© Lloyd

http://www.pnas.org/content/suppl/2014/05/08/1401150111.DCSupplemental/pnas.1401150111.sapp.pdf

SUPPORTING INFORMATION APPENDIX to:
– Chronological evidence fails to support claim of an isochronous widespread layer of cosmic impact indicators dated  to 12,800 years ago
– David J. Meltzer,1 Vance T. Holliday,2 Michael D. Cannon3 and D. Shane Miller4
– Introduction and notes
– page 2 Site specific discussions 1. Abu Hureyra
– Group 3a page 3 16. Lingen
– Group 2a page 37 2. Arlington Canyon
– Group 3a page 6 17. Lommel
– Group 3c page 38 3. Barber Creek
– Group 3a page 10 18. Melrose
– Group 3a page 39 4. Big Eddy
– Group 3b page 14 19. Morley
– Group 1a page 40 5. Blackville
– Group 3a page 20 20. MUM7B
– Group 2a page 41 6. Blackwater Draw
– Group 3a page 22 21. Murray Springs
– Group 3a page 41 7. Bull Creek
– Group 3c page 27 22. Newtonville
– Group 2a page 45 8. Carolina Bays
– Group 2a page 27 23. Ommen
– Group 3c page 4 6 9. Chobot
– Group 1a page 2 8 24. Paw Paw Cove
– Group 1a page 46 10. Daisy Cave
– Group 3d page 29 25. Playa Basins
– Group 2a page 47 11. Gainey
– Group 2a page 30 26. Sheriden Cave
– Group 3d page 48 12. kangerlussuaq
– Group 2a page 31 27. Talega
– Group 3a page 50 13. Kimbel Bay
– Group 3a page 32 28. Topper
– Group 3a page 53 14. Lake Cuitzeo
– Group 3a page 34 29. Wally's Beach
– Group 2a page 58 15. Lake Hind
– Group 3c page 37
_References cited page 64 List of Tables
(Tables 1-16 are included in the site specific discussions; Tables 17-19 follow the last site entry, beginning on page 59)
_Table S1. Radiocarbon ages for Abu Hureyra
_Table S2. Radiocarbon ages for Arlington Canyon
_Table S3. OSL ages for Barber Creek
_Table S4. Radiocarbon ages for Barber Creek
_Table S5. Age/Depth chronology for Barber Creek site generated by MC Age Depth
_Table S6. Radiocarbon ages for Big Eddy
_Table S7. Age/Depth cal BP chronology for Big Eddy site generated by MC Age 1 Department of Anthropology,  Southern Methodist University, Dallas, TX 75275 2 School of Anthropology and Departmen t Geosciences, University  of Arizona, Tucson, AZ 85721 3 SWCA Environmental Consultants, Inc., Salt Lake City, UT 84111 4 School of  Anthropology, University of Arizona, Tucson, AZ 85721 SI: p.2
_Table S8. OSL ages for Blackville
_Table S9. Radiocarbon ages for Blackwater Draw
_Table S10. Radiocarbon and OSL ages for Kimbel Bay
_Table S11. Radiocarbon ages for Lake Cuitzeo
_Table S12. Radiocarbon ages for Murray Springs
_Table S13. Radiocarbon ages for Sheriden Cave
_Table S14. Radiocarbon ages for Talega site
_Table S15. OSL ages from Area D of the Topper site
_Table S16. Radiocarbon ages for Wally's Beach
_Table S17. Radiometric data used in attempts to replicate age/depth models of YDIH proponents
_Table S18. Regression coefficients from age/depth model replication analyses of Group 3a sites
_Table S19. r 2 values for age/depth model regressions from replication analyses and p values for the regression  coefficients given in Table S18 for Group 3a site s List of figures (all figures are included in the site specific  discussions)
_Figure S1. Graphs of replicated regression-based age/depth models for Abu Hureyra
_Figure S2. Graphs of replicated regression-based age/depth models for Arlington Canyon
_Figure S3. MC Age Depth plot Arlington Canyon
_Figure S4. Graphs of replicated regression-based age/depth models for Barber Creek
_Figure S5. Graphs of replicated regression-base d age/depth models for Big Eddy
_Figure S6. MC Age Depth plot Big Eddy ALL dates
_Figure S7. MC Age Depth plot Big Eddy NO outliers and no SD >100
_Figure S8. Graphs of replicated regression-based age/depth models for Blackville.
_Figure S9. Graphs of replicated regression-based age/depth models for Blackwater Draw.
_Figure S10. Graphs o f replicated regression-based age/depth models for Kimble Bay.
_Figure S11. Graphs of replicated regression-based age/depth models for Lake Cuitzeo.
_Figure S12. Graphs of replicated regression-based age/depth models for Melrose.
_Figure S13. Graphs of replicated regression-based age/depth models for Murray Springs.
_Figure S14. Graphs of replicated regression-based age/depth models for Talega.
_Figure S15. Graphs of replicated regression-based age/depth models for Topper.
_INTRODUCTION AND NOTES
_Site-by-site details are provided in this document.
_We do not discuss individual site contexts or finds in detail, save as relevant to the analysis of the chronology.
_Sites are listed in alphabetical order and identified by group, as per Figure 1 of the main text.
_Tables S1 to S16 provide the radiocarbon and OSL data for the individual sites, both the ages used and relevant ages  omitted from analyses by YDIH proponents.
_All radiocarbon dates are listed as 14C years before present.
_Tables S17 to S19 provide the data we used in replicating the original regression SI: p.3 based age/depth models of  YDIH proponents (Table S17), the regression coefficients from our replicated age/depth models as given by GraphPad  Prism (Table S18), and r 2 values for the regression models and p values for the regression coefficients as given by  GraphPad Prism (Table S19).
_Figures S1-S2, S4-S5, and S8-S15 are regression graphs for the sites in Groups 3a and 3b, based on the dates in Table  S17.
_These were produced in GraphPad Prism; regression model equations are provided in the discussion as necessary.
_In the age/depth model graphs, regressions of depth on age are shown on the left, and regressions of age on depth are  shown on the right; error bars represent 1 SD dating error terms, dotted lines represent unweighted regressions, and  solid lines represent weighted regressions (which are only possible for the regressions of age on depth shown on the  right because the variable whose standard deviation is used f or weighting should be the Dependent Variable [DV]).
_Figures S3, S6-S7 provide the MC Age Depth plots for the Arlington Canyon and Big Eddy sites.
_See the main text for discussion of the methods used to generate these plots.
_1. ABU HUREYRA GROUP 3A
_Abu Hureyra (Syria) is a large and complex Late Pleistocene/Early Holocene archaeological tell located in the  Euphrates Valley, which has yielded a rich record of early plant and animal domestication along with evidence of the  transition from foraging to farming and early village life in the Middle East.
_Occupied virtually continuously for more than 4000 years, the site was drowned behind a dam in 1974 and is  currently beneath >25 m of water (1, 2).
_Analysis for impact indicators was conducted by Bunch et al.(1) using archived sediment samples from arbitrarily  numbered excavation levels in Trench E at the site, in which there were multiple pit houses, structures and  discontinuous zones of archaeological occupation (ref.1, its SI Figure S3).
_In order to develop a chronology for the supposed YDB layer, Bunch et al.(1) used "Linear interpolation... to develop  an age-depth model based on 13 accelerator mass spectrometry (AMS) radiocarbon dates from (2)."
_All of the ages come from Abu Hureyra 1 Trench E, which has three phases of occupation that span the terminal  Pleistocene from
____11,500-10,000 14C yrs BP (2).
_The primary source on the radiocarbon ages from the site (2) lists 38 radiocarbon ages for Abu Hureya 1 but used only  16 of SI: p.4 those to ascertain the ages of the three Abu Hureyra 1 subphases as discussed in Moore et al. (2; the full  list of radiocarbon ages from the site is in ref.2, its Appendix 1).
_Only 15 of those ages are listed here (Table S1) (BM-1121 is not included as it is a composite sample from multiple  levels).
_Also listed here is an additional radiocarbon age provided in Bunch et al. (ref.1, its Table S2).
_Table S1. Radiocarbon ages for Abu Hureyra from Moore et al. (ref.2, its Table A.1).
_Dates are in stratigraphic order, as per the Harris Matrix in Moore et al. (ref.2, its Figure 5.28).
_Ages omitted without explanation by Bunch et al. (1) and Wittke et al. (3) are highlighted in gray.
_Radiocarbon ages on charcoal obtained from multiple levels are not included here.
_Lab no. Level 1 Depth (masl) 14C ± Comments
OxA-170 405 285.33 ____10600 200 Carbonized grain
OxA-8718 418 285.13 ____11140 100 Carbonized grain; omitted
OxA-407 419 285.13 ____10050 180 Charred bone
OxA-8719 419 285.13 ____10610 100 Charred bone; omitted
OxA-386 420 285.12 ____10800 160 Carbonized grain
OxA-473 425 284.95 ____10000 170 Charred bone
OxA-397 430 284.91 ____10420 140 Carbonized grain
OxA-434 430 284.91 ____10490 150 Charred bone
UCIAMS-105429 445 284.70 ____11070 40 Charcoal; Bunch et al. identify this as YDB layer (ref.2, its SI Table 2)
BM-1718R 447 284.67 ____11140 140 Charcoal; among "dates closest to the YDB layer" Bunch et al. (ref.2, its SI  Table 2)
OxA-171 457 284.72 ____10600 200 Carbonized grain
OxA-6685 455 284.72 ____10930 120 Carbonized grain; omitted
OxA-430 460 284.56 ____11020 150 Charred bone; among "dates closest to the YDB layer" Bunch et al. (ref.2, its SI  Table 2)
OxA-172 470 284.29 ____10900 200 Carbonized grain; among "dates closest to the YDB layer" Bunch et al. (ref.2,  its SI Table 2); Wittke et al. (ref.3 its SI Table S.1) identify this as YDB layer
OxA-468 470 284.29 ____11090 150 Charred bone; among "dates closest to the YDB layer" Bunch et al. (ref.2, its SI  Table 2)
OxA-883 470 284.29 ____11450 300 Carbonized grain from same layer as OxA-172 and OxA-468, but Bunch et al.  (ref.2, its SI Table 2) do not identify as among "dates closest to the YDB layer" 1
_Absolute depth of Levels 418-421 is the same; absolute depth of Levels 450, 455 and 457 is the same as per Moore et  al. (ref.2, its Figure 5.10) SI: p.5
_For reasons unspecified, 3 of these 16 ages are excluded from the regression analyses of Bunch et al.(1) and Wittke et  al.(3), although both claim to have "adopted the chronology of Moore et al. 2000" (ref.2, its SI p.2; ref.3, its SI p.2).
_The sample of radiocarbon ages they selected includes dates from all three phases of the occupation.
_There is a discrepancy as to the position of the supposed YDB layer.
_Bunch et al.(1) identify Layer 445 (depth 284.7 masl) in Phase 1 deposits as containing impact indicators and  therefore "consistent with the YDB layer," with a radiocarbon age of
____11,070 ± 40 14C yrs BP (UCIAMS-105429).
_They identify four other radiocarbon ages as "dates closest to the YDB layer," though it appears this is based on the  ages themselves and not the stratigraphic position of the sample relative to that layer (1).
_In contrast, Wittke et al. (ref.3, its SI Table S1) identify an age from Layer 470 (depth 284.29 masl) in overlying Phase  2 as the YDB layer, associated with a radiocarbon age of
____10,900 ± 200 14C yrs BP (OxA-172).
_There are two additional and older radiocarbon ages available from Layer 470 (Table S 1), yet neither of these is  identified in (3) as bearing on the YDB layer.
_The Abu Hureyra radiocarbon samples were collected from non-contiguous archaeological levels in Trench E "from  various locations across 1.66 m of sediment ranging from about 284.24 to 285.90 m asl" (ref.1, its SI p.2).
_The relative position of the levels is provided in a Har ris Matrix illustration (ref.2, its Fig 5.10, Fig.5.28), which in  turn is linked in an unspecified manner to absolute elevation (in masl).
_Our attempt to replicate the "linear interpo lation" age/depth model discussed by Bunch et al.(1) and Wittke et al.(3)  used the same 13 dates employed by those authors and used linear regression models (DV = BO + B1*IV).
_Graphs of these models are shown in Figure S1.
_Our result for the predicted depth of YDB-age deposits at this site is 284.63 masl, which varies somewhat from each  of the 284.29 masl and 284.7 masl values, discussed above, identified as the supposed YDB layer by the different  authors.
_A weighted regression analysis is statistically significant (see Table S19) and produces a predicted age of
____13,044 cal BP for the 284.29 masl layer and a predicted age of
____12,763 cal BP for the 284.7 masl layer; each of these ages falls slightly outside of the IntCal04
____12,900 ± 100 YDB interval, one in either direction.
_SI: p.6 Figure S1.
_Graphs of replicated regression-based age/depth models for Abu Hureyra.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_It can be seen in the graph of the weighted regression (Figure S1 right, solid line) that the influence of two relatively  young dates with relatively large standard deviations is reduced in com parison to the unweighted age/depth model,  leading to a lower regression slope (i.e. the weighted model leads to somewhat older ages at higher depths, though the  difference is small in the vicinity of the two reported YDB layer depths).
_2. ARLINGTON CANYON GROUP 3A
_The Arlington Canyon locality is on Santa Rosa Island, one of the California Channel Islands.
_It is a 5 m thick geological section located ~1.2 km upstream from the Clovis-age Arlington Springs site which  yielded human skeletal remains of terminal Pleistocene age (ref.4, pp.2534-2535).
_The section was selected after preliminary work suggested that deposits spanning the Younger Dryas interval might  occur here (4).
_Of particular interest were two dark layers.
_The lower was a 44 cm thick "distinctive, organic carbon rich" mud that occurred at the base of the section above a  basal gravel, a layer that appeared "similar in character" to a black mat, and subsequently was found to yield apparent  evidence of burning and impact indicators of the sort that "occur widely in the YDB layer" (ref.4, pp.2536, 2538; ref.5,  pp.12624-12625).
_A second dark silt layer, 20 cm thick, was observed higher in the section, with a ~60 cm cobble layer between the two.
_It too was reported to be carbon rich and contain charcoal and apparent impact indicators.
_SI: p.7 Sixteen radiocarbon ages were obtained on the Arlington Canyon section from depths of ~1 to 5 m below  surface (Table S 2), of which six were rejected by Kennett et al.(4): five were deemed "slightly older," possibly as a  result of an "old wood effect," and a sixth was rejected as being "out of stratigraphic sequence" (ref.4, its Table 4).
_Based on the remaining radiocarbon ages, they concluded that the supposed YDB layers (they lumped the two lower  carbon rich layers) "accumulated rapidly at
____~13.0 – 12.9 ka" (ref.4, p.2538).
_Table S2. Radiocarbon ages for Arlington Canyon as originally reported by Kennett et al.(4, and not Kennett et al. 5,  as erroneously noted in Wittke et al. ref.3, its SI Table S1).
_Radiocarbon ages rejected by Kennett et al.(4) as subject to 'Old wood' effect included here if included in the analysis  by Wittke et al.(3).
_Ages omitted without explanation by Wittke et al.(3) are highlighted in gray.
_Lab no. Depth (cmbs) 14C ± Comments
UCIAMS-47235 95-99 ____11040 30 Charcoal; omitted
UCIAMS-47236 179-183 ____12095 40 Charcoal; rejected by Kennett et al.2008 (4)
UCIAMS-47237 215-217 ____10895 35 Charcoal; omitted
UCIAMS-47238 267-270 ____11105 30 Charcoal; omitted
UCIAMS-47239 392-396 ____11105 30 Charcoal; sample depth at 394 in Wittke et al.(3)
UCIAMS-42816 403-406 ____11095 25 Wood; sample depth at 404.5 in Wittke et al.(3)
UCIAMS-36308 464-469 ____11095 25 Wood; sample depth at 461.5 in Wittke et al.(3)
UCIAMS-36307 469-475 ____11070 25 Wood; sample depth at 471 in Wittke et al.(3)
UCIAMS-36959 480-485 ____11075 30 Charcoal; sample depth at 487 in Wittke et al.(3)
UCIAMS-36960 480-485 ____11185 30 Glassy Carbon; rejected by Kennett et al.(4), used by Wittke et al.(3) who put  sample depth at 487 UCIAMS-36961 480-485 ____11440 90 Carbon Sphere; rejected by Kennett et al.(4), used by  Wittke et al.(3) who put sample depth at 487 UCIAMS-36962 480-485 ____11110 35 Carbon Elongate; rejected by  Kennett et al.(4), used by Wittke et al.(3) who put sample depth at 487 BETA-161032 480-485 ____10860 70 Wood;  omitted
UCIAMS-36306 485-491 ____11375 25 Wood; rejected by Kennett et al.(4), used by Wittke et al.(3) who put sample  depth at 495 UCIAMS-36305 493-498 ____11235 25 Wood; rejected by Kennett et al.(4), used by Wittke et al.(3)  who put sample depth at 495.5 UCIAMS-36304 498-503 ____11020 25 Wood; YDB Layer according to Wittke et  al.(3) who put sample depth at 500.5
_In their analysis and discussion of Arlington Canyon, Wittke et al.(3) "adopted the chronology of Kennett et al.(4) who  obtained accelerator mass spectrometry (AMS) 14C dates on charcoal, wood, carbon spherules, and glassy carbon,"  which yielded "12 [sic] accelerator mass spectrometry (AMS) 14C SI: p.8 dates."
_Eleven of the Arlington Canyon ages are provided in Wittke et al. (ref.3, its SI Table S1), on the basis of which,  evidently, they concluded that the supposed YDB layer "dates close to
____12.8 ka by linear interpolation" (ref.3, its SI p.2).
_Although Wittke et al.(3) also explicitly state that the YDB encompasses both dark layers, they highlight the  radiocarbon sample obtained from a depth of 500.5 cm below surface (UCIAMS-36304) – the deepest in the section,  the youngest of the eleven ages they tally, and perhaps not coincidentally, the one closest to
____12,800 ± 150 cal BP – as the age of the supposed YDB layer (ref.3, its SI Table S1).
_That conclusion cannot be supported.
_When the radiocarbon ages from Arlington Canyon utilized by Wittke et al.(3) are reanalyzed, the predicted depth of  YDB-age sediments falls at a depth of 441.1 cm below surface, not at 500.5 cm below surface, the depth that Wittke et  al.(3) identify as being associated with a YDB-age radiocarbon date.
_Moreover, a weighted linear regression analysis designed to estimate the age of the 500.5 cmbs layer places the age of  this layer at
____13,108 cal BP, over a century and a half prior to the earliest calibrated age range for the YD onset (using the  IntCal09 dates for the YDB interval).
_Of even greater concern is that the weighted regression is not statistically significant (p value of slope coefficient =  0.257) and, as can be seen in Figure S2, any relationship that does exist between age and depth for these dates is driven  almost entirely by the two dates from highest in the section, which are separated from the rest of the dates by a depth of  over 50 cm.
_Figure S2. Graphs of replicated regression-based age/depth models for Arlington Canyon.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_SI: p.9 We return below to this issue of the relationship, or lack thereof, between age and depth among the Arlington  Canyon dates.
_We also note that the center points of the dates that Wittke et al.(3) use for Arlington Canyon all predate
____12,900 cal BP; thus, predicting the depth of the
____12,800 cal BP layer requires extrapolation beyond these data on the age scale, not actually interpolation as they  state.
_There is, further, an unaccountable contradiction regarding the chronology of the supposed YDB layer.
_In response to criticisms by van H oesel et al.(6) – who, incidentally, said nothing of Arlington Canyon – Wittke et al. (7) insist "The radiocarbon dates from Arlington Canyon were never used to date the YDB " (ref.7, p.E3897, emphasis  ours).
_That is either a gross misstatement or their "linear interpolation" is based on some other unspecified chronological  data.
_It is difficult to imagine what those data might be.
_Their post hoc rationalization for not having used the Arlington Canyon radiocarbon ages is that they "are  systematically older," possibly the result of an "old wood effect."
_Here too they contradict their own claims.
_For while Kennett et al.(4) identified a possible old wood effect, they did so only in regard to five of the ages from the  site.
_Inexplicably, Wittke et al.(3) include those five ages in their tally of dates for the site, making no mention of possible  old wood concerns (ref.3, its SI Table S 1).
_As is apparent from the suite of radiocarbon ages from Arlington Canyon (Table S 2) – excluding those rejected by  Kennett et al.(4) – the "dates from upper and lower parts of the sequence are statistically similar suggesting rapid  accumulation of fluvial deposits shortly after
____~12.95 ± 0.05 ka" (ref.5, pp.12624-12625; also ref.3, its SI pp.2-3).
_That is certainly the case: although separated by 4 meters, the radiocarbon ages at the top
____ (11,040 ± 30 14C years BP at 95-99 cm below surface) and bottom
____ (11,020 ± 25 14C years BP at 498-503 cm below surface) of the section are statistically indistinguishable, as  shown by chi-square analysis (8).
_Indeed, all but two (UCIAMS-47237, BETA-161032) of the ages deemed acceptable by Kennett et al.(4) are  statistically part of the same population, and average to
____11,070 ± 10 14C years BP.
_This virtual uniformity in age across a 4 m vertical section is problematic in terms of resolving a precise age for the  supposed YDB layer.
_MC Age Depth analysis, which excludes only those ages rejected by Kennett et al.(4, a more reliable listing than in  ref.3), indicates that all layers from 92-97 cm below SI: p.10 surface, 203-263 cm below surface, and 472-500+ cm  below surface fall within the temporal span of
____12,800 ± 150 cal BP.
_When upper and lower 99% confidence intervals are included, the YD onset encompasses all layers from 196-350 cm  below surface, and 378-500+ cm below surface.
_The age/depth curve at Arlington Springs is essentially a flat-line in which terminal Pleistocene ages apply to much of  the section from about 200 to 500 cm below surface (Figure S3).
_Figure S3. MC Age-depth model for Arlington Canyon on all dates in Table S5 accepted by Kennett et al.(4), with  confidence interval derived from 1000 bootstrapped chronologies.
_Although both the lower and upper of the supposed YDB layers fall within the proper temporal range, it raises the  question of why apparent impact indicators were only found in those two layers, given that there are several meters of  sediment of the same age in that section.
_In order to better resolve the precise level and age of any supposed impact indicators at Arlington Springs, it will be  necessary to examine a more precisely dated section, and one without so much overlap in time.
_3. BARBER CREEK GROUP 3A
_Barber Creek is located in the Tar River basin on the North Carolina coastal plain (9, 10).
_The site has yielded Early Archaic through Early Woodland period cultural materials, found relatively well SI: p.11  stratified in a relict aeolian sand-sheet (with minor contributions from fluvial deposition) paralleling Barber Creek, a  tributary of the Tar River.
_Site investigators place the beginning of aeolian sand-sheet deposition at "just before or during the Younger Dryas  stadial event (ca.
____12,900-11,500 CALYBP)" (ref.10, p.26).
_Wittke et al.(3) report there is no black mat at the site, although at a depth of ~100 cm below surface "the sediments  abruptly change from alluvial to eolian deposition, producing a clear lithologic break and color change.
_The age-depth model indicates that this shift corresponds to the onset of Younger Dryas cooling" (ref.3, its SI p.4).
_They collected three 2.5-cm-thick sediment samples across the interval from 97.5 to 105 cm below surface, and  determined that the supposed YDB layer was at 100 cm below surface.
_Their age/depth model is reportedly based on the chronology of Moore and Daniel (10); however, Wittke et al.(3) do  not make full use of that chronology, omitting four additional OSL ages (Table S 3) as well as six radiocarbon ages  available from the site (Table S4; see ref.9, its Table 1; ref.10, its Table 1-1).
_Table S3. OSL ages for Barber Creek from Moore and Daniel (ref.10, its Tables 1-1 and 1-2).
_Ages omitted without explanation by Wittke et al.(3) are highlighted in gray.
_Lab no. Depth (cmbs) OSL ± Comments
FS2476 60 ____9740 590 Single aliquot – omitted by Wittke et al.(3)
FS2797 77 ____12800 710 Single aliquot – same sample as UW1963; Moore & Daniel reject mean age in favor of  minimum age model result ____ (10390 ± 620) for this sample; omitted by Wittke et al.(3)
UW1963 77 ____9100 700 Single grain – same sample as FS2797; omitted by Wittke et al.(3)
UW1907 80 ____9200 700 Single grain
UW1908 100 ____12100 700 Single grain – identified as the YDB layer
UW1909 140 ____14500 1000 Single grain
FS2511 315 ____16800 1900 Single aliquot – omitted by Wittke et al.(3)
_Their "Interpolation by second-order polynomial regression... [is] based on three optically stimulated luminescence  (OSL) dates," which they assert puts the age of the supposed YDB layer at
____ "12.1 ± 0.7 ka (range of 11.4 to 12.8 ka)" and thus is "determined to be close to 12.8 ka" (ref.3, its SI p.4).
_However, it seems apparent that the age they give for the supposed YDB layer is merely the OSL age of SI: p.12  sample UW1908, which comes from 100 cmbs.
_As such, it has a wide uncertainty (1400 calendar years), only a fraction of which overlaps with the target time span of
____12,800 ± 150 cal BP.
_That this estimate is not actually the result of interpolation by second-order polynomial regression is readily evident in  a visual inspection of their age/depth plot (ref.3, its SI Figure 2d); although presumably based on that interpolation, it is  obvious in this plot that YD onset-age sediments lie closer to a depth of ~110 cm, some 10 cm below their supposed  YDB layer.
_Table S4. Radiocarbon ages for Barber Creek from Daniel et al. (ref.9, its Table 1).
_Ages are reported from 10 cm excavation levels (measured below surface).
_Absolute depths are not provided; depths listed here are midpoint of the level.
_Ages omitted without explanation by Wittke et al.(3) are highlighted in gray.
_Lab no. Depth (cmbs) 14C ± Comments
Beta-188955 55 ____8950 40 wood charcoal, Level 6 – omitted by Wittke et al.(3)
Beta-166239 65 ____8440 50 wood charcoal, Level 7 – omitted by Wittke et al.(3)
Beta-150188 75 ____8940 70 wood charcoal, Level 8 – omitted by Wittke et al.(3)
Beta-166237 75 ____9280 60 wood charcoal, Level 8 – omitted by Wittke et al.(3)
Beta-166238 95 ____9860 60 wood charcoal, Level 10 – omitted by Wittke et al.(3)
Beta-188956 105 ____10500 50 wood charcoal, Level 11 – omitted by Wittke et al.(3)
_Our re-analysis using those same three OSL ages confirms this (employing a 2nd order polynomial equation in the  form of DV = BO + B1*IV + B2*IV 2), as the predicted depth of deposits that date to the onset of the Younger Dryas  is ~109.5 cm below surface (Figure S4).
_Figure S4. Graphs of replicated regression-based age/depth models for Barber Creek.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_SI: p.13 In estimating the age of the 100 cmbs depth Wittke et al.(3) identify as the YDB layer, both weighted and  unweighted regressions give identical and not especially interesting results due to the fact there are only three data  points involved: both return the
____12,100 cal BP age derived directly from the 100 cmbs layer as the age of this layer.
_Regression significance cannot be calculated in this re-analysis because the regression line is a perfect fit to the data,  as will generally be the case when a 2nd order polynomial equation is fit to only three points.
_As a check on the results of regression analyses described above, which as noted were based only on three of the OSL  dates from the site, we examined the available radiocarbon ages from Barber Creek (Table S4) using MC Age Depth  analysis.
_There is warrant to do so, for as Moore and Daniel state,"Both OSL age estimates from the upper meter at the Barber  Creek Site are in close agreement with the previously established radiocarbon dating sequence" (ref.10, p.24).
_The result s are shown in Table S5.
_Table S5. Age/Depth chronology for Barber Creek site generated by MC Age Depth.
_Ages calibrated with IntCal09.
_Levels shaded and in bold have median ages within span of
____12,800 ± 150.
_Supposed YDB layer is at 100 cm below surface (bold, outlined).
_Depth (cmbs) median cal age cal age – upper confidence interval cal age – lower confidence interval
98.00 ____11679 11802 11528
99.00 ____11769 11882 11607
100.00 ____11862 11963 11688
101.00 ____11956 12051 11770
102.00 ____12051 12146 11846
103.00 ____12146 12243 11923
104.00 ____12243 12340 12000
105.00 ____12339 12438 12078
106.00 ____12436 12537 12155
107.00 ____12532 12641 12232
108.00 ____12627 12744 12309
109.00 ____12722 12847 12386
110.00 ____12816 12948 12462
111.00 ____12909 13048 12538
112.00 ____13000 13147 12612
113.00 ____13089 13243 12686
114.00 ____13175 13338 12759
115.00 ____13260 13432 12830
116.00 ____13342 13523 12900
117.00 ____13421 13611 12969
_SI: p.14 These results reveals that layers from 109-111 cm below surface have median ages that fall within the span of
____12,800 ± 150 cal BP; if upper and lower confidence intervals are considered, that vertical range expands to all  layers from 108-116 cm below surface.
_In contrast, the supposed YDB layer at 100 cm is 1000 years younger, with a chronological range from
____11,963-11,688 cal BP.
_4. BIG EDDY GROUP 3B
_Big Eddy is a rich, multi-component archaeological site in southwest Missouri that has yielded a stratified sequence of  cultural materials that range from the Paleoindian through the Late Prehistoric periods (11-13).
_Testing for spherules was done by Wittke et al.(3) on five samples at depths of 311-318, 319-326, 327-335, 336-343  and 344-351 cm below surface.
_Only two of those samples yielded supposed impact markers: one from 319-326 cmbs (15 spherules/kg) and one from  327-335 cmbs (100 spherules / kg).
_Accordingly, Wittke place the "proxy-rich YDB layer from 327 to 335 cmbs" (ref.3, its SI p.5).
_To determine the age of that layer, Wittke et al.(3) "adopted the chronology of Lopinot et al. who acquired 30 AMS  radiocarbon dates based on charcoal from this sequence" from which an age-depth model was generated using  logarithmic interpolation that put its age "close to 12.8 ka" (ref.3, its SI p.5).
_Although a scatter plot of those ages with depth is provided, the full list of ages is not (ref.3, its SI Fig.3d; Table S1).
_Instead, only seven "key dates" are listed and, inexplicably, their depth data is omitted (sample depth data is provided  for virtually all other sites in that table, and that information is readily available for Bi g Eddy).
_The seven key dates are listed in order of increasing age (ref.3, its SI Table S1), though when depth data is included it  is apparent that ages are not in chronological order with depth.
_The 30 radiocarbon ages used in the Big Eddy scatter plot in Wittke et al. (ref.3, its SI Fig.3d) is provided in Table S6.
_There are several points to make in regard to the data.
_First, although there are three radiocarbon ages available from the proposed YDB layer they span ~1700 14C years  (thus, are not statistically part of the same population and cannot be averaged).
_Second, only one of those three ages
____ (10,710 ± 85 14C years BP) is within of
____12,800 ± 150 cal years BP, but only at 2 standard deviations.
_Finally, the layers above and below the supposed YDB layer all yielded YD onset ages (Table S6).
_SI: p.15 Table S6. Radiocarbon ages from Big Eddy, from Hajic et al.(11); Lopinot e t al.(12-13), as provided by  Lopinot.
_The solid lines enclose the five layers sampled for impact indicators.
_Lab no. Depth (cmbs) 14C ± Comments
AA-35462 283 ____9835 70 bark charcoal
AA-72611 285 ____9751 64 unspecified
AA-72609 286 ____9924 50 unspecified
AA-72610 294 ____10440 160 unspecified
AA-26653 298 ____10185 75 indeterminate charcoal
AA-75719 303 ____10506 53 unspecified
AA-27487 306 ____10400 75 indeterminate charcoal
AA-27480 308 ____10340 100 wood charcoal
AA-29022 313 ____10430 70 wood charcoal
AA-75720 315 ____10896 54 overlaps YD onset at 1 SD; material unspecified
AA-72607 317 ____9960 920 unspecified
AA-27488 321 ____10470 80 wood charcoal
AA-27485 322 ____11280 75 wood charcoal
BETA-230984 322 ____10940 60 overlaps YD onset at 1 SD; material unspecified
AA-72612 322 ____10959 54 overlaps YD onset at 1 SD; material unspecified
AA-27481 326 ____11160 75 overlaps YD onset at 1 SD; Bark or wood charcoal;
AA-25778 328 ____10260 85 Supposed YDB layer, does not overlap YD onset; wood charcoal
AA-27486 331 ____11900 80 Supposed YDB layer, does not overlap YD onset; bark or wood charcoal;
AA-26654 333 ____10710 85 Supposed YDB layer, overlaps YD onset at 2 SD; Indeterminate charcoal
AA-27482 338 ____11190 75 overlaps YD onset at 2 SD; Wood charcoal
AA-26655 347 ____10940 80 overlaps YD onset at 1 SD; indeterminate
AA-72608 347 ____12450 300 unspecified
AA-34586 358 ____12320 130 conifer wood charcoal
AA-34587 364 ____11930 110 Alder(?) wood charcoal
AA-72613 373 ____11960 270 unspecified
AA-34588 375 ____12250 100 Conifer wood charcoal
AA-34589 383 ____11375 80 Conifer wood charcoal
AA-27483 384 ____11910 440 indeterminate charcoal
AA-34590 386 ____12590 85 conifer wood charcoal
AA-27484 396 ____12700 180 indeterminate charcoal Wittke et al. suggest there is a problem with "Accurately dating  individual layers... because some charcoal fragments have clearly moved up and down within the sequence, and some  older charcoal may have been introduced by flood-induced redeposition from sources upstream" (ref.3, its SI p.5).
_But if that is so, why is the 327-335 cm layer considered to date to the YD onset as it has but a single date that  overlaps the YD onset (at 2SD), while layers above and below have multiple ages that overlap at 1 SD, SI: p.16 and in  the case of the layer above a tighter temporal range?
_Or is there relatively little age difference in the five layers sampled for spherules?
_If that is the case and all roughly overlap in time, then why was there a spike in spherules in only one of the layers?
_Or, is the dating of the entire section is problematic?
_Wittke et al. (ref.3, its SI p.5) put the age of the supposed YDB layer at "close to 12.8 ka," and as we found in our  replication of their statistical analyses, they appear to be correct in their general point.
_In our replication (Figure S5), we used the 30 dates in Table S6.
_Figure S5. Graphs of replicated regression-based age/depth models for Big Eddy.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_However, because those dates were appropriately reported with calibrated 2 SD ranges, rather than with symmetrical  calibrated ± SD error terms comparable to the ones re ported by Wittke et al. (ref.3, its SI Table S1), we recalibrated the  dates to obtain the calibrated ± SD terms necessary for weighted regression.
_Recalibration was done with OxCal version 4.2 (//c14.arch.ox.ac.uk/oxcal/OxCal.html), which can provide 1 SD error  terms for calibrated dates, and the IntCal04 calibration curve, following the original calibration used for this site.
_We used the median calibrated ages from IntCal04, along with the recalibrated 1 SD error terms that we obtained from  OxCal.
_Even though the calibrated 1 SD error terms that we obtained may not precisely correspond to those that generated the  plot in Wittke et al.(3), these error terms – the only ones available – should provide an accurate relative-scale measure  of the uncertainty associated with individual dates and thus be suitable for use in weighted regression.
_Our replication uses logarithmic regression, with the log of the depth variable used (i.e., Log 10 (DV) = BO + SI: p.17  B1*IV, when depth is the DV, and DV = BO + B1* Log 10 (IV), when depth is the IV) because this produces the best  match with the age/depth model shown in Wittke et al. (ref.3, its SI Fig.3).
_The regressions produced in our reanalysis are highly statistically significant and result in a predicted depth for YDB- age deposits of 330.2 cmbs, which falls squarely within the 327-335 cmbs zone identified by Wittke et al.(3) as the  YDB layer, as well as a weighted regression-predicted age of
____12,952 cal BP for this layer.
_Though this age differs from the 12.8 ka date that Wittke et al.(3) state the layer's age is close to, it falls within the  YDB interval for sites where IntCal04 was used
____ (12,900 ± 100 cal BP).
_The chronology generated by the MC Age Depth analysis (Table S7) reveals that all layers from 322.5-332.5 cm  below surface have median ages that fall within the span of
____12,900 ± 100 cal BP; if upper and lower confidence intervals are considered, that range expands to all layers from  318.5-342 cm below surface.
_The resulting age/depth curve also reveals there are multiple outliers (that is, ages outside the 99% confidence interval  of the age/depth spline) and, of course, ages with wide uncertainties (Figure S6).
_Figure S6. MC Age-depth model for Big Eddy based on all dates in Table S9, with 95% confidence interval derived  from 1000 bootstrapped chronologies.
_SI: p.18 These outliers contribute little to the resolution of the age/depth curve, and indeed potentially bias the result.
_Removing the problematic outliers (the 14C ages at depths of 321, 328, 331, 347
____ [12,450 ± 300] and 383 cm below surface), as well as the ages with a SD >100 years, as per the suggestion of  YDIH proponents (14) (14C ages at depths of 294, 317, 358, 364, 373, 384, 396 cm below surface), changes the  age/depth curve, most notably in revealing there was a sharp increase in the sedimentation rate tha t peaked
____~12,900 cal BP, but which began several centuries earlier and lasted several centuries later (Figure S7).
_That essentially flattened the age/depth curve on either side of the YD onset, indicating that all layers from depths of  320-348 cm below surface have median ages that fall within the span of
____12,900 ± 100 cal BP, with upper and lower confidence intervals within that window that encompass all layers  from >316 to 355 cm below surface (Table S7).
_Figure S7. MC Age-depth model for Big Eddy after the elimination of outliers in Figure S6 (ages outside the 95%  confidence interval) and ages with standard deviations >100 14C years (shown in gray).
_The 95% confidence interval derived from 1000 bootstrapped chronologies.
_SI: p.19 Table S7. Age/Depth cal BP chronology for Big Eddy site generated by MC Age Depth analysis, as calculated  after removal of outliers (ages outside 99% confidence intervals) and ages with SD > 100 radiocarbon years.
_Levels shaded and in bold have median ages within span of
____12,900 ± 100.
_Supposed YDB layer is at 327-335 cm below surface (outlined).
_Depth (cmbs) median cal age cal age – upper confidence interval cal age – lower confidence interval
316 ____12700 12798 12626
317 ____12731 12826 12656
318 ____12761 12852 12685
319 ____12788 12874 12712
320 ____12812 12894 12737
321 ____12834 12916 12760
322 ____12853 12938 12780
323 ____12868 12956 12797
324 ____12880 12970 12808
325 ____12889 12980 12811
326 ____12895 12987 12812
327 ____12899 12992 12810
328 ____12901 12994 12807
329 ____12901 12994 12803
330 ____12900 12993 12798
331 ____12899 12991 12793
332 ____12897 12991 12790
333 ____12896 12992 12787
334 ____12896 12994 12786
335 ____12896 13001 12787
336 ____12897 13008 12790
337 ____12899 13015 12794
338 ____12901 13023 12795
339 ____12904 13031 12787
340 ____12907 13040 12782
341 ____12911 13049 12778
342 ____12916 13058 12776
343 ____12923 13069 12776
344 ____12931 13081 12779
345 ____12941 13094 12784
346 ____12953 13108 12792
347 ____12968 13124 12803
348 ____12985 13144 12817
349 ____13004 13169 12834
350 ____13026 13196 12854
351 ____13050 13224 12877
352 ____13076 13253 12902
353 ____13105 13287 12930
354 ____13135 13322 12961
355 ____13168 13360 12992
_SI: p.20 In effect, then, although the supposed YDB layer at 327-335 cm below surface falls within the proper  temporal range, that is also true of all the layers sampled for impact indicators at the Big Eddy site.
_In order to better resolve the precise level and age of any supposed impact indicators at Big Eddy it will be necessary  to examine a continuous and directly dated sedimentary section, rather than obtain a sample of supposed indicators  from one location on the site which is then dated by a conglomeration of ages from several other locations on site, all  of which were deposited at more or less the same time.
_5. BLACKVILLE GROUP 3A
_The Blackville locality (South Carolina) is a site on a sand rim of one of the Carolina Bays.
_According to Bunch et al.(1), samples for analysis were cored by hand auger 5 from a section of aeolian and alluvial  sediments ~2 m thick, which lay unconformably atop an apparent Miocene marine clay (ref.1, p.E1905).
_A 15-cm thick peak in apparent impact indicators was reported to occur from 175-190 cm below surface; the  supposed YDB layer was placed at 183 cm below surface (ref.1, its SI p.4).
_It was not possible to obtain radiocarbon ages "because of sediment mixing by deep-rooted plants" (ref.1, its SI p.4).
_Instead, three OSL dates (Table S 8) were obtained at depths of 107, 152 and 183 cm below surface (ref.1, its SI Table  S2), and the age of the supposed YDB layer was subsequently based on "linear interpolation" of the upper and lower of  the OSL ages.
_The OSL age in between at a depth of 152 cm below surface
____ (18,540 ± 1680 years BP) was rejected because of the "large magnitude of [its] age reversal" which was argued  to be the result of older sediments lying stratigraphically higher than younger sediments (ref.1, its SI p.5).
_But given the problem of sediment mixing, there is no way to know which date or dates in the section should be  rejected.
_No explanation is given as to why the lowest OSL age at 183 cm below surface was not rejected as being the result of  younger sediments mixed into older strata.
_Presumably it was kept for reasons other than its coincidence with the presumed age of the supposed YDB layer.5
_Strictly speaking, hand augers do not bring up 'cores' per se, but rather 10-20 cm of sediment (the amount depending  on the size of the auger and the resistance of the sediment), which is invariably churned by the twisting and cutting  motion of the auger bit.
_Nor is it clear how OSL ages were retrieved from a bucket auger.
_SI: p.21 Table S8. OSL ages from Blackville, from Bunch et al. (ref.1, its Table S2).
_Lab no. Depth (cmbs) OSL ± Comments
LB862 107 ____11500 1030 LB861 152 18540 1680 Rejected by Bunch et al.(1)
LB859 183 ____12960 1190 YDB layer identified by Bunch et al.(1)
_Given that an OSL age was available directly on the supposed YDB layer, the necessity of a linear interpolation is  unclear.
_Nor is it obvious that the date for the supposed YDB layer is actually based on a linear interpolation: first, the  interpolated date on the supposed YDB layer is reported to be
____12.96 ka, coincidentally the same age as the OSL date itself from that depth (which was accompanied, of course,  by an 1190 year uncertainty).
_Second, when an attempt to replicate the linear interpolation is made (Figure S8), the predicted depth of YDB-age  deposits is 174.67 cm below surface, slightly above the depth of the supposed YDB layer.
_Further, the age of the 183 cmbs layer that is predicted by weighted regression is
____12,960 cal BP, and given that OSL ages are comparable to calibrated radiocarbon ages and that the calibrated age  of the YD onset is put at
____12,800 ± 150 cal BP, the modelled date for the Blackville supposed YDB layer predates the YD onset.
_Figure S8. Graphs of replicated regression-based age/depth models for Blackville.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_In deriving these re-analysis results, we exclude the same date that Bunch et al.(1) reject.
_We also note that, based as they are on only two data points, it is not possible to calculate significance for the SI: p.22  regressions that use the data of Bunch et al.(1).
_We observe, however, that the great deal of overlap in the error terms of the two unrejected dates (see Figure S8)  indicates that in fact there is not a strong relationship between age and depth even for these two dates.
_Regardless, given that the supposed YDB layer and t he OSL sampling was from a zone 15cm thick in a section known  to be subject to mixing, and given the wide uncertainty in the actual OSL age from this layer of
____12,960 ± 1190 – which could pre- or post-date the YD onset (with an OSL age range of
____14,150-11,770 years BP) – there is little precision in the dating of the supposed YDB layer at Blackville.
_6. BLACKWATER DRAW GROUP 3A
_Blackwater Draw (also known as the Clovis site, but more correctly, Blackwater Locality 1) is the type locality for the  Clovis artifact style.
_The site is situated in a large quarry pit created by long-term gravel mining that has destroyed most of its deposits.
_Much of the stratigraphic research and radiocarbon dating at the site has focused on the "North Bank" of the quarry  (on exposures mostly destroyed in the 1960s) (15-17) and along the margins of the "South Bank" (on exposures still  present) (18).
_These sections a re ~360m apart.
_Two key points about the stratigraphy, archaeology and geochronology at the site, apparently not realized by the YDIH  proponents, are that the lithostratigraphy is not isochronous across the site, and Clovis archaeological materials are only  dated in the North Bank area of the site.
_The key strata are Units B, C and D.
_Unit B is sand and gravel alluvium deposited by springs around the basin margin; Unit D is diatomite, diatomaceous  earth, and organic-rich mud deposited in lacustrine and palustrine environments.
_Unit C, however is a much more complex unit.
_It includes alluvium, slopewash, and mudflow deposits that interface with both Units B and D.
_The stratigraphic position of Clovis archaeological remains on the South Bank was unclear until Haynes re-examined  the 1936 work of Cotter (19; see the discussion in 16).
_Initially, Haynes (ref.18, p.370) was uncertain whether the South Bank Clovis material was in Unit B or C.
_His subsequent re-examination clearly showed that Clovis finds from 1936 were from the upper contact at the top of  Unit B (ref.16, its Fig.17).
_Unit B produced two radiocarbon ages
____ (11,810 ± 90 and ____11,380 ± 150 14C years BP), which are in reverse SI: p.23 stratigraphic order and  considered to be minimum ages (ref.18, p.367).
_On the North Bank all of the Clovis material is in Units B2 and B3.
_Between the two "banks" and ~175m north of the South Bank was an area excavated in 1956 and 1957 (the so-called  Jelinek excavations).
_Both mammoth and bison were found in what appears to be Unit C in this area (ref.18, p.370, 19).
_According to Firestone et al.(21-23), LeCompte et al.(24), and Wittke et al.(3), supposed impact markers are found on  the South Bank at the base of Unit D, just above or resting on Unit C.
_More specifically,"YDB markers are concentrated in a ~2-cm layer of fine-grained fluvial or lacustrine sediment that  lies at the base of the black mat in the uppermost stratigraphic horizon [Unit C] containing in situ mammal bones and  Clovis artifacts" (ref.22, its SI p.18). Wittke et al.
_2013 (ref.3, its SI p.6) likewise state "The thin contact between Units D and C represents the YDB layer."
_As noted, however, there is no Clovis material in Unit C just below Unit D on the South Bank.
_In their initial publication on the YDI H, Firestone et al.(21) claim that the South Bank contained a ledge "jammed  with spears, tools, and bone" 18 inches above the Clovis level (in upper Unit C) and the supposed YDB extinction zone  (contact of Units C and D).
_They further assert that "Eight radiocarbon dates indicated that no humans had visited Blackwater Draw for more than  1000 years" after the impact "event" (ref.21, p.73).
_Ignoring the fact that radiocarbon dating cannot indicate the presence or absence of humans, no such artifact-laden  ledge is reported from the South Bank or anywhere else in the site.
_More importantly, the eight radiocarbon dates are not identified.
_Kennett and West (25, p.E110) argue that "Folsom-age materials occur above the Clovis materials but a hiatus of  ~500 years is suggested."
_These statements indicate a lack of understanding of the site and are contradicted by ample evidence for essentially  continuous archaeological occupation from Clovis time into the Holocene (15-18).
_Likewise, the assertion in Wittke et al. (ref.3, its SI p.6) that "Haynes (1995) demonstrated that the Clovis-aged C  stratum and YD-aged D stratum occurs across ≈50% of the area" is flawed by the fact that Haynes' cited work was not  at the Clovis site proper but instead in the outlet channel between the site and Blackwater Draw proper, as Haynes  makes clear.
_SI: p.24 In their discussion of the South Bank, Firestone et al.(22) focus on three radiocarbon ages for the Late  Pleistocene strata (B, C and D) at the site (18).
_No rationale was provided to explain the selection of these three dates among the many available from this site on  those strata.
_We note that two of the three cited 14C ages, including a critical age of
____~12.98 ka, were collected and dated in the 1960s from the site's North Bank ~360m distant (26).
_The original radiocarbon age of that ~12.98 ka cal age, moreover, has a standard deviation of 500 years
____ (11,040 ± 500 14C years BP), and therefore provides little precision beyond placing it in the Late Pleistocene.
_Moreover, destruction by quarrying makes it impossible to precisely correlate the two distant areas of the site.
_Further, radiocarbon dates on the lower diatomite are available from just ~13m west of the sample section of  Firestone et al.(22) and Wittke et al.(3) but they make no mention of these
____ (10,740 ± 100 [AA-1362] and
____10,470 ± 580 [A-4701] 14C years BP).
_Although apparently aware that there are many radiocarbon ages available from the site, Wittke et al.(3) estimate the  age of the supposed YDB layer at their spherule sample section, the same section on the South Bank sampled by  Firestone et al.(21) and Surovell et al.(27), using just five radiocarbon ages on Units B, C, D and E at the site (Table S  9).
_For inexplicable reasons, they include only one age each for Units C and D – the units that incorporate their supposed  YDB layer.
_There are many more available from those strata, as shown in Table S9 – which is itself not intended to be  comprehensive (a fuller listing is available in 16-18, among other sources).
_Nonetheless, it highlights the availability of other ages from the strata at the site specific to the supposed YDB layer  (units E-B), and more importantly emphasizes the variation in the age of those units across the site.
_Such should suffice to show that the data provided by Firestone et al.(22) and Wittke et al.(3) are inadequate to  determine the age of their supposed YDB layer.
_By Wittke et al.'s own admission four of the five radiocarbon ages they utilize are from "~60 m east" of where they  sampled for spherules, and the fifth from "~175 m northeast" of their sample location (ref.3, its SI, pp.6-7).
_The four ages from ~60 m distant are from a stratigraphic section studied by Haynes (18); an additional 16  radiocarbon ages are available from that same study on the same stratigraphic units, yet those were ignored by Wittke et  al.(3), who provided no explanation or rationale for the selection of just the four dates they used.
_SI: p.25 Table S9. Radiocarbon ages from Blackwater Draw, from Wittke et al.(ref.3, its Table S.1), also Haynes (18)  and Haynes and Warnica ref.16, its Table 4).
_Ages omitted without explanation by Wittke et al.(3) are highlighted in gray.
_Note: this table does not include all ages from all those stratigraphic units or the site, only those deemed "most  reliable" for their stratigraphic position, as specified by Haynes and Warnica (16)
_Lab no. Level Depth (masl) 14C ± Comments
A-4703 E 3 1237.97 ____10000 910 South Bank, measured depth from Haynes 1995
A-4705 E 1 1237.86 ____9260 320 South Bank, measured depth from Haynes 1995
AA-87338 E ____9889 50 North Bank; omitted
AA-87337 E ____9820 110 North Bank; omitted
A-1372 E/D ____10250 200 South Bank; omitted
AA-86575 D 2a ____10281 58 North Bank; omitted
AA-87335 D 2a ____10376 50 North Bank; omitted
AA-2261 D 2 1237.76 ____9950 100 South Bank, measured depth from Haynes 1995
AA-1370 D 2z ____10260 230 South Bank; omitted
AA-1364 D 1g ____10210 110 South Bank; omitted
AA-1363 D 1g ____10160 120 South Bank; omitted
AA-1362 D 1e ____10740 100 South Bank; omitted
A-4701 D 1e ____10470 580 South Bank; omitted
AA-39843 D 1b ____10526 70 North Bank; omitted
AA-89168 C 1 ____10884 67 North Bank; omitted
SMU-1880 C ~1237.56 ____10780 110 ~175m north of the South Bank, arbitrary depth from Wittke et al. 2013  using unspecified method
AA-1360 C ____10580 100 South Bank; omitted
AA-30454 B 3 ____10914 72 North Bank; omitted
AA-87917 B 3 ____10933 56 North Bank; omitted
AA-2262 B 1b2 1237.48 ____11810 90 South Bank, measured depth from Haynes 1995
AA-1375 B 1a ____11380 150 South Bank; omitted Wittke et al.(3) also failed to note that Haynes (ref.18, p.367)  considered AA-2262 as possibly contaminated and a n age reversal (ref.18, its Fig.8E).
_The fifth radiocarbon date they used (SMU-1880), reported by Johnson and Holliday (20), came from a sample  recovered in a jacketed section of possible Unit C material recovered during excavations in 1956.
_This is an area of the site long ago separated from the South Bank by extensive quarrying, precluding any  microstratigraphic correlation (especially given SI: p.26 the Stratum C complexities).
_The absolute elevation of this sample cannot be estimated, let alone with the centimeter-scale precision give n in  Wittke et al. (ref.3, its SI Table S1).
_For that matter, they provide no explanation of the means by which they "integrated the three locations" – that is, their  sample section, and the two locations ~60 m and ~175 m distant from which the radiocarbon ages were obtained – on  to a common depth scale in order "to produce a generalized composite age-depth model" (ref.3, its SI p.7).
_Their depth profile is arbitrary both internally and especially with respect to their sampling section on the South Bank  (where the supposed YDB layer is "centered at an elevation of 1,238.32 meters above sea level (masl)" identified as the  D/C interface (ref.24, p.E2961).
_As a result, the "logarithmic interpolation" that is said to date their supposed YDB layer in Unit C to ~12.8 ka (ref.3,  its SI p.7) is based on a meaningless depth scale, and hence lacks statistical and stratigraphic merit.
_The logarithmic regression-based age/depth model of Wittke et al.(3) is indeed reproducible, which may help to  explain why these particular dates were chosen and/or assigned to the depths that they were.
_Using the same five dates as Wittke et al.(3) and those authors' depths for those dates, we estimate the depth of YDB -age deposits to be within 3 cm of their reported depth for the YDB layer, and a statistically significant weighted  regression provides an estimated age of
____12,866 cal BP for this depth, well within the YDB interval (Figure S9; the log of the age variable is used in our  regressions because this results in the best match with the age/depth model shown in ref.3, its SI Fig.4).
_Figure S9. Graphs of replicated regression-based age/depth models for Blackwater Draw.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_SI: p.27 The weighted regression for this site also illustrates the value of taking dating uncertainty into account in  age/depth modeling, which YDIH proponents do not do, as the one date in this dataset with a very large error term  exerts less influence on the weighted regression than on the unweighted regression (see Figure S9).
_Nonetheless, because of the errors in stratigraphic interpretation, the highly problematic and selective use of available  dates, and the arbitrariness of the depths used, there is little r eason to accept the results of the age/depth model of  Wittke et al.(3).
_7. BULL CREEK GROUP 3C
_Bull Creek, in the Oklahoma Panhandle, is one of the three sites first reported with high levels of nanodiamonds  (NDs) at the YDB (ref.28, its SI Fig.1b).
_The original source identifies an associated date of
____~13,000 ± 100 cal BP but provides no stratigraphic context; this appears to be the date of
____11,070 ± 60 14C years BP reported by Bement et al.(29).
_The date has little accuracy as a moment in time, however, because the sample was 9 cm thick collected from the  middle of the A horizon of a buried soil (29), and A-horizons represent the mixing of organic carbon through the  duration of soil formation.
_In a subsequent paper, based in part on re-sampling of the section (30), the supposed ND peak is from a sample 5cm  thick collected below the zone that yielded the radiocarbon date.
_The supposed ND zone, therefore, pre-dates the YDB.
_But then Bul l Creek also has and spike in Late Holocene to modern sediments as well (30), making it evident that  nanodiamonds are not restricted to the YD onset.
_8. CAROLINA BAYS GROUP 2A
_The Carolina Bays on the Atlantic Coastal Plain are "are a group of ~500,000 highly elliptical and often overlapping  depressions scattered throughout the Atlantic Coastal Plain from New Jersey to Alabama....
_All of the Bay rims examined were found to have, throughout their entire 1.5-5-m sandy rims, a typical assemblage of  YDB markers" (ref.23, pp.40-42).
_The age and origins of these surficial depressions have been debated and discussed for decades (31-32), including the  proposal they formed from impacts (33).
_Firestone et al.(21) present data from the Carolina Bays as supporting the YDIH, arguing that they date to ~12.9k cal  years: "All of the evidence fits our theory that the rims and bays formed all at the same SI: p.28 instant" around
____12,900 cal BP; they also attribute to Ivester and colleagues ages on two bay rims of
____11,300 and ____12,630 years ago using OSL (ref.21, p.127).
_Firestone et al. also report an OSL date of
____11,400 ± 6100 years BP as close to the age of Clovis and therefore use it to date the YDB (ref.21, p.127).
_Yet, OSL ages are calendar years, and ages of
____~11,300 and ____~12,630 are too young for Clovis and by themselves postdate the onset of the Younger Dryas.
_To be sure, the OSL age of
____11,400 ± 6100 years BP overlaps the onset of the Younger Dryas, but then it also includes much of the first half  of the Holocene.
_Given this span of time, it cannot usefully determine the age of supposed impacts indicators at these sites.
_Moreover, the OSL ages on bay rims reported by Ivester et al.(34) do not pertain to the initial formation of any bay.
_Extensive OSL dating has shown that these bays formed at different times in the past, though generally between
____140,000-120.000,
____80.000-60,000 and
____50,000-12,000 years ago (35).
_Firestone et al.(22) subsequently realized the ages of the Carolina Bays vary, and sought to assert that the supposed  YDB layer found in 15 of the bays dated to
____12,900 years BP.
_They based this assertion on the fact that the markers found therein were identical to those found elsewhere dated to
____12,900 years BP (ref.22, p.16019).
_That circular argument cannot be used as chronological evidence, as it assumes what it ought to demonstrate.
_9. CHOBOT GROUP 1A
_The Chobot site is in lake sediments on the Canadian prairie near Edmonton.
_Little is reported for the site.
_Both Firestone et al. (ref.22, its SI Section C) and Wittk e et al. (ref.3, its SI p.7 and its SI Fig.5) report an increase in  "impact spherules" immediately above a Clovis occupation zone.
_Firestone et al. (ref.22, its SI Section C) state that "the Clovis level is capped by the YDB layer" (also ref.23, p.38).
_Wittke et al. (ref.3, its SI Fig.5) likewise refer to "observed Clovis artifacts located at the base of the black layer," but  admit they "were unable to date the site radiometrically because of bioturbation by plant roots.
_However, the stratigraphic position of the spherule layer is immediately above the uppermost level containing  abundant Clovis points and artifacts" (ref.3, its SI p.7).
_Yet, neither publication presents or cites data based on archaeological excavations.
_Ives and Froese (36), archaeologists familiar with the site, SI: p.29 report that Clovis artifacts from the site are neither  abundant (only three are known) nor found in place (the three points were recovered from the surface).
_Wittke et al.(37) in response agreed "with Ives and Froese that the Chobot site is challenging because it is undated.
_We also agree that some lithics at the site are non-Clovis, but Chobot has three acknowledged Clovis points, which  are more than at many Clovis sites."
_Of course, the relevant issue is not the absolute number of Clovis points, but their numbers relative to the entire lithic  assemblage at the site, which is overwhelmingly of much younger age, and the irrefutable fact that no Clovis points  have been found in situ or in a manner that can be related to the site stratigraphy.
_Wittke et al. nonetheless assert that while "The evidence from Chobot may seem unpersuasive as a single site... [it] is  highly consistent with the multicontinental YDB record.
_Similar coeval marker peaks occur at ∼ 30 dated YDB sites in 10 countries on four continents.
_Thus, the best explanation is that Chobot contains the YDB layer where indicated" (37).6
_The logic here is flawed: each of these localities must be dated independently of one another, and independently of a  proposed layer of supposed impact markers, else (as with the Carolina Bays) the reasoning is circular.
_In addition, the assertion that Clovis points were found in place or in great abundance or can be used to date the age of  the supposed YDB layer at the site is incorrect.
_Although Wittke et al.(3) do not report radiometric ages from Chobot, such are in fact available from this site:  Firestone (39) reports dates of 3600 and 1520 14C years BP on material from the site, though asserted these ages were  invalid.
_As Ives and Froese (36) observe, those dates stratigraphically bracket the supposed YDB Layer, although they are  more than 9000 years younger than the onset of the Younger Dryas.
_10. DAISY CAVE GROUP 3D
_Daisy Cave is a multicomponent archaeological site on the northeast coast of San Miguel Island off the coast of  southern California (40).
_Firestone et al.(22) report the occurrence of impact markers from Stratum I, described as a "dark brown cave soil,"  from which a radiocarbon age of
____11,180 ± 130 14C years 6
_In that same response Wittke et al.(37) assert there is a nonalgal black mat present at the Folsom (NM) site.
_That is neither correct (38), nor relevant.
_SI: p.30 BP
____ (13,219-12,913 cal BP at 1 sigma [IntCal09]) was obtained on a small carbonized twig fragment by Erlandson et  al.(41).
_Firestone et al. (ref.23, pp.38, 40) noted that "several markers were found, but others, including Ir[idium] were not  found, possibly because the protected cave shelter prevented accretion."
_Although Firestone et al. (ref.22, its SI Figs. 16 and 17) indicate additional samples other than the one from Stratum I  were obtained, no in formation is provided on the stratum (strata) in which those occur, or their ages.
_11. GAINEY GROUP 2A
_The Gainey site is the type locality for the Gainey point, considered related to the Clovis style.
_The site has no black mat, no reliable radiocarbon dates, nor a distinct Paleoindian level.
_As its investigators observe, Gainey is "a plowed site with few undisturbed deposits" (ref.42, p.267).
_The Paleoindian artifacts are on or just below the surface.
_Firestone et al. (ref.22, its Table B-2) report a thermoluminescence (TL) date of
____12,400 ± 1000 for the locality, but the cited source (42) presents no such date, but instead a radiocarbon age of  2880 ± 175 14C years BP for Feature 7 at the site (ref.42 p.266).
_Wittke et al. (ref.3, its SI p.9), likewise erroneously citing Simons et al.(42), report a single OSL (sic) age from the  site of
____12,360 ± 1,240 years BP obtained at ~30 cm depth, and imply it is the "only" age available for the YDB layer.
_Presumably this is the same date as reported by Firestone et al.(22).
_There is, however, a second TL date of
____11,420 ± 400 years BP from the site (ref.43, p.110).
_Both of these TL dates are on burned chert artifacts, and their position is unspecified, save for a published hand-drawn  diagram of Feature 37, from which the younger of the two TL dates was obtained (ref.43, its Fig.4).
_Hence, the position of these dated artifacts in what Wittke et al. admit is an "extensively bioturbated" deposit (ref.3, its  SI p.9) is not known.
_Further, that these are TL as opposed to OSL dates makes the ages problematic.
_They were never formally published with the lab data necessary for independent evaluation and they were determined  some time before ~1990 and thus did not employ modern methodologies.
_At that time, the viability of TL dating of burned chert had not been demonstrated.
_Regardless, like OSL ages, TL ages are read on the SI: p.31 same temporal scale as calibrated radiocarbon ages, and  hence the Gainey TL dates are younger than the onset of the Younger Dryas.
_With the inclusion of their standard deviations they overlap with the onset of the Younger Dryas, but those standard  deviations are so large as to essentially lack any precision – certainly none sufficient to date an impact event.
_For the record, the Gainey site excavators consider the
____12,400 age "earlier than expected" for Gainey and believe the site was occupied closer to the younger end of the  standard error for that age (ref.44, p.28).
_Finally, one of the YDIH proponents supplied 16 carbon microspherules said to be from the YDB layer at Gainey to  Mark Boslough for AMS radiocarbon dating.
_One of the microspherules returned an age of 207 ± 87 14C years BP (ref.45, p.23).
_Firestone (39), aware there were several modern and "future" ages on carbon spherules from Gainey, sought to  explain these based on as-yet unverified cosmogenic processes for radiocarbon enrichment.
_A simpler and more obvious answer is that these supposed YDB impact indicators are instead Late Holocene particles  resulting from normal terrestrial processes or the result of contamination during sample preparation by YDIH  proponents.
_12. KANGERLUSSUAQ GROUP 2A
_In cooperation with PBS in late 2008, Kurbatov and others (46) conducted a pilot field and laboratory experiment to  determine if impact indicators were present in the Greenland ice sheet.
_Samples of sufficient size were not available from the deep ice cores of central Greenland (GISP2, GRIP and  NorthGRIP), and so they sampled at a locality on the margin of the ice sheet east of Kangerlussuaq, West Greenland,  ~1km inland from the ice margin.
_Due to deformation and shearing, the ice sheet here is thinner, the Younger Dryas age ice varies "unevenly in  thickness," and portions of the stratigraphic record are possibly not preserved (ref.46, pp.750-751).
_Kurbatov et al.(46) are appropriately circumspect about the precise ages of the sampled section, noting that "If the  stratigraphic interpretation we have adopted is correct, the continuous sequence of ice samples we collected represents  ~6000 years of ice, spanning the YD at a resolution of about every 50-100 years" (ref.46, p.751).
_The section they examined "displayed dust stratigraphy and isotopic values similar to deep ice-core records...  suggesting that the YD layer is in the correct stratigraphic position at that site," and was possibly "deposited after the  end of the last glacial SI: p.32 episode
____ (14.6 ka) and prior to the earliest Holocene (11.6 ka)" (ref.46, pp.751, 757).
_However, as they concluded, the results for both the chronology and the analytical procedures for detecting impact  indicators need to be significantly refined (ref.46, p.757).
_13. KIMBEL BAY GROUP 3A
_Kimbel Bay (North Carolina) is also a Carolina Bay, in this case one incised by a drainage that breached its rim.
_The site stratigraphy was viewed in an "exploratory trench" (of unspecified size and depth) and a hand auger was used  to "extract a 490-cm deep core [sic]" (ref.3, its SI p.10 [see the note above regarding augering]).
_The upper 440 cm of sediment is reported to be a mix of aeolian and colluvial sands with discontinuous silty clay  bands, unconformably underlain by a fine-grained marine clay of apparent Cretaceous age.
_Eight 15-cm thick samples were collected from the 76 cm between 297 and 373 cm below surface, but only the six  deepest were examined for impact indicators.
_Wittke et al. identified an apparent peak of impact indicators centered at 358 cm below surface (ref.3, its SI p.10).
_In order to determine the age of the supposed YDB layer,"discontinuous samples from different locations in the  trench" were collected over a 303-cm thick sediment sequence (ref.3, its SI p.10).
_It is not specified how far apart laterally the samples were located.
_From those samples, three OSL and three radiocarbon ages were obtained (Table S1 0).
_"Logarithmic interpolation" was used to develop an age-depth model for the supposed YDB layer using all OSL and  radiocarbon ages – despite the obvious chronological reversal of the OSL age at 436 cm below surface – and is said to  show that the supposed YDB layer dates to ~12.8 ka (ref.3, its SI p.10).
_Table S10. Radiocarbon and OSL ages from Kimbel Bay from Wittke et al.(ref.3, its Table S.1)
_Lab no. Depth (cmbs) 14C/OSL ± Comments
UCIAMS-52613 147 ____1195 15 charcoal
LB-863 375 ____25500 2720 OSL
LB-864 406 ____26080 2940 OSL
AW-SKB-6 436 ____21640 2630 OSL
UCIAMS-52622 445 ____27250 130 charcoal (lab number as in [3])
UCIAMS-52622 450 ____39690 710 charcoal (lab number as in [3])
_SI: p.33 It is difficult to account for that conclusion.
_For one, the supposed YDB layer is in the lower sand, just 10 cm above the youngest OSL age of
____25,500 years BP, with no indication of any sort of depositional or weathering hiatus, nor erosion.
_This obvious time gap is not considered in Wittke et al.'s (3) age estimate nor is this hiatus ever discussed or  otherwise considered.
_Although our re-analysis of the age / depth data puts the predicted depth of YDB-age sediments at 353.56 cm below  surface (Table 3 in main text), which is within the vertical span of the 15 cm level that yielded the purported impact  indicators, that re-analysis, like the original, was based on single point age estimates.
_Interpolating an age/depth curve using solely point estimates is highly problematic at Kimbel Bay, where statistical  uncertainties on the OSL ages range up to almost 3000 years and those with radiocarbon ages are an order of magnitude  lower.
_Accordingly, it is critical in developing an age/depth model for this site to account for the differential weight of these  uncertainties.
_Doing so yields results that diverge greatly from the conclusion reached by Wittke et al.(3) about the age of the  supposed YDB layer at Kimbel Bay, even using the same data and regression model type as they do (Figure S10; the log  of t he age variable is used in our regressions for this site because this results in the best match with the age/depth  model shown in ref.3, its SI Fig.8).
_A weighted regression model predicts the age of the 358 cmbs depth to be
____12,094 cal BP, over 500 years younger than the
____12,800 ± 150 interval.
_Figure S10. Graphs of replicated regression-based age/depth models for Kimble Bay.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_SI: p.34 The weighted regression is significant, giving reason to place statistical confidence in this age estimate, and as  can be seen in Figure S10, the weighted regression is far less influenced than the unweighted regression by the three  OSL dates with the very large dating error terms in the middle of the age sequence.
_We also note that while the use of a logarithmic model may be appropriate in this case, additional geoarchaeological  evidence is necessary to support its use.
_The Kimbel Bay age/depth models (Figure S10) are noteworthy for being highly curved.
_Such curves may well be appropriate for describing the relationship between age and depth at this site, if there is  independent reason to suspect variability in depositional rates across the profile such that the distances between ages  would be compressed at greater depths in a manner that leads to a logarithmic relationship.
_However, no such evidence for or discussion of variable depositional rates in the higher levels of this site (i.e. above  the sand-clay interface at 440 cmbs) is given by YDIH proponents.
_As such, the use of a logarithmic model in this case may simply amount to another instance of inductively connecting  dots.7
_Under the circumstances, it cannot be claimed that the supposed YDB layer at Kimbel Bay is remotely close in time to  the YD onset.
_14. LAKE CUITZEO GROUP 3A
_Lake Cuitzeo is a large lake in central Mexico from which a 27 m sediment core was extracted.
_Found within that core was a "10 cm thick carbon rich layer" occurring at a depth between 2.82 and 2.50 meters  below surface, which yielded supposed impact indicators.
_Sediment samples were examined for indicators at 5 cm intervals between 2.65 and 2.8 meters below surface, and at  10 cm intervals above and below that span between 2.2 and 3.6 meters below surface (ref.47, p.E739 and its Table 5).
_A total of 22 bulk sediment radiocarbon ages are reportedly available from the core (Table S1 1), although only 21 are  provided (ref.47, p.E739 and its SI Appendix, Table 1).
_Site investigators developed an age/depth model using a 5th order polynomial regression, but only after rejecting six  of the radiocarbon dates because they are "older than the [age/depth] interpolation predicts" (ref.47, p.E739).
_However, as Blaauw et al.(48) observed,"the rejected dates were in stratigraphic order" and, more 7
_This may also be the case for Big Eddy and Blackwater Locality 1, the other two sites where logarithmic age/depth  models have been used (3), and it is certainly the case for the sites where 2nd order polynomial models have been used,  as noted in the main text.
_SI: p.35 problematic, the six rejected dates provide the only chronological control from the portion of the core that  includes the supposed YDB layer (the ages came from 2.05 to 3.1 meters below surface).
_The rejection of those ages was subsequently justified by a spike in total organic carbon at 2.75 meters below surface,  said to indicate major contamination by "radiocarbon-dead or very old carbon" (ref.47, p.E739).
_Curiously, this apparent major contamination of very old or dead carbon occurs amidst a "major rise in charcoal" in  the core said to be the result of "a major episode in biomass burning" that marked the onset of the Younger Dryas  (ref.47, p.E741).
_Table S11. Radiocarbon ages from Lake Cuitzeo, from Israde-Alcantara et al. (ref.47, its SI Table 1; also three  additional ages reported by Wittke et al. ref.3, its Table S.1).
_Lab no. Depth (mbs) 14C ± Material
AA-9351 0.70 ____930 55 bulk carbon
AA-9352 0.85 ____1755 115 bulk carbon
AA-9353 1.35 ____6165 70 bulk carbon; used by Wittke et al.(3)
AA-9354 1.95 ____8830 215 bulk carbon; used by Wittke et al.(3)
WW-3361 2.05 ____14270 50 bulk carbon; rejected by Israde-Alcantara et al.(47)
T7-M31 2.25 ____17605 215 bulk carbon; rejected by Israde-Alcantara et al.(47)
WW-3362 2.45 ____21730 70 bulk carbon; rejected by Israde-Alcantara et al.(47)
OS-7133C 2.55 ____21600 100 bulk carbon; rejected by Israde-Alcantara et al.(47)
WW-3363 2.75 ____27360 130 bulk carbon; rejected by Israde-Alcantara et al.(47)
WW-3375 3.10 ____32940 190 bulk carbon; rejected by Israde-Alcantara et al.(47)
T11-M47 3.35 ____15500 130 bulk carbon; used by Wittke et al.(3)
WW-6422 3.65 ____23870 100 bulk carbon
WW-3576 3.75 ____28289 120 bulk carbon
WW-6423 3.80 ____29490 190 bulk carbon
WW-8454 4.00 ____22780 120 bulk carbon
WW-8455 4.40 ____21450 100 bulk carbon
AZ-120 4.70 ____26800 900 tephra
WW-8456 5.35 ____29890 280 bulk carbon
AA-9359 6.10 ____32565 2885 bulk carbon
WW-3364 6.65 ____28600 140 bulk carbon
AA-9770 9.10 ____42400 1000 bulk carbon
_Regardless, it is claimed that the 5th order polynomial regression "predicts that the 12.9-ka YD onset is at a depth of  approximately 2.9 to 2.7 m" below surface (ref.47, p.E739), although the YDB is also illustrated – without regard to  the discrepancy – as occurring from 2.82 to 2.5 m below surface (ref.47, its Fig.1).
_Regardless, the conclusion regarding its age is not supported by the regression equation SI: p.36 provided in Israde- Alcántara et al. (ref.47, p.E739 and its SI Table 1: wherein y = −5E−07 x 5 + 6E−05 x 4 − 0.0025 x 3 + 0.0366 x 2 −  0.010 8 x + 0.512).
_Solving for y with that equation if x is 12.9 k cal BP returns a solution that puts the supposed YDB layer at 2.58 m  below surface, 12 cm above the top of their reported YD onset zone.
_When a 5th order polynominal regression line is re-calculated using their same dates 8, the resulting regression  equation (y =-6E-20 x 5 + 8E-15 x 4-3E-10 x 3 + 6E-06 x 2-0.0205 x + 93.278 [equation is in units of cm]) puts the  predicted depth of sediments dating to
____12,900 cal BP at 2.59 m below surface (Figure S11), also above the depth of the onset of the YD as stated by  Israde-Alcántara et al.(47).
_Further, a weighted regression model (again using the same dates, as well as a 5th order polynomial model) predicts  that the age of sediments at 2.82 m below surface — the depth that Israde-Alcántara et al. (ref.47, p.E739) identify as  marking the onset of the layer with impact indicators — is
____15,916 cal BP, approximately 3,000 years too early.
_Because this regression is not significant (just a s our replication of the depth-prediction regression presented by  Israde-Alcántara et al.[47] is not significant), limited stock should be placed in this result.
_Figure S11. Graphs of replicated regression-based age/depth models for Lake Cuitzeo.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_We observe, however, that our result in roughly consonant with the one independently derived by Blaauw et al.(48).
_Using an alternative method for age/depth modeling that also takes dating uncertainty 8 Specifically, the 15 unrejected  dates shown in Israde-Alcántara et al. (ref.47, its SI Table 1).
_Wittk e et al. (ref.3, its SI Table 1) show only three of these dates, and they give a calibration result for one (T11-M47)  that differs from that given by Israde-Alcántara et al. (ref.47, its SI Table 1).
_SI: p.37 into account, and again excluding the six dates rejected by Israde-Alcántara et al. (ref.47, p.E739), those  authors propose an age range of
____~21,200-16,000 cal BP for the core depth of 2.8 meters below surface (48).
_Of course, any attempt to determine a precise age for a supposed YDB layer in a portion of the core for which all dates  have been a priori rejected is at best problematic.
_We agree with Blaauw et al. that the layer investigated by Israde-Alcántara et al. "is not demonstrably or securely dated  to the start of the YD, and indeed... is most likely several millennia older" (48).
_15. LAKE HIND GROUP 3C
_Firestone et al.(22-23) attribute to Boyd et al.(49) the finding that "at
____~12.76 ka, the ice dams on the lake failed catastrophically as part of a regional pattern of glacial lake drainages.
_At the YDB, the failure rapidly transformed the lake from deep to shallow water." However, Boyd et al. only state that  " low water levels were established in the southern glacial Lake Hind basin by at least
____10,400 yr B.P." (ref.49, p.601); they had no earlier ages to provide more precise chronological control.
_Firestone et al. (ref.22, its SI Table 2) provide a radiocarbon age of
____10,610 ± 25 14C years BP which they calibrate as
____12,755 ± 87 cal BP.
_Although presumably calibrated with IntCal04, that calibration cannot be precisely replicated (IntCal04 returns an age  of
____12,757-12,661 cal BP at 1 SD).
_Moreover, even at two standard deviations – calibrated as ranges of
____12,791-12,617 (96% area under the curve) and
____12,442-12,416 (4% area under the curve) – their YDB layer falls outside the presumed temporal window of the  YD impact.
_It is perhaps noteworthy that Boyd et al. (ref.49, p.602), who are cited by Firestone et al.(22) on the chronology of  draining of Lake Hind, observe that Folsom groups were in the region soon after the lake drained, indicating that  despite claims to the contrary (22), there was no apparent hiatus in post-Clovis occupations (see also 26).
_16. LINGEN GROUP 2A
_Although "abundant charcoal" at Lingen is reported and illustrated from the supposed YDB layer (ref.3, its SI Fig.9),  evidently none of it was submitted for radiocarbon dating.
_Instead, as noted in the text, there is only an age available from nine cm below the layer the yielded the supposed  impact markers, at a depth of 52.5 cm below surface (the age of
____11,310 ± 60 14C years BP, as noted in the main text).
_The SI: p.38 supposed impact markers were found "at a depth of 43.5-47.5 cmbs in a 3-cm-thick [sic] dark, charcoal- rich layer at the top of the Usselo sand," which according to Wittke et al. correlates with the onset of the YD dated at  ≈12.8 ka" (ref.3, its SI pp.10-11).
_According to Wittke et al. (ref.3, its SI Table S1) this age estimate is derived from an age-depth model, but none is  shown or otherwise mentioned.
_Instead, it is based on the "position of the YDB layer at the top of the Usselo layer," but as we discuss in the main text  that is from a single radiocarbon date at the Usselo type site (50), which cannot be generalized to an otherwise undated  specific layer at site in which that horizon occurs, as van Hoesel et al.(51) have shown.
_We note that Wittke et al. (ref.3, its SI p.11) reject the arguments of van Hoesel et al.(51) regarding the variable age of  the Usselo layer on the grounds that van Hosel and colleagues neglected to address calibration issues.
_As van Hoesel et al.(6) state in rejoinder, that "comment is simply not true."
_17. LOMMEL GROUP 3C
_As at Lingen, the age of the supposed YDB at Lommel (Belgium) is in part based on the age of the Usselo soil at the  type site.
_However, at Lommel "One AMS 14C date was acquired on charcoal extracted from the YDB layer at 48.5 cmbs,  providing a date of
____11.48 ± 0.10 14C ka BP
____ (13.39 ± 0.12 cal ka BP)" (ref.3, its SI p.12).
_Wittke et al. claim this date is "consistent with an age of ≈12.8 ka" (ref.3, its SI p.12).
_It is impossible to agree that this is so even on their terms since that age of
____13,400 cal BP – which we emphasize is directly on the supposed YDB layer – is well outside the chronological  range they themselves grant to the onset of the Younger Dryas.
_Although they add that "Van Geel et al.(50) acquired an AMS 14C date of
____10.95 ± 0.05 14C ka BP (12.86 ± 0.07 cal ka BP) from nearby at the same site," (ref.3, its SI p.12) that fact is  questionable and irrelevant.
_Van Geel et al.(50) worked at the type Usselo section, ~160 km distant from Lommel, and as earlier noted the Usselo  soil represents many hundreds of years of accumulation (52-53).
_The timing of deposition of the impact indicators at Lommel is known, and known not to date to the YD onset.
_18. MELROSE GROUP 3A
_The Melrose site consists of colluvium from the surface to 38 cm below surface, which is resting on glacial till that is  exposed in a shallow (<50 cm) trench (ref.1, its SI p.5).
_Five contiguous samples SI: p.39 were taken from 5-48 cm below surface, with apparent impact indicators found in  an 8 cm thick interval from 15-23 cm below surface; that marks the supposed YDB layer (ref.1, its SI p.5).
_If an 8-cm-thick zone represents the YDB, then the upper 15 cm represents the entire Holocene.
_Yet, no evidence of prolonged weathering is indicated for the surface layer.
_The stratigraphy therefore indicates that a significant part of the Holocene section is likely missing.
_Owing to "a dearth of datable charcoal... and because of sediment mixing by deep-rooted plants," it "was not possible  to acquire direct radiometric dating of the sedimentary profile" (ref.1, its SI p.6).
_Accordingly, Bunch et al.(1) acquired a single OSL sample at 28 cm below surface 9 which returned an age of
____16,400 ± 1600 years BP.
_To determine the age of the supposed YDB layer, they used "linear interpolation" to connect the OSL age and the  surface, which they assumed to be modern; this reportedly "dates the proxy-rich YDB layer at a depth of 21 cmbs to
____12.9 ± 1.6 ka" (ref.1, p.E1905 and its SI p.6).
_However, their interpolation seemingly involved little more than drawing a straight line between a single point at
____16,400 years ago and 0 years ago: moreover, although the surface of the site might be "modern," the sediment  comprising that surface might not be, hence using a value of 0 is arbitrary and meaningless.
_Replicating this "linear interpolation" of two data points through regression (Figure S12) results in a predicted depth  for YDB-age sediments of about 22 cmbs.
_Assuming (after ref.1, p.E1905 and its SI p.6) that 21 cmbs is indeed the depth to be used for the proposed YD B layer  (and not say, 19 cmbs, which is the midpoint of the 15-23 cmbs zone), this result is about 1 cm lower than that depth,  though still within 15-23 cm interval identified as containing impact indicators.
_More important, an unweighted regression of age on depth 10 predicts an age of
____12,300 cal BP for the 21 cmbs depth (Table 3 in main text), which falls well outside the "12.9 ± 1.6 ka" interval  that Bunch et al. ascribe to this depth (ref.1, p.E1905 and its SI p.6).9
_Wittke et al. (ref.3, SI Table 1) give a depth for the site's OSL date of 30 cm.10
_A weighted regression model is not possible for Melrose because one of the data points (the ground surface at 0) is  not a date with an error term.
_Regression significance cannot be calculated for this site because there are only two data points, unavoidably resulting  in a perfect model fit.
_SI: p.40 Figure S12. Graphs of replicated regression-based age/depth models for Melrose.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_Beyond the problems with the questionable age/depth interpolation presented for this site, Bunch et al. (ref.1, its SI  p.6) suggest that the humic colluvium is redeposited glacial till, in addition to being subjected to bioturbation.
_That plus the lack of precision in the sing le OSL date available from the site indicates that the age of the zone with  supposed YDB indicators is unknown, other than likely being post-glacial in age.
_19. MORLEY GROUP 1A
_The Morley site is a drumlin west of Calgary formed by the Cordilleran ice sh eet, the age of which is said to be  "constrained by the end of local deglaciation to
____~13 ka" (ref.22, its SI Table 2).
_Dating of the YDB appears to be based primarily on the supposed presence of impact indicators, reported to occur on  top of the drumlin.
_The impact indicators therefore must be younger than
____~13,000 cal BP, but there is no age control for this actual local it y.
_The estimate of
____~13.ka for the Morely drumlin is based on the fact that "The largest drumlin field near Ontario (5000 km 2)  contains 3,000 drumlins that date to shortly after
____13 ka, and the age of the Morley drumlin field appears to be similar" (ref.23, p.3838).
_However, no evidence is provided to show in what manner the drumlins are similar or necessarily the same age, which  of course must be shown given that the drumlins were deposited by two different continental ice masses – the  Cordilleran and Laurentide – and the Ontario drumlin field is ~2600 km southeast of Morley.
_SI: p.41
_20. MUM7B GROUP 2A
_The MUM7B site is a high elevation locality in the northwestern Venezuelan Andes that reportedly "dates to within  the range of the proposed [YDB] event" (ref.54, p.49).
_The site section has several radiocarbon dates, yet all predate the Younger Dryas onset (ref.54, its Table 1).
_The basis for t he claim that the site is within the temporal range of interest rests on the identification of a 3-cm thick  black mat found ~20 cm above the youngest dated stratum (at
____11,440 ± 100 14C years BP, or
____13,397-13,208 cal BP [IntCal09]).
_That black mat is not dated.
_At the time of the original investigation, the black mat was attributed to "an alpine grass fire or groundwater  phenomena" (55).
_It was subsequently offered as a candidate for inclusion among sites showing YD-age impact markers, an d al though  undated its occurrence above sediments dated by AMS to
____13,100 cal BP was said to place it "well within the YD window," and to suggest that the MUM7B 'black mat'  "sediments are coeval with 'black mat' sites in North America" (ref.54, p.53).
_That chronological reasoning is problematic, however, since not all black mats are the same age (ref.26, its Fig 3; 56).
_As we have previously noted many 'black mats' "have base dates younger than the onset of the YD, and others have  base dates older than the onset of the YD" (ref.26, p.582).
_In fact, Quade et al.(57) document their formation throughout the Holocene in the Great Basin (see also 58).
_21. MURRAY SPRINGS GROUP 3A
_Murray Springs is a well-known archaeological site along Curry Draw, a tributary of the San Pedro River (59).
_Clovis-age kills of mammoth and bison were buried by an algal mat, the type locality for the "black mat" used as a  marker bed by the YDIH proponents.
_In the kill area, the black mat (Unit F2) clearly buries the Clovis-age remains (which are in and resting on Unit F1  channel sands).
_Firestone and colleagues (ref.22, its SI Section C; ref.23, p.34) state "a thin layer (<2 cm) that contains YDB markers  lies at the base of the black mat and immediately overlies the bones."
_Wittke et al. also note that "YDB proxies were deposited at the F1/F2 contact on top of bones and artifacts but  beneath the black mat within a brief temporal window of a few weeks" (ref.3, its SI page 13).
_SI: p.42 However, no evidence has been presented by any investigator for supposed YDB markers overlying the bones  at Murray Springs.
_Further, the sampling reported by Firestone et al.(21-22) was along Trench 22 which has never been radiometrically  dated (59).
_The sampling section discussed by Wittke et al. is not specified and has not been dated either (ref.3, its SI p.14) so we  assume it likewise is from exposures in Trench 22.
_Yet, the archaeological features buried by the black mat, the Clovis age-channel deposits (Unit F1) under the black  mat, and the marl (Unit E) are not present at the south end of the Trench 22 exposure where the samples for impact  indicators were taken.
_The black mat at that section represents a compressed and mixed facies that covers a disconformity on top of Unit D.
_The age of the surface of Unit D has not been determined but based on the age of units F1 and E, it must have been  exposed for at least several hundred years if not thousands of years before the Clovis occupation.
_As to how the supposed YDB was dated, Firestone et al.(22) cite eight 14C ages averaging
____10,980 ± 80 14C years BP (ref.22, its SI Table 2); this would appear to be a misstatement of the average of eight  ages from F1 calculated as
____10,900 ± 50 14C years BP (ref.59, its Table A.1).
_Wittke et al.(3) recognize that Haynes and Huckell (59) "did not directly date the strata at our sampling location," and  in the absence of direct dating they "integrated several locations to produce a generalized composite age-depth model  that is consistent with the geochronology of Haynes and Huckell (59) for strata F2 through strata E.
_For the 46-cm interval, we utilized the seven dates available on sub-strata that match the stratigraphic designations  provided by Haynes and Huckell" (ref.3, its SI p.14).
_They fail to mention, however, there are ~70 radiocarbon ages available that match those stratigraphic designations  (for the full listing see ref.59, its Table A.1).11
_We grant that not all of those 70 ages are of comparable quality or relevance (and some are different fractions of the  same sample), but Wittke et al.(3) do not explain or justify why they selected 11
_The tally of Murray Springs radiocarbon dates is not intended to be comprehensive (for a full listing see 59), but is  intended to highlight the availability of other ages from the relevant stratigraphic units and their variation across the  site.
_Such should suffice to show that the data provided by Firestone et al.(22) and Wittke et al.(3) are inadequate to  determine the age of their supposed YDB layer.
_SI: p.43 the seven they did.
_When compared with a list of 48 ages from the relevant strata (Table S1 2), including many identified by the site  investigators as reliable, it is obviously not the case that dates were selected by Wittke et al.(3) based on proximity to  their sampling locality or to other ages – none are even from the same area of the site – or based on the material dated  or the degree of uncertainty accompanying the age.
_Table S12. Radiocarbon ages for Murray Springs from Wittke et al. (ref.3 its T able S.1) and Haynes and Huckell  (ref.59, its Table A.1).
_The latter lists 70 radiocarbon dates derived from ~50 samples (multiple fractions were dated on some) from unit F2,  basal F2 contact, F1 and E.
_The list below includes single ages as well as average ages from multiple fractions.
_Ages omitted without explanation by Wittke et al. (ref.3, its Table S.1) are highlighted in gray.
_Lab no. Level 14C ± Comments
A-969 F2c ____9020 360 Area 1; residue & humates (average); omitted
A-977 F2b ____10250 170 Area 1; carbonate
A-989b F2a ____10360 90 Area 1; humates; omitted
TX-1239 F2 ____9310 150 Trench 13N; carbonate; omitted
TX-1252/1253 F2 ____9600 150 Trench 13N; residue & humates (average); omitted
TX-1238 F2 ____9810 150 Trench 13N; carbonate
TX-1237 F2 ____9660 150 Trench 13N; carbonate; omitted
TX-1184/1185 F2 ____9820 110 Trench 13N; humates (average); omitted
I-4566 F2 ____8830 170 Profile B; carbonate; omitted
AA-26210 F2a4 ____9823 46 Profile B; residue; omitted
AA-26211 F2 ____10325 44 Profile B; residue; omitted
AA-26212 F2a1 ____10628 60 Profile B; residue
SMU-130/133 F2 ____9400 30 North of RR; humates (average); omitted
TX-1181-1182 F2 ____9370 140 Area B; humates (average); omitted
TX-1460/1461 F2 ____9820 110 Area 8; humates (average); omitted
A-1045 F2 /D ____10760 100 Area 4; charcoal +
F2 TX-1044 F2 /D ____12600 2440 Area 4; charcoal + F2; omitted
TX-1045 F2 /D ____10260 140 Area 4; charcoal + F2; omitted
SMU-19 F2 /F1 ____10740 190 Trench 28; humates; omitted
SMU-29 F1 ____10790 150 Trench 28; humates; omitted
TX-1462 F1 ____10930 170 Trench 28; charcoal
SMU-1463 F1 ____10900 200 Trench 28; humates; omitted
A-805A/805B F1 ____11220 330 Area 1; charcoal (average); omitted
SMU-17 F1 ____8770 80 Area 1; charcoal; omitted
TX-1406 F1 ____12940 390 Area 1; charcoal; omitted
TX-1413 F1 ____11080 180 Area 1; charcoal; omitted
SMU-190 F1 ____12820 450 Area 1; snail shell CO 2; omitted
SMU-42 F1 ____10840 140 Area 2; charcoal; omitted
SMU-18 F1 (E2) ____11190 180 Area 2; charcoal
SMU-41 F1 ____10840 70 Area 2; charcoal; omitted
SMU-43 F1 ____11160 110 Area 2; humates; omitted
TX-1459 F1 ____10710 160 Profile B; charcoal; omitted
SMU-27 F1 ____10890 180 Trench 20; charcoal; omitted
SMU-28 F1 ____11210 200 Trench 20; humates; omitted
SI: p.44
I-4565 E ____10430 160 Profile A; carbonate; omitted
I-4563 E ____9780 140 Profile A; carbonate; omitted
I-4562 E ____12310 170 Profile A; carbonate; omitted
I-4564 E ____19620 380 Profile A; carbonate; omitted
A-897 E ____21200 500 Profile A; carbonate; omitted
SMU-33 E ____11880 250 Profile Y; marl residue
SMU-34 E ____13980 190 Profile Y; marl residue; omitted
SMU-35 E ____18060 150 Profile Y; marl residue; omitted
SMU-36 E ____16180 420 Profile Y; marl residue; omitted
SMU-37 E ____27560 2300 Profile Y; marl residue; omitted
SMU-38 E ____19650 1400 Profile Y; marl residue; omitted
TX-1234 E ____10480 200 Trench 1; carbonate; omitted
TX-1235 E ____13310 190 Trench 1; carbonate; omitted
TX-1236 E ____10750 170 Trench 1; carbonate; omitted
_Just as problematic as the long list of omitted ages is the fact that, Wittke et al. admit, the seven ages they use are from  different locations across an "approximately 300×400 m excavation area" (ref.3, its SI p.13).
_This raises the question as to how those ages were "integrated" into a common depth scale for purposes of their  "interpolation by second order polynomial regression used to develop an age-depth model," thereby enabling them to  precisely determine the age of the "1-cm-thick YDB layer [that] was found at a depth of about 2.46 m at the contact  between strata F1 and F2" (ref.3, its SI p.13).
_No explanation is provided, and given that the elevation of these stratigraphic units varies across the site, the only  conclusion to be reached is that the depths assigned to those seven widely scattered dates are arbitrary and unreliable.
_None of those samples actually occur at those depths.
_Hence, the statistical analysis based on them must be as arbitrary and unreliable well.
_Leaving these problems aside for the moment, the age/depth model results that Wittke et al. (ref.3, its SI Fig.11)  present for Murray Springs are nonetheless reproducible.
_Using the same seven ages and depths from Wittke et al. (ref.3, its SI p.14 and its SI Tables S1 and S3), we estimate  the depth of the supposed YDB deposits to be within 1 cm of the 2.46 m depth that they report for the YDB layer  (Figure S13; a 2nd order polynomial model is used following Wittke et al.[3]).
_Further, a weighted regression provides an estimated age of
____12,809 cal BP for this depth, which falls within the YDB interval.
_The weighted regression, however, is not statistically significant.
_SI: p.45 Figure S13. Graphs of replicated regression-based age/depth models for Murray Springs.
_L : regression of depth on age; R : regression of age on depth. Error bars represent 1 SD.
_Dotted lines represent unweighted regressions, and solid lines represent weighted regressions.
_Ultimately, however, and like the case at Blackwater Draw, because of the problematic selection of dates and the  arbitrary and unclear depth scale used, 12 there is little reason to accept the age/depth model Wittke et al.(3) present for  Murray Springs.
_22. NEWTONVILLE GROUP 2A
_The Newtonville (New Jersey) exposure is a sandpit with two layers of loamy sand colluvium (the upper an ashen  gray, the lower a yellowish brown) separated by a clear wavy boundary, with the lower of the two atop a Miocene-age  sand and gravel, at the top of which a fragipan formed (ref.60, its SI Fig.S1b).
_Two 10-cm thick samples were collected for analysis of impact indicators, just above and just below the boundary of  the two loamy sand strata.
_A single OSL age of
____16,800 ± 1700 years BP was obtained from matrix in a sand-filled thermokarst involution below the lower loamy  sand, indicating that this coversand is Late Wisconsin in age (ref.60, its SI p.1).
_No age control is available for the upper loamy sand.
_Curiously, Wu et al.(60) report that the underlying, Late Wisconsin age sediment yielded 2000 magnetic  microspherules per kilogram, while the upper and younger loamy sand yielded only 1800 microspherules per  kilogram.12
_Murray Springs is one of several sites in Wittke et al. (ref.3, its SI Table S1) in which depths are not listed in the depth  column.
_Given that the actual depths for the Murray Springs radiocarbon samples are not on the same vertical scale, that is  perhaps not surprising.
_The depths used in Wittke et al.(3), however, can be derived from information scattered through the text and tables  cited.
_SI: p.46
_23. OMMEN GROUP 3C
_As at Lingen and Lommel, the stratigraphic section at Ommen, Netherlands, is tied to the Usselo soil, and with that a  presumption of a
____12,800 cal BP date.
_Like those sites, Ommen yielded a charcoal-rich layer at the top of the Usselo sand, at a depth of 117.5 cmbs, which  YDIH proponents identify as the supposed YDB layer.
_But as at Lommel, their radiocarbon results undermine that claim.
_Charcoal from the supposed YDB layer at Ommen produced an age of
____11,440 ± 35 14C years BP (which they calibrate as
____13,300 ± 50 cal BP) (ref.3, its SI Table S1).
_Wittke et al. (ref.3, its SI pp.14-15) admit this dat e "is somewhat older than typical charcoal in the YDB," and  attribute that result to "bioturbation, redeposition, or the 'old wood' effect." The possibility of mixing begs the  question of the stratigraphic integrity of the supposed YDB layer.
_In light of w hat they deem an older than acceptable age, they suggest that "the stratigraphic position of the YDB in  this site at the top of the Usselo sand is consistent with an age of ≈12.8 ka" (ref.3, its SI p.15).
_We have already discussed the flaws in this line of reasoning, and here again observe that a more straightforward  conclusion is that this radiocarbon date merely confirms the fact that the Usselo soil represents ~1400 years of surface  stability, and that the impact markers could date to any time within t hat age range.
_24. PAW PAW COVE GROUP 1A
_Paw Paw Cove is an exposure situated along the western side of Tilghman Island in the Chesapeake Bay, Maryland.
_Archaeological investigations conducted by Darrin Lowery resulted in the recovery of Clovis-age artifacts "located at  the contact of the 2ABtxb soil horizon and the overlying Btx [loess] soil horizon" (ref.61, p.53-54).
_The underlying 2ABtxb soil horizon has a radiocarbon age of
____17,820 ± 170 14C years BP (ref.61, p.54), and the presence of Clovis-age artifacts on top of it suggested to  Lowery that "a significant portion of terminal Pleistocene geologic time is missing from the region's upland  depositional sequence.
_It is assumed that the region's upland areas had been eroded and deflated during and slightly after the Clovis  occupation" (ref.61, p.56).
_The radiocarbon age on the underlying soil provides a terminus post quem for those artifacts.
_SI: p.47 Surovell et al.(27) sampled the section in the area where the artifacts were found in an effort to look for YDB  indicators.
_LeCompte et al. (ref.24, p.E2961) obtained a sample from a stratigraphic section estimated to be "within less than a  few hundred meters of the site reported in Surovell et al." LeCompte et al. state that "Surovell et al. assumed they were  sampling the YDB layer, and we have not questioned that assumption" (ref.24, p.E29610).
_But since no ages are available from either section of material from the surface on which the artifacts were found,  there is no evidence that either section contains an in situ YDB-age zone.
_Neither study makes reference to the underlying radiocarbon age, or the issue of deflation of the surface on which the  Clovis-age artifacts were found.
_25. PLAYA BASINS GROUP 2A
_The playa basins of the Great Plains are presented as having been "blown out of the soft earth by flying debris" (ref.21,  p.217) from the "extra terrestrial event" at
____12,900 cal BP.
_But the chronology fails here, for Firestone et al. (ref.21, p.216-217) misstate the results of the playa studies reported  by Holliday et al.(62).
_Firestone et al. (ref.21, p.216) state that "Holliday and coworkers (62) took core samples from beneath twelve dry  lake beds in the area and found radiocarbon dates ranging from
____16,000 to ____20,000 years ago.
_He concluded that because these dates are from the underlying formation, the Salinas [playas] formed more recently."
_Firestone et al. further state that "the age of the basins is consistent with the date of the Event," that is, the proposed  YDB impact (ref.21, p.218).
_In fact, however, no radiocarbon ages
____16,000-20,000 were recovered from any underlying formation.
_Holliday et al.(62) clearly state that some playa basins were present throughout the Late Pleistocene, but others  formed between
____13,000 and ____10,000 14C yrs BP, and that the basins were formed by erosion.
_Holliday's more recent data (63) shows that most playas are
____>14,000 cal BP and thus substantially older than the Younger Dryas and not 'caused' by it.
_And as noted in the main text, there is a Clovis mammoth bone bed – the Miami site (64-65) – in a pre-YD-age playa  basin.
_Of course, by the YDIH Clovis should not occur in a post-impact feature.
_26. SHERIDEN CAVE GROUP 3D
_Sheridan Cave is on the side of a karstic sinkhole.
_Excavations yielded two bone points, a small Clovis point, and some microdebitage (66).
_A "1.5 cm thick, charcoal-rich layer at 45.3 cm below the SI: p.48 cave floor" was identified as a YDB layer and  produced apparent impact indicators (ref.3, i ts SI p.15).
_That charcoal lens, according to site's original investigators, is in the upper 20 cm of their Stratum 5a (ref.66, p.514).
_Wittke et al.(3) infer the age of the supposed YDB layer from three samples that "returned dates of
____10.84 ± 0.08 14C ka BP
____ (12.80 ± 0.07 cal ka BP),
____10.92 ± 0.03 14C ka BP
____ (12.83 ± 0.05 cal ka BP), and
____10.96 ± 0.06 14C ka BP
____ (12.87 ± 0.08 cal ka BP)."
_Together these three ages "indicate that the impact proxy-rich YDB layer at 45.3 cm dates to ≈12.8 ka, consistent with  the age of the YDB at other sites" (ref.3, its SI p.15).
_In fact, however, only two of those dates are from the charcoal lens / supposed YDB layer: the third,
____10.92 ± 0.03 14C ka BP (rounded from
____10,915 ± 30 14C BP), was on Bone Point 2, which was recovered 10 cm above the charcoal lens, at the contact  between Stratum 5a and overlying Stratum 5b.
_Wittke et al. (ref.3, its SI Fig.13) claim that the "Clovis projectile points made of chert... and bone... were found in the  charcoal-rich YDB layer."
_That is not the case.
_As is clearly illustrated by Redmond and Tankersley (ref.66, Fig.9) and affirmed by Waters et al.(67),"the charcoal  level within Stratum 5a is 10-15 cm below the artifact bearing level" (ref.67, p.108).
_Moreover, the chronology of the charcoal lens is stratigraphically anomalous.
_There are 10 radiocarbon ages from Stratum 5a, which range from
____10,840 to ____12,840 14C yrs BP (Table S1 3).
_The overlying unit, Stratum 5b, has 5 radiocarbon ages, four of which range from
____10,470 to ____11,710 14C yrs BP (the fifth,
____13,120 14C yrs BP, is rejected as an outlier by Redmond and Tankersley, ref.66, p.514).
_Leaving aside the two radiocarbon ages that are directly from the charcoal lens, the four youngest ages from upper  Stratum 5a in which the lens is located form a statistical population (as determined by chi-square analysis [8]) with an  average age of
____11,550 ± 30 14C years BP.
_The overlying Stratum 5b has several statistically distinct age groups, only two of which (CAMS-10349 and CAMS- 33970) can be averaged, which results in an age of
____11,100 ± 40 14C years BP.
_In effect, the charcoal lens is significantly younger than the stratum in which it occurs, as well as being younger than  three of the four ages available from the overlying stratum.
_The charcoal lens is closest chronologically to Stratum 5c, which is atop Stratum 5b.
_In fact, had there been no dated samples SI: p.65
_17. Hester J (1972) Blackwater Locality No.1: a stratified early man site in eastern New Mexico. Fort Burgwin  Research Center, Dal las.
_18. Haynes CV, Jr.(1995) Geochronology of paleoenvironmental change, Clovis type site, Blackwater Draw, New  Mexico, Geoarchaeology 10:317-388.19.
_Cotter JL (1937) The occurrence of flints and extinct animals in pluvial deposits near Clovis, New Mexico, Pt. IV:  Report on the excavations at the Gravel Pit in 1936. Proceedings of the Philadelphia Academy of Natural Sciences  89:1-16.20. Johnson E, Holliday VT (1997) Analysis of Paleoindian bonebeds at the Clovis site: new data from old  excavations. Plains Anthropol 42:329-352.
_21. Firestone R, West A, Warwick-Smith S (2006) The cycle of cosmic catastrophes: how a Stone-Age comet changed  the course of world culture. Bear, Rochester, VT.
_22. Firestone R, et al.(2007) Evidence for an extraterrestrial impact 12,900 years ago that contributed to the  megafaunal extinctions and the Younger Dryas cooling. Proc Natl Acad Sci USA 104(41):16016 – 16021.
_23. Firestone R., et al.(2010) Analysis of the Younger Dryas impact layer. Journal of Siberian Federal University.  Engineering and Techno logies 3 (1) :30 – 62.
_24. LeCompte MA, et al.(2012) Independent evaluation of conflicting microspherule results from different  investigations of the Younger Dryas impact hypothesis. Proc Natl Acad Sci USA 109(44):E2960 – E2969.
_25. Kennett JP, West A (2008) Biostratigrap hic evidence supports Paleoindian population disruption at ∼ 12.9 ka.  Proc Natl Acad Sci USA 105:E110.
_26. Holliday VT, Meltzer DJ (2010) The 12.9ka ET impact hypothesis and North American Paleoindians. Curr  Anthropol. 51:575-607 _27. Surovell TA, et al.(2009) An independent evaluation of the Younger Dryas extraterrestrial  impact hypothesis. Proc Natl Acad Sci USA 106(43):18155 – 18158.
_28. Kennett DJ, et al.(2009a) Nanodiamonds in the Younger Dryas boundary sediment layer. Science 323(5910):94.
_29. _Bement LC, Carter BJ, Varney RA, Cummings LS, Sudbury JB (2007) Paleo-environmental reconstruction and  bio-stratigraphy, Oklahoma Panhandle, USA. Quat Int. 169 : 39-50.
_30. Bement LC, et al.(2014) Quantifying the distribution of nanodiamonds in pre-Younger Dryas to Recent Age  deposits along Bull Creek, Oklahoma Panhandle, USA. Proc Natl Acad Sci USA 111 (5) :1726-1731.
_31. kaczorowski RT (1977) The Carolina Bays: A Comparison with Modern Oriented Lakes. Coastal Research  Division Technical Report No. 13-CRD, University of South Carolina, Columbia.
_32. Thom BG (1970) Carolina Bays in Horry and Marion counties, South Carolina. Geol Soc Am Bull. 81:783-814.
_33. Melton FA, Schriever W (1933) The Carolina "Bays": Are they meteorite scars? J Geol 58:128-134.
_34. Ivest er AH, Godfrey-Smith DI, Brooks MJ, Taylor BE (2003) Concentric sand rims document the evolution of a  Carolina Bay in the middle coastal plain of South Carolina. Geol Soc Am Abstracts with Programs 35 :169. SI: p.66
_35. Brooks M, Taylor B, Ivester A (2010) Carolina Bays : time capsules of culture and climate change. Southeastern  Archaeology 29:146-163.
_36. Ives J. and D. Froese (2013) The Chobot site (Alberta, Canada) cannot provide evidence of a cosmic impact  12____800 y ago.
_Proc Natl Acad Sci USA 110:E3899.
_37. Wittke J, et al.(2013) Reply to Ives and Froese: Regarding the impact-related Younger Dryas boundary layer at  Chobot site, Alberta, Canada Proc Natl Acad Sci USA 110:E3900.
_38. Meltzer DJ (2006) Folsom: new archaeological investigations of a classic Paleoindian bison kill. University of  California Press, Berkeley.
_39. Firestone R.(2009) The case for the Younger Dryas extraterrestrial impact event: mammoth, megafauna, and  Clovis extinction, 12____900 years ago. Journal of Cosmology 2, 256 – 285.
_40. Rick T, Erlandson J, Vellanoweth R (2001) Paleocoastal fishing along the Pacific Coast of the Americas:  evidence from Daisy Cave, San Miguel Island, California. Am Antiq. 66:595 – 614.
_41. Erlandson J, et al.(1996) An archaeological and paleontological chronology for Daisy Cave (CA-SMI-261, San  Miguel Island, California. Radiocarbon 38:355-373.
_42. Simons D, Shott M, Wright H (1984) The Gainey site: variability in a Great Lakes Paleo-indian assemblage.
_Archaeology of Eastern North America 12:266-277.
_43. Simons D (1997) The Gainey and Butler sites as focal points for caribou and people. In Caribou and reindeer  hunters of the northern hemisphere, L. Jackson and P.Thacker, eds, pp.105-131. Avebury, Brookfield, VT.
_44. Simons D, Shott M, Wright H (1987) Paleoindian research in Michigan: current status of the Gainey and Leavitt  projects.
_Current Research in the Pleistocene 4:27 – 30.
_45. Boslough MB, et al.(2012) Arguments and evidence against a Younger Dryas Impact Event. Geophys Monogr Ser  198:13 – 26.
_46. Kurbatov AV, et al.(2011) Discovery of a nanodiamond-rich layer in the Greenland ice sheet. J Glaciol 56:749 –  759.
_47. Israde-Alcántara I, et al.(2012) Evidence from central Mexico supporting the Younger Dryas extraterrestrial  impact hypothesis.
_Proc Natl Acad Sci USA 109(13):E738 – E747.
_48. Blaauw, V.T. Holliday, J. Gill and K. Nicoll (2012) Age models and the Younger Dryas Impact Hypothesis. Proc  Natl Acad Sci USA 109:E2240.
_49. Boyd M, Running GL IV, Havholm K.(2003) Paleoecology and geochronology of Glacial Lake Hind during the  Pleistocene-Holocene transition: A con text for Folsom surface finds on the Canadian Prairies. Geoarchaeology 18  :583-607.
_50. van Geel, B, Coope G, van der Hammen T (1989) Palaeoecology and stratigraphy of the lateglacial type section at  Usselo (The Netherlands). Rev. Palaeobot. Palynol. 60:25-129.
_51. van Hoesel A, et al.(2012) Nanodiamonds and wildfire evidence in the Usselo horizon postdate the Allerød –  Younger Dryas boundary. Proc Natl Acad Sci USA 109(20):7648 – 7653.
_52. Hoek WZ (1997) Late-glacial and early Holocene climatic events and chronology of ve getation development in  the Netherlands.
_Veg. Hist. Archaeobot. 6:197-213.
_53. kaiser K, et al.(2009) Palaeopedological marker horizons in northern central Europe: characteristics of Lateglacial  Usselo and Finow soils. Boreas: an Int. J. Quat. Res. 38:591-609.
_54. Mahaney WC, et al.(2010) Evidence from the northwestern Venezuelan Andes for extraterrestrial impact: The  black mat enigma. Geomorphology 116(1-2):48 – 57. SI: p.67
_55. Mahaney WC, et al.(2008) Evidence for a Younger Dryas glacial advance in the Andes of northwestern Venezuela.  Geomorphology 96:199 – 211.
_56. Haynes CV, Jr.(2008) Younger Dryas "black mats" and the Rancholabrean termination in North America. Proc  Natl Acad Sci USA 105(18):6520 – 6525.
_57. Quade J, Forester R, Pratt W, Carter C (1998) Black mats, spring-fed stream s, and Late-Glacial-Age recharge in  the southern Great Basin. Quaternary Research 49:129 – 148.
_58. Pigati JS, et al.(2012) Accumulation of impact markers in desert wetlands and implications for the Younger Dryas  impact hypothesis. Proc Natl Acad Sci USA 109(1 9):7208 – 7212.
_59. Haynes CV, Jr., Huckell BB (2007) Murray Springs: a Clovis Site with multiple activity areas in the San Pedro  Valley, Arizona.
_University of Arizona Press, Tucson.
_60. Wu, Y, Sharmaa M, Le Compte M, Demitroff M, Landisa J (2013) Origin and proven ance of spherules and  magnetic grains at the Younger Dryas boundary. Proc Natl Acad Sci USA 110:E3557 – E3566.
_61. Lowery D (2009) Geoarchaeological investigations at selected coastal archaeological sites on the Delmarva  Peninsula: the long term interrelationshi p between climate, geology and culture. PhD dissertation, Department of  Geology, University of Delaware.
_62. Holliday VT, Gustavson T, Hovorka S (1996) Stratigraphy and geochronology of playa fills on the Southern High  Plains. Geol Soc Am Bull. 108 :953-965.
_63. Holliday VT, Mayer JH, Fredlund G (2008) Geochronology and stratigraphy of playa fills on the Southern High  Plains. Quaternary Research 70 :11-25.
_64. Holliday VT, Haynes CV, Hofman JL, Meltzer DJ (1994) Geoarchaeology and Geochronology of the Miami  (Clovis) Site, Southern High Plains of Texas Quaternary Research 41:234-244.
_65. Sellards EH (1938) Artifacts associated with fossil elephant. Geol Soc Am Bull. 49:999-1009.
_66. Redmond BG, Tankersley K (2005) Evidence of early Paleoindian bone modification and use at the She riden  Cave site (33WY252), Wyandot county, Ohio. Am Antiq. 70:503-527.
_67. Waters MR, Stafford TW Jr, Redmond BG, Tankersley KB (2009) The Age of the Paleoindian Assemblage at  Sheriden Cave, Ohio. Am Antiq. 74:107-111.
_68. Goodyear AC (2005) Evidence of pre-Clovis Sites in the eastern United States. Paleoamerican Origins: Beyond  Clovis, eds.
_Bonnichsen R, Lepper B, Stanford D, Waters M (Texas A&M University Press, College Station) pp.103-112.
_69. Waters MR, Forman S, Stafford TW, Foss J (2009) Geoarchaeological investigations at the Topper and Big Pine  Tree sites, Allendale County, South Carolina. J Archaeol Sci. 36:1300-1311.
_70. Miller DS (2010) Site formation processes in an upland Paleoindian Site : The 2005  – 2007 Topper Firebreak  excavations. South Carolina Institute of Archaeology and Anthropology, Columbia.
_71. Smallwood AM, Miller DS, Sain S (2013) An overview of the Clovis lithic assemblage from the Topper Site,  South Carolina.
_In the Eastern Flu ted Point Tradition, ed. Gingerich J (University of Utah Press, Salt Lake City) pp.280-298.
_72. Goodyear AC (2013) Update on the 2012-2013 Activities of the Southeastern Paleoamerican Survey. Legacy  17:10-12. SI: p.68
_73. Miller DS (2007) Site Formation Processes at an U pland Paleoindian Site: The 2005-2007 Topper Firebreak  Excavations.
_Master's Thesis, Department of Anthropology, University of Tennessee, Knoxville.
_74. Sain D (2011) Clovis blade technology at the Topper Site (38AL23). Master's Thesis, Department of Anthropo  logy, Eastern New Mexico State University, Portales.
_75. Smallwood AM (2011) Clovis Technology and Settlement in the American Southeast. Ph.D. Dissertation,  Department of Anthropology, Texas A&M University, College Station.
_76. Larsen CE, Schuldenrein J (1990) De positional history of an archaeologically dated floodplain, Haw River, North  Carolina.
_Archaeological Geology of North America, eds. Lasca NP, Donahue J (Geological Society of America Centennial  Special Volume 4, Boulder, Colorado), pp.161-181.
_77. Rawling JE (2000) A review of clay lamellae. Geomorphology 35:1-9.
_78. Kooyman B, et al.(2001) Identification of horse exploitation by Clovis hunters based on protein analysis. Am  Antiq. 66: 686-691.