Climate Audit

Hormes et al 2001 is noted up in passing by the NAS panel, but is given short shrift relative to Thompson. It is not mentioned at all in the IPCC 4AR second draft. This article is one of the underpinnings for the view that there have been 8 warm intervals in the Alps during the Holocene. Rather than worrying about these interesting matters, which I may get to on another occasion, I want to glean some information about how glacial sediments are formed in this particular case and why stratigraphy should be mapped. Let me note in passing that stratigraphy is mapped carefully in mining geology with attention paid to both cross-sections and level plans.

Hormes et al 2001 reported on the discovery in the late 1990s of finds in the outwash plain of the Unteraargletchser as follows:

In the summer of 1995, tree stems and disk-like rounded Cyperaceae fen peat fragments were found on the Unteraarglacier’s outwash plain at an altitude of 1920 m asl.; these were clearly eroded from the glacier’s bed and transported by an enormous jàƒ⵫ulhlaup (glacier outburst flood) event in context with the actual retreat…

Here is a map from the article showing the location of the finds. They report other data from outwash plains of several other nearby glaciers.

Figure 1. Location map from Hormes et al 2001, showing location of finds in outwash plain of the Unteraargletscher.

I don’t know much, if anything, about what happens in a jàƒ⵫ulhaup (the word comes from Iceland). They also note that this particular glacier had experienced a decadal advance from 1979-1991, which seems to be related to how the final delivery of the subfossil trees to the outwash plain eventually took place:

As the present glaciers can be considered as the decaying remnants of larger “‹Å“Little Ice Age’ glaciers, at least partial reworking of our samples during that advance is likely. Final delivery of the samples to the glacier forefield, however, occurred only with the 1979–91 readvance and subsequent retreat.

The outwash plain itself is reported to have several terraces with a total width of about 1.7-3.2 m.

The proglacial area of the Unteraargletscher is divided into three terrace levels: 0.5–1 m (level I), 1– 2 m (level II), or .2 m (level III) above the main Aare meltwater stream. The floodplain (level I) is rapidly changing due to the shifting braided river. Level III contains complex moraine and fluvioglacial deposits. Clasts of organic material were flushed out of the glacier by a jàƒ⵫ulhlaup-type flood. Subglacial channels and water bodies are subject to constant changes of a continuous and discontinous nature (plastic deformation of the ice and collapse of drainage systems). Due to drainage processes, water passages are created, widened or destroyed because of variations in hydrostatic water pressure (Iken et al., 1983). As a consequence of such processes, sudden and rapid draining of water from within a glacier may occur (jàƒ⵫ulhlaups or “glacier burst” floods) (Maizels, 1997; Paterson, 1994). From levels I and II, 112 wood and 166 clasts of peat were sampled in October 1995 and 13 wood samples in September 1996. From these 125 wood samples, which were up to 126 cm long and 30 cm in diameter, 36 pieces were radiocarbon dated.

Similar descriptions are provided for samples from the areas of the other glaciers under study. When the wood samples and peat were subjected to radiocarbon testing, they had dates ranging over nearly 10,000 years, which Hormes et al grouped into 8 intervals (although impressionistically the interval definition is not tight.) They compared the dates of these intervals to intervals previously estimated in Scandinavia and reported considerable, but not exact overlap. While the wood samples contained numerous specimens from the Roman Warm Period (and the authors report historical evidence of Roman passes now engulfed by glaciers), they did not report any samples from the Medieval Warm Period – or for that matter from subsequent period, although the LIA advance must obviously have engulfed some wood from a more recent period.

By examining the surface of the wood and peat, they concluded that it had been transported and reworked and was not in situ. They concluded that the wood and peat had grown upslope of the find location and had been transported downslope over the years.

A crucial factor is the original growing location of the subfossil trees and of the peat: did the trees and peat grow where glaciers now fill the basin? This is undoubtedly the case because the reworked peat is glacially compressed and in geological terms we are dealing with a sedimentary basin confined by the Holocene lateral moraine complexes. Peat clasts from Unteraargletscher and Vadrec del Forno point to the existence of in-situ organic material in former proglacial areas which are now occupied by the glaciers (Figure 4). In particular, the steep slopes of the environment are far from suitable habitats for substantial peat growth and suggest that the peat grew in the former proglacial area itself and, therefore, cannot have been fallen onto the glaciers. Radiocarbon dates on wood and peat clasts therefore correspond with the time of glacial recessions and tree growth at higher elevations than at present. The reworked and mechanically deformed condition of the wood samples indicate that glacial readvance did actively override trees and peat. This is supported by palaeovegetational evidence that the tree-line was up to 100–200 m higher than present during the Holocene (Burga, 1988; Tinner et al., 1996).

Following is their Figure 4. If one multiplies this by 8 repetitions with different advances and retreats, one would be getting the idea.

I also presume that (not shown here) there is a continuous downslope movement of the ice at lower levels. In ice cores at summits (e.g. Quelccaya, Kilimanjaro), the observed thickness of the annual layers decreases more or less in an negative exponential. In order for this to occur, a percentage of each layer has to be "extruded" downslope every year. I presume that this would be happening throughout the glacier so that, in equilibrium, the annual accumulation over the entire glacier (net of surface sublimation) would be balanced by a type of "undertow" delivery to the downslope tongue. (I’ve not consulted texts on this, abd it’s probably explained more clearly in glacier texts, but this mechanism seems to be required to explain the negative exponential annual layers at the summit.)

Here is my understanding of what’s happening with stratigraphy in these small alpine glaciers, based on reading a few articles and not on independent expertise. It appears that glaciers advance and recede on a variety of scales, presumably ranging from annual to millennial. What makes interpretation rather difficult is that the most recent glacier surge (the Little Ice Age) surge was the largest since the last Ice Age. Moraines (consisting of till bulldozed by the glaciers) from the LIA advance are readily discerned on surface photographs. Unfortunately for people attempting to recover the past history, the most recent advance bulldozed older tills (say from a first millennium advance) which typically no longer exist. As a type of image, think of the advances as being types of wave breaks on a beach, with the most recent LIA advance being a big wave that wiped out sand ripples from earlier wave breaks. It’s not a precise analogy, but it puts one in an appropriate frame of mind, I think.

One can picture a lot of different possibilities in how tree stems, even in situ stems, could be exposed. For example, in Alberta, the stratigraphical evidence is that the recently exposed 2900-year old in situ stumps were a bed of about 5 m thickness, presumably some sort of paleosol, but we don’t have any cross sections or level plans. The warmth in the Holocene Optimum was clearly more than in the MWP. Could one have had a situation where the 2900-year old trees were destroyed (but not uprooted) by the 2900-year advance; they were covered up by paleosol in a subsequent retreat, but without trees re-advancing to that elevation in the Roman WP or MWP. Then the big LIA glacier advance re-arranged everything leaving it possible for the trees to be re-exposed following retreat of the LIA glacier.

I don’t think that the present evidence in Alberta enables one to say that the glacier did not retreat to its present position during the RWP or MWP – I would say that that question is open- all one can say right now is that it’s likely that the treeline did not advance to the 3000-year old treeline during either the RWP or MWP. Also neither the treeline nor the glacier are in equilibrium with the present climate and there is both more glacier retreat and more treeline advance in inventory even with no further temperature increase. It would be interesting to see what the present estimates are for what the treelines and glacier lines would be under present conditions.

If we go from these two cases to Quelccaya, what do we know about finds there? All we know is that they found a 5000-year old fossil plant. What has been reported about the stratigraphy of the find? Nothing. Is this the only organic discovered at Quelccaya? Is it the only one that’s dated? Both of these possibilities seem pretty unlikely. When one sees the complications of the stratigraphy at other alpine glacier sites that have been reported in greater detail (although still hardly adequately), what can we conclude from the isolated Quelccaya find? Can we conclude that the glacier has retreated further than at any point for 5000 years? In the other two cases, this is not necessarily the case. So how can anyone draw any conclusions from a find which has been so poorly reported?

References:
Hormes, A., MàƒÆ’à‚⻬ler, B.U. and Ch. SchlàƒÆ’à‚⻣hter (2001): The Alps with little ice: evidence for eight Holocene phases of reduced glacier extent in the Central Swiss Alps.- The Holocene 11/3: 255-65. url
Anne Hormes website