Applicato per la prima volta un approccio completamente nuovo per studiare le variazioni del margine orientale della calotta. La ricerca si è basata sull’analisi di una carota di ghiaccio di 584 metri di profondità prelevata presso il sito di Renland nella regione di Scoresby Sund.
In uno studio pubblicato su Nature Communications il 3 Ottobre 2019, che ha visto coinvolto un team internazionale di ricercatori tra i quali Barbara Delmonte e Giovanni Baccolo del laboratorio Eurocold del dipartimento di Scienze dell’ambiente e della terra dell’Università di Milano-Bicocca, è stato per la prima volta applicato un approccio completamente nuovo per studiare le variazioni del margine orientale della calotta di ghiaccio della Groenlandia.
La ricerca si è basata sull’analisi di una carota di ghiaccio di 584 metri di profondità prelevata presso il sito di Renland (71.30°N, 26.72°W, 2315 m s.l.m.), ubicato nella regione di Scoresby Sund, periferica rispetto alla calotta e quindi molto sensibile alle oscillazioni del margine glaciale. Attraverso l’analisi delle polveri minerali provenienti dalle regioni proglaciali, sollevate e trasportate dal vento negli ultimi 120.000 anni, è stato possibile distinguere in modo inequivocabile i periodi in cui la calotta copriva queste aree marginali dai periodi in cui le aree proglaciali erano libere da ghiaccio e sottoposte all’azione dei venti.
La ricostruzione dell’estensione delle calotte glaciali polari nel passato in risposta a cambiamenti climatici repentini o di lungo periodo è di fondamentale importanza per i modelli climatici, in quanto è un parametro fondamentale per la stima del contributo delle calotte all’innalzamento del livello del mare. Solitamente, questo tipo di ricostruzione si basa su diversi parametri, come l’età di esposizione delle rocce nelle aree deglaciate circostanti i ghiacciai oppure la datazione di reperti organici.
La calotta glaciale della Groenlandia è oggi particolarmente sensibile alle variazioni climatiche in atto. Tuttavia la ricostruzione dei limiti della calotta in epoche antecedenti l’Ultimo massimo glaciale – il periodo dell’ultima glaciazione, conclusosi circa 18.000 anni fa, durante il quale si ebbe la massima espansione dei ghiacci – è molto difficile a causa della rimozione e del rimaneggiamento dei materiali ad opera dei ghiacciai in avanzata durante i periodi freddi. Ciò inevitabilmente impedisce di ricostruire il comportamento della calotta prima dell’ultima grande avanzata glaciale e durante le variazioni climatiche repentine verificatesi più di una ventina di volte nell’ultimo periodo glaciale, tra circa 116.000 e 18.000 anni dal presente.
Grazie alla precisa datazione della carota è stato possibile affermare che il margine orientale della calotta era in fase di avanzata tra 113400±400 e 111000±400 anni dal presente, ovvero durante l’inizio dell’ultima era glaciale. Al contrario, era in fase di progressivo ritiro tra 12100±100 e 9000±100 anni dal presente, ovvero durante la prima parte dell’interglaciale in cui viviamo, l’Olocene.
Non sono state però rilevate evidenze significative di cambiamenti al margine della calotta durante l’ultimo periodo glaciale e specialmente in corrispondenza degli eventi di “Dansgaard-Oeschger”, caratterizzati da un riscaldamento climatico repentino seguito da un più graduale ritorno a condizioni glaciali.
Non bisogna dimenticare che la polvere minerale depositata sulla calotta glaciale rappresenta uno dei fattori che contribuisce all’annerimento della neve superficiale (riduzione dell’albedo) con conseguente fusione del ghiaccio e perdita di massa. L’importanza di questo studio è quindi duplice: da un lato ha permesso di stimare i tempi di risposta della calotta nei confronti delle variazioni climatiche naturali avvenute nel passato, dall’altro ha fornito elementi che permetteranno in futuro di quantificare il contributo delle polveri minerali nei processi di “feedback” (retroazione) tipici del sistema climatico.
Accurate estimates of the past extent of the Greenland ice sheet provide critical constraints for ice sheet models used to determine Greenland’s response to climate forcing and contribution to global sea level. Here we use a continuous ice core dust record from the Renland ice cap on the east coast of Greenland to constrain the timing of changes to the ice sheet margin and relative sea level over the last glacial cycle. During the Holocene and the previous interglacial period (Eemian) the dust record was dominated by coarse particles consistent with rock samples from central East Greenland. From the coarse particle concentration record we infer the East Greenland ice sheet margin advanced from 113.4 ± 0.4 to 111.0 ± 0.4 ka BP during the glacial onset and retreated from 12.1 ± 0.1 to 9.0 ± 0.1 ka BP during the last deglaciation. These findings constrain the possible response of the Greenland ice sheet to climate forcings.
Although ice cores are geographical point measurements, they represent a record of air, water and aerosols transported to the ice over regional or even hemispheric scales. In contrast, reconstructions of past ice sheet limits are typically limited to the locations of the individual measurements1,2. These measurements include dating of moraines and subglacial rocks by cosmogenic surface-exposure methods and radiocarbon dating of exposed organic material3. Although East Greenland is mountainous and relatively inaccessible, the deglacial timing and location of the ice sheet margin has been intensively studied, particularly in the locality of Scoresby Sund and Milne Land. It is a challenge to investigate changes in ice sheet extent prior to the LGM, due to the removal and/or reworking of chronological features such as moraines and erratics. Hence there are large dating uncertainties regarding glacial advance after the Eemian4. Ice core dust records may complement this research because ice caps and ice sheets are sensitive recorders of aeolian dust, such as that deflated from glacial outwash plains5, and ice cores typically feature accurate chronologies over multimillennial timescales6.
Records of past dust deposition have been reconstructed from central Greenland ice cores, yet no single record covers the last glacial cycle entirely. Representative dust fluxes for the Holocene have been reported as 24 ± 9 mg m−2 yr−1 from the South Greenland DYE-3 core7,8 and Steffensen9 reported fluxes of 7–11 mg m−2 yr−1 from the central Greenland GRIP core. During the last glacial period, the ice core dust concentration was 10–100 times greater than in the Holocene due to enhanced continental aridity, increased wind strength, lower snow accumulation and longer atmospheric particle lifetime9,10,11. Around 90% of the dust mass in central Greenland during the Holocene and 95% during the glacial comes from particles smaller than 4 μm, as large particles are depleted during transport due to gravitational settling9.
The provenance of dust in Greenland ice cores has been primarily assigned by comparing geochemical parameters with likely dust sources in arid zones of the Northern Hemisphere. Mineralogy and strontium/neodymium isotope ratios12,13 provide the classic means of establishing dust provenance, with central Asian deserts (the Gobi and Taklamakan in particular) providing the best geochemical matches to the dust found in central Greenland during both the Holocene and last glacial. Bory et al.14, also investigated late Holocene ice from two small coastal Greenland ice caps, Renland and Hans Tausen, identifying greater dust fluxes with distinctly less-radiogenic Sr and Nd isotope ratios (i.e. lower εNd, higher 87Sr/86Sr) compared to central Greenland ice cores. Although no representative source was identified, Bory et al. speculated that a potential contributor was the Caledonian fold belt, a formation that comprises most of North and East Greenland14 and for which less-radiogenic Sr isotopic signatures have been reported15.
The dust record of the RECAP ice core was obtained from the Renland ice cap in the Scoresby Sund region of central East Greenland. The RECAP ice core (71.30°N, 26.72°W, 2315 m asl) was drilled in June 2015 less than 2 km from the site of the 1988 Renland ice core16. The surface elevation at the drill site is comparable to the DYE-3 (2490 m asl) and NEEM (2450 m asl) ice cores. The core reaches 584 m to bedrock, and contains a complete climate record back to 120 ka b2k (before 2000 CE) (see Supplementary Note 1, Supplementary Figs. 1–3 for information regarding the time scale). It was drilled into a topographic valley, resulting in a thick and well-resolved Holocene sequence (533 m), a strongly thinned glacial sequence (20 m) and a partially-preserved Eemian sequence (8 m) above 23 m of stratigraphically disturbed ice. The coastal location of the Renland ice cap provides important geographic climate information that can be compared with central Greenland ice cores as well as providing a sensitive indicator of changes at the margins of the Greenland ice sheet.
We use the RECAP large dust particle record (larger than 8 μm) as an indicator of the presence of local dust sources and present new isotope geochemistry data constraining the likely sources of dust found on the Renland ice cap. The RECAP dust record also provides an independent age constraint on changes to local relative sea level and the location of the East Greenland ice sheet margin, which are both important controls on the presence of dust deflation sources through the onset of the glacial and the deglaciation. During both the Holocene and the previous interglacial period (the Eemian) the ice core dust record was dominated by coarse particles likely to originate from Kong Christian X Land in central East Greenland. We infer the East Greenland ice sheet margin advanced from 113.4 ± 0.4 to 111.0 ± 0.4 ka BP during the glacial onset and retreated from 12.1 ± 0.1 to 9.0 ± 0.1 ka BP during the last deglaciation. These findings provide important constraints for ice sheet models used to investigate the sensitivity of the Greenland ice sheet to climate forcing parameters.
RECAP dust record
The RECAP dust record (Fig. 1, Supplementary Fig. 4) confirms previously reported features of central Greenland dust records such as extreme concentration variability during stadial/interstadial cycles. The dust record also displays unusually high concentrations during the interglacials. Dust records from central Greenland ice cores (DYE-3, GRIP, GISP2, NGRIP) consistently feature low concentrations in the Holocene (<102 µg kg−1) and high glacial stadial concentrations (>103 µg kg−1)17. In contrast, the RECAP dust record displays intermediate concentrations during the Holocene (5–11.7 ky b2k, 305 ± 117 µg kg−1, 1σ error bounds) and late Eemian (119.0–120.8 ka b2k, 861 ± 402 µg kg−1, 1σ error bounds) with higher concentrations in the glacial stadials (>103 µg kg−1) and lower concentrations in the glacial interstadials (<200 µg kg−1). We evaluate the glacial and interglacial sections of the RECAP dust record separately below.
Considering the glacial section (11.7–119.0 ka b2k) of the RECAP dust record, we find similar features of the dust record (concentrations, particle size mode, scales of variability) to central Greenland ice cores. RECAP dust concentrations varied by a factor of 10–100 between mild interstadial and cold stadial periods commonly known as Dansgaard-Oeschger events18. These changes are attributed to changes in aridity and dust storm activity in central Asian deserts as well as hemispheric-scale atmospheric circulation patterns19,20. We compare the RECAP dust record to the NGRIP ice core, which is the longest continuous central Greenland dust record available10. We observe a Pearson correlation coefficient of 0.95 ± 0.01 between the log-scaled 100-year mean RECAP and NGRIP glacial dust records. The NGRIP dust concentration is 1.7 ± 0.2 times greater than the RECAP dust concentration. Assuming an identical dust flux to the two sites, the different dust concentrations can be explained by the different snow accumulation rates at NGRIP (19 cm ice equivalent yr−1)21 and RECAP (45 cm ice equivalent yr−1). The accumulation difference alone explains the higher dust concentrations in NGRIP compared to RECAP, without the need to invoke differences in source activity or atmospheric dust transport. The mode of the RECAP glacial dust size distribution, i.e. the particle size contributing most to the total mass (see Methods for calculation of the mode), is also in good agreement with those reported for central Greenland ice cores. The RECAP glacial dust size distribution mode is 2.22 ± 0.02 µm, compared to 1.73 µm for NGRIP (Fig. 2)10. The close similarities of dust concentrations and particle size distributions strongly suggest the glacial dust deposited at RECAP, GRIP, GISP2 and NGRIP originated from a common source18. In the absence of geochemistry data for RECAP glacial dust, we assume that the central Asian dust source determined for other Greenland ice cores12,13 also provided dust to Renland ice cap throughout the glacial.
The RECAP dust record also confirms previous findings of surprisingly high interglacial dust concentrations at coastal Greenland ice core sites compared to central Greenland ice core sites. Bory et al.14, reported elevated late Holocene dust concentrations in coastal Greenland ice cores from Renland (1360 µg kg−1, dated 1604–1662 CE) and Hans Tausen (476 µg kg−1, dated ca. 1000 CE) ice caps, which are similar to those reported here for RECAP interglacial samples. In comparison, dust concentrations in contemporaneous central Greenland ice core samples ranged from 35 to 123 µg kg−1. The elevated interglacial dust concentrations found at Renland and Hans Tausen suggest an alternate or additional source of dust impacts on coastal East and Northeast Greenland but not the central Greenland ice core sites.
Further support for an additional coastal Greenland interglacial dust source is provided by particle size distributions in the RECAP ice core, which reveal a much larger particle mode than those found in central Greenland ice cores. The mode of the RECAP Holocene size distribution is 19.6 ± 1.0 µm, compared to 1.47 µm for NGRIP (Fig. 2). Furthermore, the NGRIP and RECAP glacial dust size distributions are concave with power law tails, whereas the Holocene RECAP size distribution is bimodal and therefore indicative of contributions from two distinct sources. A power law was fitted to the tail of the RECAP Holocene particle size distribution between 4 and 6 µm and the concentration of excess particles in the size range 0.9–2.5 µm was determined. The mode of these small RECAP Holocene particles is 1.88 ± 0.04 µm (Fig. 2, see Methods for details on separating the two distributions), consistent with that of NGRIP early Holocene ice. The large RECAP Holocene particle mode (19.6 ± 1.0 µm) suggests a dust source local to Renland ice cap, as such large particles are rapidly sedimented and typically reside in the atmosphere for less than a day19.
To better characterize changes in large and small particle deposition in the RECAP dust record, we identify two particle size ranges representative of these modes and determine their concentrations over time. The small (1.25–2.9 µm) and large (8.13–10.5 µm) particle size ranges are shown in Fig. 2 and, respectively, correspond to size ranges assigned to ‘distal’ and ‘local’ dust sources in a comparable study22. The record of RECAP large particles (Fig. 1) shows elevated values during interglacials (Holocene 66 ± 28 µg kg−1 and late Eemian 205 ± 170 µg kg−1, durations as previously defined) and low values during the glacial (14 ± 21 µg kg−1). During the Holocene, RECAP large particle concentrations peak (135 µg kg−1) at 7.8 ± 0.1 ka b2k coinciding with the Holocene climatic optimum at Renland23. This coincidence may mark the maximum extent and/or activity of the local East Greenland dust production zone, which is the product of both the retreating ice sheet margin and the lowering relative sea level. Northern Hemisphere insolation was at a maximum during the Holocene climatic optimum, implying a maximum rate of meltwater runoff from the Greenland ice sheet even if the ice sheet continued to lose elevation (and therefore mass) until ~7 ka b2k23. Available reconstructions indicate that the central east Greenland deglacial response in relative sea level change was almost complete by 8 ka b2k24,25,26. This suggests that relative sea level decrease may have had a greater influence than ice margin retreat with respect to the establishment of dust production areas local to Renland ice cap. Three instances of high concentrations of large particles are found in the glacial, and all correspond to the ages of tephra layers previously identified in central Greenland ice cores (26, 55 and 81 ka b2k)27, although geochemical measurements have not yet been undertaken to confirm a volcanic source for these large particles. Otherwise, RECAP large particle concentrations vary by a factor of just 2.6 ± 0.5 throughout the glacial stadials and interstadials (Supplementary Note 2, Supplementary Fig. 5), which is consistent with a factor 2 variability in the snow accumulation rate in central Greenland28,29. If the snow accumulation rate at Renland follows the same pattern, the glacial large particle concentration variations can be explained solely by variations in snow accumulation, suggesting the location of the central East Greenland ice sheet margin did not change significantly through the rapid stadial/interstadial cycles. Furthermore the data suggest a small but constant flux of dust from local sources was transported to Renland ice cap throughout the glacial.