Ice cores extracted from polar regions and high-altitude/low-latitudes glaciers are archives of past climatic-environmental histories as they record physical, chemical and biological characteristics of the past atmosphere. To interpretate this information, it is of fundamental importance to link the englacial depth of the various ice sections to the timing of the original snow deposition. This provides a function that is commonly known as chronology.

Counting annual layers, establishing time markers (e.g. volcanic horizons, ¹⁰Be, ³H peaks etc.), synchronizing with other dated paleo-records, and developing ice flow models are widely used methods to precisely and accurately date ice cores from polar regions (Parrenin et al., 2007; Svensson et al., 2020), where the negligible/slow horizontal flow at the drilling sites and the largely below freezing englacial temperatures allow a full physical-chemical preservation of the ice stratigraphy and the time markers embedded within the accumulated snow layers. In contrast, dating ice layers from high-altitude/low-latitude glaciers is more challenging because of: (i) their smaller ice thicknesses, implying that counting annual layers is usually limited to the upper 50 to 100 m (typically covering only a few centuries) (Schwikowski et al., 2014); (ii) a larger horizontal flow at the drilling sites that can quickly alter/disrupt the original ice stratigraphy (Thompson et al., 2000), and; (iii) increasingly widespread post depositional processes that are linked to modern warmer air temperature and summer meltwater percolation which can overprint (or prevent the full conservation of) the time markers embedded in the glaciers (Gabrielli et al., 2010).

In 2011 we extracted four 60–75 m long ice cores located within 10 m from each other from the Alto dell'Ortles glacier (3859 m), near the summit of Mt. Ortles (3905 m) in the Eastern Alps, Italy. By using the three ∼ 75 m longest cores (from now on cores #1, #2 and #3; with core #4 of only ∼ 60 m being archived for future analyses) we demonstrated that the upper firn portion of the Alto dell'Ortles glacier was temperate, with intense summer meltwater percolation that possibly affected this drilling site since the 1980s. However, the underlying ice was cold (-2.8 °C near the bedrock) (Gabrielli et al., 2012) which preserved stratigraphy that was thousands of years old (Gabrielli et al., 2016). Our discovery of millennial-age ice in the Eastern Alps was corroborated by another ice core record extracted from the nearby drilling site of Weißseespitze (3500 m, Austria) where ¹⁴C dating above the basal ice resulted in an age of 5.9 ± 0.7 kyr cal BP (Bohleber et al., 2020). Together these two studies expand findings from the Western Alps, where even older bottom ice was discovered at Colle Gnifetti (¹⁴C date of > 15 kyr cal BP; Jenk et al., 2009) and at Dôme du Goûter (9.4 ± 0.4 kyr cal BP, with ice below estimated to be at least 12 kyr; Legrand et al., 2025), to the Eastern sector of the Alps.

In order to develop an initial chronology, we previously identified well-defined ³H and beta activity peaks at 41 m depth from atmospheric thermonuclear testing attributed to 1963 and used the activity of ²¹⁰Pb to date the Alto dell'Ortles cores to 59 m depth (∼ 1930 CE) (Gabrielli et al., 2016). In addition, determination of a ¹⁴C age of a larch needle embedded near the basal ice, along with ¹⁴C dating of several ice samples (using the water insoluble organic carbon fraction entrapped; WIOC-14; Uglietti et al., 2016; Jenk et al., 2009; Hoffmann et al., 2018) allowed us to obtain absolute time markers from the deepest ice, the age of which was calculated to date back to ∼ 7000 years. All the age constraints obtained were used to build an initial chronology (TC2016 from now on) by means of Markov chain Monte Carlo simulation (COPRA model; Breitenbach et al., 2012), which also provided time uncertainty as a function of depth from a few years in the upper modern firn layers to ±500 years in the deepest portion of this ice core (Gabrielli et al., 2016).

A large uncertainty in TC2016 not only exists for the basal layers of the ice core but also at intermediate depths (60–72 m), dated between 1900 and 200 CE (e.g., ∼ ±50 years at 1850 CE, ∼ ±100 at 1650 CE and ∼ ±400 at 500 CE). The relative time uncertainty of the intermediate portion of the record (50 %–70 %) was higher than both for the modern (1900–2011 CE, 5 %–25 % uncertainty) and the oldest (200 CE–5000 BCE; 10 %–30 % uncertainty) sections. The large uncertainty of this intermediate section essentially arises from (i) a lack of time markers such as volcanic horizons; and (ii) the availability of only two WIOC ¹⁴C ages (1355 ± 205 and 429 ± 286 CE) above 72 m depth, with the younger of them characterized by a relatively large uncertainty due to the limits of the applied ¹⁴C method for ages of less than 500 years (combination of increased analytical uncertainty for small samples and the relative flatness of the atmospheric ¹⁴C calibration curve during this period; Uglietti et al., 2016). Altogether, the large absolute and relative uncertainty posed questions on the suitability of TC2016 for the reconstruction of meaningful climatic and environmental histories of the Alps and Central Europe from these cores.

Here we report additional stratigraphic/time markers, which, while confirming TC2016 within its uncertainty, allow its improvement. Therefore, using the same approach as used for TC2016, a first intermediate timescale was built, in the following named CP2025/1, which already gaining from an improved depth alignment of the three Alto dell'Ortles cores used (by an increased number of tie points obtained from the δ18O records, and newly available Pb records), as well as from new, additional age constraints from: (1) ¹⁴C dating of an additional organic fragment (charred spruce needle; 232 ± 126 BCE) discovered near the bottom ice (72.82 m) in core #1; and (2) the automated counting of annual layers with Straticounter (Winstrup et al., 2012; Winstrup, 2016) based on pollen, δ18O and dust records from core #1 from 2011 to 1927 CE (51 m depth in core #1) and visual counting from 1926 to 1900 CE (57 m depth in core #1). To further refine the chronology, the CP2025/1 timescale then served as the basis to apply a temporal synchronization between the high resolution non-crustal Pb concentration (ncPb) record from core #3 with a well-dated (±5 years) Arctic record (McConnell et al., 2019) for the period from 1755 CE back to 175 BCE. Thereby, allowed shifts in age were constrained by the requirement to be consistent within the CP2025/1 dating uncertainty, largely defined by the absolute dating (i.e., the calibrated ¹⁴C age probability range). Last, for a model based assessment and discussion, the finally derived, empirical chronology CP2025/2, of improved accuracy and precision, was compared to results from the Dansgaard-Johnsen ice flow model (Dansgaard and Johnsen, 1969) applied to the Alto dell'Ortles drilling site.

The first step to develop a more accurate chronology was to revise the previously published (Gabrielli et al., 2016) common depth scale for the three Alto dell'Ortles cores, using core #2 for depth reference as we did for TC2016 (see Sect. S1 in the Supplement). Depth realignments of core #2 with cores #1 and #3 were performed in two steps using the Analyseries 2.0.8 software (Paillard et al., 1996). First, the δ18O records were more finely matched using additional tie points between cores #1 and #2 (for a total of 122 points) and between cores #3 and #2 (87 points). In this case the 122 tie points between cores #2 and #1 provide a Pearson correlation r = 0.78 (significant at p < 0.01) between all the δ18O values aligned on core #2 depth while 87 tie points between cores #2 and #3 give r = 0.79 (p < 0.01). For comparison, in TC2016 the 17 tie points between cores #2 and #1 provided r = 0.72 (p < 0.01) while 14 tie points in cores #2 and #3 gave r = 0.67 (p < 0.01). The increase in r underlines the improvement in the depth alignments compared to the one performed previously for TC2016 as also illustrated in Fig. 1.

While a more detailed alignment was obtained (Fig. 1, blue vertical lines), a lack of tie points persisted within the 60–73 m depth interval between cores #2 and #3. In TC2016 this was an unresolved issue only realized after publication, resulting in a depth misalignment of a few tens of cm, and consequentially in a significant time lag (up to ∼ 300 years) between the three cores (see Sect. S1 in the Supplement). To also refine the depth alignment of cores #2 and #3 in this interval, 31 additional tie points were considered in this study (Fig. 1, red vertical lines). Those were determined by matching the main features in the high resolution Pb concentration records from independent analysis in cores #1 and #3; by discrete Inductively Coupled Plasma Mass Spectrometry (ICP-MS) and continuous flow analysis (CFA) ICP-Sector Field MS (ICP-SFMS) at the University of Venice and at the Ohio State University, respectively (Gabrieli, 2008; Gabrieli and Barbante, 2014) (Fig. 2). Details on the Pb concentration analysis are provided in Sect. S2 in the Supplement, which also includes a brief discussion of the method induced offset in absolute concentrations observed between cores #1 and #3. We note that, since the Pb records are used exclusively for depth alignment in this section and for synchronization in Sect. 5 – with the primary goal of improving the chronology, particularly in terms of uncertainty – a detailed environmental, geochemical, or historical interpretation lies beyond the scope of this manuscript.

With Pb not determined in core #2 due to insufficient ice volume, depth alignment based on the Pb records between 60 and 73 m was performed in three steps: (i) by matching the Pb concentration records in cores #1 and #3 (31 tie points obtained) using Analyseries 2.0.8 software (Paillard et al., 1996) (Fig. 2); (ii) by then transferring the depths of the 31 Pb points in core #1 to core #2 using the depth map obtained by matching cores #1 and #2 using δ18O, see above; and (iii) by linking the δ18O stable isotope records in cores #2 and #3 over this interval using as a guide the depth of the 31 supplemental Pb tie points. The final correlation obtained within this depth interval between the two Pb records is r = 0.91 (p < 0.01). We note that the significant correlations between the δ18O and Pb records from the different Alto dell'Ortles cores indicates a high degree of reproducibility, which confirms the absence of any observed stratigraphic disturbance at this drilling site (Gabrielli et al., 2016).

By use of the improved depth alignment of the three Alto dell'Ortles cores, the already existing time markers from TC2016 and new markers, obtained from the respective analyses in the three cores (Table 1), could be combined on a revised common depth scale (Tables S1–S3; subsequently used to construct the revised time scale, see Sects. 4 and 6). A new time marker in the deepest portion was obtained from an additional ¹⁴C age (Sect. 3.1). To refine the chronology of the younger part from 1900 to 2011 CE, three additional time constraints (2006, 1995, 1986 CE; Table 3) and 113 new annual time markers obtained by instrumental and visual annual layer counting (ALC) of pollen, dust and δ18O were considered, while also still relying on the beta and ³H activity peaks related to 1955 and 1963 CE, respectively, previously used also in the construction of TC2016 (Sect. 3.2).

By combining the time markers established in TC2016 with the new, additional markers presented in Sect. 3 on the revised common depth scale (Sect. 2), a preliminary revised version of the previously published TC2016 timescale was constructed; in the following denoted as CP2025/1. In line with the development of TC2016, the continuous depth-age relationship of the Alto dell'Ortles cores (Fig. 5) was built by fitting all the empirical time markers (defined in m w.e. depth and years before 2012; Table S2) by means of Markov chain Monte Carlo simulation (COPRA model, 2000 simulation runs), thereby accounting for their linked uncertainty ranges. This provides a depth-age relationship with an associated depth-time dependent uncertainty (Breitenbach et al., 2012). For the most recent portion of the chronology (2011–1927), the original uncertainty of the single annual layers provided by StratiCounter (1–10 years) was retained. Overall, 113 annual time markers provided by StratiCounter from 2012 to 1900, anchored by five distinct time horizons (Table 3) were combined with seven ¹⁴C dated layers in the deep part of the record (651 CE back to around 7000 years before present) (Fig. 5).

Comparison of this updated Alto dell'Ortles timescale (CP2025/1) with the previously published TC2016 chronology is displayed in Fig. 5. As a result of replacing the dating by ²¹⁰Pb with the new time markers from the annual layer counting (constrained by the three additional time markers of 2006, 1995, 1986; Sect. 3.2), the two timescales differ significantly during the most recent 20 years. Also, CP2025/1 is around 20-100 years older between 1920 and 1600 CE, while elsewhere, CP2025/1 and TC2016 agree within their uncertainties (Fig. 5).

Six well-dated ncPb records from Arctic ice cores were shown to document the spatial and temporal changes in Arctic lead pollution from 200 BCE to 2010 CE, before the 19th-century Industrial Revolution, primarily caused by European emissions related to past mining operation in Central Europe and the Mediterranean region (McConnell et al., 2019). They display periods of increase in atmospheric ncPb from the Roman period (see also McConnell et al., 2025), through the Early Middle Ages to the 1970s industrial peak. Superimposed on these major variations, distinct multiannual to multidecadal variations exist, marked by increases coincident with the exploitation of mining regions for Pb and Ag metallurgical production, decreases coincident with climate disruptions, famines, major wars, and plagues occurring in Europe and the Mediterranean region during the last two millennia (McConnell et al., 2019).

Using Pb records for synchronizing alpine and Arctic ice core chronologies has previously been applied (Preunkert et al., 2019; Osman et al., 2021). The underlying geochemical assumption is that an anthropogenic Pb signal from emission sources located in the European region and recorded even in Arctic ice, will most likely be imprinted, at essentially the same time, also in ice archives located much closer to these sources. Because the Alps – with Alto dell'Ortles being part of them – represent a central archive of past atmospheric pollution in Europe (e.g., Preunkert et al., 2019; Eichler et al., 2023), we adopted this synchronisation approach, choosing one of the above mentioned Arctic records characterized by their imprinted signal of European Pb emission.

Among the six Arctic records, the one from the Akademii Nauk ice cap in Severnaya Zemlya, Russian Arctic (AN) was selected for the following reasons. While, like the other Arctic records, it shows similarities with the high-resolution ncPb record from the Alto dell'Ortles ice core (Fig. 6), (i) it contains one of the longest continuous chronologies from 200 BCE to 1999 CE; (ii) this record is characterized by the largest Pb amplitude signal; and, (iii) it is suggested by atmospheric modelling to be the record most strongly influenced by Pb emissions from silver mining and metallurgical activities in central Europe, the region close to Alto dell'Ortles, during the past millennia (McConnell et al., 2019). Also, the uncertainty of the AN chronology was estimated to be low; up to a maximum of ±5 years with less than 2 years uncertainty at the known volcanic time horizons within the Greenland reference record (McConnell et al., 2019). Considering the relatively small dating uncertainty over the entire time covered by the AN record, when compared to the typical dating uncertainty for high altitude ice cores, the potential benefit of synchronisation becomes evident.

In detail, the synchronisation of ncPb from AN was performed with the continuous flow analysis Pb record from Alto dell'Ortles core #3 because of its higher resolution when compared to the record from core #1. Alto dell'Ortles ncPb was obtained from Pb concentrations corrected for mineral dust contribution by using a Pb / Rb ratio of 0.51 (value obtained for a section between 69.55 and 69.94 m depth, characterized by the lowest Pb concentrations) which is reasonably consistent with the ratio of 0.15 provided for the upper crust by Wedepohl (1995). Values below the detection limit were replaced by half the analytical detection limit of Pb (Sect. S4 in the Supplement). Synchronisation was performed with the records on their respective time scales, i.e., the intermediate CP2025/1 timescale for Alto dell'Ortles (see Sect. 4), using the Analyseries 2.0.8 software (Paillard et al., 1996). Thereby, to avoid over-interpretation of individual data points and to account for the reported dating uncertainty of AN (≤ ±5 years), the records were averaged to a 10-year time resolution. For subsequent matching of the midpoints of the 10-year averages, only the most prominent features (obvious peak maxima and minima) were matched (see below). In addition, multiple constraints were applied: (i) final shifts in age should not exceed the initial CP2025/1 dating uncertainty range (i.e., overlap at least within 2σ uncertainties of CP2025/1 and the final uncertainty estimate), and also remain within 2σ of the calibrated ¹⁴C age probability range at the respective sampling horizons; (ii) compression or expansion of time intervals between tie-points should not produce larger/smaller amplitudes in layer-thickness variations than those derived from the initial CP2025/1 age scale (feature implemented for control in Analyseries). Age uncertainty estimates for the tie points were derived by combining the reported dating uncertainty of AN with a visual, best-possible assessment of how well the assigned peak maxima and minima are represented in either record. For example, a broad minimum, or several nearly equally low values occurring within close (i.e., decadal) proximity, would result in a larger uncertainty, estimated as approximately the half-width of the feature(s).

Following the above methodology, 11 prominent features between 175 BCE and 1755 CE were selected for synchronizing (Fig. 6). For the selected tie-points, the shift in age (CP2025/1 – AN age) was ∼ 10–20 years or less back to 1600 CE; around 40 years back to 1550 CE, between 120 and 270 years from 1250 to 250 CE; and between -280 to -240 years between 0 and 200 BCE. Importantly, 7 out of the 11 tie points were within the 1σ uncertainty of the initial CP2025/1 timescale based on the absolute ages from ¹⁴C, with the other 4 being within the 2σ range (see Figs. 6 and 7). Further we note that: (a) the ¹⁴C age (659 ± 102 BCE) of the larch needle (Table 2) was not shifted in the revised timescale which remains unchanged before 175 BCE, and (b) the final Alto dell'Ortles timescales for the 20th century is fully independent, since no synchronization tie-points were selected for this most recent period where the ice was dated based on the annual layer counting and the time horizons described in Sect. 3 (showing already high correlation with AN as indicated by a Pearson's r of 0.61 for annual values, p < 0.001). A brief discussion of the industrial-period section of the ncPb record from the Alto dell'Ortles record – particularly regarding the absolute concentrations which are somewhat lower than expected for this period if compared to other records from the Alps (Eichler et al., 2023; Gabrieli et al., 2014; Preunkert et al., 2019) – is provided in the Supplement (Sect. S3). Additionally, an illustrative example of the applied tie-point selection used for synchronization is described and visualized in (Fig. S4). In Sect. 6, an assessment of the synchronisation using the final Alto dell'Ortles timescale is provided.

Finally, we emphasize that incorporating the applied constraints in the synchronization procedure was a crucial part to exclude subjective or arbitrary age shifts during wiggle matching. We therefore recommend adopting such a quantitatively controlled, absolute-dating-based approach as a standard for future synchronization (wiggle matching) of ice cores from high-altitude (mountain) glaciers, whether aimed at establishing chronologies or supporting the interpretation of paleo-records.

The 113 annual time markers provided from the counting of annual layers from 2012 to 1900 (Sect. 3.2), the 11 tie points obtained from the match of the Alto dell'Ortles core #3 with the AN Arctic Pb record (1755 CE–175 BCE; Sect. 5), and four ¹⁴C ages from the deepest part below (659 BCE back to ∼ 7000 years; Sect. 3.1) were finally used to develop the final Alto dell'Ortles time scale CP2025/2. Using these data points (with depths in m w.e. and the age as years before 2012; see summary in Table S3), the exact same methodology as applied to construct TC2016 and CP2025/1 was applied (Markov chain Monte Carlo simulation, 2000 simulation runs, using COPRA; Breitenbach et al., 2012). Again, fitting the time markers with their given 1σ uncertainty, allowed to propagate the dating uncertainty to any depth/age. The resulting final Alto dell'Ortles depth-age relationship, CP2025/2 with its linked uncertainty is shown in Fig. 7. Note that although the uncertainty of this revised chronology was substantially reduced compared with the initial chronology TC2016, the actual age shift is relatively small – on the order of only a few decades – and remains within the previous uncertainty estimates (see Fig. 5).

With the final time scale at hand , the performed synchronisation with the AN ncPb record was assessed, comparing it to the Alto dell'Ortles record, now on the CP2025/2 timescale (Fig. 8). As a sort of reference, the independently dated, slightly shorter Pb record from Colle Gnifetti in the Western European Alps (CG03B, Switzerland;) was similarly compared to both the AN and Alto dell'Ortles record (see Sect. S3 in the Supplement and Figs. S2–S4). In general, the match between Alto dell'Ortles and AN is remarkable and provides additional evidence of the stratigraphic and temporal continuity of this ice core record (i.e., no evidence of age reversals or ice flow disturbance). By applying the synchronization, the correlation over the entire record increased from 0.26 to a significant value of 0.45 (r, p < 0.001), for the record before 1750 CE from 0.09 to 0.46 (r, p < 0.001). Also, the agreement between Alto dell'Ortles on CP2025/2 and the fully independently dated CG03B ice core is good back to ∼ 1500 CE until when the dating of CG03B is reasonably precise (r = 0.43, p < 0.001; Fig. S6). We also note that CG03B agrees well with AN back to at least 1500 CE (Fig. S7), confirming the underlying assumption of a contemporary pre-industrial atmospheric Pb signal in the Arctic and the Alps form common European emission sources (Sect. 5). However, the records of CG03B and AN start to diverge further back in time, which may not be too surprising considering the rapid increase in dating uncertainty of CG03B (from 7 years at 1763 CE to around 300 years at ∼ 0 CE/BCE; see Fig. S5).

Finally, we applied the Dansgaard-Johnsen (DJ) glaciological ice flow model (Dansgaard and Johnsen, 1969) to investigate the revised age-depth relationship CP2025/2 at the Alto dell'Ortles drilling site (see also Sect. S5 in the Supplement). The DJ model (one-dimension; 1D), uses an approximation of the ice flow to calculate the vertical strain rate (thinning of annual ice layers with depth). It results from assuming the horizontal velocity (vx) described by Glen's law to be constant from the glacier surface down to a given height above bedrock (h; shear zone thickness), where it decreases linearly with depth to zero at the bed (no-basal ice flow, i.e., ice frozen to bedrock; blue curve in Fig. 9). It should be noted that the term h is the result of the DJ model-approximation for vx and subsequent mathematical reformulations, rather than being descriptive of a distinct separation between two zones existing in reality.

A general limitation of modelling the age-depth relationship, independent on the complexity of the applied ice flow model, arises for the assumption of steady-state conditions. Unless additional glacier mass balance related information far into the past were available, this assumption is unavoidable. However, information about the age of a layer at a certain depth can provide invaluable constraints for model parameters that are not accessible otherwise (equally for models of lower or higher complexity). For reasonable and meaningful model output, the choice of parameter values is crucial. Here, we used two independent approaches (A1 and A2) for the determination of annual net accumulation rate (b) and h. A1 is based on present-day glaciological observations (Sect. 7.1), and A2 is based on the time information provided by the empirical dating described in this study (Sect. 7.2). We further used parameters from A1 in a mixed approach (B) with A2 (Sect. 7.3 and Table 4). In all cases steady-state conditions were assumed, which for the applied model (DJ) can be defined as: (i) constant glacier thickness (H), (ii) no-basal ice flow, (iii) constant annual net accumulation rates (b), and consequently, (iv) a constant shear zone thickness (h). For the glacier thickness H and the condition of non-basal flow, we relied on present-day observations. For H, we adopted the present value of 74.88 m (61.15 m w.e.) (Gabrielli et al., 2016), implying that the ice thickness had not changed significantly since glacier build-up at the drilling site ∼ 7 kyr BP. The “no-basal ice flow” assumption seems reasonable considering the ice temperature at bedrock of -2.8 °C in 2011 (Gabrielli et al., 2012), despite the recent warming.

An intermediate revised chronology (CP2025/1) was obtained for the Alto dell'Ortles ice cores by an improved depth realignment of cores #1, #2, #3, incorporating an absolute time marker from a newly discovered ¹⁴C dated macro-organic fragment (232 ± 126 BCE); and new data points from counting annual layers (from pollen, dust and δ18O records) visually and by means of an automatic algorithm (StratiCounter) between 1900 and 2011. To derive CP2025/1, all the time markers were combined and fitted for a continuous depth-age scale by Markov chain Monte Carlo simulation. For the youngest section of the core, this already resulted in an improved and more accurate chronology when compared to the previously published version (TC2016), which only relayed on absolute dates from ²¹⁰Pb in the youngest part (1900–2012 CE). The intermediate age scale was then used as the basis for subsequent synchronization of a new, highly-resolved ncPb concentration record from the Alto dell'Ortles Pb ice core with a well-dated polar ice core record from Akademii Nauk (AN, Russian Arctic) used as the age reference from 1755 CE back to 175 BCE. 11 additional tie points for model input were selected, thereby applying constraining criteria based on age and associated uncertainty of absolute ¹⁴C dates. By finally combining all available time markers (only tie-points for the synchronized range), the subsequent age model output yielded the final revised and fine-tuned new Alto dell'Ortles chronology CP2025/2. The absolute-dating-based and thus quantitatively controlled synchronization procedure described is recommended as a standard for future applications to ice cores from high-altitude (mountain) glaciers, whether aimed at establishing chronologies or supporting the interpretation of paleo-records. The new ¹⁴C age (232 BCE; stratigraphically consistent with the previous ¹⁴C dates), and the correlation between the ncPb CP2025/2 and AN records confirmed previous findings that stratigraphic disruption caused by disturbance in ice flow can be ruled out, at least for age inversions exceeding the current dating uncertainty and the resolvable spatial sampling resolution.

Finally, an ice flow model investigation of CP2025/2 was performed by adopting a simple 1D Dansgaard-Johnsen model, the results of which suggest that non-steady state conditions (e.g. changes in snow accumulation rate over time) need to be considered to explain the resulting CP2025/2 age-depth relationship. This particularly concerns the Little Ice Age (1250–1850 CE) when a decrease in snow accumulation rate was inferred, perhaps because of an increase of wind erosion efficiency. Clearly, non-steady state approaches that are currently used in polar regions should be considered and applied to model the ice flow at high altitude-low latitude ice core drilling sites.

In conclusion, this revised more accurate time scale (CP2025/2) will allow to provide additional detailed climatic environmental histories of Central Europe during the second half of the Holocene. In addition, CP2025/2 has the potential to become an important high-altitude term of comparison for multiple paleoclimate archives extracted at lower elevation in Central Europe.