Climate: Past, Present & Future

Climate: Past, Present & Future

Levoglucosan, the witness of past fires

Levoglucosan, the witness of past fires
Name of proxy


Type of record

Biomass burning


Lake and marine sediments and ice cores

Period of time investigated

Present to approximately 130,000 years ago

How does it work?

Levoglucosan is a molecule that is exclusively formed during the combustion of vegetation at low-temperature. It is therefore considered to be a source-specific tracer for biomass burning. During these fire events, levoglucosan is emitted into the atmosphere and can be transported over hundreds of kilometres. The extent of its atmospheric degradation is currently under debate, however, several studies have demonstrated that levoglucosan remains stable in the atmosphere for several days under most atmospheric conditions. It has been extensively used as a tracer for biomass burning in aerosols in numerous air-quality studies. Levoglucosan can provide information on the occurrence and origin of biomass burning, given that the source area of the levoglucosan is known. So far, levoglucosan is usually interpreted in terms of increased or decreased occurrence of biomass burning.

Figure 1: Illustration of levoglucosan transport from the land to the bottom of the ocean or lakes.

Recently, the use of levoglucosan as a biomass burning proxy in geological archives has gained increasing interest. Indeed, levoglucosan has been analysed in lake sediments (Battistel et al., 2017), in marine sediments (Lopes dos Santos et al., 2013) and in ice cores (You et al., 2016). It can be transported to these environments by atmospheric transport and by rivers (Fig. 1), where the biomass burning history of the source area is preserved. Seminal advances in the development of this proxy have occurred e.g. on the effect of transport and deposition on the levoglucosan. These molecules are likely transported to the ocean floor attached to marine biogenic particles in the water and do not substantially degrade during settling in the water column (Schreuder et al., 2018). However, they seem to be partially degraded in the top layer of the sediment and therefore changes in preservation conditions over time might influence the levoglucosan record. Other factors that might also influence levoglucosan accumulation are changes in wind strength and direction. This can result in decreasing or increasing transport of levoglucosan to the specific environment where the cores/samples are taken. This illustrates the importance to constrain the factors influencing the levoglucosan record in the context of a multi proxy approach.

What are the key findings that have be done using this proxy ?

So far, levoglucosan studies have mainly focused on reconstructing fire history of the last few centuries and of the Holocene (e.g. Shanahan et al., 2016; Zennaro et al., 2014), but it has also been detected in sediments up to 130 kyrs (Lopes dos Santos et al., 2013). It shows that levoglucosan has the potential to reconstruct fire history on short time scales (i.e. yrs to kyrs) as well as on long time scales (i.e. kyrs to myrs). The fire biomarker is frequently used together with proxies for vegetation and climate to get a better understanding of the interactions between fire, humans, vegetation and climate. For example, Lopes dos Santos et al. (2013) used the levoglucosan proxy in a marine sediment core offshore Australia to reconstruct past levels of biomass burning on the Australian continent over the last 130 kyrs. They also studied biomarkers for vegetation composition and archived information on human arrival and extinction of animals heavier than 40 kg (megafauna). They found out that around 44-42 kyrs, vegetation change was the consequence of the extinction of megafaunal browsers and led to the build-up of fire-prone vegetation in the Australian landscape, as illustrated in figure 2.

Figure 2: Interactions between fire and the environment in Australia around 44-42 kyears.

                                                                                          This article has been edited by the editorial board



            Battistel D., Argiriadis E., Kehrwald N., Spigariol M., Russell J.M. and Barbante C. (2017) Fire and human record at Lake Victoria, East Africa, during the Early Iron Age: Did humans or climate cause massive ecosystem changes? The Holocene 27, 997-1007.

Lopes dos Santos R.A., De Deckker P., Hopmans E.C., Magee J.W., Mets A., Damsté J.S.S. and Schouten S. (2013) Abrupt vegetation change after the Late Quaternary megafaunal extinction in southeastern Australia. Nature Geoscience 6, 627-631.

Schreuder L.T., Hopmans E.C., Stuut J.-B.W., Damsté J.S.S. and Schouten S. (2018) Transport and deposition of the fire biomarker levoglucosan across the tropical North Atlantic Ocean. Geochimica et Cosmochimica Acta.

Shanahan T.M., Hughen K.A., McKay N.P., Overpeck J.T., Scholz C.A., Gosling W.D., Miller C.S., Peck J.A., King J.W. and Heil C.W. (2016) CO2 and fire influence tropical ecosystem stability in response to climate change. Scientific reports 6, 29587.

You C., Xu C., Xu B., Zhao H. and Song L. (2016) Levoglucosan evidence for biomass burning records over Tibetan glaciers. Environmental Pollution 216, 173-181.

Zennaro P., Kehrwald N., McConnell J.R., Schüpbach S., Maselli O.J., Marlon J., Vallelonga P., Leuenberger D., Zangrando R. and Spolaor A. (2014) Fire in ice: two millennia of boreal forest fire history from the Greenland NEEM ice core. Climate of the Past 10, 1905-1924.

Pollen, more than forests’ story-tellers

Pollen, more than forests’ story-tellers
Name of proxy

Sporomorphs (pollen grains and fern spores)

Type of record

Biostratigraphy and Geochronology markers, Vegetation dynamics


Terrestrial environment

Period of time investigated

Present to 360 million years

How does it work?

The sporomorphs (pollen grains and fern spores) are cells produced by plants involved in the reproduction. They are microscopic (less than a fifth of a millimeter) and contain a molecule called sporopollenin in their cell wall, which is very resistant to degradation. The sporopollenin molecule allows sporomorphs to be preserved in sedimentary archives such as lake sediment or peat deposits.

These reproductive structures appeared during the Paleozoic (570 million years ago) but the first spores looked rather similar and were indistinguishable among species. Later speciation of plants promoted the diversification of the reproductive cells between species and brought the opportunity to relate the fossil sporomorphs found in the sedimentary archives to the parental plant that produced them.

Figure 1. Plant communities are different depending on a wide range of environmental conditions. Above: Andean grasslands (páramo) in Ecuador. Below: Swamp forest in Orinoco Delta (Venezuela).

Plants are immobile organisms, and each species has its own tolerance range to the existing environmental conditions. The occurrence of certain plant communities in a specific environment depends on their different tolerance ranges. For instance, we do not observe today the same plants growing in the tropical rainforests of South America than in the polar tundra (Figure 1). Paleopalynology is the discipline that helps characterizing which plant species have occurred at a specific location during a particular time period. This provides information on the environmental conditions of the studied region. To identify the different species, palynologists have to analyze under the optical microscope the specific features of the sporomorphs’ cell walls. They look at e.g. the presence of spines or air sacs, or the number of apertures that the pollen grain has (Figure 2). These features are specific for each plant, which allows relating the pollen grain found in the sedimentary archive to the plant that produced it at the study location at a particular period of time.

Figure 2. Pollen grains have very different morphologies that allow identification of the plant that produce them. A: Byttneria asterotricha (Sterculiaceae); B: Triplaris americana (Polygonaceae); and C: Calyptranthes nervata (Myrtaceae). Bar scales in the pictures represent 25 micrometers.

What are the key findings made using this proxy?

Paleopalynology has a wide range of applications in geoscience. For instance, the presence of specific sporomorphs has been used as chronological markers to pinpoint several geological periods, especially in the far past biostratigraphy (million years ago)  (Salard-Cheboldaeff 1990).

In palaeoecology (the ecology of past ecosystems), the analyses of fossil sporomorphs help in specifying the dynamics of vegetation communities through time. This type of work started a century ago by Lennart van Post (1916) and provided the opportunity to study plants population and community natural trends within the appropriate temporal frame for long-lived species (i.e. tree species such as pines or oaks can live several centuries). Moreover, it provides a unique empirical evidence of the actual responses of vegetation to disturbances that occurred in the past, e.g. natural hazards, human populations land use and other anthropogenic impacts, or climatic shifts.

For instance, regarding past climates, paleopalynology allowed us to:

i)               understand the independent behavior of the species during glacial cycles (i.e., when a single species responded to changes, but the plant community as a unity did not respond) in forming new plant communities each time (Davis 1981; Williams and Jackson 2007);

ii)             map the re-colonization events and the assemblages formed during the last deglaciation until the vegetation communities we observe today (Giesecke et al. 2017).

In addition, in some characteristic environments, such as mountain regions, the occurrence and disappearance of specific species can allow the estimation of the temperature change with respect to present-day conditions. Another example has been developed in the last decade: the study of organic compounds contained in the sporopollenin of the sporomorphs’ walls. It has been identified as an accurate proxy that registers UV-B rays’ signals (Fraser et al. 2014). As UV-B rays’ are related to solar irradiation trends through time, reconstructing the organic compound variations in the sporomorphs’ walls allows reconstituting past solar irradiation trends in continuous archives such as lake and peat deposits.

This all shows that despite being a tiny structure, pollen grains are the story tellers of how the planet has been changing through history and can provide a wide range of outcomes essential for geosciences.


Davis, M.B. (1981). Quaternary history and the stability of forest communities. In: West, D.C., Shugart, H.H., Botkin D.B. (Eds.) Forest succession. New York, NY: Springer-Verlag.

Fraser, W.T., Lomax, B.H., Jardine, P.E., Gosling, W.D., Sephton, M.A. (2014). Pollen and spores as a passive monitor of ultraviolet radiation. Frontiers in Ecology & Evolution 2: 12.

Giesecke, T., Brewer, S., Finsinger, W., Leydet, M., Bradshaw, R.H.W. (2017). Patterns and dynamics of European vegetation change over the last 15,000 years. Journal of Biogeography 44: 1441-1456.

Salard-Cheboldaeff, M. (1990). Interptropical African palynostratigraphy from Cretaceous to late Quaternary times. Journal of African Earth Sciences 11: 1-24.

Von Post, L. (1916). Om skogsträdpollen i sydsvenska torfmosslagerföljder. Geol.   Fören. Stockh. Förhandlingar 38, 384–390.

Williams, J.W., Jackson, S.T. (2007). Novel climates, no-analog communities, and ecologica lsurprises. Frontiers in Ecology and the Environment 5: 475-482.

                                                                                                                           Edited by Célia Sapart and Carole Nehme

How to reconstruct past climates from water stable isotopes in Polar ice cores ?

How to reconstruct past climates from water stable isotopes in Polar ice cores ?

Ice cores are a favored archive to study past climates, because they provide a number of indications on the history of the climate and of the atmospheric composition. Among these, water stable isotopes are considered as a very reliable temperature proxy. Yet, their interpretation is sometimes more complicated than a simple one-to-one correspondence with local temperature and requires intercomparison with other proxy records, as various processes affect the signal found in the ice cores.

How does it work?

All water molecules are not equal: some are heavier due to one of their atoms being substituted with a heavier counterpart (the standard oxygen molecule, 16O, can be substituted by an 17O or an 18O, whereas the hydrogen (H) can be substituted by a deuterium (D=2H)). These molecules (e.g. H216O, H218O and HD16O) are called isotopologues (or isotopes in short, but it’s technically inaccurate), and have each different physical properties. As a result, the molecules react differently to external factors, leading to fractionation (processes that affect the relative abundance of isotopes). The isotopic composition (commonly referred to as δ18O, δ17O and δD) of snow is governed by fractionation from the evaporation site, where the moisture first enters the atmosphere, to the precipitation site where it is deposited to the ground (Dansgaard 1953)(see video below). First, over the ocean, the heavier isotopes are less likely to take part in the formation of moisture, leading to lower concentrations of the heavy isotopes in the clouds compared to the mean oceanic water isotopic composition (lower concentration means that the δ18O is more negative). Then, as the air masses move toward the poles, temperature decreases leading to precipitation (either liquid or solid). The heavier isotopes will be preferentially found in the condensed phase than the light ones, which depletes even more the cloud from its heavy isotopes. Finally, in remote areas of the Polar Regions, the isotopic content of the final precipitation results from successive precipitation events. Since, as we saw, the precipitation contents in each isotope are different, each successive precipitation event will have a different isotopic composition, with the final one having the least heavy isotopes – a process called distillation. At each step of the moisture’s path from ocean to cloud to precipitation, the isotopic fractionation is strongly influenced by temperature. This leads to a temperature signal in the isotopic composition of both vapour and precipitation.

This video shows the isotopic fractionation at each step of the water cycle (from the evaporation over the ocean to the location where the precipitation occurs) are integrated, giving the temperature and humidity sensitivity of the isotopic content of the precipitation (modified from Casado (2016)).

Over glaciers and ice sheets, snow accumulates and can remain preserved for hundred of thousands of years. Thus, analysing the isotopic composition of these successive layers enables us to retrieve past temperatures. Considering the very low amount of water necessary to obtain a measurement, analysing an ice core can provide continuous and high-resolution time series of past climatic variations.

A classic way to retrieve temperature from isotopic composition data is to use the spatial relationship between δ18O of surface snow and surface temperature (e.g. Lorius and Merlivat (1975) for Antarctica). That is, to measure simultaneously the present-day temperature and the δ18O of surface snow across the study area and to make use of their linear spatial relationship to infer past temperatures from the δ18O of the ice core. However, one should keep in mind two main limitations when using such a method. First, it assumes that the spatial relationship between δ18O and temperature is a good surrogate for the temporal δ18O versus temperature relationship. However, this link is known to change with time (and hence depth of the snow deposit). Second, processes occurring after the snow has fallen, such as sublimation or blowing wind, can affect the way the snow is layered in the ice core, as well as its isotopic composition.

Resolution and noise

The local accumulation (expressed in cm of snow per year) is a determining factor for both the extent of an ice core record and the maximal resolution that can be achieved. As the present-day ice thickness is capped between 3 and 4 km, it is necessary to choose a site with low accumulation to obtain an ice record spanning several glacial/interglacial (i.e. warmer and colder) cycles (Fischer et al. 2013). For instance, the NGRIP ice core in Greenland was retrieved in one of the thickest part of the Greenland ice sheet (see Figure 1. b)). Similarly, the Dome C ice core (the ice core spanning the furthest in the past to date which is 800,000 years before present) was retrieved in an area of Antarctica where the ice sheet was very thick (more than 3000 meters) and the accumulation very low (roughly 2.5 cm per year).

Figure 1: Greenland and Antarctic ice core sites: (a) Isotopic signal from the NGRIP, (b) and (c) maps of ice thickness in Greenland and Antarctica, respectively, and (d), isotopic signal from the Dome C. The isotopic signal for both sites is presented against the depth (left) and with the associated age model (right), warm periods are shown in grey to indicate the correspondence between age and depth in the ice cores (Casado et al, 2017).

For sites with low accumulation, the snow stays exposed at the surface for a long time. Hence, the initial precipitation signal is modified by local processes occurring after the snow has fallen (Ekaykin et al. 2002). This prevents proper recording of the signal at timescales below several years, for sites with accumulation lower than 8 cm per year (Münch et al. 2016). Deeper in the ice, diffusion processes smoothen the isotopic composition time series, erasing part of the climatic signal (Johnsen 1977). This limits the interpretation of ice core records at time scales smaller than a few decades. Finally, retrieving a temperature signal at high resolution for longer time scales remains a challenge because of the varying relationship between δ18O and the temperature deeper in the ice. The first limitation is the accumulation rate itself, which is typically lower during glacial periods. The temporal resolution also gets lower with depth as the ice thins due to the increasing pressure exert by the overlying ice layers. As illustrated in Fig. 1, the number of years per meter globally increases with the depth of the record, from roughly 20 years per meter at the top of the core at Dome C up to 1,400 years per meter for glacial periods at the bottom. Overall, the variability found in single ice core records combines both the climate variability and several signatures from other local processes affecting the snow.

Isotope-temperature calibration

The temperature signal retrieved from δ18O can be tested against independent temperature time series, such as borehole temperature measurements at the ice core site, to aid in reconstructing the correct δ18O versus temperature relationship (the so–called “calibration” process). The temperature of the borehole from which the ice core was extracted is measured at different depths. Small variations in these temperatures provide a reliable but low-resolution measurement of past temperature changes as the ice is a good thermal insulator. Measurements performed in Greenland have suggested that the use of the spatial δ18O versus temperature relationship described above underestimates by a factor of two the magnitude of the temperature change between the last glacial maximum (LGM, the last period when ice sheets were at their peak extension) and the present-day (Cuffey et al. 1994). Jouzel et al. (2003) confirmed this using computer simulations, and further showed that the δ18O versus temperature relationship does not remain constant over time. This large variability can be due to differences in the large-scale atmospheric circulation, vertical structure of the atmosphere, seasonality of precipitation, modification of location of the moisture source regions or of their climatic conditions.

Figure 2: Relationships between isotopes and temperature for different locations and timescales (indicated along the horizontal axis). A higher value would lead to higher temperature difference estimate for the same δ18O difference (Casado et al, 2017).

Calibration of the isotopic paleothermometer is therefore essential, and is realised through different methods. Station data can be used at the seasonal and interannual scales, isotopes of other gases at decadal to centennial scales (Guillevic et al. 2013) and borehole temperature measurements at millennial scales (Orsi et al. 2017). Finally, climate models which represent isotopic processes (called “isotope-enabled”) can be used to infer the isotope-temperature relationship with a direct control on the time scale and on the period. For instance, Sime et al. (2009) highlighted that for warm interglacial conditions, the isotope-temperature relationship can become non-linear whereas it is not the case for cooler (glacial) conditions.

The δ18O versus temperature relationships found in the literature (Fig. 2) span values ranging between 0.2 ‰/°C to 1.5 ‰/°C. From this compilation, it is clear that a more complex framework than simple linear regression to temperature is necessary to interpret the isotopic signal.


If water isotopes from ice core records are insightful tools to reconstruct past climates, there are fundamental limits to their power of reconstruction.

The above therefore calls for a careful use of isotopic records when these time-series are used for general inferences about the climate system (e.g. Huybers and Curry (2006)). A possible way forward is to use isotope-enabled global climate models (Sime et al. 2009). A complementary approach is to undertake process field studies (Casado et al. 2016), which can help to evaluate how the isotopic signal is modified after the deposition, and how the relationship between isotopes and temperature is altered at the seasonal and inter annual timescales.

The Beyond EPICA – Oldest Ice project plans to retrieve an ice core in Antarctica in which over 1.5 million years of climatic record could be retrieved. This will enable to go further back than the 800, 000 years old ice core obtained at Dome C and thus, would be a breakthrough into studying the changes in orbital forcing during the mid-Pleistocene transition (900 to 1,200 thousand years ago) during which the glacial-interglacial cycles shifted from lasting 41, 000 years on average to 100, 000 years.

Mountain glacier variations: natural thermometers and rainfall gauges

Mountain glacier variations: natural thermometers and rainfall gauges
Name of proxy

Fluctuations of mountain glaciers

Type of record

Geomorphological features


Continent – High mountain areas

Period of time investigated

From historical periods (c.a. 300 years ago) to the end of the Pleistocene (up to 200 000 years back in time)

How does it work?

Mountain – or “alpine” – glaciers are small ice bodies (from 1 to 10 000 km2). Although they represent only 0.3% of the total volume of the present-day cryosphere, they contain a large amount of useful information on the past climatic conditions on the continents. The position of the glacier margins, and its volume, is defined by its mass balance (the balance between ice accumulation and ice melt), that is mainly controlled by two important climate variables: air temperature and snow precipitation. Their internal behaviour is very sensitive to climate change, and they respond rapidly (<50 years) to climatic fluctuations. This gives alpine glaciers the ability to record climate change with a high temporal resolution.

The peculiarity of this climatic proxy is that it brings together several fields of research: glaciology, geomorphology, geochronology, and numerical modeling. In order to interpret mountain glacier fluctuations as climatic proxies, it is necessary to:

(i) Observe and interpret past glaciated landscapes (including landforms such as moraines and roches moutonnées) to infer past mass balances. This task often requires several days of field work in remote high area locations (Fig. 1a);
(ii) Establish the age of formation of these landscapes, most of the time using carbon-14, or other rare isotopes (such as beryllium-10 or helium-3). The concentrations of these rare isotopes increases with the time spent by a rock at the surface (Fig. 1b);
(iii) Perform computer-based numerical model simulations to infer the main climatic parameters (temperatures and precipitation) from the variations of glacial volumes (Fig. 1c).

Figure 1: Summary of the methodology used to derive paleoclimatic conditions from the fluctuations of mountain glaciers a) Sampling of a boulder sitting on a moraine for surface exposure dating using cosmogenic helium-3 (Altiplano, Tropical Andes), b) Principles of dating using in situ cosmogenic nuclides, c) Numerical modeling to interpret past glacial extent in paleotemperatures and paleoprecipitation conditions (from Blard et al., 2007)

Note that the position of a glacier may be considered as a “2 unknowns – 1 equation” problem, making useful any independent inputs from other continental paleoclimatic proxies, when available. These could for example be pollen-based reconstructions, lake level fluctuations or isotopic tools (measured in speleothems, lacustrine inorganic deposits or biogenic carbonates) that can bring quantitative constraints on temperature, precipitation or both (see previous posts for more information). If such complementary proxies are not available in the studied areas, it is necessary to remain cautious and propose a range of possible paleotemperature and paleoprecipitation reconstructions (Fig. 1c).

What are the key findings made using this proxy?

In some high altitude areas, Alpine glaciers, and changes in their mass balance over time, are the only indicators of paleoclimatic change. They have e.g. allowed us to understand that:
(i) The end of the Last Glacial Maximum (c.a. 18,000 years ago) was broadly synchronous (Schaefer et al., 2006; Clark et al., 2009);

(ii) The lapse rate (i.e. the vertical temperature gradient in the atmosphere – temperatures are lower higher up a mountain) was steepened during the Last Glacial Maximum (Blard et al., 2007), although this is controversial in some regions (Tripati et al., 2014);

(iii) The Little Ice Age (XVIIth-XIXth centuries) was a global event (Rabatel et al., 2006);

(iv) Because glaciers respond to both temperatures and precipitation amount, glacier fluctuations have also been used to reconstruct the changes in past precipitation or ‘paleoprecipitation’ at a high spatial resolution. The typical size of glacier watersheds is few hundreds to thousands square kilometers, which make them ideal paleorainfall gauges. This allows us to determine the paleoprecipitation variability at the regional scale (e.g. Martin, 2016 used paleoglaciers to establish the spatial pattern of rainfall in the Tropical Andes during the Heinrich 1 event, 16,000 years ago).

Blard et al. (2007) - Persistence of full glacial conditions in the central Pacific until 15,000 years ago, Nature 449 (7162), 591.

Clark et al. (2009) - The last glacial maximum, Science 325 (5941), 710-714.

Martin, PhD Thesis, Université de Lorraine, 2016

Rabatel et al. (2006) - Glacier recession on Cerro Charquini (16 S), Bolivia, since the maximum of the Little Ice Age (17th century), Journal of Glaciology 52 (176), 110-118.

Schaefer et al. (2006) - Near-synchronous interhemispheric termination of the last glacial maximum in mid-latitudes, Science 312 (5779), 1510-1513.

Tripati et al. (2014) - Modern and glacial tropical snowlines controlled by sea surface temperature and atmospheric mixing, Nature Geoscience 7 (3), 205-209.

Edited by the editorial board