Model–data comparison for the 8.2 ka BP event: confirmation of a forcing mechanism by catastrophic drainage of Laurentide Lakes
Introduction
During the Holocene, the coldest period in the North Atlantic region occurred around 8.2 cal. ka BP and lasted for ∼300 yr. In Greenland ice-core records, this 8.2 ka BP event is characterized by an ∼7.4 °C reduction in temperature (Leuenberger et al., 1999), a decrease in ice accumulation rate, increasing wind speeds and a drop in atmospheric methane levels (e.g. Alley et al., 1997; Spahni et al., 2003). Proxy records containing evidence for this event, ranging from the North Atlantic to monsoonal domains, suggest at least a semi-global impact of the event (e.g. Gasse and Van Campo, 1994; Hughen et al., 1996; Klitgaard-Kristensen et al., 1998).
A slowing down of the ocean thermohaline circulation (THC) as a result of a freshwater perturbation has been proposed as the cause for the event (e.g. Alley et al., 1997; Barber et al., 1999). The slowdown resulted in a decrease of the northward heat transport by the North Atlantic Ocean, leading to a pronounced cooling in the region. Most likely, the huge proglacial Laurentide Lakes (Lake Agassiz and Ojibway) in front of the Laurentide Ice Sheet (LIS) were the source of this freshwater (e.g. Leverington et al., 2002; Teller et al., 2002; Clarke et al., 2004), which drained into the Hudson Bay when the LIS disintegrated (e.g. Vincent and Hardy, 1979; Veillette, 1994; Barber et al., 1999). Model studies support the theory that a freshwater perturbation in the North Atlantic can slow down the THC (e.g. Rooth, 1982; Broecker et al., 1985, Broecker et al., 1988, Broecker et al., 1989; Stocker and Wright, 1991; Manabe and Stouffer, 1995, Manabe and Stouffer, 1997; Rind et al., 2001; Vellinga and Wood, 2002).
Other studies, however, have pointed to the concurrence of a reduction in solar irradiance at 8.3 ka BP and the 8.2 ka BP event (Bond et al., 2001; van Geel et al., 2003), arguing that changes in solar activity could have triggered the observed climate changes, most likely involving changes in ocean circulation. Muscheler et al. (2004) investigated the change in Δ14C around the 8.2 ka BP event by comparing them to changes in the 10Be flux, and concluded that there is no convincing evidence that solar forcing has caused the 8.2 ka BP event. However, they speculate that a period of decreasing solar forcing at the start of the 8.2 ka BP event could have been involved in triggering the climatic changes, but was probably not the main cause.
A detailed understanding of the 8.2 ka BP event is important, as it may provide us with information on the response of the climate system to a disturbance of the THC under interglacial conditions. This is essential because most climate models suggest that a disturbance of the THC is likely to occur in greenhouse climate scenarios, triggered by an increased freshwater flux (more precipitation and runoff) and warming of the surface ocean which both lead to a reduction of the surface water density (Cubasch et al., 2001).
To study the response of the early Holocene climate to a freshwater pulse from the Laurentide Lakes, Renssen et al., 2001, Renssen et al., 2002 performed several freshwater perturbation experiments using a global atmosphere–sea-ice–ocean model. In these experiments, solar irradiance was kept constant. A freshwater pulse equivalent to the estimated volume of the Laurentide Lakes draining in 20 yr produced a weakening of the THC for ∼300 yr and showed general agreement between simulation results and proxy evidence for the 8.2 ka BP event. However, a detailed comparison of these model results with proxy evidence has not yet been carried out, and could provide additional information on the 8.2 ka BP event.
A model–data comparison could shed light on several important questions concerning the event that are still insufficiently answered:
- 1.
What is the geographical distribution of the event?
- 2.
How is the event expressed at different geographical locations?
- 3.
What are the underlying mechanisms, which explain the observed expression?
Answering these questions is important because it helps identify the forcing mechanism for the event. A predominantly circum-North Atlantic distribution of the event would support a THC weakening as the cause (Rind and Overpeck, 1993). A global cooling, on the other hand, may point to the involvement of solar forcing. Thus, by looking at the geographical distribution, possible causes can be excluded. As well as identifying the forcing mechanism, it is important to look at the expression of the event, i.e. the types of climatic response (cooling, drought, windy), their magnitude and seasonal expressions. This gives information on the climatological processes involved in the spreading of the event. Finally, information on the timing and teleconnections provide us with insight in the climatological processes involved in a disturbance of the THC during interglacial climatic conditions.
To address these questions, we perform an extensive model–data comparison, utilizing the simulation results of Renssen et al., 2001, Renssen et al., 2002 and published proxy evidence around the globe. We also compare transient simulation results with two high-quality records. This method gives us the opportunity to compare observed and simulated evolution of the event at different locations. Both the geographical distribution and evolution of the event allow identification of the forcing mechanism (Rind and Overpeck, 1993). A general agreement between simulation results and the proxy data would support a freshwater-induced THC weakening as a cause for the event, as no other forcing is applied in the experiments of Renssen et al. (2002).
Section snippets
Model and experiment
The numerical simulation experiments were performed with the ECBilt–CLIO three-dimensional atmosphere–sea-ice–ocean model (Goosse et al., 2001). The atmospheric part is version 2 of ECBilt, a spectral T21, three level quasi-geostrophic model described in detail by Opsteegh et al. (1998). The sea-ice–ocean component is the CLIO model, consisting of a primitive-equation, free-surface ocean general circulation model (OGCM) coupled to a comprehensive thermodynamic–dynamic sea-ice model (Goosse and
Proxy evidence
In this section, we discuss the published proxy data that record an event around 8.2 ka BP. The records are discussed for specific regions, starting with evidence from Greenland. After the regional discussions, a comparison is made with the modeled temperatures and hydrological response. The different locations and the inferred paleoclimatic responses in proxy-data and simulation results are summarized in Table 1. All dates are in calendar years BP, unless otherwise indicated.
Geographical distribution and expression of data
If we look at the geographical distribution pattern of the locations where a change in climate during the 8.2 ka BP event can be observed (Fig. 7), we see a clear concentration of records in the Northern Hemisphere. Moreover, records reflecting a cooling are mainly concentrated in the circum-North Atlantic area, although no records from the western North Atlantic Ocean have recorded the event. The North Atlantic region has a climate that is directly influenced by the North Atlantic surface water
Conclusions
The observed expression and distribution of the 8.2 ka BP event in proxy data is captured reasonably well by the model. Cooling mainly occurs in Europe, Greenland, North America and the eastern North Atlantic Ocean, while reductions in precipitation are mainly observed in Europe, Greenland, North Africa and East Asia. Moreover, the general agreement between the high-resolution records and the simulation results suggests that the forcing applied in the model and in the ‘real-world’ were similar.
Acknowledgements
We thank Jim Teller and an anonymous reviewer for their useful comments, which helped to improve the paper. Both authors are sponsored by the Netherlands Organization for Scientific Research. Alex Wright is thanked for checking English writing and Dick Kroon for his useful comments. This research has benefitted from the ESF/HOLIVAR program that provided financial support to attend the HOLIVAR training course held at UCL in London.
References (160)
- et al.
Holocene environmental variability—the record from Lago Grande di Monticchio, Italy
Quaternary International
(2002) The Younger Dryas cold interval as viewed from Central Greenland
Quaternary Science Reviews
(2000)- et al.
Early to mid-Holocene Atlantic water influx and deglacial meltwater events, Beaufort Sea slope, Arctic Ocean
Quaternary Research
(2004) - et al.
Final stages in the collapse of the Laurentide Ice Sheet, Hudson Strait, Canada, NWT: 14C AMS dates, seismic stratigraphy, and magnetic susceptibility logs
Quaternary Science Reviews
(1995) - et al.
Palaeoclimate and the formation of sapropel S1: inferences from Late Quaternary lacustrine and marine sequences in the central Mediterranean region
Palaeogeography, Palaeoclimatology, Palaeoecology
(2000) - et al.
The North Atlantic atmosphere–sea surface 14C gradient during the Younger Dryas climatic event
Earth and Planetary Science Letters
(1994) - et al.
The Eastern Mediterranean paleoclimate as a reflection of regional events: Soreq Cave, Israel
Earth and Planetary Science Letters
(1999) - et al.
High resolution paleoceanography of the Guaymas Basin, Gulf of California, during the past 15000 years
Marine Micropaleontology
(2004) - et al.
High resolution UK37 sea surface temperature reconstruction in the Norwegian Sea during the Holocene
Quaternary Science Reviews
(2002) Patterns and processes of Late Quaternary environmental change in a montane region of southwestern Europe
Quaternary Science Reviews
(2002)