Final Report Summary - CLIMICE (Late Holocene climate and sea ice variability in the southwestern Labrador Sea)
Project context and objectives
Inspired by the current debate on the contributions of natural processes and anthropogenic emissions to the present climate warming and Arctic sea-ice decline, this project aimed to reconstruct late Holocene changes in sea-ice variability and regional sea surface temperatures (SST) in the SW Labrador Sea using high-resolution marine sediment records. The records were also expected to provide information on the past variability of ocean currents in the study area and to contribute to a better understanding of Arctic sea-ice export by the Baffin-Labrador Current (BLC). The study area is climatologically crucial, since melt water from sea ice and icebergs entrained by the Labrador Current directly affects the formation of North Atlantic Deep Water, which drives the redistribution of heat between the hemispheres. We proposed to use an innovative multi-proxy approach using established (diatoms, alkenones), developing (d18O from diatom silica) and highly novel (IP25) sedimentary proxies.
We analysed two short cores south-east (SE) and north-east (NE) of Newfoundland covering the last ca. 100 years in order to, especially, assess the novel geochemical sea-ice proxy IP25 against observational data (monitored sea-ice concentrations, air temperature and SSTs) and against an established sea-ice proxy (abundance of sea-ice diatoms). The short cores were also used to validate the alkenone-derived SSTs against measured SSTs in the area. Focusing on the NE side of Newfoundland - the site most affected by sea ice and fresh water transport from the Arctic via the BLC - we then analysed a ca. 5-metre long core at a high resolution for alkenones (SST, ocean current variability), IP25 (sea ice) and diatoms (qualitative sea surface temperature and sea-ice proxy, as well as a proxy for ocean current variability). In addition we decided to analyse a variety of sterols, indicative of the main algal groups, in order to assess the community structure of primary producers, the changes in productivity over time and the potential effect sea-ice exerts on these. The short core sterol results have already been further analysed, whereas the interpretation of the long-core results is still ongoing. A five-metre core from SE Newfoundland was only analysed for alkenones (SST). We proposed to carry out experimental analyses of d18O from diatom silica as an additional SST proxy. However, based on the clay mineral analysis of the sediment material, mineral mixtures including biogenic opal were present. A separation of opal or individual clay minerals for determining oxygen isotope analyses was deemed not to be possible from these muddy sediments.
IP25 has, to date, been used in a few studies to produce sea-ice reconstructions in the Canadian Arctic, the Fram Strait and North Icelandic Shelf area over the last millennia. These records have then been compared to other climate-related proxy-data. However, studies enabling the comparison with modern observational data are still lacking. Such studies are crucial for the further establishment of this highly promising proxy. In addition, IP25 has never been analysed this far south; all previous IP25 records originate from the High Arctic. The results from our short cores show the following;
- significantly higher IP25 concentrations NE of Newfoundland, which is consistent with modern sea-ice observations;
- a statistically significant correlation between IP25 and observed sea-ice concentrations;
- an overall agreement between our proxy data and the measured regional air temperature data and the large-scale atmospheric pressure systems (North Atlantic Oscillation - NAO - and Northern Annular Mode - NAM), with a positive NAO/NAM indicating strong NW winds, colder conditions and more sea ice in the area.
The trends, both in instrumental data and IP25, indicate colder decades during the early 20th century, a warmer period starting in the 1920s and culminating in the 1960s, with even colder winters than in the early 20th century and more sea ice prevalent from the mid-1980s to the mid- 1990s. Interestingly, although the sea-ice diatom records from both short cores show the same overall trends - clearly more sea-ice diatoms on the NE site of Newfoundland - their abundance over the last 100 years show little trend, and do not correlate with observed temporal changes in sea-ice concentrations, surface air temperatures or the NAO/NAM indices. It seems that in this particular environmental setting, IP25 is a more sensitive measure of past sea ice. These results lend support to the validity of our late-Holocene IP25 record. The alkenone-based sea surface temperature reconstruction agrees with the IP25 data and shows a strong cooling trend in the late 1980s. Around the same time the sterol record suggests increased primary production, which is likely to be caused by the increased duration of sea ice in the study area, affecting the nutrient cycling and the intensity of the spring bloom.
The sea ice observed around Newfoundland is predominantly advected from the Arctic - mostly from the Baffin and Hudson Bays. Hence the long-term record from NE of Newfoundland can be seen as a record of Arctic sea-ice export to the Labrador Sea. Our IP25-based reconstruction of sea-ice export shows increasing sea-ice concentrations starting from ca. 2 700 years ago in concert with the neo-glacial cooling, a clear increase from ca. 1 500 years ago, culminating in the Little Ice Age (LIA) around 1 750 AD. Present sea-ice concentrations are clearly lower corresponding to sea-ice conditions during the Medieval Climate Anomaly (MCA). Our sea-ice reconstruction agrees well with other IP25 records from the High Arctic, which all show a distinct increase in sea ice during the last ca. 3 000 years with lower sea-ice concentrations evident in most of these records during recent times. The ice-diatom record shows similar trends to the biomarker-based reconstruction, but is again much more muted, thus underlining the sensitivity of IP25.
The alkenone-based SST reconstruction from NE of Newfoundland, the site more sensitive to changes in the BLC, indicates decadal-scale variability and a distinct colder interval between ca. 1000-1350 AD, loosely concurrent with the MCA. During the LIA, SSTs are generally somewhat warmer with intermittent colder episodes, whereas the SSTs in the 20th century show an overall slight decreasing trend with lowest temperatures occurring in the 1990s. The SST history offshore of SE Newfoundland shows a different pattern, with decreasing SSTs until ~1300 AD, after which the temperatures show little change. These results suggest a clear link between the Baffin-Labrador Current strength and the North Atlantic Oscillation/Northern Annular Mode, supporting a more persistent positive NAO/NAM system and stronger BLC during the Medieval Climate Anomaly, whereas more persistent negative NAO/NAM conditions and a weaker BLC prevailed during the Little Ice Age. The suggestion of a stronger BLC during the MCA is supported by our clay mineral analysis, which revealed lower clay content and hence coarser sediment, indicating increased current activity. Also the diatom data corroborates this finding, as species associated with the BLC increase at the time. While the NE Newfoundland oceanic variability appears to be primarily controlled by overall shifts in NAO/NAM, the hydrographical changes observed off SE Newfoundland in contrast reflect propagated (sub)tropical ocean signals embedded in the Gulf Stream.
Main results
Our results suggest that the late-Holocene decrease in summer insolation (neo-glacial cooling) has an overriding influence on the long-term sea-ice export from the Arctic via the Baffin-Labrador Current, with an observable recent reduction concurrent with rising 20th century temperatures. The SSTs on the other hand provide a record of changes in the atmospheric pressure systems over the North-Atlantic and their effects on the strength and position of surface ocean currents.
We have initiated collaboration with ocean modellers, to fine-tune existing sea-ice and ocean models using our high-resolution sea-ice, SST and ocean circulation data from the NW Labrador Sea, and to use the improved models to produce future scenarios of NW Atlantic Ocean and ultimately climate trends. The outcome of these exercises will be made available both to the scientific community and to environmental policymakers. Our work will feed into programmes and projects run by well-established international scientific organisations involved in Arctic and palaeo-environmental reconstruction research such as PAGES, which is responsible for assimilating climate proxy data for use in global climate reconstruction exercises like the Intergovernmental Panel on Climate Change (IPCC) report, and is the primary source of information on the science of climate change for policymakers in governments worldwide.
Inspired by the current debate on the contributions of natural processes and anthropogenic emissions to the present climate warming and Arctic sea-ice decline, this project aimed to reconstruct late Holocene changes in sea-ice variability and regional sea surface temperatures (SST) in the SW Labrador Sea using high-resolution marine sediment records. The records were also expected to provide information on the past variability of ocean currents in the study area and to contribute to a better understanding of Arctic sea-ice export by the Baffin-Labrador Current (BLC). The study area is climatologically crucial, since melt water from sea ice and icebergs entrained by the Labrador Current directly affects the formation of North Atlantic Deep Water, which drives the redistribution of heat between the hemispheres. We proposed to use an innovative multi-proxy approach using established (diatoms, alkenones), developing (d18O from diatom silica) and highly novel (IP25) sedimentary proxies.
We analysed two short cores south-east (SE) and north-east (NE) of Newfoundland covering the last ca. 100 years in order to, especially, assess the novel geochemical sea-ice proxy IP25 against observational data (monitored sea-ice concentrations, air temperature and SSTs) and against an established sea-ice proxy (abundance of sea-ice diatoms). The short cores were also used to validate the alkenone-derived SSTs against measured SSTs in the area. Focusing on the NE side of Newfoundland - the site most affected by sea ice and fresh water transport from the Arctic via the BLC - we then analysed a ca. 5-metre long core at a high resolution for alkenones (SST, ocean current variability), IP25 (sea ice) and diatoms (qualitative sea surface temperature and sea-ice proxy, as well as a proxy for ocean current variability). In addition we decided to analyse a variety of sterols, indicative of the main algal groups, in order to assess the community structure of primary producers, the changes in productivity over time and the potential effect sea-ice exerts on these. The short core sterol results have already been further analysed, whereas the interpretation of the long-core results is still ongoing. A five-metre core from SE Newfoundland was only analysed for alkenones (SST). We proposed to carry out experimental analyses of d18O from diatom silica as an additional SST proxy. However, based on the clay mineral analysis of the sediment material, mineral mixtures including biogenic opal were present. A separation of opal or individual clay minerals for determining oxygen isotope analyses was deemed not to be possible from these muddy sediments.
IP25 has, to date, been used in a few studies to produce sea-ice reconstructions in the Canadian Arctic, the Fram Strait and North Icelandic Shelf area over the last millennia. These records have then been compared to other climate-related proxy-data. However, studies enabling the comparison with modern observational data are still lacking. Such studies are crucial for the further establishment of this highly promising proxy. In addition, IP25 has never been analysed this far south; all previous IP25 records originate from the High Arctic. The results from our short cores show the following;
- significantly higher IP25 concentrations NE of Newfoundland, which is consistent with modern sea-ice observations;
- a statistically significant correlation between IP25 and observed sea-ice concentrations;
- an overall agreement between our proxy data and the measured regional air temperature data and the large-scale atmospheric pressure systems (North Atlantic Oscillation - NAO - and Northern Annular Mode - NAM), with a positive NAO/NAM indicating strong NW winds, colder conditions and more sea ice in the area.
The trends, both in instrumental data and IP25, indicate colder decades during the early 20th century, a warmer period starting in the 1920s and culminating in the 1960s, with even colder winters than in the early 20th century and more sea ice prevalent from the mid-1980s to the mid- 1990s. Interestingly, although the sea-ice diatom records from both short cores show the same overall trends - clearly more sea-ice diatoms on the NE site of Newfoundland - their abundance over the last 100 years show little trend, and do not correlate with observed temporal changes in sea-ice concentrations, surface air temperatures or the NAO/NAM indices. It seems that in this particular environmental setting, IP25 is a more sensitive measure of past sea ice. These results lend support to the validity of our late-Holocene IP25 record. The alkenone-based sea surface temperature reconstruction agrees with the IP25 data and shows a strong cooling trend in the late 1980s. Around the same time the sterol record suggests increased primary production, which is likely to be caused by the increased duration of sea ice in the study area, affecting the nutrient cycling and the intensity of the spring bloom.
The sea ice observed around Newfoundland is predominantly advected from the Arctic - mostly from the Baffin and Hudson Bays. Hence the long-term record from NE of Newfoundland can be seen as a record of Arctic sea-ice export to the Labrador Sea. Our IP25-based reconstruction of sea-ice export shows increasing sea-ice concentrations starting from ca. 2 700 years ago in concert with the neo-glacial cooling, a clear increase from ca. 1 500 years ago, culminating in the Little Ice Age (LIA) around 1 750 AD. Present sea-ice concentrations are clearly lower corresponding to sea-ice conditions during the Medieval Climate Anomaly (MCA). Our sea-ice reconstruction agrees well with other IP25 records from the High Arctic, which all show a distinct increase in sea ice during the last ca. 3 000 years with lower sea-ice concentrations evident in most of these records during recent times. The ice-diatom record shows similar trends to the biomarker-based reconstruction, but is again much more muted, thus underlining the sensitivity of IP25.
The alkenone-based SST reconstruction from NE of Newfoundland, the site more sensitive to changes in the BLC, indicates decadal-scale variability and a distinct colder interval between ca. 1000-1350 AD, loosely concurrent with the MCA. During the LIA, SSTs are generally somewhat warmer with intermittent colder episodes, whereas the SSTs in the 20th century show an overall slight decreasing trend with lowest temperatures occurring in the 1990s. The SST history offshore of SE Newfoundland shows a different pattern, with decreasing SSTs until ~1300 AD, after which the temperatures show little change. These results suggest a clear link between the Baffin-Labrador Current strength and the North Atlantic Oscillation/Northern Annular Mode, supporting a more persistent positive NAO/NAM system and stronger BLC during the Medieval Climate Anomaly, whereas more persistent negative NAO/NAM conditions and a weaker BLC prevailed during the Little Ice Age. The suggestion of a stronger BLC during the MCA is supported by our clay mineral analysis, which revealed lower clay content and hence coarser sediment, indicating increased current activity. Also the diatom data corroborates this finding, as species associated with the BLC increase at the time. While the NE Newfoundland oceanic variability appears to be primarily controlled by overall shifts in NAO/NAM, the hydrographical changes observed off SE Newfoundland in contrast reflect propagated (sub)tropical ocean signals embedded in the Gulf Stream.
Main results
Our results suggest that the late-Holocene decrease in summer insolation (neo-glacial cooling) has an overriding influence on the long-term sea-ice export from the Arctic via the Baffin-Labrador Current, with an observable recent reduction concurrent with rising 20th century temperatures. The SSTs on the other hand provide a record of changes in the atmospheric pressure systems over the North-Atlantic and their effects on the strength and position of surface ocean currents.
We have initiated collaboration with ocean modellers, to fine-tune existing sea-ice and ocean models using our high-resolution sea-ice, SST and ocean circulation data from the NW Labrador Sea, and to use the improved models to produce future scenarios of NW Atlantic Ocean and ultimately climate trends. The outcome of these exercises will be made available both to the scientific community and to environmental policymakers. Our work will feed into programmes and projects run by well-established international scientific organisations involved in Arctic and palaeo-environmental reconstruction research such as PAGES, which is responsible for assimilating climate proxy data for use in global climate reconstruction exercises like the Intergovernmental Panel on Climate Change (IPCC) report, and is the primary source of information on the science of climate change for policymakers in governments worldwide.