.5 354.2 1.20% dissolved 9866.5 315.5 1.07% the San Andres Formation waters and Totals 29562.5 100.00% Table 4.17. Weight % of minerals from those of the Artesia Group is in the ion Section 32.424333, T10S-R25E. 172

Normative Mineral Reconstruction

90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0

10.0 0.0 —AV

Sec. 32.424333, TIOS - R25E

Figure 4.46. Normative mineral chart Section 32.424333, T10S-R25E. Artesia Group.

concentrations of the remaining dissolved portions after mineral reconstruction. Each sample from either group can be classed as having a Mg' and cr residual or a Na and

SO42- residual. Both ratios are important to understanding the water-rock reaction history, and for observing other influences that may play an important role in fashioning the present-day waters of a carbonate-evaporite basin.

Modern sea water has a ratio of Na : Cl" of 0.854:1 moUKg (Faure, 1991). Having

0.146 mol/Kg of Cl" greater than Na suggests that not all cr in modern sea water be derived from the 1:1 molar dissolution of halite. Because an excess of over Na+ occurs in modern sea water, the supposition that a similar chemical relationship prevailed 173 in Permian time may be reasonable. The converse of a Cl" excess is having an excess of

Early in this portion of the study it became apparent that there were areas that tended to show different general ranges in the ratio of Na : cr. Where the ratio is greater than

0.854:1 (modern sea water), the water sample has been referred to as having an excess of

Na 4 . There is no logical geological reasoning that should attribute all Na and cr solely to the dissolution of halite. Likewise, not all of the Na.- and can logically be allocated to connate sea water as a sole source. Some chlorides along with sulfates were probably originally included in the marine connate waters trapped within the carbonates, gypsums and clays deposited in the Permian oceans and near-shore areas. The Na-smectite clays are commonly associated with marine environments and are frequently found to contribute considerable sulfate and chloride to fresher groundwaters (Mazor, 1991).

Sodium excesses do occur in the remaining dissolved portion of the samples used in this study after the minerals have been reconstructed. These sodium excesses are discussed in the following section discussing calcium-sodium replacement and the resulting associated minerals' stability.

Appropriately, when an excess of cr appears in the mineral reconstructions of this study there is usually an excess of Mg'. More interestingly, the ratio of Cl to Mg' in

,umol/L for the remaining dissolved portion is very frequently 1:2.00. The overall average of all samples used in this study is Cl" : Mg' is 1:82. The ratio of 1:2 is the mole ratio of MgCl2. Based upon my own experience with oil field brines, and as cited by

Hurlbut (1952), MgCl2 is a common constituent of brine waters. Brine waters of these 174 high concentrations are found close to the Mineral p mol/L ing/L - Weight %

albite 141.5 37.1 0.40% basin Roswell groundwater (Havenor, halite 110002.1 6992.8 74.67% dolomite 1086.3 200.3 2.14% 1968). The intrusion of saline waters into the anhydrite 10398.3 1415.2 15.11% K2Mg/S0 4)3 1378.9 240.3 2.57% Roswell groundwater basin in T10S-R25E NaNg 8.1 0.7 0.01% MgSO4 3339.1 401.9 4.29% (refer to Figure 4.52, page ?) has been dissolved 1819.9 76.0 0.81% Totals 9364.4 100.00% discussed earlier and is of significance to Table 4.18. Weight % of minerals from Section 29.443333, T10S-R25E. show that subsurface brines can and do exist at shallow depths in the area.

Reconstructions for the Artesia Group (Table 4.18 and Figure 4.47) portray the minerals that could have been dissolved to form the waters analyzed from Section

Normative Mineral Reconstruction

Figure 4.47 Normative mineral chart Section 29.443333, T10S-R25E. Artesia Group. 175

29.443333, T10S-R25E. The original Mineral p mol/L meL Weight %

albite 224.7 58.9 0.17% ,umol/L Na :Cl" ratio is 1.014:1 and shows halite 428684.0 27251.4 78.87% dolomite 0.0 0.0 a Na excess is present in the sample. That anhydrite 34278.7 4665.3 13.50% K2MgiSO 4) 3 5325.0 927.9 2.69% indication is supported in the mineral NaNCs 0.0 0.0 MgS 04 4263.3 513.2 1.49% reconstruction with a residual dissolved dissolved 26157.4 1135.7 3.29% Totals 34552.4 100.00% Na of 1352 pmol/L. Also include in the Table 4.19. Weight % of minerals from Section 14.312, T10S-R25E. remaining dissolved unreconstructed portion is 468 prnol/L SO42- . Again, a probable source of the excess SO42" is release from clay surfaces where they were adsorbed during original deposition in the sulfate-rich ocean waters of the Permian.

Normative Mineral Reconstruction

80.0

70.0

60.0

50.0

40.0

30.0

20.0

10.0

0.0

`13

Figure 4.48. Normative mineral chart Section 14.312, T10S-R25E. Artesia Group. 176

The sample collected from Section 14.312, T10S-R25E is presented in Table 4.19 and

in Figure 4.48. The mineral reconstruction from this location has no carbonate formed,

partly because all the HCO3" was consumed in the reconstruction of albite. This is a

similar condition to that in the highest halite sample (Table 4.1 and Figure 4.44), but this

water had less Na' available. The original fitnoUL ratio of Na+:C1- in the sample was

0.950:1.

The remaining dissolved ion concentration contains 3380 ,umoUL SO 42 , 22644

,umol/L Cr and 134 ,umol/L of NO3 . These three ions show in the remaining dissolved

portion because the sodium, potassium and magnesium were limited and progressively

removed in the formation of albite, halite and langbeinite. This occurs in only one other

sample (Section 2.442, T11S-R26E, a spring sample). In the latter spring sample, the

cation/anion electrical ratio was extremely unbalanced (16.2%). The Table 4.17

cation/anion electrical imbalance was 2.4%, slightly higher than is desirable.

Although the electrical balance raises a question as to the use of the sample, the chart

(Figure 4.48) of the minerals reconstructed shows a high degree of correlativity with the

others from this Township. More important, it shows mineral consistency with the

Artesia Group water signature, and distinct differences with the San Andres Formation

water signatures.

The last sample from the T10S- R25E collection, and the last for the Artesia Group

comparison, is shown in Table 4.20 and Figure 4.49. This sample is from the

northwestern part of the Township. It is farthest removed from both the Y - 0 fault and

the area of greatest known saline encroachment into the artesian carbonate aquifer. 177

In the eastern part of this Township the .Mineral 1.i mol/L rng/L Weight %

albite 0.0 0.0 samples have the least percentage of halite 31017.6 1971.8 52.35% dolomite 632.3 116.6 3.10% carbonate whereas halite and anhydrite anhydrite 8891.5 1210.1 32.13% K2 Mg2(SO 4) 3 397.8 69.3 1.84% predominate. The presence of Mg' and NaNg 0.0 0.0 MgSO4 2390.0 287.7 7.64% dissolved 2345.6 111.1 2.95% Cl - in the ratio of 1:2 in the southwestern Totals 3766.6 100.00% part of the Township, coincident with an Table 4.20. Weight % of minerals from Section 5.3, T10S-R25E. area of known saline encroachment, further establishes the usefulness of using the mineral reconstruction process as both a correlation tool and an interpretative tool.

Normative Mineral Reconstruction

60.0

50.0

• 40.0 o

• 30.0 o 20.0

10.0

0.0 o C/3

Sec. 5.3, TIOS-R25E

Figure 4.49. Normative mineral chart Section 5.3, T10S-R25E. Artesia Group.

178

Quaternary Alluvial Aquifer Signature Mineral a nrol/L nrg/L Weight %

albite 99.9 26.2 0.19% Water samples from Section 36, T11S- halite 104360.4 6634.2 47.13% dolomite 1945.7 358.8 2.55% R25E have been previously identified as anhydrite 3901.7 531.0 3. 77%

K2Mg00 4) 3 1328.5 231.5 1.64% producing from the shallow Quaternary NaNq 0.0 0.0 MgSO4 50651.6 6096.7 43.31% alluvial aquifer (see Fingerprint diagram, dissolved 4329.0 197.7 1.40% Totals 14076.0 100.00% Figure 4.27, and figures in Appendix B). Table 4.21. Weight % of minerals from Section 36.144, T11S-R25E waters. The mineral reconstruction for the first of

these samples, Section 36.144, T11S-R25E, is presented in Table 4.21. The graphical

presentation of the mineral data is in Figure 4.50. The mineral reconstruction describes

the waters as having dissolved small amounts of dolomite, anhydrite and kieserite, but

Normative Mineral Reconstruction

50.0 45.0 40.0 35.0 1 5 30.0 a p. 25.0 20.0

it`fl 15.0 10.0 5.0 0.0 _AriOrr o il

Sec. 36.144, T11S-R25E Figure 4.50. Normative mineral chart Section 36.144, T11S-R25E. Quaternary sedimentary and alluvial deposits. 179 large and nearly equal volumes of halite Mineral ,u mol/L mg/L Weight %

albite 108.2 28.4 0.16% and langbeinite This assemblage suggests halite 137439.9 8737.1 50.73% dolomite 3279.4 604.7 3.51% an evaporite sequence deficient in anhydrite anhydrite 5018.9 683.1 3.97%

K,Mg2 (80 4) 3 1702.7 296.7 1.72% (compared with the San Andres Formation NaNQ 0.0 0.0 MgSCA 56749.3 6830.6 39.66% or Artesia Group). The Salado and Rustler dissolved 1318.7 40.6 0.24% Totals 17221.2 100.00% Formation meet this general description Table 4.22. Weight % minerals from Section 36.14234, T11S-R25E. and were locally present above the Artesia

Group-upon which much of the Quaternary alluvium is deposited (Kelley, 1971;

Kinney et al., 1968).

The second Quaternary alluvial aquifer water sample is reconstructed in Table 4.22

Normative Mineral Reconstruction

45.0 ., ii 40.0 35.0 30.0 II 25.0 20.0 15.0 10.0 - 5.0 ., figiir,il 0.0 , , :lin:i''' ' o f rn

Sec. 36.14234, T11S-R25E

Figure 4.51. Normative mineral chart Section 36.14234, T11S-R25E. Quaternary sedimentary and alluvial deposits. 180 and in Figure 4.51. Both figures are Mineral p mol/L mg/L Weight % correlative with the previous set and albite 108.2 28.4 0.16% halite 13 7439.9 8737.1 50.73% dolomite 3279.4 604.7 3.51% further describe and illustrate the signature anhydrite 5018.9 683.1 3.97%

Kpilg,2 (SO 4) 3 1702.7 296.7 1.72% of the Quaternary alluvial waters. NaNg 0.0 0.0 MgSQ 56749.3 6830.6 39.66% The last of the set of Quaternary dissolved 1318.7 40.6 0.24% Totals 17221.2 100.00% alluvial water samples is from the same Table 4.23. Weight % of minerals from Section 36.14234A, T11S-R25E. section and is shown in Table 4.23 and

Figure 4.52. The graphical chart shows a high degree of correlativity with the previous two samples of Quaternary alluvial aquifer waters. This correlativity is further displayed by using the three graphs, Figures 4.50, 4.51 and 4.52, combined and rotated to resemble

Normative Mineral Reconstruction

Sec. 36.I4234A, TI1S-R25E

Figure 4.52. Normative mineral chart Section 36.14234A, T11S-R25E. Quaternary sedimentary and alluvial deposits. 181

three side-by-side sample

quantity charts as is shown in ,...... ,,,...... ,,,,, : 4 Figure 4.53. This combination =..

chart shows that even where gm . : . there are high percentage r,.. , : , . .4 • • L ,III„„,,,,,,, • 9maim.--.- .....:',..0 .. 'II ', EgazgamoIII . differences in the quantities of ; .. WOE I NC . : the minerals present, there is a ------• ------—

high degree of correlation of Figure 4.53. Correlation using normative minerals the relative concentrations charts shown in Fig. 4.48, 4.49, and 4.50. between samples. The tech-

niques of correlation are commonly used, fundamental tools of the geologist for visual

comparisons to learn relationships that are often hidden from statistical, numerical

analyses.

Summary of Normative Mineral Reconstructions

Table 4.9 on page 162 is an example of the normative mineral reconstruction process.

The minerals selected for reconstruction are discussed at the beginning of this section.

During the presentation of the tables and charts that show the reconstructed minerals representing the distinctive Fingerprint diagram water signatures seen earlier, several analysis considerations were mentioned. These included a cation/anion electrical balance plus the remaining dissolved portion that was not reconstructed into minerals. 182

To do the normative mineral reconstructions the cation/anion electrical imbalance was corrected. Table 4.24 is a compilation of mineral reconstruction prno1/1, concentration and weight percentage of all the samples used in this study. Table 4.24 also shows the cation/anion balance ratio.

Table 4.25 details the composition of the remaining dissolved ion portion after mineral reconstruction. Table 4.25 also shows the Ne/C1- ratio before mineral reconstruction. Additionally, the ratio of Mg2+/C1.- in the remaining dissolved ion portion is shown in Table 4.25.

The information in Table 4.24 and Table 4.25 below is defined by "sheet" numbers

(i.e., 27-1). This number denotes the township group followed by the well or spring sample number for the purposes of tracking computations. The sample location is given as Township, Range, Section, and section subdivision as shown in the attached

Appendix A, Figure 5.1, page 219.

183

Normative nouerai calculations and data Calculation dissol ed dt

II iI7EK 2642% 56.91% 7.03% 2° /00.06% 373.5.42 3820 7.113 111 .. , . •°. A o sr ,

3- 1 0° 73A 7 0 509 0.797 -1- J1 4. 7 0. t2 Milliaiii iMilillainialra I ', . 9

43 [jA1*IiliZE)WI 0.266 morizzaintannaisaimmarantitawamottaui • 6141.1tIrMAMMargimuigamaxanzamminfammuionniagnal 170.46 0.878 tsfrm.awprttsrqmignaaztummuwjgmrswmmmtkamuaugtwwmtj 4-4 103.23 .1 . 4 P MAE= 1. 495 .1 ' . 7% 1.101MINESI . NOMA , 5 1.11' 4-5 10S. 5 •.I . I 111111-61MEI 153096 . A 100.0046 .57.31,g 1135.67 2.367 - 5 8 i • III' , . I'.. 3- 1 O.24E.S 3 .443 JP . 053% 70.70% 2. 5 1 .1 ,S .6 A . % 100.06% 5400.19 170.08 0.377 IMINI.R11, hilllifilffilliEl 5. A, 5- 3 MI 2.0696 lO000% Iii 107 09 1.164 MMUS iiirinuorou mmikvaiwzattigiumgaiuminuanstifinuainusra 1. 5=5 10S. 4 . . 1 RILIMIltiatiltalikilnIlLtilantiginiiiallffillgi ,.. 9 Ili1110150la NM 0. 39 0.4O% MEE 1.704 UMAIIIMM, iii 1. MiliangainWatia, :01 1.885 0.1.4 0.671 7-3 095.24E14.121 WIGG 1.78% 32.91.86 46.33% 1.0815 9.0115 3 7 % 100.0096 2854.13 90.82 1.303 7-4 9,S248.1 1.3 JCWIO 81% 28.44% 49.31% 0.93% 9.9196 5.8396 100.06% 3947.0 12642 1.189 7- thrgallargargaalliMaaMailMilMIL-Mal 39 3.88 1 4. : 1.560 Willa t . : , r - ' - • . . • :' . 111MtalliElligUalnaaIfignal.MMI .: 7-7 97-T24E02.42312 H I 73% M.= 4 . A .4 •. 05,009% 9 . . 6 . OT 7-8 05 31 SU .4 A MU 3.309% .6 A 0.97% REM 100.06% 2709.63 8619 0.680 • ' lallifini . • 4 .: i

• IIMA '5 • • 111110=111.11111Malligigibi . •s 8-2 0-9V.1.3E.36 . 133111 . A 46 . A . 111 155.55% 9. 5 : • . • 1. 42 9 E 20 I Ai 33 8116 100.06% 2110.04 8608 2.708 laillialaillillatillil t s rt

i f Awn MIMI iffeiriMillinina •, wlisigagigungmimmutuausigairami , , , . • 10-2 611;VV-AlliATIOLVIV.Rital 1 • ' ISMAILka likaaridiMMALLIUZIMENA 1.1 1.55 10-3 PW191' "4""nr-Vill • 94 WffiligntliMIUMIW41n11UTAMBOWSLOOMMUI HQ." 5.022 10-4 1.413 10-5 Piffillitall 2: 7,4 malifrojamm motimmilfid 871.5 0.190 5. 1191P1161011PHOU1111111 ' 1 ' I ' ' I ItIU g 1Lata.1.1=1111111111=111115MngAU allige= ' ' ' Will r -4. 1.089 (385 3 100.00% 12686 1.523

08S.24E.17.143P CO 41.04% 3. 2% . 4 1.4 A Ma=a A .62 4.86 1.46 11-2 08S 24E. 18.241 P CO 39.44% 40 56% . %11111.1=61 . A 100.00% 4 121.43 0.819 Il- 085.24E. /3.111 J CO Ewa 0% 23'. 0.41_ Il-4 INERILLIFISIIPLEMIliningitallIMEMILLIMILUNAMtkil/ 1100% n=12 Il.' 1.1. 3 - N. zsaturnistitinfillWaintmugramm iiii:gijuggiugaimmuczajuwam , .. . frM-6 4E.0 .4 C FA I. A 5 . 6 100.0 -A 109 .3 r CO' .377 086% 10 8116 0 12° 0 0 38 0.462

f 11111L1.11.2111.1111111.11MMILMealailkiligliaMitkalgeMallWAta 1. 0.55. l'i I4, ' 0.98 13 67928E.30.431BL4N 0.6756 17.606 1.52% 41.86% 1.27% 1055% 6.60% 100.0095 1290620 ''9.8 1.106 13-2 07, .26E.19.243 ELAN 23.875 1.6396 43.8796 0.9516 24.4196 5.2496 100 0096 5064.14 239.17 1.623 _. Table 4.24. Mineral reconstruction weight % calculation data.

Table 4.24 is continued on the next page. 184

Normative mineral calculations and data '. c ation MINIIIIMINIMINIIIIMIIIII IIIIMINIMMIIIIIMMILIZZLICIEZEI'.t • Amen Sheet ovms ip Range -ction Ilki 'II ite Mai le., Mg,(SO4)5 NaNC), MgSQ Disso ved Toble lirM png/L :4 ,lerence MMMIMIMIMINNISIMi,.0,0 IMMilMillMMN