Keywords

1 Introduction

A small change in the atmospheric condition over a long time (climate) has always altered the sea ice condition over the poles (Alekseev et al. 2001). The existence of these, so-called refrigerators of the globe (poles) are very vital for the proper sustenance of Earth’s ecosystem (Gawor et al. 2016). The atmospheric circulation; mainly polar cells highly play a role in maintaining the Earth’s global temperature (Qian et al. 2015). In Antarctica, it is the ice or the glaciers that are formed over the land, while over the Arctic it is the seawater which freezes to form sea ice (Hellmer, 2004). In the Arctic, these ice are highly delicate as small changes in the atmospheric-oceanic interaction can alter their existence altogether. Some of the factors like temperature, wind and current highly weigh a major responsibility in causing variation in the formation and decay process of sea ice conditions (Laidler et al. 2008). On the contrary, over Antarctica, it is only the interaction with the atmosphere that needs to be considered. However one cannot overlook the fact that this continent has ice anchored to its shores—known as fast ice (Fraser et al. 2012). But the distinct characteristics of these ice are that they cannot be moved by winds or currents (Heil 2006). Therefore in the case of Antarctica, the oceanic factors influencing the condition of ice/glaciers can be neglected at least to a small extent. Due to the aforementioned concerns raised over the Arctic region as a whole, the regions over this circle are considered due to the scientific and potential attention it demands.

Therefore to understand the interaction of developing sea ice with the atmosphere above it, parameters like sea surface temperature (SST), temperature above 2 m (T2M) and outgoing longwave radiation (OLR) are considered throughout this paper. The Arctic climatic condition has already taken a one-eighty-degree turn due to the increase in greenhouse gases which are emitted by various human intervening anthropogenic activities (Wang and Overland 2012). As the amount of ice cover over the sea deteriorates, more appearance of open water enhances the warming effect by absorbing more heat into the system (Bintanja et al. 2011). The latent heat capacity of water is higher, it holds on to this heat for a prolonged period of time causing the openings to widen further. This is one of the reasons for the rise in winter warming and the weakening of the seasonal cycle (Bintanja and Van der Linden 2013). Such changes also attribute to the concurrent anomaly patterns of SST. SST over the Arctic are highly influenced by the formation of the North Atlantic oscillation (NAO) (Singh et al. 2013). This oscillation basically develops due to the pressure difference between the Azores and Iceland (Chepurin and Carton 2012). These NAO have the potential to alter the temperature at the surface to a great extent.

Further, the change in air temperature is also seen to impact the melting of sea ice to a greater extent. The rate of change in temperature above 2 m is often governed by the horizontal advection and diabatic heating of the atmosphere (Hurrell et al. 2003). Such horizontal advection is even likely to aggravate extratropical cyclones which may cause the transport of heat and energy in a poleward direction. Apart from the heat that prevails over the surface, it is vital to look into the radiative processes which also govern the atmosphere. The atmospheric energy budget is often maintained by the emission of longwave radiation back into the atmosphere (Serreze et al. 2011). A series of moisture and heat advection into the atmosphere alters the longwave fluxes of the Arctic (Hartmann 1994; Shah et al. 2020) causing large-scale variability in atmospheric circulation and thereby influencing the radiation budget over the region (Kapsch et al. 2016).

Considering all the changes in the atmosphere which are evidently visible, the study attempts to understand the influence of these parameters on the sea ice condition. Therefore the two seas, namely the Laptev Sea and the Greenland Sea are taken and their sea ice conditions are associated with the atmospheric variables to understand the degree of influence during both; the growth stage and decay stage of sea ice for the spans; 2012–16 and 2017–21 (five years each). The study yields a general understanding of the current state of these seas with respect to their cryospheric entities.

2 Study Domain

Detailed topography of the two considered marginal seas; the Laptev Sea and the Greenland Sea is mentioned in the upcoming section.

2.1 Laptev Sea

Laptev Sea is a marginal sea of the Arctic Ocean off the coast of Northern Siberia bounded by the Taymyr Peninsula and the islands of Severnaya Zemlya on the west and by the New Siberian Islands and Kotelny Island on the east (Cox et al. 2016), with coordinates 105–150°E and 63–79°N (Fig. 1). Laptev Sea is known as the birthplace of iceFootnote 1 and is, therefore, the chief source of Arctic sea ice. The sea has an average outflow of 483,000 km2/year (estimated during the period 1979–95) (Ghosh 2021). It is therefore considered one of the significant regions of net ice production and export (Itkin and Krumpen 2017, Zakharov 1966). Laptev is extremely shallow, with water depths between 15 m and 200 m (Dethleff et al. 1998), and comprises an area of ~500 × 103 km2. The sea ice over the sea is formed towards the coast during early winter, later it drifts west across the Arctic, before breaking up in the spring into the Fram Strait located between Greenland and Svalbard.

Fig. 1
A, a map marks the Laptev Sea and Greenland Sea over the Arctic Circle. B, a map of Greenland Sea. C, a map of Laptev Sea.

Region of interest a Location of Laptev and Greenland Sea over the Arctic Circle, b Laptev Sea and c Greenland Sea

Full size image

If ice is formed after winter then it will be thinner compared with the previous one (during winter) and thus is more likely to melt before it reaches the Fram Strait. The topography of the sea is also seen to have an intense impact. The sea experiences heat waves from its neighbouring part of Siberia. During this phase, the surface air temperature was reported to be 38 °CFootnote 2 over the south of Laptev Sea. Often this heat was an aftermath of the winds that predominantly brought hot continental air towards the seaFootnote 3 further making the sea ice push itself north.

2.2 Greenland Sea

The Greenland Sea is the outlying portion of the Arctic Ocean. It has an area of 1,205,000 km2 and a depth of 1450 m. The sea is surrounded by Greenland in the west, Svalbard Island in the east, the Arctic Ocean in the north and the Norwegian Sea and Iceland in the south,Footnote 4 with coordinates 30°W–10°E and 63–79°N (Fig. 1). Greenland Sea has an acute role in deep ocean convection (Timokhov 1994). The intensity of these convections is often governed by the heat and freshwater fluxes contributing to buoyancy and the buoyancy advection of the region (Brakstad et al. 2019). The aftermaths of such processes are the release of brine during the formation of ice from water, causing the salinity of the underlying water to increase (Rudels, 1990). The freezing process may indirectly cause water to sink into deeper layers, which is a precondition since the fresher water at the surface requires higher density to penetrate the denser water below (Visbeck et al. 1995).

3 Data and Methods

To understand the ocean—atmospheric coupled with cryospheric conditions of the two seas—the Laptev Sea and the Greenland Sea, certain atmospheric and cryospheric parameters are considered for the span of 2012–21. The data acquisition sources along with the methodology adopted are described in subsequent sections.

3.1 Data Sources

3.1.1 Sea Ice Data

The sea ice parameters used for the current study are: sea ice area (SIA), sea ice extent (SIE) and sea ice concentration (SIC). The timeline of 2012–21 is divided into two spans, i.e. 2012–16 and 2017–21. This is in order to compare the latter time period with the former one. Additionally, the recent decade is selected to understand the current influence of temperature on sea ice growth and decay, as it is already recorded by the scientist that a significant incline in temperature is observed during the selected span compared with the past 40 years. All three variables are retrieved from Near-Real-Time DMSP SSMIS Daily Polar Gridded Sea Ice Concentrations (NSIDC)Footnote 5 having a spatial resolution of 25 × 25 km.

  • SIA: It is the percentage of sea ice within each data cell and is added up to report how much of the Arctic is covered by ice; the area typically uses a threshold of 15%,Footnote 6 measured in km2.

  • SIE: It defines a region as ice-covered or not ice-covered. For each satellite data cell, the cell is said to either have ice or to have no ice, based on a threshold. The most common threshold used by NSIDC is 15%, meaning that if the data cell has greater than 15% SIC, the cell is considered ice-covered; less than that is considered to be ice-free.Footnote 7

  • SIC: This describes the relative amount of area covered by ice, compared with some reference area. Thus, concentration describes how much of a 25 × 25 km box is covered by sea ice. Ice concentration typically is reported as a percentage (0% to 100% ice).

3.1.2 Atmospheric Reanalyses

The reanalyses atmospheric data of SST, T2M and OLR are downloaded from ERA-5 at a spatial resolution of 27.75 × 27.75 km for the spans 2012–16 and 2017–21.

  • SST: It is the temperature of seawater near the surface. In ERA-5, SST is given by two external providers. Before September 2007, SST is from HadISST2 and from September 2007 onwards, it is from OSTIA. This parameter has units in kelvin (K).Footnote 8

  • T2M: This parameter is the temperature of the air at 2 m above the surface of the sea. 2 m temperature is calculated by interpolating between the lowest model level and the Earth's surface, taking account of the atmospheric conditions. This parameter is also measured in kelvin (K) (See Footnote 8).

  • OLR: The thermal (longwave) radiation emitted to space at the top of the atmosphere is commonly known as outgoing longwave radiation (OLR). This parameter is accumulated over a particular time period which depends on the data extracted. For the reanalysis, the accumulation period is over 1 h ending at the validity date and time. The unit of the data at the time of retrieval is J m−2 (See Footnote 8). But for the current study, it is converted to Wm−2.

3.2 Methodology

After the acquisition of cryospheric (SIA, SIE and SIC) and atmospheric (SST, T2M and OLR) datasets from the above-mentioned domains, the method of interpolation and extrapolation is done to bring each dataset to one spatial resolution, in this case, 27.75 km. Data are only retrieved for the specific latitude–longitude of Laptev and Greenland seas using shape files of these regions. NaN values are removed, and the void pixels are assigned 0 value for the prompt averaging of data points. The daily data are averaged to obtain monthly data for each variable. The time frame 2012–21 is divided into two spans as; 2012–16 and 2017–21. Splitting the time domain into two gives a better understanding of the trend of events happening in five years each. Researchers in the past have already contributed a lot to the decline of sea ice over the entire region of the Arctic. It is established that ~13% is the rate of decline every decade (Meehl et al. 2018; Simon et al. 2022; Zhang et al. 2016). Further, the differences between the two periods are taken for the parameters, SIA and SIE to understand the significant rise/fall in sea ice conditions prevailing over the seas. Later, regression analyses of these parameters with the atmospheric parameters are done, to understand the influence of temperature and radiation in sea ice growth/decay. The decay phase consists of months from April–September and the growth phase consists of months from October–March. Separate analyses of these two seasons help one understand the life cycle of sea ice better.

4 Results and Discussion

4.1 Spatial Variability in Sea Ice Condition

4.1.1 Laptev Sea

The spatial plot (Fig. 2) shows SIC over Laptev during the decay phase (April–September). For the months from June to August, SIC steadily decreases. In June, the sea is seen to experience SIC greater than 61%. However, in July the sea shows SIC smaller than 73%. Ultimately in August, the SIC is found to be less than 25% and open waters are also seen appearing. This pattern of melt in sea ice is quite normal because only during these months, summers are seen intensifying with the rise in temperatures (rise in both SST and T2M). While comparing the two time zones, i.e. 2012–16 with 2017–21, it is understood that SIC during the latter period is less than that during the former years. While comparing the month of June (Fig. 2a, b) shown in the two timelines, it is found that during 2012–16 (Fig. 2a), the concentration of sea ice is above 85% at a majority of the locations. However, only a small area at the centre of the sea has a concentration below 37%. Even at locations above Taymyr Peninsula (78–76°N, 110°E), denser sea ice is formed.

Fig. 2
6 spatial plots present the percentage variations of sea ice concentration over Laptev Sea during June 2012 to 2016, June 2017 to 2021, July 2012 to 2016, July 2017 to 2021, August 2012 to 2016, and August 2017 to 2021, respectively.

Spatial variation of sea ice concentration (SIC) over Laptev Sea during a June, 2012–16, b June, 2017–21, c July, 2012–16, d July, 2017–21, e August, 2012–16 and f August, 2017–21

Full size image

On the contrary, for the year 2017–21 (Fig. 2b) in June, higher latitudes exceeding 76°N have denser and homogenous SIC with 97% SIC. But over the sea, SIC of less than 37% is abundantly present when compared with that in June 2012–16. The greater decline in SIC values in June 2017–21 is due to two reasons. Firstly, the SST value in 2012–16 (June) is 271.95 K and that during 2017–21 is 272.24 K, showing a significant rise in the latter period. Therefore rise in SST can cause a decline in SIC. Secondly, there exists ~ 1 K rises in T2M values. T2M in 2012–16 is 280.03 K and that during 2017–21 is 281.87 K. The warmth evolved at the surface of the sea will definitely cause the air above it to heat up, ultimately contributing to the rise in T2M. By July 2012–16 (Fig. 2c), majority of sea ice have come down to concentrations below 85%. While comparing it with the past month (June), the concentration has significantly decreased over the entire sea which is as expected during the melt. Some locations within the sea show SIC dropping down even below 25%, regions such as south of Severnaya Zemlya, north of New Siberian Islands and to the north of Lena River Delta (LRD). While comparing this span with the recent years 2017–21 (Fig. 2d) it is visible that the sea has tremendously lost ice. Regions east of LRD however remain somewhat similar (which is an exceptional case). An arc-like contour (connecting the LRD and the New Siberian Islands) which is also found during the former span has considerably lost a huge amount of ice, which has caused SIC to fall below 13%. Similar is the situation with sea ice located towards the east of Taymyr Peninsula. During this month the south part of the sea shows a temperature of 279 K and the north part exhibits temperature almost above the freezing point. Additionally, during this season, the shores observe temperatures of 295–297 K, due to which sea ice towards the land region (Russia) is less dense than the portions above it. However, July which is the primary month of the decay season is demarcated for bringing in excessive fog and snow squalls, which may contribute to the development of sea ice at times.

Further in August 2012–16 (Fig. 2e), the entire SIC of the sea drops to range below 25%. The region around New Siberian Islands is found to be completely ice-free, even when the location is close to the central Arctic. Additionally, the region around LRD also shows a slight reduction of sea ice and a gradual bend towards ice-free conditions. August being very close to September (which is demarcated as the month with Arctic sea ice minimum condition), such a large decline in SIC can be anticipated. Much more alarming is the state of sea ice during the latter span, 2017–21 for the month of August. From the spatial plot (Fig. 2f) it is clearly visible that the entire sea is absent of any concentrations above 13%. Researchers in the past have observed low-pressure systems getting developed over the Arctic during this month. Due to the existence of such low-pressure systems, the ice tends to move towards the southward direction. When this ice gets too low to come in contact with the waters which have water temperatures below the freezing point, the prevailing sea ice tends to melt contributing more to ice-free regions. This may be the causative factor attributing to the decline in sea ice conditions. Apart from the spatial comparison of sea ice during the decay phase for the two spans; 2012–16 and 2017–21, quantitative calculations are also made and are visible in Table 1. Table 1 displays the difference in SIA and SIE for each month. Here, both SIA and SIE are represented in million square kilometres. Therefore, when the differences are zero (or null), it does not indicate that the values of these parameters in both spans are the same. Instead, it shows that the difference is not quite significant.

Further in August 2012–16 (Fig. 2e), the entire SIC of the sea drops to range below 25%. The region around New Siberian Islands is found to be completely ice-free, even when the location is close to the central Arctic. Additionally, the region around LRD also shows a slight reduction of sea ice and a gradual bend towards ice-free conditions. August being very close to September (which is demarcated as the month with Arctic sea ice minimum condition), such a large decline in SIC can be anticipated. Much more alarming is the state of sea ice during the latter span, 2017–21 for the month of August. From the spatial plot (Fig. 2f) it is clearly visible that the entire sea is absent of any concentrations above 13%. Researchers in the past have observed low-pressure systems getting developed over the Arctic during this month. Due to the existence of such low-pressure systems, the ice tends to move towards the southward direction. When this ice gets too low to come in contact with the waters which have water temperatures below the freezing point, the prevailing sea ice tends to melt contributing more to ice-free regions. This may be the causative factor attributing to the decline in sea ice conditions. Apart from the spatial comparison of sea ice during the decay phase for the two spans; 2012–16 and 2017–21, quantitative calculations are also made and are visible in Table 1. Table 1 displays the difference in SIA and SIE for each month. Here, both SIA and SIE are represented in million square kilometres. Therefore, when the differences are zero (or null), it does not indicate that the values of these parameters in both spans are the same. Instead, it shows that the difference is not quite significant.

Table 1 Monthly variation in SIA, SIE along with the difference in the two parameters for the two spans; 2012–16 and 2017–21 over Laptev Sea
Full size table

The highest difference between the two is observed in the month of July and October with values of 1.11 × 105 km2 and 1.17 × 105 km2 respectively. SIE during 2012–16 is almost seen to be steady with a value of 8.72 × 105 km2 throughout winter. The lowest value in SIE is observed during September (1.19 × 105 km2). Similar to 2012–16, in 2017–21 also the highest value remains static. An understanding of this steadiness is that in millions this might appear constant, but there might be very slight differences if we lower the place values. The minimum value is during September (1.05 × 105 km2). While comparing the difference in SIE, it remains almost insignificant during DJFMA. Here also the highest difference in SIE is observed in July and October with values of 1.57 × 105 km2 and 1.55 × 105 km2 respectively.

4.1.2 Greenland Sea

The Greenland Sea is extremely further away from the Central Artic region and closer to the North Atlantic Ocean, the Atlantic waters with temperatures ~287.85 K are seen to greatly influence the sea ice conditions over the region compared with the Arctic waters. In Fig. 3 the SIC is found to exhibit a layering effect with locations closer to the Norwegian Sea showing low SIC and the region near Greenland showing higher SIC. During June 2012–16 (Fig. 3a), starting from the north near Svalbard, the sea is devoid of SIC above 65%. However as one moves towards the south, the region of sea closer to Greenland experiences higher SIC than compared with those near the Norwegian Sea (Table 2).

Fig. 3
6 spatial plots present the percentage variations of sea ice concentration over Greenland Sea during June 2012 to 2016, June 2017 to 2021, July 2012 to 2016, July 2017 to 2021, August 2012 to 2016, and August 2017 to 2021, respectively.

Spatial variation of sea ice concentration (SIC) over Greenland Sea during a June, 2012–16, b June, 2017–21, c July, 2012–16, d July, 2017–21, e August, 2012–16 and f August, 2017–21

Full size image
Table 2 Monthly variation in SIA, SIE along with the difference in the two parameters for the two spans; 2012–16 and 2017–21 over Greenland Sea
Full size table

The Greenland region excessively rich in snow and glaciers would have facilitated the sea ice to develop at least to a few magnitudes. Again in June 2017–21 (Fig. 3b) the concentrations are seen to be greater in the north; towards the south, the decline is visible.

The locations near Iceland have SIC of less than 49% and the width of the trail is also seen to be low. Overall during June, some regions show high SIC during the recent span compared with that during the latter span, but this does not suffice that the number of pixels during 2017–21 is considerably less compared with that during 2012–16. By July 2012–16 (Fig. 3c), SIC has significantly deteriorated from higher SIC values by June 2012–16 and this trend is obvious as we have considered the decay process of sea ice. Here SIC greater than 85% is confined to the northern locations. The area closer to Arctic Circle might add to an advantage for the growth in sea ice towards the north of the Greenland Sea. Towards the south and close to the Iceland region, the SIC is always found to be less than 37%. Comparing this with the latter span of 2017–21 (Fig. 3d), the area of spread of SIC in the south of the Greenland Sea has noticeably decreased. However, the region around Svalbard is showing slight development of SIC in the recent span than in the previous span. By August 2012–16 (Fig. 3e), the entire sea is devoid of any sea ice over the south. In the north, near the Arctic, SIC is greater than 85%. However with time, by 2017–21 (Fig. 3f), it has significantly decreased.

4.2 Daily Variability in Temperature (T2M) Over Laptev and Greenland Seas

As we now know how sea ice condition over Laptev and Greenland Seas prevails, therefore it is vital to look into the behaviour of T2M over the considered seas. Figures 4 and 5 show the daily temperature variation for the considered years; 2012–16 and 2017–21 over the Laptev Sea and Greenland Sea. Over Laptev (Fig. 4), from a glance at the plot, one can highlight that the summers during 2017–21 are much hotter than those during 2012–16. June, July and August are found to show extremely high temperatures.

Fig. 4
The profile of daily temperature variation over the Laptev Sea is represented in various colors, for the years 2012 to 2016 and 2017 to 2021.

Daily variability profile of temperature (T2M) over Laptev Sea during 2012–16 and 2017–21

Full size image
Fig. 5
The profile of daily temperature variation over the Greenland Sea is represented in various colors, for the years 2012 to 2016 and 2017 to 2021.

Daily variability profile of temperature (T2M) over Greenland Sea during 2012–16 and 2017–21

Full size image

Overall the range of T2M over Laptev lies within the range of 238–285 K. By the end of June in 2012–16, T2M starts to rise from 280 to 281 K. Soon 9th July onwards, the value again increases to 282 K and further increases on 18th July to 283 K after which, a gradual decline to subsequent values are observed. Apart from July, the hike in the value of T2M is also observed during June, but here the T2M values are limited to 281 K. While comparing these temperatures with those during 2017–21, it is evident that the values during this span are significantly higher than those of the former span. When the temperature reaches 280 K by the end of June 2012–16, the value has already reached 284 K in 2017–21. The highest fluctuation in the value of T2M here in the latter span is also observed during June–July like that of the former span. Sea ice decay is usually from April to September. Here in the temperature profile, it is distinctly visible that during these months of the year, the air above the sea is extremely warm when compared with other months. The heat exchange between the atmosphere and the sea further contributes to the melt away of the newly formed sea ice over the region. It is also understood that the latent heat capacity of water is much higher than that of soil/land, therefore the water molecules over the sea hold up the heat for a longer duration and gradually emit it in the form of ice-albedo feedback or longwave infrared radiation into the atmosphere which again facilitates the air above it to gain more and more heat (Meehl el al. 2018). This cyclic exchange of heat between the two systems hampers not only the formation of new ice but also causes the melt away of multiyear ice contributing to the appearance of more open water over the region.

While understanding the variation of T2M over the Greenland Sea (Fig. 5) it is found that the range of T2M is significantly higher than those of the Laptev Sea. The lower range of T2M over the Laptev Sea is about 238 K, but over the Greenland Sea, it has increased to 260 K, however when observing the highest boundary limit, it is found that the temperature is confined to 279 K. During 2012–16, the summer T2M lies between 270 and 279 K. A seasonal cyclic pattern is observed for T2M with the temperature slowly building up in July–August and further subsiding. In 2017–21, the summer is seen to exist for a prolonged period, this can be justified by the higher values (above 273 K) that exist in May and October which is quite absent in the previous years; 2012–16. Also during July–August the temperatures are seen to be higher than that in other months during 2017–21 when compared with 2012–16. Rest the cyclic pattern of T2M with highs and lows are also visible during this period as that of 2012–16. Similar to the Laptev Sea, the Greenland Sea has higher values of T2M during 2017–21 which is a reason for the low values of SIC, SIA and SIE.

4.3 Linear Regression Analyses of Sea Ice Area (SIA) with Sea Surface Temperature (SST), Temperature at 2m (T2M) and Outgoing Longwave Radiation (OLR)

4.3.1 Laptev Sea

Monthly variability in temperatures during the summers for the considered years over the Laptev Sea provides the need to extend the study quantitatively to establish the role of these in the sea ice retreat process. Therefore linear regression analyses are performed between SIA and atmospheric entities such as; SST, T2M and OLR and are displayed in Table 3. Also the byproducts of the regression analyses such as; the slope, intercept and correlation coefficient values are all mentioned in Table 3, which gives one valid grounds to associate the two parameters quantitatively. From the previous section (Sect. 4.1), the spatio-temporal pattern of SIC clearly shows the recent stratification which is happening over the sea around the year. The stratification is primarily due to the gush of fresh waters from the Lena River and the sea ice melt waters. Additionally, the presence of an ice shelf inhibits the exchange of energy and momentum between the ocean and the atmosphere (Simon et al. 2022). It is also known that the deep shelf region of the sea consists of low temperature saline waters which are not significantly affected by seasonal changes/fluctuations. However, the interaction of the temperature above the surface and the air above it with the sea ice can be understood by associating SST and T2M with SIA.

Table 3 Results of regression analyses of SIA with SST, T2M and OLR during the growth and decay of sea ice for the two spans; 2012–16 and 2017–21 over Laptev Sea
Full size table

During the growth of sea ice in both periods, SST lies within the range between 271–275 K. Majority values are seen highly concentrated around lower SST and higher SIA values. Regression analyses of SST with SIA yields r2 = 0.69 (p < 0.05) during 2012–16 and r2 = 0.92 (p < 0.05) during 2017–21, showing SST has a greater influence on sea ice during the recent span (however the p values state that the two r values are statistically insignificant). Further during the decay stage of sea ice decrease in SIA is found with an increase in SST for all the years. Here, the range of SST is found to be slightly higher than that during winters, which also adds to the excess decay of sea ice during summers. The correlation coefficient (r2) values are 0.92 (p > 0.05) and 0.89 (p > 0.05) for 2012–16 and 2017–21, respectively. During winters, it is found that the impact of SST on SIA is greater in 2017–21 than in 2012–16, however during summers, it is the opposite, i.e. the influence of SST on SIA is more in 2012–16 than in 2017–21. The results directly point out the recent winter amplification which is prevailing over the Arctic as a whole.

While understanding the relationship of SIA with T2M during winters, it is found that the area remains almost constant (8–9 × 105 km2) with a decrease in T2M from 235 to 255 K. This trend is the same for all the considered periods. The regression analyses performed between the two quantities show r2 = 0.74 (p < 0.05) and r2 = 0.76 (p < 0.05) for the years 2012–16 and 2017–21, respectively. Further, during summer (decay of sea ice), an exact fit of these parameters is not found, which is validated by the correlation coefficient values mentioned in Table 3.

For the years 2012–16 and 2017–21, r2 = 0.37 and 0.49 (p < 0.05), respectively. In the case of OLR, SIA is seen to increase with the increase in OLR during the growth stage and during decay, SIA is found to decrease with a decrease in OLR, indicating that both the parameters are directly associated; as one increases the other also increases. Correlation coefficient values are r2 = 0.71 and 0.76 (p > 0.05) during growth and r2 = 0.19 and 0.34 (p < 0.05) during decay, for years 2012–15 and 2017–21, respectively.

Overall the parameters SST and T2M are inversely associated with SIA while OLR is positively related. The high degree of influence among SIA with SST and T2M is observed during winter months; precisely the influence is more during the recent span than the previous years. Similarly larger influence of OLR on SIA is also found during the decay period than in the growth stage.

4.3.2 Greenland Sea

Greenland Sea located towards the south of Arctic Circle is greatly influenced even by a slight hike in temperature. Additionally, over this region, the NAO is observed to show high impact. One of the reason by which the oscillation is developed is through the pressure difference which originates between the Azores Island and locations closer to Iceland. Azores Island shows high pressure when Iceland observes low pressure or vice versa. The other causative factors may be associated with wide range of physical and biological responses including; the variation in sea ice cover, wind speed, sensible heat flux, latent heat flux, temperatures, OLR, etc. After 1990s it is known that SST is highly correlated with NAO. It is also believed that the determination of NAO helps one know the Arctic atmospheric condition to one third extent (Panicker et al. 2021).

Here, Fig. 6 regression analyses are performed between SIA and atmospheric variables; like SST, T2M and OLR during both the decay and the growth months of sea ice for all the considered spans; 2012–16 and 2017–21. During the growth of sea ice, gradual development of sea ice is observed in Fig. 6a (as one moves from right to left). When the temperature falls from 279 to 275 K, the plot clearly shows the inverse relationship between the two variables. SST and SIA are negatively correlated for both the timelines; 2012–16 and 2017–21 (r = −0.94). The value r2 = 0.88 (p > 0.05) and r2 = 0.89 (p > 0.05) during the aforementioned periods clearly demonstrates that the two quantities are highly negatively correlated (as; r2 > 0.5). The information related to intercept, slopes, etc. are clearly addressed in Table 3. However during the decay stage of sea ice, primarily when it is summers, the range of SST is seen to be noticeably high than that during winters. Here the highest point in the range reaches up to 282 K additionally, the observed slope in Fig. 6b is much steep than those in winters. During 2012–16, r2 = 0.90 and 2017–21, r2 = 0.87, which shows that during the former span the sea ice is vastly influenced by the change in SST than during the latter years. While relating T2M with SIA (Fig. 6c and d), the concentration of higher SIA values are observed in places where the temperature is low. The relationship between the two entities is again expected to be strongly opposite with r2 = 0.70 and r2 = 0.84 (p > 0.05) in 2012–16 and 2017–21, respectively. Further relating SIA with OLR gives new insight to the study.

Fig. 6
4 scattered and line graphs labeled from A to D. A and B, depict S I A times 10 to the power 5 versus S S T. C and D, depict S I A times 10 to the power 5 versus T 2 M. Each graph plots 2 lines and 2 types of dots, with decreasing trends. The dots are plotted for the years 2012 to 2016 and 2017 to 2021.

Regression analyses of Sea Ice Concentration (SIA) with Sea Surface Temperature (SST) during a growth phase, b decay phase of sea ice, Sea Ice Concentration (SIC) with Temperature at 2m (T2M) during c growth phase and d decay phase over Greenland Sea during the years 2012–16 and 2017–21  

Full size image

Even though OLR is a function of temperature, it can yield the amount of radiation emitted by ice or snow when treated as a black body (Zhang et al. 2016; Shah and Srivastava 2020). The correlation coefficient value clearly demonstrated that the existence of sea ice is directly related with OLR. As one increases, the development of the other also occurs. During sea ice growth the range of OLR lies between −180 Wm−2 and −152 Wm−2 and SIA between 1 × 105 km2 and 6 × 105 km2. When OLR increases the growth of sea ice also occurs. The correlation coefficient values of OLR and SIA are not strong which are visible from the values in Table 4, r2 = 0.37 for the both the periods. However, during the decay phase of sea ice the range of OLR has dropped down from −204 Wm−2 to −168 Wm−2. During 2012–16, r2 = 0.20 and during 2017–21, r2 = 0.16, which states that the relationship between the two parameters are not exceedingly strong.

Table 4 Results of regression analyses of SIA with SST, T2M and OLR during the growth and decay of sea ice for the two spans; 2012–16 and 2017–21 over Greenland Sea (r2 value have p < 0.05)
Full size table

Overall it can be stated that over Greenland Sea, SIA holds negative correlation between SST and T2M and positive correlation with OLR (at least to a small extent). The notable feature is that better correlation values are observed more during the growth phase when compared with that during the decay phase. SST and T2M highly show negative influence on SIA during winter than in summers. This relation suggests Greenland Sea experiences more warming during winter, than in summer. This occurrence is not unique to this sea, but there exists a well-known concept termed as Arctic Amplification (AA), which is experienced by majority of the seas in the Arctic Circle.

The results represented in the paper demonstrates the spatio temporal variation in sea ice conditions over the two seas of Arctic, namely Laptev Sea and Greenland Sea for the spans; 2012–16 and 2017–21. Laptev being a marginal sea of the Arctic shows high variability in SIC, SIA and SIE compared with Greenland Sea which is only an outlying portion of the Arctic Ocean. Even while quantitatively comparing the amount of sea ice over Laptev Sea and Greenland Sea, the range of SIA and SIE is noticeably high for Laptev when compared with Greenland Sea. The differences between the two seas in SIA and SIE are ~ 3 × 105 km2 and ~ 1 × 105 km2, respectively. Additionally when comparing the recent span with the former span, over all the considered seas the decrease in sea ice condition is evidently visible with the passage of time. Over Laptev the highest difference in SIA between the two periods; 2012–16 and 2017–21 is observed in the month of July and October with values 1.11 × 105 km2 and 1.17 × 105 km2, respectively. Here also the minimum value is during September (1.05 × 105 km2). The difference in SIE is also found to be in July and October with values 1.57 × 105 km2 and 1.55 × 105 km2, respectively. However the difference in SIA/SIE over Greenland Sea is less than that over Laptev Sea. The highest difference in SIA is observed during September, 0.54 × 105 km2. The SIE trend is similar to that of SIA. The difference in SIE is highest in September and lowest in March.

While associating SIA with atmospheric parameters (SST, T2M and OLR), over both the regions, SST and T2M are inversely associated with SIA while OLR is positively related with SIA. The high degree of influence among SIA with SST and T2M is observed during winter months; precisely the influence is more during the recent span than the previous years. Similarly larger influence of OLR on SIA is also found during the decay period than in the growth stage. This clearly makes the sea suspicious of have been experiencing winter warming more than summer warming. This is not only in this case; this effect of Arctic Amplification (AA) is a known concept which is already being experienced by majority of the seas over the Arctic Circle. Over Laptev, in 2017–21 r(SIA, SST) = −0.96 and r(SIA, T2M) = −0.87 which is higher compared with r(SIA, SST) = −0.83 and r(SIA, T2M) = −0.86 during 2016–21. Similarly over Greenland Sea, r(SIA, SST) = −0.94 and r(SIA, T2M) = −0.91 during 2017–21 and r(SIA, SST) = −0.94 and r SIA, T2M) = −0.84 during 2012–16. Since outgoing longwave radiation has a positive influence on both the seas during both the spans; r(SIA, OLR) = 0.85 for Laptev and r(SIA, OLR) = 0.61 for Greenland Sea.

Given the relationship of SIC, SIA and SIE with atmospheric conditions over Laptev and Greenland Sea, it is now understood that small change in these entities may be impacting the ice condition in inverse manner causing more warming of the region as this feedback mechanism moves in a cyclic manner. Additionally, there is extensive change experienced in the sea ice condition over both the seas during the recent decade.

5 Conclusions

The results represented in the paper demonstrate the spatio temporal variation in sea ice conditions over the two seas of Arctic, namely Laptev Sea and Greenland Sea for the spans; 2012–16 and 2017–21. Laptev being a marginal sea of the Arctic shows high variability in SIC, SIA and SIE compared with Greenland Sea which is only an outlying portion of the Arctic Ocean. Even while quantitatively comparing the amount of sea ice over Laptev Sea and Greenland Sea, the range of SIA and SIE is noticeably high for Laptev when compared with Greenland Sea. The differences between the two seas in SIA and SIE are ~3 × 105 km2 and ~1 × 105 km2, respectively. Additionally when comparing the recent span with the former span, over all the considered seas the decrease in sea ice condition is evidently visible with the passage of time. Over Laptev the highest difference in SIA between the two periods; 2012–16 and 2017–21 is observed in the month of July and October with values 1.11 × 105 km2 and 1.17 × 105 km2, respectively. Here also the minimum value is during September (1.05 × 105 km2). The difference in SIE is also found to be in July and October with values 1.57 × 105 km2 and 1.55 × 105 km2, respectively. However the difference in SIA/SIE over Greenland Sea is less than that over Laptev Sea. The highest difference in SIA is observed during September, 0.54 × 105 km2. The SIE trend is similar to that of SIA. The difference in SIE is highest in September and lowest in March.

While associating SIA with atmospheric parameters (SST, T2M and OLR), over both the regions, SST and T2M are inversely associated with SIA while OLR is positively related with SIA. The high degree of influence among SIA with SST and T2M is observed during winter months; precisely the influence is more during the recent span than the previous years. Similarly, larger influence of OLR on SIA is also found during the decay period than in the growth stage. This clearly makes the sea suspicious of have been experiencing winter warming more than summer warming. This is not only in this case; this effect of Arctic Amplification (AA) is a known concept which is already being experienced by majority of the seas over the Arctic Circle. Over Laptev, in 2017–21 r(SIA, SST) = −0.96 and r(SIA, T2M) = −0.87 which is higher compared with r(SIA, SST) = −0.83 and r(SIA, T2M) = −0.86 during 2016–21. Similarly, over Greenland Sea, r(SIA, SST) = −0.94 and r(SIA, T2M) = −0.91 during 2017–21 and r(SIA, SST) = −0.94 and r(SIA, T2M) = −0.84 during 2012–16. Since outgoing longwave radiation has a positive influence on both the seas during both the spans; r(SIA, OLR) = 0.85 for Laptev and r(SIA, OLR) = 0.61 for Greenland Sea.

Overall, the paper demonstrates how the variability of atmospheric temperature inherently affects the condition of sea ice over the two seas of the Arctic. Knowing the relationship and the extent of the effect of one on the other variable helps the scientific community to predict and make necessary changes at the global level to limit the excess melt away of sea ice over the seas of the Arctic. Additionally, further scientific studies are required to understand other atmospheric variables which hamper the pace and rate of sea ice melt and growth over the region.