Cryospheric Sciences


Image of the Week – Apocalypse snow? … No, it’s sea ice!

Image of the Week – Apocalypse snow? … No, it’s sea ice!

Sea ice brine sampling is always great fun, but sometimes somewhat challenging !

As sea water freezes to form sea ice, salts in the water are rejected from the ice and concentrate in pockets of very salty water, which are entrapped within the sea ice. These pockets are known as “brines”.

Scientists sample these brines to measure the physical and bio-geochemical properties, such as: temperature, salinity, nutrient, water stable isotopes, Chlorophyll A, algal species, bacterial number and DNA, partial pressure of CO2, dissolved and particulate Carbon and Nitrogen, sulphur compounds, and trace metals.  All of this helps to better understand how sea ice impacts the atmosphere-ocean exchanges of climate relevant gases.

In theory, sampling such brines is very simple: you just have to drill several holes in the sea-ice ensuring that the holes don’t reach the bottom of ice and wait for half an hour. During this time, the brine pockets which are trapped in the surrounding sea ice drain under gravity into the hole. After that, you just need to sample the salty water that has appeared in the hole. Simple…

…at least it would be if they didn’t have to deal with the darkness of the Antarctic winter, blowing snow, handling water at -30°C and all while wearing trace metal clean suits on top of polar gear…hence the faces!

This photo won the jury prize of the Antarctic photo competition, organised by APECS Belgium and Netherlands as part of Antarctica Day celebrations (1st of December).

All the photos of the contest can be seen here.

Edited by Sophie Berger and Emma Smith

Jean-Louis Tison is a professor at the Université libre de Bruxelles. His activities are focused on the study of physico-chemical properties of « interface ice », be it the « ice-bedrock » (continental basal ice) , « ice-ocean » (marine ice) or « ice-atmosphere » (sea ice) interface. His work is based on numerous field expeditions and laboratory experiments, and on the development of equipments and analytical techniques dedicated to the multi-parametric study of ice: textures and fabrics, stable isotopes of oxygen and hydrogen, total gas content and gas composition, bulk salinity, major elements chemistry…


Image of the Week – For each tonne of CO2 emitted, Arctic sea ice shrinks by 3m² in summer

Image of the Week – For each tonne of CO2 emitted, Arctic sea ice shrinks by 3m² in summer

Declining sea ice in the Arctic is definitely one of the most iconic consequences of climate change. In a study recently published in Science, Dirk Notz and Julienne Stroeve find a linear relationship between carbon dioxide (CO2) emissions and loss of Arctic sea-ice area in summer. Our image of this week is based on these results and shows the area of September Arctic sea ice lost per inhabitant due to CO2 emissions in 2013.

What did we know about the Arctic sea ice before this study?

Since the late 1970s, sea ice has been dramatically shrinking in the Arctic, losing 3.8% of its area per decade. Sea-ice area is at its minimum in September, at the end of the melting season.

The main cause of this loss is the increase in surface temperature over the recent years (Mahlstein and Knutti, 2012), which has been more pronounced in the Arctic compared to other regions on Earth (Cohen et al., 2014). The use of statistical methods involving both observations and climate models shows that the recent warming in the Arctic can be attributed to human activity, i.e. mainly greenhouse gas emissions (Gillett et al., 2008). This suggests a direct link between human activity and Arctic sea-ice loss, which is confirmed in the Fifth Assessment Report of the Intergovernmental Panel on Climate Change (IPCC).

How exactly is sea-ice loss related to CO2 emissions ?

Notz and Stroeve (2016) relate the Arctic sea-ice decline to cumulative CO2 emissions since 1850 (i.e. the total amount of CO2 that has been emitted since 1850) for both observations and climate models. Cumulative CO2 emissions constitute a robust indicator of the recent man-made global warming (IPCC, 2014).

The two quantities are clearly linearly related (see Figure 2). From 1953 to 2015, about 3.5 million km² of Arctic sea ice have been lost in September while 1200 gigatonnes (1 Gt = 10e9 tonnes) of CO2 have been emitted to the atmosphere. This means that for each tonne of CO2 released into the atmosphere, the Arctic loses 3 m² of sea ice.

Fig 2: Monthly mean September Arctic sea-ice area against cumulative CO2 emissions since 1850 for the period 1953-2015. Grey circles and diamonds show the results from in-situ (1953-1978) and satellite (1979-2015) observations, respectively. The thick red line shows the 30-year running mean and the dotted red line represents the trend of 3 m² sea-ice area loss per tonne of CO2 emitted. [Credit: D. Notz, National Snow and Ice Data Center ]

Starting from the relationship between cumulative CO2 emissions and sea-ice area, it is then easy to attribute to each country in the world their own contribution to sea-ice loss based on their CO2 emissions per capita. The countries that stand out in the map are thus the countries emitting the most in relation to their population.

Could the Arctic be ice-free in the future?

If this relationship holds in the future (in other words, if we extend the red dotted line to zero sea-ice area in Figure 2), adding 1000 Gt of CO2 in the atmosphere would free the Arctic of sea ice in September. Since we are currently emitting about 35 Gt CO2 per year, it would take less than 30 years to have the Arctic free of sea ice in the summer (which confirms previous model studies (e.g. Massonnet et al., 2012)).

Edited by Clara Burgard and Sophie Berger

Further reading

DavidDavid Docquier is a post-doctoral researcher at the Earth and Life Institute of Université catholique de Louvain (UCL) in Belgium. He works on the development of processed-based sea-ice metrics in order to improve the evaluation of global climate models (GCMs). His study is embedded within the EU Horizon 2020 PRIMAVERA project, which aims at developing a new generation of high-resolution GCMs to better represent the climate.

Image of the Week: Atmospheric CO2 from ice cores

Image of the Week: Atmospheric CO2 from ice cores

The measurements of atmospheric CO2 levels at Manu Loa, Hawaii read 401.01ppm on the 7th of December this year. To understand the significance of this number, you just need to look at the figure above from the 4th IPCC report. It shows the changes in CO2 concentrations during the past 800,000 years based on ice core measurements. Values have fluctuated between 190ppm and 280ppm. In other words, both the level of present-day atmospheric CO2 and the rapidity of the increase is unprecedented.

The figure also shows the projections from the IPCC AR4 report for different emission scenarios. Which scenario will turn out to be the most likely might be determined at COP21 in Paris right now.

Read more:

Measurements at Manu Loa, Hawaii

Camping on the Svalbard coast

Camping on the Svalbard coast

In early April 2015, a small team of 2 Belgian and 2 French researchers went to Svalbard. The goal? Testing new methods to measure sea-ice thickness and ice algal biomass, but also measuring greenhouse gases in the sea ice in relation with the ‘STeP’ (Storfjorden Polynya multidisciplinary study) campaign. With funding from the French Polar Institute (IPEV) and IPSL and logistical arrangements by the Laboratoire d’Océanographie et du Climat (LOCEAN, Paris), we had the opportunity to conduct a short field campaign, long enough to perform instrumental tests and ice coring.

The expedition was arranged with Stefano Poli – a tourist guide in Svalbard. People and equipment were driven on snow-mobiles to Agardhbukta, 100 km South East of Longyearbyen. The conditions for this expedition were quite rudimentary; a tent, a burner and sleeping bags. There are no human settlements in this remote location, so Stefano chose a camping spot, as safe as possible with respect to polar bears, right in front of the fjord, our working place…

Quite exciting, isn’t it? Let’s take a look at what we got up to:

Outside the tent (credit: A. Lourenço).

Setting up the Camp

How do you set up a camp in the Arctic? First, you look for a hidden place, ideal for bear watching (in our case we chose a place with a small hill on our back and open on a wide and flat area). Then the hard work starts:

  1. Build the body and the membrane of the tent.
  2. Dig a hole in the snow right under the entrance to allow carbonic gases to escape.
  3. Set up the oil burner circuit: the oil tank positioned outside the tent is sent to the burner through a pipe covered by snow to avoid spilling accidents, and another pipe made from superposition of aluminum cylindrical cans links the burner to the air above the tent. A hole in the membrane of the tent is designed for that purpose.
  4. Circle the tent with a bear alarm. This was totally handmade and consisted of a gun firecrackers guided by a thread, not really sufficient to stop a polar bear!

The daily life

Eating, sleeping, working, everything was adapted to Arctic conditions. The meals – morning, lunch and dinner – were just dry food in hermetic bags that you fill in with boiling water. Better choose the orange bags, chili con carne is the best. To sleep, reindeer skins were placed directly on the ground (i.e. snow) as mattresses, and sleeping bags were in natural bird feathers. The ideal position is when you find the perfect distance between your feet and the burner.

Inside the tent (credit: M. Kotovitch)

Inside the tent (credit: M. Kotovitch)

As the bear alarm might not be totally reliable, our guide offered us (well, without the possibility to decline or give up) a memorable nocturnal experiment, a series of 2-h bear-watching shifts, with a survival kit consisting of a flare gun, 2 tea thermos and 1 teddy polar bear for superstition.

In the Field

The objectives of this short campaign were (i) to sample early spring sea ice, snow and seawater in the Storfjorden region; (ii) to calibrate non-destructive methods for ice thickness and biomass retrievals in sea ice; and (iii) to measure greenhouse gases in sea ice in relation with the ‘STeP’ campaign. This cruise is scheduled for next summer in Storfjorden, led by IPSL-Paris and involves paleo-oceanographers, physical and chemical oceanographers as well as biogeochemists from several countries.

Why is it so important to develop a non-destructive method while working on sea ice? Because the general and only way known currently to sample sea ice in its entire thickness consists of coring, which destructs the site and can alter sea ice biogeochemical conditions.

With these goals in mind, the initial plan was to operate 2 or 3 stations per day on coastal landfast ice in Storfjorden. Agardhbukta was chosen for its situation (not too far from Longyearbyen) and as one of the locations in Storfjorden where we had good expectations to find practicable sea ice in this season, which was required to carry out our work. Our guide Stefano mentioned he saw a satellite image with new sea ice on March 23 in that location. And indeed, the sampled ice was probably not older than 2-3 weeks (Figure 4). Regarding the sampling planning, our expectations where a little bit overestimated. The weather conditions were so snowy and windy that we hardly had the time to sample one full station a day… This is how Polar Regions surprise us.


Bear watching (credit: A. Lourenço)

Edited by Sophie Berger and Nanna Karlsson

Marie Kotovitch is a PhD student at the Chemical Oceanography Unit, University of Liège, supervised by Bruno Delille. She is working with sea ice and gas transport (mostly greenhouse gases like CO2 and N2O). She has a collaboration with the Laboratory of Glaciology at the Université Libre de Bruxelles and was involved in this campaign in Svalbard to analyze the biological aspect of this study.

Do Beers Go Stale in the Arctic? – Jakob Sievers

Do Beers Go Stale in the Arctic? – Jakob Sievers

A story about CO2 -fluxes between sea-ice and the atmosphere

What’s it all about?

Whenever I have had to describe my PhD research project to people outside of my research community, I have always found it useful to use an analogy most people are familiar with, namely beers. Now that I have the full attention of the entire class, allow me to explain. Say you were to find yourself at an outside café, grabbing a beer in the beautiful weather with your friends. You get caught up in the conversation and soon your beer is lukewarm and stale and you struggle with the last few drops while gesturing somewhat frantically at the waitress to bring you a new beer. What has happened? Well, the partial pressure of CO2 (pCO2) in the beer is much larger, relative to that of the atmosphere. This leads to gradual diffusion (or flux) of CO2 from the surface of the beer into the atmosphere, and eventually the establishment of an equilibrium pCO2 in the beer, similar to that of the atmosphere. Of course, you have probably been waving at the waitress long before the equilibrium is reached. Now imagine that you had performed this experiment in the cold of winter. The waitress brings you a beer and from the moment it reaches your table a thin layer of ice starts forming on the top. What happens with the CO2 levels now?

Though a bar would probably have been a nice place for fieldwork, obviously my fieldwork took place in a much colder environment, namely the arctic sea-ice in northeast Greenland. The kingdom of the sort of curious four-legged fella shown in the photo above. And the answer to the above question might be “nothing” in the case of a beer. I.e., the ice works as a sort of lid on the beer, preventing any fluxes of CO2 from occurring, but what about in a sea-ice environment? Until very recently the beer analogy would have applied equally well here, as most scientists would have told you that no CO2 fluxes could take place in this sort of environment. Accordingly, all climate models currently treat sea-ice covered regions as regions of no exchange in terms of CO2. Lately, however, a number of papers have been published, suggesting that fluxes can occur at different stages of the annual sea-ice cycle. The paradigm shift occurred when it was discovered that CO2 may be transported through brine channels within the sea-ice. Brine channels form within sea-ice because of the salt in the sea-water, and they tend to be larger if the temperature is higher. So, at the bottom of the ice where the ambient temperature is a balmy 0°C, the channels are largest (Fig. 1).

Figure 1: An actual photo of brine channels within sea-ice [Junge et al., 2001, Annals of Glaciology]

Figure 1: An actual photo of brine channels within sea-ice [Junge et al., 2001, Annals of Glaciology]

Processes – scratching the surface

It is thought that the bulk of fluxes occur primarily in the spring during sea-ice melting. To understand why, we have to go back to when the sea-ice is formed during wintertime. As the ice forms, salts and CO2 are concentrated in the brine channels, and are expelled into the ocean through the larger brine channels in the bottom of the ice. The process is called gravity drainage and refers to the sinking of the highly saline and cold brine water which is denser relative to the underlying sea-water. Some of the expelled CO2 will continue to the deeper ocean column and as such lead to a CO2 loss of the original surface system. Hence, upon springtime melting, the upper water column has less CO2 than the atmosphere, which then causes an uptake of CO2 from the atmosphere. I.e.: a re-carbonization of the beer, so to speak. The mechanism has been coined the sea-ice driven carbon pump and has been estimated to drive an annual uptake of 50 million tons of carbon annually in the arctic alone, constituting a significant fraction of the total CO2 uptake of the arctic ocean which is estimated to be 66-199 million tons of carbon annually. Hence, understanding the impact of this carbon pump is important, particularly because the impact of climate change is felt more dramatically in the arctic compared to the rest of the world. Sea-ice cover is becoming increasingly ephemeral and glacial freshwater runoff, which inherently has a low partial pressure of CO2, is becoming increasingly ubiquitous in the fjord systems.

What are the challenges in this field of study?

To incorporate the carbon pump into existing climate models we need first to understand the physical and biogeochemical processes which drive sea-ice carbon fluxes in both coastal and open water conditions as well as during the entire life-cycle of sea-ice. Sounds simple, right? Of course not. Like most experimental investigators in cryospheric sciences we are struggling with considerable logistical challenges in a vast and unforgiving environment, where temperature conditions often have instrument-developers shaking their head in disbelief. “We didn’t test the instrument for that sort of thing”. How reassuring. Secondly, there are substantial risks involved, both for people and instruments, when conducting experiments during periods of thin sea-ice, which also happen to be the periods in which fluxes are most pronounced, and thus all the more important to understand. Fortunately we are equipped with fairly unique vehicles for transportation during these conditions (Fig. 2). Finally, the fluxes that are reported on sea-ice are often significantly smaller than what is typically reported in terrestrial environments, leaving investigators at times struggling to discern actual measurements from artificial noise.

Figure 2: Our polar air-boat for safe and fast transportation on thin sea-ice. During the experiment this bad boy was typically referred to as a gasoline-to-noise converter. In this particular picture the air-boat is pictured on thick sea-ice, hence the use of normal winter clothing instead of marine safety suits. Photo: Jakob Sievers.

Figure 2: Our polar air-boat for safe and fast transportation on thin sea-ice. During the experiment this bad boy was typically referred to as a gasoline-to-noise converter. In this particular picture the air-boat is pictured on thick sea-ice, hence the use of normal winter clothing instead of marine safety suits. Photo: Jakob Sievers.

What did I focus on during my PhD?

A common method for flux-observation is the micrometeorological Eddy Covariance method. It involves setting up a 2-5m high tower on a surface of interest and at the top installing:

(1) A three-dimensional ultrasonic anemometer, which measures 3D wind patterns,

(2) A gas analyzer, which measures the atmospheric concentration of e.g. H2O, CO2 or CH4 depending on which type of flux you are interested in.

Figure 3: Our eddy covariance tower on thin ice (20cm) in the outermost region of Young Sound (NE Greenland). Photo: Jakob Sievers.

Figure 3: Our eddy covariance tower on thin ice (20cm) in the outermost region of Young Sound (NE Greenland). Photo: Jakob Sievers.

Together they allow for calculation of the average flux in an upwind area in front of the tower. This area will vary in size depending on the height of the tower and wind-conditions. One of our instrument towers are shown in Fig. 3. From a mathematical standpoint the method is very simple, but it requires a number of quite complicated assumptions concerning the characteristic wind-patterns of the environment in question. Often these assumptions are not discussed in detail in papers. During my PhD I found that because most fluxes in a sea-ice environment are so small the critical assumptions began breaking down. It seemed that nearly all observations reflected large-scale meteorological patterns and flow-characteristics of the topography instead of fluxes in the local area of interest. Realizing this, I successfully developed and tested a very comprehensive data-analysis method to disentangle contributions of interest and contributions which were not reflective of the local environment of interest. Due to the technical nature of this new method I will not elaborate here, except to say that it was recently published in the Journal of Atmospheric Chemistry and Physics. Subsequently I was able to analyze simultaneous observations of CO2-fluxes and environmental parameters such as site energy balance and wind-speed to achieve a better understanding of some basic causal relationships for CO2 fluxes in a sea-ice environment.

Because this field of study is so new, and in-situ experiments are so challenging, much work is still needed but hopefully we will soon have a sufficiently detailed understanding of the physical and biogeochemical factors driving CO2 fluxes in the sea-ice environment, to be able to develop a model relationship and upscale that to all polar regions in which sea-ice exist.

Figure 4: An unexpected challenge during fieldwork: invasion by sled dog pups. Photo: Karl Attard.

Figure 4: An unexpected challenge during fieldwork: invasion by sled dog pups. Photo: Karl Attard.

I would like to leave you with one final picture illustrating a somewhat unexpected challenge during our experiment: having to reclaim our instruments and tools from the sled dog pups (Fig. 4).

Jakob Sievers ( Arctic Research Centre / Aarhus University) has just completed his PhD at Aarhus University in Denmark as part of the DEFROST project. He is interested in the physical and biogeochemical drivers of CO2 cycling in sea-ice environments and the development of flux models and flux observation methods under these challenging circumstances. He is currently seeking funding for the purpose of continuing his research in northern Norway.