Evidence for European presence in the Americas in AD 1021

The study addresses the longstanding question of when Norse (Viking) activity occurred in North America, specifically at L'Anse aux Meadows, Newfoundland—the only confirmed Norse site in the Americas. Previous chronological estimates relied on stylistic analyses of structures and artifacts and interpretations of Icelandic sagas, which were written down centuries after the events and suggest a brief occupation near the end of the first millennium. Conventional radiocarbon dating at the site produced more than 150 dates (55 linked to Norse occupation), but calibration uncertainties, methodological limitations of earlier decades, small sample sizes, potential inbuilt age in samples, and large measurement errors yielded wide calibrated ranges spanning or exceeding AD 793–1066. The purpose of this study is to establish a precise calendar-year date for Norse presence by leveraging a globally synchronous radiocarbon anomaly (AD 993) recorded in tree rings, thereby providing a secure chronological anchor for Norse presence in North America and for broader late Viking Age chronology.

The paper reviews prior archaeological, historical, and radiocarbon evidence related to Norse presence at L'Anse aux Meadows and beyond. Archaeological campaigns have detailed the site’s structures and environment and suggest it served as a base for expeditions further south. Interpretations of Icelandic sagas have informed hypotheses about the duration and extent of Norse activity, generally pointing to brief occupation, but their late transcription limits reliability. Past radiocarbon efforts (largely from the 1960s–1970s) suffered from larger uncertainties, small samples, potential inbuilt age, and sometimes inappropriate handling of offsets, leading to broad date ranges incompatible with a short occupation scenario. Broader literature documents globally coherent radiocarbon anomalies caused by cosmic-ray events in AD 775 and AD 993, which provide potential absolute time markers in dendrochronological series. These anomalies have been successfully used for precise dating in other archaeological contexts, motivating their application here.

The study employs a high-precision radiocarbon approach combining annual tree-ring sampling, accelerator mass spectrometry (AMS), wiggle matching, and χ²-based pattern matching to the AD 993 atmospheric 14C anomaly.

• Materials and sampling: Four wooden items modified by metal tools (indicative of Norse workmanship and not made by Indigenous groups of the period) were sampled from L'Anse aux Meadows (find numbers 4A 59 E3-1, 4A 68 E2-2, 4A 68 J4-6, 4A 70 B5-14). Microscopic examination of transverse and radial sections enabled ring counting; samples were collected ring-by-ring starting at the waney edge (numbered 0 for the outermost ring, -1 for penultimate, etc.). Three items contained identifiable waney edges and sufficient ring counts; one (4A 70 B5-14) spanned only nine rings and lacked the AD 993 anomaly, thus was excluded from precise dating.
• Taxonomy: Two items were identified as fir (Abies, likely A. balsamea), and one as juniper/thuja type (Juniperus/Thuja), all conifer softwoods without vessels; identification used microscopic features (ray height, crossfield pits, resin canals, axial parenchyma) with up to ×400 magnification.
• Pretreatment and measurement: In total, 127 14C measurements were made on 83 individual rings (115 at Centre for Isotope Research, Groningen; 12 at Curt-Engelhorn-Center Archaeometry, Mannheim). Wood was pretreated to α-cellulose (or holocellulose as applicable) using acid–base–acid and bleaching protocols; PEG consolidation (present in all but 4A 68 E2-2) was removed by prolonged hot-water rinses. Samples were combusted in elemental analyzers, CO2 converted to graphite, and measured on MICADAS AMS instruments. Stable isotope ratios were assessed for quality control and normalization; laboratory standards and blanks were run concurrently. Interlaboratory agreement showed no significant offset (5.1 ± 7.9 14C yr).
• Calibration and modeling: Initial wiggle matching of each item’s multi-year 14C series to the Northern Hemisphere IntCal20 calibration curve (OxCal 4.4, D_Sequence) provided 95% ranges for waney edges between AD 1019 and AD 1024. To pinpoint the exact felling year, χ² pattern matching was applied to six consecutive rings (-31 to -26) expected to contain the AD 993 anomaly, using a single-year reference dataset (Büntgen et al. 2018; B2018). For each item, trial waney-edge years within AD 1016–1026 were tested, and the χ² minimized to identify the best-fit alignment of the sharp AD 993 anomaly.
• Seasonality: Earlywood/latewood development at the waney edge was observed to assess felling season where possible; PEG consolidation hindered season determination in one sample.

• Three independently sourced wood items (from at least two species and different trees) from L'Anse aux Meadows contain the AD 993 radiocarbon anomaly exactly 29 rings in from the waney edge, allowing precise felling-year determination.
• Exact cutting year: AD 1021 for all three items based on χ² best fits to B2018; alternative fits within AD 1019–1024 passed at 95% probability, but the optimal alignment consistently indicated AD 1021.
• Seasonality: One item (4A 68 J4-6) shows a spring felling; another (4A 68 E2-2) indicates summer/autumn; season for 4A 59 E3-1 could not be determined due to past PEG consolidation.
• Measurement corpus: 127 total 14C measurements on 83 rings (115 Groningen; 12 Mannheim). Individual results typically better than ±2.5‰, some averaged results better than ±1.5‰ (~12 14C yr). No significant interlaboratory offset (5.1 ± 7.9 14C yr).
• Wiggle-matched waney-edge ranges initially AD 1019–1024; precise χ² matching yielded AD 1021 for all three items.
• Taphonomic and contextual evidence argue against driftwood or later modification: presence of waney edges on all three items, local wood availability at the time, and consistent tool marks from metal tools support in situ Norse modification around AD 1021.
• This constitutes the earliest secure calendar year for European presence in the Americas and a fixed point indicating that by AD 1021 human migration had encircled the globe.

The findings directly resolve the long-debated timing of Norse presence in North America by providing an exact calendar year anchored to a globally synchronous radiocarbon anomaly. By identifying the AD 993 event within tree-ring series and counting to the waney edge, the study overcomes prior limitations of radiocarbon dating in this period and the uncertainties of saga-based chronologies. The convergence of three independently sourced wood items on the same cutting year (AD 1021) strongly indicates Norse activity at L'Anse aux Meadows in that year. Contextual factors—metal tool marks, preservation of waney edges, and local wood abundance—support contemporaneous Norse modification rather than reuse of driftwood or scavenged material. The date provides a robust chronological anchor for late Viking Age studies, enabling improved assessments of site occupation duration and synchronizing archaeological sequences across the North Atlantic. It also serves as a reference point for investigating subsequent cultural, biological, and epidemiological consequences of Norse–Indigenous interactions, though genetic data from Norse Greenlanders thus far show no evidence of gene flow from the Americas.

The study provides precise evidence that Norse were active on the North American continent in AD 1021 by exploiting the AD 993 radiocarbon anomaly recorded in tree rings from Norse-modified wood at L'Anse aux Meadows. This secure date refines late Viking chronology, marks the earliest confirmed European presence in the Americas, and represents the earliest known point by which humans had encircled the globe. The result establishes a definitive tie point for future research into the scale and consequences of Norse transatlantic activity, including cultural exchanges, potential pathogen transmission, and biotic introductions. Future work can apply similar anomaly-based dating to other sites and materials to resolve the timing of migrations and interactions with exact-year precision and to investigate the broader temporal pattern of Norse exploration beyond AD 1021.

• Only three wood items from distinct trees contained sufficient ring counts and the AD 993 anomaly for exact dating; a fourth item was excluded due to too few rings and absence of the anomaly.
• Determination of felling season was not possible for one sample due to prior PEG consolidation.
• While the cutting year is established as AD 1021, the overall duration, frequency, and extent of Norse activity in North America remain unresolved and require further investigation.
• The precise dating approach depends on the presence of a globally recognizable radiocarbon anomaly (e.g., AD 993) and a preserved waney edge; contexts lacking these features cannot be dated to exact years using this method.