Using volcanic ash in climate reconstruction

Tephra is another name for volcanic ash. With a powerful volcanic eruption, tephra can travel large distances and cover huge areas. For my project, the important idea was that a blanket of tephra covered a large area simultaneously. If it is possible to identify the same volcanic ash in two sets of old soil from different locations, then it is possible to say that the two sets of soil were created at the same time as the tephra.

Tephra is a term used to describe solid volcanic ejecta and it includes materials of varying composition and structure. Tephrochronology is a geochronological technique based on the location and identification of chemically distinct ash layers within a sediment sequence. The dominant control on the chemical composition and morphology of fresh ejecta is the chemistry of the source magma, although eruptive processes also influence the final chemistry and morphology. Tephrostratigraphy is a technique which exploits tephra layers as means of stratigraphical correlation. Tephrochronology is an extension of this technique, whereby dates determined for the source event are applied to distal tephra deposits, permitting the layers to be used as independent chronostratigraphical horizons. Broadly speaking, there are four steps in the application of tephrochronology.

The first step is to determine whether tephrochronology has potential as a dating approach for the location and temporal range in question. This involves identifying possible sources of volcanic ash and identifying the large-scale explosive events associated with those eruptive centres. It is advisable to determine an approximate age range for the sequence, so timings of known volcanic events can be taken into consideration. An understanding of the interaction between depositional environments and tephra chemistry may also useful, as this may provide information regarding the preservational potential of the sedimentary environment. The Black Sea region is within the fallout range of volcanic centres in Italy, Greece, Turkey, Armenia, Syria and Georgia (Figure 3.2), although current evidence suggests that not all of these centres experienced large-scale explosive volcanic activity during the Late Pleistocene or Holocene (Aksu et al., 2008; Narcisi and Vezzoli, 1999; Karapetian et al., 2001; Lebedev et al., 2006). The most active and well-documented volcanic centres are the Hellenic Arc in Greece and the Campanian Province in Italy. A recent study identified eight tephra layers in the Marmara Sea (Aksu et al., 2008), all of which originated from Italian or Greek provinces. Other likely sources of tephra include the extinct volcanoes Elbrus and Kazbek in the Caucasus, although unfavourable winds and lack of recent activity render these volcanoes unlikely sources of tephra. Studies of volcanic centres in Turkey suggest explosive events during the past 100 kyr, although no distal deposits have been found in the Black Sea region, suggesting that eruptions were either of low magnitude or their ejecta was dispersed away from the Black Sea Region (Kuzucuoglu et al., 1998). The chemical classification of the Italian, Hellenic and Anatolian volcanic provinces are shown in Figure 3.3

The second step in the application of tephrochronology is the location of tephra within a sediment sequence or sample. If the tephra is visible and relatively thick, then this stage is straightforward. If the layer is invisible to the naked eye (also known as cryptotephra layer), then detection becomes more problematic. A systematic, non-automated investigation can be labour intensive, and if the probability of the sediment containing tephra is low, then alternative dating strategies are worth considering. If the probability of abundant tephra layers is high, the sedimentary context is favourable and other dating approaches are unfeasible, then a systematic approach may be warranted. Several automated tephra detection techniques exist, including magnetic separation (Mackie et al., 2002), magnetic susceptibility, X-radiography, reflectance spectrometry and luminescence spectrometry (Gehrels et al., 2008), although some of these techniques have limited applicability. It is also important to consider the probable chronostratigraphical significance of the deposit, as it is possible that some tephra deposits may be reworked and cannot therefore be used as a chronological control.

The third step in tephrochronological analysis is the separation of the tephra from the non-volcanic sedimentary components. Again, a broad suite of techniques have been developed to facilitate this process. An ideal deposit would be a thick stratum of pure tephra, whereas a more challenging deposit would be a bioturbated cryptotephra layer in a chemically aggressive, Si mineral-rich sedimentary context. When deciding on the experimental strategy, some of the key considerations include the Si mineral-content of the sediment, the degree of bioturbation and reworking, the chemical stability of the tephra shards (Pollard and Barron, 2003), the chemistry of the depositional environment and the age of the tephra (Federman, 1984). There are too many variables to be considered in a theoretical sense here, but a comprehensive discussion can be found in Turney and Lowe (2001) and Gehrels et al. (2008).

Distribution of tectonic plate boundaries, active volcanoes and volcanic provinces in the Mediterranean and Black Sea region. `A’ indicates the Calabrian volcanic province, `B’ indicates the Hellenic volcanic province, `C’ indicates Anatolian province, including Caucasian volcanoes (van der Meijde et al., 2003; McKenzie, 1970).

The final step in most tephrochronological analyses is the determination of shard chemistry. The aim of this step is to attribute the shard to a volcanic province, and ideally a single volcanic event. Chemical composition can be determined via a number of instruments such as semi-quantitative scanning electron microscopes, or laser-ablative inductively-coupled plasma mass spectrometers, but the most commonly applied technique is electron microprobe analysis (EMPA). Depending on the machine and the analytical parameters, bulk and trace element composition can be established for individual tephra shards. The high precision targeting afforded by EMPA permits the measurement of interior surfaces of tephra shards, bypassing the problem of hydration rims (Federman, 1984; Froggatt, 1992), as well as confirming tephra homogeneity. In a period of intense eruptive activity, a volcanic source may eject tephra in several successive phases. In this case, tephra chemistry may exhibit only limited variation, precluding discrimination via bulk-chemical analysis. In this situation, trace or rare-earth element analysis may provide a means of further discrimination. Again, the full complexities of quantitative chemical analysis are beyond the scope of the discussion presented here, but many papers (e.g. Davies et al., 2002) discuss the relative merit of the various techniques. It should be noted that in order to ensure that results from different laboratories are comparable, standard conventions should be adopted and the relevant details presented alongside any analyses (Froggatt, 1992).

Once results have been obtained, the data may require further interpretation before incorporating them into an age model. An example of a weak correlation would be the chemical profiling of only the bulk elements of a low number of shards from a sample within tephrostratigraphically complex region, which can only be correlated at the province level. An example of a strong correlation would be bulk, trace and rare-earth element analyses from numerous highly abundant shards, occurring in a well-stratified sediment sequence, which are then attributed to a volcanic province which experienced only one significant eruptive event. The former would be considered an unreliable datum, perhaps indicative of a chronostratigraphically insignificant tephra layer, whereas the latter could be reliably used as a chronological tie-point and may even provide a potential means of calibrating inaccuracies in radiometrically determined age points (Kwiecien et al., 2008).

To read a full account of how I used volcanic ash in my research, take a look at my Ph.D thesis here. I've also uploaded a spreadsheet that contains a lot of geochemical information for European volcanoes.