Glacierclimate relationships Little Ice Age LIA scale

'The term "Little Ice Age" relates to the behaviour of the glaciers not directly to climate' (Grove, 2001, p. 76; after Luckman). What was once a well accepted, if poorly understood, interval of Earth history is now in debate, as is its companion, the Medieval Warm Period (MWP) (Bradley & Jones, 1993; Hughes & Diaz, 1994; Broecker, 2001). Instrumental data for climatic variables do not exist prior to the 17th century, therefore the atmospheric and ocean climates (Figs 21.1 & 21.2) have to be inferred from a variety of proxy data. With few exceptions the evidence for glacial response in this time period is no longer couched in terms of mass or energy balances but usually in climatic terms (Fig. 21.2). Within the LIA period, or approximately the past 600yr, the primary record of glacial response to climatic perturbations are end moraines that are located down-valley from present-day glaciers. It appears from the literature (Grove, 2001) that the LIA was underway by the 13th and 14th centuries. The boundary conditions for LIA glacier-climate systems are simpler than the response and interactions on longer time-scales (Fig. 21.6a versus 6c). On the longer time-scales, even during the Holocene, the numbers of tidewater glaciers and marine-based margins with their particular dynamics are more evident. On LIA time-scales the dynamics of change rest largely within the atmospheric climate system, and includes the oceans (Fig. 21.6a). On these time-scales tectonic movements, including glacial isostatic

(1) insolation = present

(2) Direction of glacial isostatic recovery

(3) Direction of relative sea level

(2) Direction of glacial isostatic recovery

(3) Direction of relative sea level

(b) "LIA" Holocene glacial events


(1) insolation = > & <present bn >0 (Fig 9)

(1) insolation = > & <present bn >0 (Fig 9)

Figure 21.6 Sketch of ice sheet with terrestrial and marine margins and controls on (a) Little Ice Age (LIA) and (b) last glacial cycle.

depression or recovery, are small, probably exist (Clark, 1977), but do not force a glacier response. Changes in relative sea level (RSL) are also small. We will consider the LIA history and neoglaciation of Iceland as representing a climate system that should integrate a substantial hemispheric signal because of its location in relationship to the NAO, its proximity to sites of deep sea convection in the Iceland and Greenland Seas (Fig. 21.3) (Broecker, 1997; Malmberg & Jonsson, 1997; Delworth & Dixon, 2001), and the known effect of sea ice on the climate (Fig. 21.4A & B) (Bjorns-son, 1969; Chapman & Walsh, 1993; Deser et al., 2000; Ogilvie & Jonsson, 2001). Barlow (2001) noted that the Greenland isotopic records for the period ad 1400-1980 might be negatively correlated with European temperature anomalies because of the climate 'see-saw' (Rogers & van Loon, 1979).

Interactions between glacial history and the surrounding land/oceans has been evaluated by Mysak & Power (1992) in terms of the feed-backs associated with the GSA of the late 1960s (Fig. 21.4A) and on a more regional scale by Stotter et al. (1999) for North Iceland (Fig. 21.4B). Although there is some dispute on the origin of the salinity anomaly in the Nordic Seas, what is not in dispute is that it probably originated 1000s of kilometres away within the borders of the Arctic Ocean, either as an excursion in the flow of the McKenzie River or by the increased transport of sea ice through Fran Strait associated with atmospheric (wind) processes (Mysak & Power, 1992; Serreze et al., 1992). An additional forcing mechanism, and one not explicitly considered by these authors, is the influence of changes in solar activity on the climate, a topic which has received considerable attention or re-emphasis in the last decade or so (Harvey, 1980; Stuiver et al., 1991; Bond et al., 2001; Shindell et al., 2001; Lean et al., 2002).

The Mysak & Power's (1992) negative feed-back loop indicates that there is some level of self-regulation in at least part of the climate system (Fig. 21.4A). Hydrological changes on the Canadian Arctic margin are considered to propagate into the Greenland and Iceland Seas and cause a freshening of the surface waters, which reduces cyclogenesis, thus decreasing precipitation and reducing the freshwater anomaly. An additional box has been added to this model to interface with longer Holocene time-scales. On longer time-scales, changes in bathymetry of the key Canadian Arctic Channel sills are associated with glacial isostatic recovery and changes in the global ice volume (Figs 21.4A & 21.6b).

Changes in sea ice have an impact on the temperature and precipitation over north Iceland (Ogilvie, 1997) so that an increase in sea ice is associated with a decrease in temperature, a decrease in precipitation, a fall in the ELA, and an increase in glacier extent (Fig.21.4B) (Stotter et al., 1999). Although there are no long-term mass balance measurements on northern Iceland, the sense of this association is that the decrease in temperature, associated with the increase in sea ice offthe coast, is more important than the decrease in precipitation (Fig. 21.4B). This assertion does not appear to explain the trends of the past 40yr, however (Fig. 21.5D).

In Vestfirdir (northwest Iceland) there are numerous semipermanent snowbanks and evidence for LIA and neoglacial moraine-building intervals (Eythorsson, 1935; Thorarinsson, 1953). The Dragnajokull Ice Cap lies on the Tertiary basalt surface with an ELA around 570m asl. However, the hypsometry of Vest-firdir is such that an ELA lowering of only 100 m would increase the glacierization of the area by 50% (Principato, 2003). The ages of LIA moraines on Iceland indicate major intervals of moraine formation over the past three centuries (Grove, 2001; Wastl et al., 2001) although there is also evidence for a glacial moraine deposition around 700yr BP.

Figure 21.2 represents a conceptual view of the glacier-climate system. It is useful in terms of present-day interactions but how well can we move backwards and infer climate from the glacial record? Figure 21.7 illustrates some of the available data, which can place glacial response on Iceland into the context of hemispheric, regional and local changes of climate. It shows the estimated hemispheric temperature anomalies for the past 600 yr (Mann et al., 1999), the changes in sea ice off Iceland since the Settlement in ca. ad 870 (Bergthorsson, 1969; Bjornsson, 1969; Ogilvie, 1991, 1992; Ogilvie et al., 2000), and inferred changes in seafloor temperature from a small North Iceland fjord (Andrews et al., 2001; Castaneda et al., 2004) (Figs 21.3 & 21.7B). Mann et al. (1999) (Fig. 21.7A) examined which climatic forcings were most important in determining climate change prior to human intervention (but see Ruddiman, 2003a). Variations in solar activ ity were most strongly associated with the 600yr record (Lean et al., 1995).

For the past 600 yr (Fig. 21.7A) the hemispheric picture is dominated by temperatures below the mean of the reference series. A pronounced cold interval appears around ad 1450 and other prolonged cooler intervals are centred around ad 1600, 1700 and between ca. 1800 and 1900. How far does climate data from Iceland mimic this record? The data from Stykkisholmur, northwest Iceland (Fig. 21.3), extends back to ad 1820. The winter (J, F, M) and summer (J, J, A) trends (Fig. 21.7C) indicate an interval of both cold winter and summer temperatures centred on ad 1860 followed by a rather dramatic increase in temperatures, led by winter values, starting ca. ad 1900. The correlation between winter and the subsequent summer temperatures is quite evident (r2 = 0.48); the correlation between the Mann et al. (1999) series and Stykkisholmur is only r2 = 0.14.

A reconstruction of changes in the Iceland Low pressure system has been presented on the basis of the Greenland ice-core data (Meeker & Maywekski, 2002). This reconstruction suggests that a fundamental mode change occurred around ad 1400 when overall high-pressure conditions changed abruptly to lower pressures, suggesting an intensification of the Iceland Low (Fig. 21.3). Based on the recent data from Stykkisholmur (Fig. 21.5) this might imply that, starting between ad 1400 and 1450, winter accumulation would have increased over Iceland.

In attempting to extend our knowledge of Iceland climate and glacial variations over the past millennium a critical question is how far changes in the offshore sea temperatures are correlated with temperatures on land, and with glacier response (Sigurdsson & Jonsson, 1995)? Iceland is only ca. 100,000km2 and studies of climatic data show a very strong correlation between weather stations across the island (Einarsson, 1991). The 50-yr time series from the Siglunes hydrographic CDT data, off north Iceland (Fig. 21.3) (Olafsson, 1999), shows a significant correlation with the Stykkisholmur MAT data (r2 = 0.55) (Fig. 21.7D), hence it is reasonable to postulate that changes in the marine realm can be used to hindcast temperature conditions on land.

A proxy for marine temperatures in the late Holocene is the stable S18O of foraminifera from marine cores, as it has been shown (Smith et al., 2005) that these primarily reflect changes in temperature. Figure 21.7B shows a reconstruction of sea-ice severity off Iceland versus the S18O record in core B997-328PC from the head of a small fjord in Vestfirdir (Fig. 21.3) (Castaneda et al., 2004). The agreement is substantial since the Settlement ca. ad 870. There was a A818O change of ca. 0.6%o between the MWP (Hughes & Diaz, 1994) and the LIA. This corresponds to a temperature change of ca.3°C, which is within the temperature variations caused by the GSA of the late 1960s when water-column temperatures fell 5°C (Malmberg, 1985). Adjustment in the chronology of B997-328, given the errors in 14C dating, would easily result in an even stronger association. The S18O results indicate a prolonged interval of cold marine conditions between ad 1500 (500yr BP) and ca. ad 1900.

0 0

Post a comment