Clara Burgard is a PostDoc at the Max Planck Institute for Meteorology in Hamburg. She investigates the evolution of sea ice in general circulation models (GCMs). There are still biases in the sea-ice representation in GCMs as they tend to underestimate the observed sea-ice retreat. She tries to understand the reasons for these biases.
Fig.1: Mooring in the fjord next to Qaanaaq [Credit: Measurement campaign team].
Many polar scientists who have traveled to Svalbard have heard several times how most of the stuff there is the “northernmost” stuff, e.g. the northernmost university, the northernmost brewery, etc. Despite hosting the four northernmost cities and towns, Svalbard is however accessible easily by “usual-sized” planes at least once per day from Oslo and Tromsø. This is not the case for the fifth northernmost town: Qaanaaq (previously called Thule) in Northwest Greenland. Only one small plane per week reaches the very isolated town, and this only if the weather permits it. And, coming from Europe, you have to change plane at least twice within Greenland! It is near Qaanaaq, during a measurement campaign, that our Image of the Week was taken…
Who, When and Where?
In January 2017, a few German and Danish sea-ice scientists traveled to Qaanaaq to set up different measurement instruments on, in and below the sea ice covering the fjord near Qaanaaq. While in town, they stayed in the station ran by the Danish Meteorological Institute. After a few weeks installation they traveled back to Europe, leaving the instruments to measure the sea-ice evolution during end of winter and spring.
What and How?
The goal of the measurement campaign was to measure in a novel way the evolution of the vertical salinity and the temperature profiles inside the sea ice, and the evolution of the snow covering the ice. These variables are not measured often in a combined way but are important to understand better how the internal properties of the sea ice evolve and how it affects or is affected by its direct neighbors, the atmosphere and the ocean. The team had to find a place remote enough from human influence, and with good ice conditions. As there are only few paved roads in Qaanaaq, cars are not the best mode of transport. The team therefore traveled a couple of hours on dog sleds (in the dark and at around -30°C!), with the help of local guides and their well-trained dogs (see Fig. 2 and 3).
Fig. 2: While the humans were working, the dogs could take a well-deserved break [Credit: Measurement campaign team].
Once on the spot, the sea-ice measurement device was introduced into the ice by digging a hole of 1m x 1m in the ice, placing the measurement device in it, and waiting until the ice refroze around it. Additionally, a meteorological mast and a few moorings were installed nearby (see Image of the Week and Fig. 3) to provide measurements of the atmospheric and oceanic conditions during the measurements. Further, a small mast was installed to enable the data to be transferred through the IRIDIUM satellite network.
Fig. 3: Small meteorological mast with dog sleds in the background [Credit: Measurement campaign team].
Finally, the small instrument family was left alone to measure the atmosphere-ice-ocean evolution for around four months. After this monitoring period, in May, the team had to do this trip all over again to get all the measurement devices back. Studying Greenlandic sea ice is quite an adventure!
Yes! The cryosphere can be found everywhere! [Credit: Max Pixel].
It is time again for the annual family meeting of the European Geosciences Union! A lot of interesting talks, posters and events are waiting for you! But this also means you will have to use your brain a lot to concentrate and understand what is going on…
Every year, around 15,000 geoscientists meet in Vienna for the EGU General Assembly (see our guide to navigating EGU 2018). It is an exciting event, where a high amount of exchange is taking place in a very intense way during one week. This also has the effect that, after a few days (or hours), you feel like you need a break!
We have a suggestion! Why not taking some time away from the huge conference center and explore Vienna for a little time? The weather is expected to be nice! But, of course, we do not want you to forget about our beloved cryosphere, so be sure to grab an ice cream at the right place!
Last but not least, do not forget about our cryo-social events!
Pre-Icebreaker Meet Up
When and Where:Sunday 8 Apr, from 16:00-18:00 at Cafe Merkur
The conference icebreaker can be a daunting experience to attend alone but it is a great event to go along to. We are organising a friendly pre-icebreaker meet up for cryospheric ECSs. We will meet up, have a chat, eat some cake and then head to the EGU conference centre together in time for the main icebreaker. Keep your eyes on the Facebook event for more details!
There will be a return of the infamous joint APECS and EGU Cryosphere division night out – come and join us for Viennese food and drinks and plenty of laughs! If you want to travel from the conference centre together, we will meet after the poster session at 18:50 at the main entrance (look for the blue and white EGU Cryosphere signs!) or you can meet us at Zwölf-Apostelkeller at 19:30. Two important things to do if you would like to come:
Please fill out Doodle poll to give us an idea of numbers!
Please remember to bring cash to pay for your own meal and drinks (it is possible to pay by card, but it will be very slow if 50+ people are trying to do it!)
Follow the Facebook event for updates and hopefully see plenty of faces old and new there 😀
EGU Cryosphere Blog and Social Media Team Lunch
When and Where: Wednesday lunchtime (12:15), on the left when looking at the main entrance.
Come along for an informal lunch meeting if you are already — or interested in getting — involved in the EGU Cryosphere team (which includes this blog and out social media channels). We will meet on the left of the main entrance to the conference centre at 12:15 and then we will decide on where to go depending on the weather. Don’t forget to bring your lunch with you. Please email Sophie Berger for more details.
When I think of snow, I tend to either think about the bright white ski slopes in the mountains or the large white areas in the Arctic. However, natural phenomena can lead to colorful snow. Our Image of the Week shows snow can be green! Snow can also turn orange, pinkish, grey and even yellow… But where do these different shades of snow come from?
The most common color of snow is white (see Fig. 2). Snow generally appears white when it is pure snow, which means that it is only an aggregate of ice and snow crystals. When sunlight meets the snow surface, all frequencies of the sunlight are reflected several times in different directions by the crystals, leading to a white color of the snowpack.
The pink-red-colored snow, commonly called “watermelon snow”, can also be caused by snow algae (see Fig. 3). The snow algae responsible for the pink color are similar to the ones responsible for green color. However, these algae use pigments of red color to protect their cells from high sunlight and UV radiation damage during the summer. Just like how we use sunscreen to protect our skin! The red pigments come either from iron tannin compounds or, more commonly, from orange to red-pigmented lipids.
There is also another origin for pink snow: Penguin poo! Indeed, the krill they eat contain a lot of carotenoids that give their poo a red color.
Yellow snow is the result of a different process (and no, it is not from Penguin pee!). Fig. 4 shows the Sierra Nevada in Spain before and after dust transported from the Sahara settled down on the snow-covered mountain tops. The dust was lifted up from the Sahara desert and blown north before ending its trip in Spain.
Fig. 4: Snow-covered Sierra Nevadas (Spain) before and after a dust deposition event [Credit: modified from NASA’s Earth Observatory]
Do these colors have an influence on snow cover?
In all cases of colored snow, the snow surface is darker than before. The darker surface absorbs more sunlight than a white surface, which causes the snow to melt faster… Therefore, although it looks artistic, colored snow is not necessarily healthy for the snow itself…
So, if you don’t like winter because everything is boring and white, just think about the variety of snow colors and try to look out for these special types! 🙂
Figure 1: Left - Mean sea-ice extent in September 2016 (month of minimum sea-ice extent) compared to the median ice edge between 1981 and 2010 (pink).
Right- Mean sea-ice extent in March 2017 (month of maximum sea-ice extent) compared to the median ice edge between 1981 and 2010 (pink). [Credit: Image courtesy of the National Snow and Ice Data Center]
The reduction in Arctic sea-ice cover has been in the news a lot recently (e.g. here) – as record lows have been observed again and again within the last decade. However, it is also a topic which causes a lot of confusion as so many factors come into play. With this Image of the Week we will give you a brief overview of the ups and downs of sea ice!
In general, Arctic sea ice is at its minimum extent at the end of the summer (September), and its maximum extent at the end of the winter (March). Our Image of the Week (Fig. 1) shows the summer and winter sea ice cover over the last year. In September 2016, the Arctic sea-ice minimum covered the second smallest extent since the beginning of satellite observations (38 years). Only 4.14 million square kilometres of the Northern Hemisphere were covered by sea ice on the day of minimum extent (September 10th). The maximum sea-ice extent was observed on March 7th 2017, only 14.42 million square kilometres of sea ice were observed: the lowest maximum since the beginning of satellite observations.
How long do we have until Arctic summer sea-ice cover is completely gone?
The Arctic Ocean is defined as ice-free, when the sea-ice area does not exceed 1 million km². Due to the close relationship between CO2 emissions and the sea-ice area (see one of our previous posts), it is likely that the summer Arctic sea-ice cover will fall below this threshold during the 21st century. Under the highest emission scenario (RCP 8.5 – IPCC, 2015), an almost ice-free Arctic in September is likely to occur before the middle of the century. It is, however, not easy to predict the exact year of an ice-free Arctic summer as the extent of the ice cover depends on many parameters influencing the freezing and melting of the ice.
On one hand, some parameters and their effect on the sea-ice cover are well understood and their future evolution can be projected quite well through climate models. For example, changes in the sea surface temperature tend to affect the starting date of the freezing period while changes in air temperature tend to affect the starting date of the melting period. As both air temperature and sea surface temperature are projected to increase in the long term, due to climate change, the period where ice can be present will be reduced more and more.
On the other hand, some parameters lead to several concurring effects, which are difficult to separate clearly and not always fully understood. Therefore, their future evolution and influence on sea ice is not totally clear. For example, the sea-ice loss leads to more open ocean areas, which absorb solar radiation, causing warming and therefore leading to faster sea-ice melting – a mechanism called “sea-ice albedo feedback”. At the same time, more open ocean areas also lead to more evaporation and therefore more clouds, which shield the ice from solar radiation and therefore lead to less warming of the ice and ocean surfaces.
Still, even if we knew the effect and long-term evolution of all these parameters, the exact date of ice-free Arctic could not be defined easily in advance. Why? The chaotic nature of the atmosphere leads to very short-term effects that influence the ice cover as well…
Be careful! A record minimum does not always mean a record maximum (and vice versa)!
On shorter time scales, sudden changes in the atmospheric circulation can have a large impact on sea-ice extent. Therefore, it is not guaranteed that a year with a record low maximum will have a record low minimum and vice versa. For example, heat waves and warm air outbreaks or high winds due to the transport of low pressure systems into the Arctic can lead to a more rapid decline of the sea-ice cover. The other way round, if the atmosphere from lower latitudes does not disturb the Arctic region, the sea-ice cover can stabilise again.
What about this year (2016/2017 season)?
Sometimes, it is not clear why sea-ice retreats rapidly. For example, the low 2016 minimum came as a surprise as the cover started with a very low minimum but then did not melt as fast as in previous years, due to average or below average temperatures. Only shortly before the minimum extent, stormy conditions came into play and led to the low extent that was observed (see Fig. 2).
In the last decades, although it recovered in some years between the record lows, the Arctic sea-ice cover has overall been declining. This is not the case on the other side of the planet, in Antarctica. Note that Antarctica is a complete different setting than the Arctic Ocean. The former being a continent surrounded by ocean and sea ice, the latter being an ocean with sea ice surrounded by continents.
Figure 3: Comparison of Antarctic sea-ice extent between different years for summer (left) and winter (right). [Credit: Image courtesy of the National Snow and Ice Data Center]
In recent decades, Antarctic sea-ice has been increasing very slowly (see Fig.3). Scientists were puzzled as such an evolution was not expected in a global warming framework. Explanations for this behaviour are that this is likely due to changing wind and surface pressure patterns around Antarctica. Contrary to this trend, this year (2016/2017) was a record low maximum and minimum in Antarctic sea-ice cover. This change is puzzling scientists even more. It remains unclear up to now if this is a permanent shift in the tendency of Antarctic sea ice or if this a single event. Be sure that the next months will be full of papers trying to explain this change in behaviour, it is going to be exciting!
Probability of a White Christmas in the USA (based on weather data from 1981-2010).
[ Credit : National Climate Data Center, NOAA. Via Wikimedia Commons]
Christmas is coming to town and in the Northern Hemisphere many of us are still dreaming of a white Christmas, “just like the ones we used to know”. But how likely is it that our dreams will come true?
What is the definition of a White Christmas ?
Usually Christmas can be defined as a “White Christmas” if the ground is covered by snow on either Christmas Eve or Christmas Day depending on local traditions. If you believe Christmas movies, it seems like Christmas was accompanied by snow much more often in the past than today! But is this really the case, or is it just the “Hollywood” version of Christmas? According to the UK Met Office White Christmases were more likely in the past. Due to climate change, average global temperatures are higher, which in many places reduces the chance of a White Christmas. However, the chances of a White Christmas also depend strongly on where you live…
Living in Western or Southern Europe, the Southern US or the Pacific coast of the US? Unlucky you!
Not too surprisingly, most of the inhabitants of Portugal, Southern Spain, and Southern Italy have probably never experienced White Christmas in their hometown. Maybe more counter intuitively the probability of a White Christmas is also low in most of France, the Netherlands, Ireland, and the Southern UK! In the US, the probability of a White Christmas increases from South to North, except on the Pacific Coast, which has a very low probability of a White Christmas.
Probability of a White Christmas in Europe (snow on the ground on 25th of December), inferred from reanalysis data (ERA Interim from 1979-2015). Probability [in %] increases from white to blue [Credit : Clara Burgard, Maciej Miernecki. We thank the ECMWF for making the data available]
What influences the probability of snowfall on Christmas?
The mean air temperature decreases with altitude and latitude, meaning that chances of a white Christmas increase the further North and at the higher you travel. However, coastal regions represent an exception. The air often has traveled over the ocean before reaching land. As the ocean is often warmer than the land surface in winter, the air in coastal regions is often too warm for snow to form. Additionally, in the Northern Hemisphere, ocean currents on the Western coast of the continents tend to carry warm water to high latitudes, while ocean currents on the Eastern coast tend to carry cold water to low latitudes. The probability of snowfall is therefore even lower in Western coastal regions (e.g. Pacific coast of the US, Atlantic coast of Europe).
Don’t despair !
If you want to increase your chances of experiencing a White Christmas, you have three solutions:
You already live in an area with high probability of White Christmas (lucky you!) – Sit tight and do a “snow dance”, here is one suggestion that we have heard works well:
We hope that you find a satisfactory solution that makes you happy this Christmas. Otherwise, remember that snow is not the only thing that defines Christmas. Enjoy the relaxed time with family and friends and prepare yourself for the coming new year! If you find yourself at a loose end, then there is always the back catalogue of EGU Cryosphere Blog posts to read – and we guarantee a healthy dose of snow and ice can be found here.
So, this is it from the EGU Cryosphere blog team for 2016. See you in 2017 – after all, the snow must go on…
MetOffice website with interesting facts around White Christmas!
Figure 1: Mean sea-ice thickness for November 2010 from CryoSat-2 satellite measurements. [Credit: The data processing of CryoSat-2 sea-ice thickness is funded by the German Federal Ministry for Economic Affairs and Energy (Funding: 50EE1008) and data of November 2010 are provided by meereisportal.de (Funding: REKLIM-2013-04) ].
Have you ever stumbled upon a satellite picture showing observations of the Arctic or Antarctic? You often see a circle where there is no data around the exact location of the geographic pole – as you can see in our Image of the Week. A few days ago, I wanted to explain this to one of my friends and turned to my favourite search engine for help. My search turned up a tremendous amount of stories and “scientific” studies about the Earth being hollow, with access to the centre of our hollow planet through these holes at the pole.
Obviously this is not the case. So here at the EGU Cryosphere blog we thought we’d better to set the record straight and explain the real reason for the “polar hole”.
Why do we need satellites?
Let’s start at the very beginning with how Earth observation data (e.g. temperature, ice cover, cloud cover, etc…) is collected. In the early days, measurements could only be collected pointwise, e.g. at weather stations (see Fig. 2) or by scientists traveling over land and by ship to specific locations. As a consequence, data coverage was very sparse and often clustered in places that were easily accessible, such as North America or Europe (Fig. 2). Additionally, measurements were even more sparse in hostile environments like the polar regions. It was therefore difficult to monitor these areas and study, for example, the evolution of polar ice sheets and sea-ice cover.
Since the 1970s, the use of satellites has greatly improved our ability to make remote observations around the world with a high spatial and temporal resolution, leading to much better monitoring of, for example, global weather and temperature. It has also allowed us to collect a vast amount of data in the difficult to access polar regions.
Figure 2: Map of the land-based long-term monitoring stations included in the Global Historical Climatology Network. Colours indicate the length of the temperature record available at each site. [Credit: created by Robert A. Rohde from published data and is incorporated into the Global Warming Art project]
Earth Observation Satellites
Satellites orbiting the Earth allow is to make remote observations and measurements of what is happening in the atmosphere and on the surface of the Earth. Earth observation satellites are divided in two categories according to the way in which they circle (orbit) the planet:
Geostationary satellites: orbit around the Earth’s Equator at an altitude of about 36000 km. They orbit in sync with the Earth (taking around 24h to complete a rotation) and therefore are always pointing at the same region (see video below). They provide observations of a given region on a high temporal resolution. However, given their location at the Equator, they do not cover the polar regions well.
Polar orbiting satellites: circle the Earth at a lower altitude around 850 km and their orbit is nearly perpendicular to the Equator. They are not in sync with the Earth’s orbit, circling the the Earth around once every 100 minutes. They therefore cross polar regions several times a day. Have a look at the video below to see how this works!
So…we have polar orbiting Satellites – why can can’t we “see” the poles?
The answer: sun-synchronous orbits!
To understand the data “hole” at the poles, we need to a little more detail about the path of polar orbiting satellites. To follow the evolution of a given point on Earth, it is useful for polar orbiting satellites to always cross that point at the same time of day – this way the angle of sunlight on the surface of the Earth is as constant as possible, resulting in a consistent series of images and observations over time . This is called a sun-sychronous orbit. To follow a sun-synchronous orbit, the orbit of the satellite has to be tilted at an angle from the geographic poles, thereby preserving the observed solar angle at the Earth’s surface .
Figure 3: These illustrations show 3 consecutive orbits of a sun-synchronous satellite with an equatorial crossing time of 1:30 pm. The satellite’s most recent orbit is indicated by the dark red line, while older orbits are lighter red. [Credit: NASA , illustration by Robert Simmon]
If you get a picture of all the trajectories of a sun-synchronous satellite, they will overlap (see video below), providing a seemingly closed picture. The only region that is not covered by the satellite is a circle (the size of the circle depends on the orbit tilt) around the geographic pole. This is the explanation for the data “hole” at the pole.
Sorry to debunk the myth but there is there is no hollow Earth that can be accessed through holes at the poles. The “Polar Hole” is a purely technical matter!
You remember last winter, when everything was white and you had so much fun building a snowman with your friends? What you see on this image above, is what you would see, if you took a tiny tiny piece of your snowman and put it under a low-temperature scanning electron microscope (SEM). The colours are called “pseudo colours”, they are computer generated based on the number of electrons reflected from a particular part of the image when scanned with a focussed beam of electrons. This is a standard technique used with SEM images to help identify patterns and structure in the image.
This week, let us take you on a journey…from the water vapour in a cloud to the snowman in your garden, to find out what leads to the complex structure you can see on our Image of the Week!
To this purpose we need to start at the very beginning, with a seemingly obvious question:
What is snow?
Snow is, very simply, precipitation in the form of ice crystals that originate in clouds. These ice crystals form when water vapour condensates directly to ice, without becoming liquid water.
From water vapour to ice crystals
For its formation, an ice crystal needs the atmosphere to be colder than the freezing point (0°C) and at least slightly humid. Additionally, like water droplets, ice crystals need a condensation nucleus, for example a dust or pollen particle, to start the growing process around – see the clip from Frozen Planet (below). During the transformation from water vapour to ice, an ice crystal always takes the form of a hexagon due to the way hydrogen and oxygen atoms bond to form water. Afterwards, the temperature and humidity of the atmosphere will shape the crystals and determine how fast they grow (see Fig. 2).
Between 0 and -60°C, the basic form (also called “habit”) of an ice crystal changes three times (near -3, -8 and -40°C). Between -3 and -8°C as well as below -40°C, the crystals take the form of six-sided plates, while between 0 an -3°C as well as between -8 and -40°C, the crystals form solid columns that are hexagonal in cross section.
Fig2: Ice Crystal Morphology diagram, indicating the basic form of ice crystals as a function of temperature and supersaturation (humidity). For more information see Libbrecht (2005) [Credit: K. Libbrecht]
In drier air, the growth occurs preferentially across flat surfaces, while in moister air, growth occurs preferentially at the tips, edges and corners. Also, the higher the moisture, the more water droplets can be absorbed, enhancing the growth rate and leading to more complex crystals. On the way between cloud and ground, an ice crystal will pass through layers of different temperature and moisture, leading sometimes to melting and refreezing on the way, changing parts of the crystal as well. All these factors together lead to the fact that finding two identical snowflakes is NEXT TO IMPOSSIBLE
From ice crystals to snowflakes
When an ice crystal becomes heavier than the surrounding air, it falls down. If it meets other ice crystals on its way to the ground, they will aggregate, forming new structures and the snowflake will grow. When reaching the ground, a snowflake can therefore be an aggregate of hundreds or sometimes even thousands of ice crystals.
From snowflakes to snowmen
Even when the snowflake has reached the ground, the journey is not finished. If the snow does not melt away rapidly, ice crystals will still modify their texture, size, and shape due to changing temperature and moisture conditions, melting and refreezing processes and/or compressing due to subsequent snowfall.
When the snow cover grows due to many subsequent snowfalls (over the winter or in cold areas over several years), a complex layered structure forms, made up of a variety of ice crystals, that reflect both the weather conditions at the time of deposition and the changes within the snow cover over time.
OR… you build a snowman out of it 🙂
Olaf the enchanted snowman from Disney’s 2013 animated feature film, Frozen. [Credit : Disney Wikia]
Figure 1: Arctic sea-ice area (75°N to 90°N) VS global annual mean surface air temperature for the complex model MPI-ESM-LR. Red dots represent a seasonal average over autumn (September/October/November) and blue dots over spring (March/April/May) of a particular year in the historical and future (RCP8.5) simulations. The dotted black lines result from a linear orthogonal regression in the ice area regime with the steepest change (Credit: subset of fig 2 from Bathiany et al, 2016 ).
Why do we care about sea ice in the first place?
Sea ice is important for several components of the climate system.
Due to its high albedo, sea ice reflects a high amount of the incoming solar radiation and is therefore relevant for the Earth’s energy budget.
Sea ice inhibits the exchange of heat, moisture and momentum between ocean and atmosphere, which usually occur at the sea surface.
Where sea ice forms, it releases heat and salt. When sea ice melts, it takes up heat and reduces the salinity of the surrounding water. Sea ice therefore redistributes heat and freshwater.
Sea ice provides habitat for plants and animals and hunting grounds for animals and indigenous populations.
Sea ice is an obstacle for shorter commercial shipping routes through the Arctic and oil and gas drilling.
The Arctic sea-ice cover is decreasing!
In recent decades, the Arctic sea-ice cover has been retreating rapidly. As we care about sea ice (see above!), scientists have been trying to understand this decline and to define a time span over which the sea-ice cover is expected to totally disappear (usually below 1 million km²). Up to now, research has mostly focused on the Arctic summer sea-ice cover, as this is expected to disappear much sooner than the winter cover. However, it is also of interest how winter sea-ice cover will evolve in the future and has evolved in the past.
What is meant by summer and winter sea-ice cover?
The Arctic sea-ice area follows a seasonal cycle with a maximum in late winter and a minimum in late summer (see figure below).
Figure 2: Arctic sea-ice concentration climatology from 1981-2010, at the approximate seasonal maximum (late winter) and minimum (late summer) levels based on passive microwave satellite data. (Credit : National Snow & Ice Data Center )
So, what about our Image of the Week?
In their study, Bathiany et al. (2016) compare the characteristics of the summer and winter sea-ice loss in the Arctic in general circulation models (GCMs). They investigate the changes in sea-ice area as a function of global annual mean surface air temperature. Summer sea-ice area (see red points) declines more linearly, “with no or a less pronounced change in slope”. Winter sea-ice area (see blue points), however, declines slowly at first and then more abruptly (this can still mean several years to decades, depending on the projection scenario used!). This abrupt decrease starts when ice volume is already very small.
How can this be?
Summer sea ice is distributed very heterogeneously over the Arctic, with very thick ice north of Greenland and Canada. It takes a given time (several years) until the thick multiyear ice (ice that has not melted during the previous summer) has melted. There can therefore still be ice in one location of the Arctic, while the rest of the area is ice-free. When these big “bunks” of ice have melted, then the summer sea-ice cover is gone. Large-scale abrupt shifts in sea ice therefore cannot occur in summer.
Winter sea ice, however, forms very homogeneously over the whole Arctic basin, when the ocean reaches the freezing temperature (the ocean temperature is relatively homogeneous over the basin). Warmer conditions in winter inhibit the growth of multiyear ice but a thin cover will always form on top of the ocean if the water is cold enough even if the ice melted in summer. Therefore, the sea-ice thickness and sea-ice volume decrease whereas the sea-ice area stays relatively constant and can still cover large areas (where the ocean is cold enough for ice to form). When the ocean does not reach the freezing temperature in winter, a large area of sea ice does not form any more and the sea-ice area declines abruptly.
What is the take-home message?
The explanation for the different behaviours in the retreat of summer and winter sea-ice is quite simple: the summer sea-ice cover disappears when all summer sea-ice has melted. The winter sea-ice cover disappears when no new ice forms in winter. As ice formation and ice melting are different processes governed by different mechanisms, the behaviour of the ice decline is different in both cases.
Note: These results are only relevant for the Arctic sea-ice cover as the Antarctic sea-ice cover is governed by different processes.