remote sensing

Article Remote Sensing of Ice Conditions in the Southeastern and in the and Validation of SAR-Based Ice Thickness Products

Igor E. Kozlov 1,2,* , Elena V. Krek 3, Andrey G. Kostianoy 3,4,5 and Inga Dailidiene˙ 2,6

1 Satellite Oceanography Laboratory, Russian State Hydrometeorological University, Malookhtinsky pr. 98, St. Petersburg 195196, 2 Marine Research Institute, Klaipeda˙ University, Universiteto ave. 17, 92294 Klaipeda,˙ ; [email protected] 3 P.P. Shirshov Institute of Oceanology, Russian Academy of Sciences, Nakhimovsky pr. 36, Moscow 117997, Russia; [email protected] (E.V.K.); [email protected] (A.G.K.) 4 Interfacultary Center for Marine Research (MARE), University of Liège, B5a Sart-Tilman, B-4000 Liège, Belgium 5 S.Yu. Witte Moscow University, Second Kozhukhovsky pr. 12, Build. 1, Moscow 115432, Russia 6 Lithuanian Business University of Applied Sciences, University of Klapeida, Turgaus g. 21, 91249 Klaipeda, Lithuania * Correspondence: [email protected]

 Received: 7 October 2020; Accepted: 13 November 2020; Published: 14 November 2020 

Abstract: Here we analyze ice conditions in the Southeastern Baltic (SEB) Sea and in the Curonian Lagoon (CL) using spaceborne synthetic aperture radar (SAR) data combined with in-situ measurements from coastal stations during four winter seasons between 2009–2013. As shown, the ice conditions in the SEB and in the CL are strongly varying from year to year and do not always correlate with each other. In the SEB, ice cover may form only within 5–15 km band along the coast or spread up to 100 km offshore covering almost the entire . The mean ice season duration here is 45 days. The CL is almost fully ice-covered every year apart of its northern part subjected to sea water inflow and active shipping. The ice regime is also more stable here, however, it also possesses multiple periods of partial melting and re-freezing. In this study we also perform a validation of three SAR-based ice thickness products (Envisat ASAR 0.5-km and 1-km, and RADARSAT-2 0.5-km) produced by the Finnish Meteorological Institute versus in-situ measurements in the CL. As shown, all satellite products perform rather well for the periods of gradual ice thickness growth. When the ice thickness grows rapidly, all products underestimate the observed values by 10–20 cm (20–50%). The best results were obtained for the RADARSAT-2 ice thickness product with the highest R2 value (0.68) and the root mean square error around 8 cm. The results of the study clearly show that multi-mission SAR data are very useful for spatial and temporal analysis of the ice regime in coastal waters and semi-enclosed shallow water bodies where the number of field observations is insufficient or lacking.

Keywords: sea ice; ice cover extent; SAR-based ice thickness; validation; satellite remote sensing; SAR imaging; hydrometeorological conditions; Baltic Sea; Curonian Lagoon

1. Introduction Every year the Baltic Sea is partially covered by ice. Normally, the ice formation begins to the north of the Bothnian Bay and east of the Gulf of Finland. In average winters, ice covers the Bothnian Sea, the Archipelago Sea, the Gulf of Finland and the Gulf of Riga, as well as the northern part of the

Remote Sens. 2020, 12, 3754; doi:10.3390/rs12223754 www.mdpi.com/journal/remotesensing Remote Sens. 2020, 12, 3754 2 of 20

Baltic proper. In severe winters, the Danish Straits and the southern Baltic Proper are also covered with ice [1]. In the Southeastern Baltic (hereinafter, SEB), ice cover also forms within brackish and shallower estuaries—the and the Curonian Lagoons—that are at least partially ice-covered every year [2,3]. Historical records show that interannual variability of the Baltic Sea ice cover is very high. For the last 60 years the area covered by ice varied from 50,000 km2 (2008) to almost 400,000 km2 (1987), and from year to year the ice cover may change for 100,000–150,000 km2. The timing of the maximum ice cover extent can vary from 31 January (in 1990) up to 30 March (in 1994), while the highest ice thickness is usually found in the northern part of the Bothnian Bay (about 70 cm) [4]. Regional climate change in the Baltic Sea during the last 30 years has led to a notable increase of sea surface temperature, which is evident in all areas and all seasons [5–7]. Accordingly, a change towards milder winters has been observed which is characterized by a decrease of maximum ice cover extent, ice thickness, and ice season duration [4,8–11]. A pronounced decrease of the ice season duration was also reported for the Southeastern Baltic Sea [12–14], where according to [12], it has decreased by 50% during the 1961–2005 period equivalent to the ice season shortening by 1 month. The largest European freshwater estuary, the Curonian Lagoon (hereinafter, CL) is also affected by regional climate change [3,15]. During the last two decades, mean winter air temperature over the CL increased by 1.7 ◦C. The ice season duration in the lagoon is very irregular, and the ice cover can form and break-up several times per season. Higher air temperatures observed during the last decade favor unstable sea ice regime in the CL characterized by higher spatial variability of sea ice cover due to more frequent ice drift events. As such, this might add certain errors to the ice season duration estimates currently obtained mainly from range-limited visual observations from several coastal stations. Such observations do not resolve a spatial spreading of the ice cover over the entire lagoon. Moreover, the adequate knowledge of the ice regime and its variability over the CL and in the SE Baltic is locally important, e.g., for shipping operations in the Port of Klaipeda,˙ local transportation and fisheries, planning commercial mussel cultivation grounds, as well as for the lagoon and coastal biota [16]. With an ongoing reduction in the number of ice monitoring coastal stations, there is a wider perspective opening and a certain need for the utilization of satellite remote sensing to perform regular observations and investigate rapidly changing sea ice regime in the SE Baltic and in the Curonian Lagoon. Such methods, primarily based on synthetic aperture radar (SAR) observations, are well elaborated and used for operational monitoring of sea ice extent and concentration, ice thickness, drift and deformation in various of the Baltic Sea [17–21]. However, their use in the SE Baltic Sea is limited until now. The only exclusion is a recent study presented in [3], where the authors used multi-mission SAR data to analyze ice phenology and dynamics in the Curonian Lagoon, importantly showing a pronounced shortening of ice season duration observed over this region during 2002–2017. Yet, this study provides no information about the ice conditions in the adjacent coastal area of the SE Baltic Sea. Another important issue is that not all ice parameters operationally retrieved from SAR data undergo comprehensive validation versus in-situ measurements because the latter are often lacking, spotty and intermittent in time, while the ice properties themselves might differ from one region to another [22]. The aim of this paper is, therefore, to investigate spatial and temporal variability of ice conditions in the SE Baltic Sea and in the Curonian Lagoon and describe their peculiarities and dependence on background meteorological conditions. An important task of the study is also to perform the validation of regularly available SAR-based ice thickness products versus in-situ measurements in this part of the Baltic Sea.

2. Study Area

The study area includes the southeastern part of the Baltic Sea bounded by 19◦E from the west and 56◦N from the north, and the Curonian Lagoon, the largest coastal lagoon in (Figure1). Remote Sens. 2020, 12, 3754 3 of 20

The area of the marine part considered here is 18,045 km2. The CL is separated from the Baltic Sea by the Curonian with a narrow Klaipeda˙ Strait in its northern part. The area of the lagoon is 1544 km2 3 withRemote maximum Sens. 2020 length, 12, x of 93 km and varying width of 0.8–46 km, volume of 6.3 km and an3 of average 19 depth of 3.8 m [5,23,24]. The total river runoff is 25.1 km3 (84% belongs to the River—21 km3), whichaverage is four depth times of of 3.8its volume. m [5,23,24 An]. The average total salinity river runoff of the is seawater 25.1 km3 in(84% the belongs SEB is about to the 7% Neman, while in the lagoonRiver— it21 varies km3), inwhich the rangeis four oftimes 0.5–7.5% of its volume. [25]. The An lagoonaverage watersalinity being of the hypereutrophic, seawater in the SEB its qualityis is controlledabout 7‰, mostly while by in physicalthe lagoon factors it varies such in asthe the range wind of regime,0.5–7.5‰ temperature [25]. The lagoon and water level variations, being hypereutrophic, its quality is controlled mostly by physical factors such as the wind regime, and transparency [24]. Shallow bathymetry and low salinity are key factors discriminating the ice temperature and level variations, and transparency [24]. Shallow bathymetry and low salinity are regime in the lagoon and the SEB. key factors discriminating the ice regime in the lagoon and the SEB.

FigureFigure 1. Map 1. Map of theof the study study area. area. Legend: Legend: 1—marine1—marine part part of of the the study study area; area; 2—the 2—the Curonian Curonian Lagoon; Lagoon; 3—boundaries3—boundaries of exclusive of exclusive economic economic zones zones (EEZ); (EEZ); 4—ice 4— monitoringice monitoring stations stations in the in theCuronian Curonian Lagoon; Lagoon; 5—D-6 oil platform. The contour lines show isobaths. 5—D-6 oil platform. The contour lines show isobaths.

3. Materials3. Materials and and Methods Methods InvestigationInvestigation of of spatial spatial properties properties ofof iceice cover in in the the study study site site was was based based on the on analysis the analysis of of spaceborne synthetic aperture radar (SAR) images acquired by Envisat ASAR (European Space spaceborne synthetic aperture radar (SAR) images acquired by Envisat ASAR (European Space Agency, Agency, ESA), RADARSAT-1 (Canadian Space Agency, CSA, Canada), RADARSAT-2 (McDonald, ESA),Dettwiler RADARSAT-1 & Associates, (Canadian MDA, Space Canada) Agency, (Table CSA, 1), Canada), distributed RADARSAT-2 by Kongsberg (McDonald, Satellite Se Dettwilerrvices & Associates,(Norway). MDA, In total, Canada) 277 SAR (Table images1), distributed during four by winter Kongsberg seasons Satellite in 2009– Services2013 were (Norway). used with In 83 total, 277 SARimages images in 2009 during–2010, four 74— winterin 2010 seasons–2011, 71 in— 2009–2013in 2011–2012 were, and used 49—in with 2012 83–2013. images For in every 2009–2010, winter, 74—ina 2010–2011,period 71—inbetween 2011–2012, 20 November and and 49—in 15 April 2012–2013. was taken For for every consideration. winter, a period between 20 November and 15 April was taken for consideration. Table 1. Main characteristics of spaceborne SAR images used in the study.

Scene Size, Spatial Resolution, Number of Satellite Band Period km m Images Envisat ASAR C 400 × 400 150 × 150 144 November 2009–April 2012 RADARSAT-1 C 300 × 300 50 × 50 63 December 2009–March 2013 RADARSAT-2 C 500 × 500 100 × 100 70 December 2009–April 2013

Remote Sens. 2020, 12, 3754 4 of 20

Table 1. Main characteristics of spaceborne SAR images used in the study.

Spatial Number of Satellite Band Scene Size, km Period Resolution, m Images Envisat ASAR C 400 400 150 150 144 November 2009–April 2012 × × RADARSAT-1 C 300 300 50 50 63 December 2009–March 2013 × × RADARSAT-2 C 500 500 100 100 70 December 2009–April 2013 × ×

The method of ice cover detection in the SAR data is based on sensitivity of radar return to the surface roughness and dielectric properties of the medium from which the radar signal is backscattered. For different wind conditions, SAR imaging geometry and ice age, the strength of the return SAR signal from the sea ice would differ by several orders of magnitude [26]. However, the ice-open water boundary can still be detected rather effectively [27,28]. In this work, we use a visual detection of the ice edge in satellite data that is often used in similar studies [3,14,29]. Detected ice fields were digitized from SAR images and the area of ice cover extent was calculated in ArcGIS 10.0. The area of the ice cover detected in subsequent SAR images was interpolated when there was a gap in observations. The main advantage of the method is the ability of SAR to perform measurements under cloudy conditions and at night that are major limitations for the optical and infrared remote sensing in the Baltic Sea [30,31]. While being effective for observing surface pollutants and signatures of upper-ocean dynamics over ice-free regions mostly under low to moderate winds [32–36], SAR observations of sea ice patterns are feasible under a wider range of wind speeds [14,37], and were successfully used for the sea ice studies in the nearby [38]. We also use cloud-free optical data from Moderate Resolution Imaging Spectroradiometer (MODIS) onboard Terra and Aqua satellites to illustrate the ice cover formation and decay in the Curonian Lagoon. Here we also perform a validation of several SAR-based ice thickness (IT) products derived from Envisat ASAR and RADARSAT-2 observations versus in-situ IT measurements. All considered products were produced by the Finnish Meteorological Institute (FMI) and obtained from Copernicus Marine Environment Monitoring Service (CMEMS, https://resources.marine.copernicus.eu). The first product is the gridded ice thickness based on the digitized Baltic ice charts produced manually by the ice analysts at FMI (FMI-BAL-SEAICE_THICK-L4-NRT-OBS data set). It is produced daily since 2010 at the nominal resolution of 1 km based on satellite SAR and optical data, and ground truth from icebreakers, ships and ice observation stations. This product is considered only for winter season of 2010–2011, when it was primarily based on Envisat ASAR data (hereinafter, ASAR 1-km IT product). The other two IT products are available since the end of 2010 and are based on automatic SAR data classification [17]. These are gridded ice thickness data based on the digitized FMI ice charts refined by Envisat ASAR (winter seasons of 2010–2012, hereinafter, ASAR 0.5-km) and RADARSAT-2 (winter seasons of 2011–2013, hereinafter, RS-2 0.5-km) at the nominal resolution of 0.5 km (FMI-BAL-SEAICE_THICK-SAR-NRT-OBS data set). The detailed description of these IT products is given in [39]. In-situ IT measurements over several locations in the Curonian Lagoon were obtained from Lithuanian hydrometeorological stations in Nida and Vente,˙ and from one Russian station in Otkrytoye (see Figure1). The in-situ IT values were obtained from daily ice-drilling measurements when the ice thickness and depth on top of it were measured manually by tape with measurement error of 1 cm. Since in-situ IT data were available only till the winter of 2012–2013 ± and satellite IT products became available from 2010, our validation is limited by 2010–2013 period. For comparison, we extracted satellite-based ice thickness values over the points nearest to the locations of the three coastal stations. For the analysis of seasonal ice cover variability, we also use wind field data obtained from Autonomous Hydro-Meteorological Station (AHMS) installed on the offshore oil platform D-6 (see Figure1) and air temperature (Ta) records from AHMS installed in Klaip eda.˙ Remote Sens. 2020, 12, 3754 5 of 20

4. Results

4.1. Ice Conditions in 2009–2013 Analysis of SAR data for the Southeastern Baltic Sea and the Curonian Lagoon shows that their ice regime is strongly varying from year to year. Figure2 illustrates a very strong di fference in the area covered by sea ice during the maximum ice cover development stage in the SEB in 2009–2013. In 2010 and 2012, the ice cover is observed along the coasts with a mean width of the ice band of about 5–15 km (Figure2a,c). To the contrary, the ice extent is anomalously wide in February 2011 when it spreads up to 100 km offshore and covers about 90% of the marine part of the study area (Figure2b). In 2013, the minimal spatial ice extent was recorded along the SEB with sea ice primarily observed along the shore and north of the Port of Klaipeda˙ (Figure2d). Notably, the Curonian Lagoon was almost fully ice-covered during all these winter seasons. Remote Sens. 2020, 12, x 5 of 19 spreads up to 100 km offshore and covers about 90% of the marine part of the study area (Figure 2b). In 2013, the minimal spatial ice extent was recorded along the SEB with sea ice primarily observed Remotealong Sens.the 2020Curonian, 12, 3754 Spit shore and north of the Port of Klaipėda (Figure 2d). Notably, the Curonian6 of 20 Lagoon was almost fully ice-covered during all these winter seasons.

FigureFigure 2. 2. Maximum sea ice extent in the SoutheasternSoutheastern Baltic Sea Sea in in winter winter seasons seasons of of 2009–2013 2009–2013 observedobserved inin spaceborne spaceborne SAR SAR data: data: (a )( RADARSAT-2a) RADARSAT on-2 8on February 8 Febru 2010ary (05:052010 (05:05 UTC); UTC); (b) Envisat (b) Envisat ASAR onASAR 24 February on 24 F 2011ebrua (09:13ry 2011 UTC); (09:13 (c) Envisat UTC); ( ASARc) Envisat on 11 ASAR February on 201211 Fe (09:10bruary UTC); 2012 (d (09:10) RADARSAT-2 UTC); (d) onRADARSAT 23 January-2 2013 on 23 (05:05 Janu UTC).ary 2013 Light (05:05 blue U contoursTC). Light mark blue the contours boundaries mark betweenthe boundaries the open between water and the theopen ice-covered water and regions. the ice-covered regions.

TheThe marine marine part part of of the the study study area area is characterized is characterized by sporadic by sporadic sea ice seagrowth ice thatgrowth clearly that depends clearly ondepends wind conditionson wind conditions and air temperature. and air temperature. In contrast, In contrast, the ice formation the ice formation is more continuous is more continuous and less sporadicand less sporadic in the Curonian in the Curonian Lagoon whereLagoon it where usually it lastsusually from lasts the from first the frosts first in frosts December–January in December– tillJanuary the end till the of winterend of seasonwinter season in March–April. in March–April Nevertheless,. Nevertheless, some some winters winters were were characterized characterized by several freezing and melting periods in the CL. As already shown in [3], the northernmost part of the lagoon near the Port of Klaipeda˙ often remains partially ice-free due to inflow of relatively warm and saline water from the Baltic Sea through the Klaipeda˙ Strait and intensive shipping (Figure3a,b). Remote Sens. 2020, 12, x 6 of 19

Remote Sens. 2020, 12, 3754 7 of 20 by several freezing and melting periods in the CL. As already shown in [3], the northernmost part of the lagoon near the Port of Klaipėda often remains partially ice-free due to inflow of relatively warm Duringand severe saline waterwinters, from ships the break Baltic the Sea first through ice forms the Klaandipėda prevent Strait the and formation intensive of stableshipping ice ( coverFigure in this3a,b). area. During Most often sever thee winters, marine ships water break penetrates the first intoice forms the northern and prevent part the of theformation lagoon of through stable ice the Klaipcovereda˙ Strait in this up area. to 25 Most km, oftenand under the marine particular water meteorological penetrates into conditions the northern and strong part of westerly the lagoon and north-westerlythrough the windsKlaipėda it mayStrait penetrate up to 25 km, up toand 40–50 under km particular down the meteorological lagoon [40]. conditions and strong westerly and north-westerly winds it may penetrate up to 40–50 km down the lagoon [40].

(a) (b)

(c) (d)

FigureFigure 3. Photographs 3. Photographs illustrating illustrating the the ice conditionsice conditions in the in Klaipthe Klaipėdaeda˙ Strait Strait taken taken on: on (a): 3(a January) 3 Janu 2010ary and2010 (b) 13 and February (b) 13 2011;Februa andry 2011 (c,d); heavyand (c, iced) accumulationheavy ice accumulation along the coasts along theof Nida coasts on of 17 Nida March on 2011. 17 PhotographsMarch 2011. taken Photographs by I. Dailidien takene ˙by and I. Dailidienė I. Kairyte.˙ and I. Kairytė.

TheThe temporal temporal variability variability of of iceice extent in in the the CL CL and and the theSEB SEB during during four winter four winter seasons seasons in 2009– in 2009–20132013 is is shown shown in inFigure Figure 4,4 , while while a a more more detailed detailed analysis analysis of of ice ice cover cover changes changes in in relation relation to to air air temperaturetemperature and and wind wind conditions conditions is is presented presented below. below. In 2009–2010In 2009–2010 ice ice season season in in the the Curonian Curonian Lagoon Lagoon has has startedstarted onon 1212 December 2009, 2009, i.e. i.e.,, 3 3days days afterafter the the air temperatureair temperature (Ta) (Ta) dropped dropped below below zero. zero. A furtherA further increase increase of windof wind speed speed above above 10 10 m /sm/s has stimulatedhas stimulated intensive intensive water mixingwater mixing and cooling and cooling of the of shallow the shallow lagoon lagoon resulting resulting in full in freeze full freeze up dated up on 19dated December on 19 December 2009. The 2009. coastal The zone coastal of the zone SEB of was the ice-freeSEB was at ice that-free moment, at that whilemoment, a thin while strip a ofthin ice wasstrip later of observed ice was along later observed the Curonian along Spit the on Curonian 22 December Spit оn and 22 disappeared December and on disappeared 24 December on 2009, 24 whenDecember the Ta became 2009, when positive the Ta (Figure became4). positive (Figure 4). FollowingFollowin ag week a week of of negative negative Ta Ta values, values, sea sea ice ice formation formation inin thethe SEB started again again on on 8 8January January 2010. Despite moderate and strong winds observed in that period, the ice cover spread rather fast 2010. Despite moderate and strong winds observed in that period, the ice cover spread rather fast along the Curonian Spit coast. A continuous freezing period lasted until 18 February 2010, and the maximum ice extent over the SEB was recorded on 8 February 2010 covering about 16 % of the study area (see Figures2a and4). Strong winds and the rise of Ta in the second half of February 2010 Remote Sens. 2020, 12, x 7 of 19

Remotealong Sens. the2020 Curonian, 12, 3754 Spit coast. A continuous freezing period lasted until 18 February 2010, and the8 of 20 maximum ice extent over the SEB was recorded on 8 February 2010 covering about 16 % of the study area (see Figure 2a and Figure 4). Strong winds and the rise of Ta in the second half of February 2010 stimulated the ice break-up and rapid melting, resulting in almost ice-free conditions observed on stimulated the ice break-up and rapid melting, resulting in almost ice-free conditions observed on 3 3 March 2010. A short period of negative Ta down to 6 C during the next week leaded to a partial ice March 2010. A short period of negative Ta down to −−6 °C◦ during the next week leaded to a partial ice formationformation that that completely completely decayed decayed on on 1919 MarchMarch 2010.2010.

FigureFigure 4. 4.Variability Variability of iceof ic extente extent in the in the Curonian Curonian Lagoon Lagoon (left (left column column) and) and in the in Southeasternthe Southeastern Baltic SeaBaltic (right Sea column (right) obtained column from) obtained satellite from data for satellite winters data of 2009–2013. for winters Legend: of 2009 1—percentage–2013. Legend: of the ice1 extent—percentage within of the the study ice extent area; within 2—daily the averagestudy area; air 2 temperature—daily average (Ta); air 3—daily temperature average (Ta); wind 3—daily speed. Zeroaverage values wind in both speed. diagrams Zero values mean in absence both diagrams of ice cover. mean absence of ice cover.

InIn the the Curonian Curonian Lagoon, Lagoon, the theice ice covercover waswas more stable stable (Figure (Figure 4).4). The The melt melt onset onset started started in inthe the north-westernnorth-western part part of of the the lagoon lagoon in in thethe beginningbeginning of March 2010. 2010. A A small small fr fractureacture of ofopen open water water near near Pervalka was well seen both in Envisat ASAR and MODIS Aqua images acquired on 10–11 March Pervalka was well seen both in Envisat ASAR and MODIS Aqua images acquired on 10–11 March 2010 2010 (Figure 5), and not captured visually at Nida and Ventė stations located nearby. Under westerly (Figure5), and not captured visually at Nida and Vent e˙ stations located nearby. Under westerly winds winds and positive Ta the ice melt moved further north- and eastward until reaching the eastern CL and positive Ta the ice melt moved further north- and eastward until reaching the eastern CL coast. Further warming led to the ice break-up in the central part of the CL. The last observation of ice during this winter season was made on 8 April 2010 along the eastern CL coast. The total ice season duration in 2009–2010 was 115 days and 65 days for the CL and the SEB, respectively. Remote Sens. 2020, 12, x 8 of 19

coast. Further warming led to the ice break-up in the central part of the CL. The last observation of Remoteice during Sens. 2020 this, 12 winter, 3754 season was made on 8 April 2010 along the eastern CL coast. The total9 of ice 20 season duration in 2009–2010 was 115 days and 65 days for the CL and the SEB, respectively.

Figure 5. Start of ice break break-up-up in the Curonian Lagoon as seen in ( a) Envisat ASAR image acquired on 1010 MarchMarch 2010;2010; ((bb)) MODISMODIS AquaAqua visible-bandvisible-band imageimage acquiredacquired onon 1111 MarchMarch 2010.2010.

The iceice seasonseason of of 2010–2011 2010–2011 started started in thein the CL CL in late in late November, November, when when early early frosts frosts and strong and strong winds abovewinds 15above m/s 15 (Figure m/s (Figure4) led to 4) a led fast to cooling a fast cooling and freezing and freezing of the lagoonof the lagoon well depicted well depicted in Figure in Figure6a–c. RADARSAT-26a–c. RADARSAT image-2 acquiredimage acquired on 30 November on 30 November 2010 shows 2010 an shows intensive an intensive ice cover formationice cover thatformation spans thethat entire spans northern the entire and northern central and parts central of the lagoon,parts of asthe well lagoon, as the as regions well as along the regions the coast along in the the southern coast in CLthe withsouthern open CL water with observed open water only observed in its central only part in its (Figure central6a). part MODIS (Figure visible 6a).band MODIS image visible acquired band onimage the next acquired day, 1 on December the next 2010, day, shows 1 December the lagoon 2010, already shows fully the ice-covered lagoon already (Figure fully6b), ice confirmed-covered by(Figure RADARSAT-1 6b), confirmed image by of R 2ADARSAT December- 20101 image (Figure of 26 Decemberc). 2010 (Figure 6c). In the marine part, the ice cover started to form two weeks later along the northern coast of the . Prior to 1 February 2011, sea ice was observed mainly under moderate winds and subzero temperaturestemperatures in in the the form form of of grease grease ice ice drifting drifting along along the coast.the coast. From From 10 February 10 February 2011, a2011, rapid a droprapid ofdrop Ta downof Ta down to 15 toC −15 led °C to led an to extensive an extensive ice formation ice forma reachingtion reaching 15,725 15,725 km2 km(about2 (about 87% 87 of% the of − ◦ marinethe marine study study area) area) by 24by February24 February 2011 2011 (see (see Figures Figure2b 2b and and4). Figure That was4). That the maximumwas the maximum ice extent ice recordedextent recorded in the SE in Baltic the SE during Baltic the during study the period. study Ice period. melting Ice started melting in the started beginning in the of beginning March, and of onMarch, 7 March and 2011,on 7 Marc onlyh grease 2011, iceonly was grease detected ice was along detected the Lithuanian along the coast Lithuanian and the coast Vistula and Spit. the TheVistula last observationSpit. The last of observation ice was recorded of ice on was 15 recorded March 2011. on 15 It total, March sea 2011. ice in It the total, SEB sea was ice observed in the SEB during was 68observed days within during six 68 episodes days within of ice six formation. episodes of ice formation. In the Curonian Lagoon, ice melt onset was dated on 10 March 2011 (Figure(Figure4 4).). TheThe iceice break-upbreak-up started in the northern part of the lagoon ( (FigureFigure 66dd),), andand thenthen spreadspread alongalong thethe entireentire westernwestern coast.coast. Varying windswinds causedcaused a a strong strong ice ice accumulation accumulation on on the the coasts coasts near near Nida Nida in somein some cases cases damaging damaging the coastalthe coastal infrastructure infrastructure (Figure (Figure3c,d). 3c,d). Later on,Later strong on, strong westerly westerly winds ofwinds 10–15 of m 10/s– caused15 m/s ancaused ice drifting an ice eastwarddrifting eastward (Figure6 e).(Figure The ice 6e cover). The remainedice cover remained longest in thelongest central in the CL central close to CL Neman close Delta to Neman (Figure Delta6f). The(Figure overall 6f). lengthThe overall of the length melting of seasonthe melting was more season than was 3 weeks,more than and 3 complete weeks, and ice decaycomplete was ice dated decay on 4was April dated 2011 on with 4 April the ice2011 season with the duration ice season equal duration to 127 days. equal to 127 days.

Remote Sens. 2020, 12, 3754 10 of 20

Remote Sens. 2020, 12, x 9 of 19

FigureFigure 6. Ice6. Ice formation formation and and decay decay in in the the Curonian Curonian LagoonLagoon during during winter winter 2010 2010–2011–2011 as as observed observed in:in: (a)( RADARSAT-2a) RADARSAT- image2 image of of 30 30 November November 2010 2010 (16:26(16:26 UTC);UTC); ( b)) MODIS MODIS Aqua Aqua image image of of 1 1D Decemberecember 2010;2010; (c) ( RADARSAT-1c) RADARSAT- image1 image of of 2 2 December December 20102010 (16:15(16:15 UTC); ( (dd)) RADARSAT RADARSAT-2-2 image image of of 20 20 M Marcharch 20112011 (16:18 (16:18 UTC); UTC); (e ()e RADARSAT-2) RADARSAT-2 image image of of27 27 MarchMarch 20112011 (16:14 UTC); UTC); (f (f) )Envisat Envisat ASAR ASAR image image of of 1 April 2011 (20:16 UTC). Light blue contours mark the boundaries between the open water and the 1 April 2011 (20:16 UTC). Light blue contours mark the boundaries between the open water and the ice-covered regions. ice-covered regions.

TheThe ice ice season season of 2011–2012 of 2011–2012 was was the shortest the shortest during during the period the period of observations of observations starting starting in second in second half of January 2012 both in the CL and in the SEB (Figure 4). A rapid drop of Ta within a half of January 2012 both in the CL and in the SEB (Figure4). A rapid drop of Ta within a week week period down to −21 °C and moderate winds triggered a fast ice formation and full freezing of period down to 21 C and moderate winds triggered a fast ice formation and full freezing of the CL the CL dated on− 6 ◦February 2012 (Figure 4). Eight days later, the first ice forms also appeared along dated on 6 February 2012 (Figure4). Eight days later, the first ice forms also appeared along the SEB the SEB coast. The ice extent grew quite rapidly and peaked on 11 February 2012 covering about 12% coast. The ice extent grew quite rapidly and peaked on 11 February 2012 covering about 12% of the of the marine study area (Figure 2c). However, the next week was characterized by a sharp increase marinein Ta studyand wind area conditions (Figure2c). that However, stimulated the ice next melting week in wasthe SEB characterized and its full decay by a sharp on 17 increaseFebruary. in In Ta andthe wind CL, conditionsthe melt onset that stimulatedwas dated iceon melting3 Marchin and the the SEB last and observation its full decay of onice 17was February. recorded In on the 31 CL, theMarch melt onset2012 along was dated the northeastern on 3 March coast and theof the last CL. observation The ice season of ice duration was recorded for the onCL 31and March the SEB 2012 alongwere the 71 northeasternand 23 days, respectively. coast of the CL. The ice season duration for the CL and the SEB were 71 and 23 days,In respectively. winter season 2012–2013, strong winds above 15 m/s and negative air temperature contributed to the ice formation in the CL in mid-December 2012 (see Figure 4). An increase of Ta in

Remote Sens. 2020, 12, 3754 11 of 20

In winter season 2012–2013, strong winds above 15 m/s and negative air temperature contributed to the ice formation in the CL in mid-December 2012 (see Figure4). An increase of Ta in the end of December and strong westerly winds led to the ice melt along the coast of CL. Later on, with the onset Remote Sens. 2020, 12, x 10 of 19 of frosts, the CL became completely ice-covered again (see Figure4). Inthe the end SEB, of December the ice seasonand strong was westerly characterized winds led by to four the ice melt formation along the episodes. coast of CL. The Later first on, fast ice was observedwith the onset along of frosts, the shore the CL of became the Curonian completely Spit ice on-covered 16 January again (see 2013. Figure The 4). maximum ice cover developmentIn the was SEB, recorded the ice season byRADARSAT-2 was characterized on by 23 four January ice formati 2013,on covering episodes. onlyThe first about fast 3.4%ice was of the marineobserved study area along (see the Figure shore 2 ofd). the However, Curonian just Spit a week on 16 later January the ice 2013. cover The completely maximum icedisappeared cover whendevelopment the air temperature was recorded rose by above RADARSAT zero. The-2 on last 23 January two ice 2013, formation covering episodes only about were 3.4% rather of the short, only 5–6marine days study long, area observed (see Figure in mid-2d). Howeve and later, just March a week 2013 later (see the Figure ice cover4). During completely these disappeared episodes, the when the air temperature rose above zero. The last two ice formation episodes were rather short, formation of fast ice was observed along the coast from Baltiysk to the northern part of the marine only 5–6 days long, observed in mid- and late March 2013 (see Figure 4). During these episodes, the studyformation area. The of last fast detection ice was observed of sea ice along was the made coast on from 26 MarchBaltiysk 2013. to the northern part of the marine Asstudy usual, area. the The ice last melt detection in the of Curonian sea ice was Lagoon made on began 26 March in its 2013. northern part. On 2 March 2013, open water reachedAs usual, Juodkrant the icee ˙melt in the in northwesternthe Curonian Lagoon part of began the CL in (seeits northern Figure1 part. for details). On 2 March Other 201 ice-covered3, open partswater of the reached lagoon remainedJuodkrantė stable in the till northwestern the end of March part of 2013. the In CLthe (see beginning Figure 1 of for April details). 2013, Other when Ta was stablyice-covered above parts zero, of ice the melting lagoon remained started near stable Vent tille ˙the and end then of March expanded 2013. alongIn the thebeginning northeastern of April coast of the2013, lagoon. when Ice-free Ta was regionsstably above strongly zero, increasedice melting southward started near towards Ventė and the then central expanded part ofalong the the lagoon, northeastern coast of the lagoon. Ice-free regions strongly increased southward towards the central as both SAR and MODIS data suggest (Figure7). A strong air temperature rise above 5 ◦C observed after 12part April of the 2013 lagoon, led toas theboth fast SAR ice and melt MODIS along data the suggest southern (Figure coasts 7). of A thestrong lagoon, air temperature where it remained rise above 5 °C observed after 12 April 2013 led to the fast ice melt along the southern coasts of the longest till 15 April 2013 (see Figure7). It total, the ice season duration for winter 2012–2013 counted lagoon, where it remained longest till 15 April 2013 (see Figure 7). It total, the ice season duration for for 24winter days in2012 the–2013 SE Balticcounted Sea for and 24 days 130 daysin the inSE the Baltic Curonian Sea and Lagoon.130 days in the Curonian Lagoon.

FigureFigure 7. Ice 7. break-up Ice break- andup and decay decay in in the the Curonian Curonian LagoonLagoon in in March March–April–April 2013. 2013. Light Light blue blue contours contours markmark the boundaries the boundaries between between the the open open water water and and thethe ice-coveredice-covered regions regions..

FigureFigure8 shows 8 shows a composite a composite map map of of ice ice cover cover probabilityprobability in in the the SE SE Baltic Baltic obtained obtained during during an an extendedextended period period of 2004–2019of 2004–2019 for for February, February, thethe monthmonth of of the the maximum maximum ice ice cover cover development development in in this region [14]. As seen, the ice cover is observed most frequently along the southwestern part of the

Remote Sens. 2020, 12, 3754 12 of 20

Remote Sens. 2020, 12, x 11 of 19 this region [14]. As seen, the ice cover is observed most frequently along the southwestern part of theCuronian Curonian Spit Spit and and north north of Klaipėda of Klaipeda ˙Strait, Strait, where where an anintense intense outflow outflow of ice of icefloes floes from from the the CL CL is isfrequently frequently observed observed during during the the ice ice season. season. In In 5% 5% of of cases cases the the ice ice extent extent covers covers about about 5 km bandband adjacent to the coast. A wider ice cover extent is observed less frequentl frequently,y, yet,yet, in some rare cases it might span almostalmost thethe entireentire SEB,SEB, asas inin FebruaryFebruary 20112011 (Figure(Figure2 2b).b).

Figure 8. Probability of ice cover observations (%) in the SE Baltic near the Curonian Spit in February according toto spacebornespaceborne SARSAR datadata duringduring anan extendedextended periodperiod ofof 2004–20192004–2019 [[14]14].. Light blue contours showshow isolinesisolines ofof 2%2% andand 5%5% probabilitiesprobabilities ofof iceice covercover observations.observations.

Figure9 9 shows shows spatial spatial maps maps of of ice ice cover cover probability probability in the in Curonianthe Curonian Lagoon Lagoon during during 2009–2013. 2009–2013. One commonOne common feature feature that is that seen is duringseen during all winter all winter seasons seasons is the highest is the highest ice cover ice probability cover probability (i.e., longest (i.e., icelongest cover ice season) cover in season) the southeastern in the southeastern part of the part lagoon. of the Depending lagoon. on Depending background on airbackground temperature air andtemperature wind conditions, and wind the conditions, area of longest the ice area observations of longest (brown ice observations color in Figure (brown8) could color be in very Figure large 8) (25could km be wide) very coveringlarge (25 km almost wide) the covering entire central almost and the entire southern central parts and of southern the lagoon parts as inof 2010–2011,the lagoon ratheras in 2010 narrow–2011, (5–8 rather km narrow wide) as (5 in–8 2011–2012, km wide) as or in have 2011 inhomogeneous–2012, or have inhomogeneous pattern as in 2009–2010 pattern as and in 2012–2013.2009–2010 andThe shortest 2012–2013. ice season The shortest is clearly ice observed season is in clearly the northern observed part in of the the lagoon,northern in part agreement of the withlagoon [3,]. in Here, agreement the location with [3] and. Here, area the ofminimal location iceand cover area regionsof minimal diff erice from cover year regions to year differ and from are mainlyyear to determined year and are by mainly the wind determined conditions by (Figure the wind9). Under conditions prevalence (Figure of 9 westerly-northwesterly). Under prevalence of winds,westerly the-northwesterly sea water from winds, the SEB the enters sea thewater lagoon from and the together SEB enters with themechanical lagoon windand together action trigger with themechanical ice break wind up along action the trigger northwestern the ice break CL coast up along and its the drift northwestern and accumulation CL coast along and theits drift opposite and northeasternaccumulation coast along (Figure the opposite9b,c). Thenortheastern higher is thecoast wind (Fig action,ure 9b,c). the The larger higher is the is areathe wind of minimal action, icethe coverlarger inis thethe northernarea of minimal lagoon partice cover (compare in the Figure northern9a–c). lagoon When part the easterly(compare and Figure southeasterly 9a–c). When winds the dominate,easterly and the southeasterly region of lowest winds ice frequency dominate, spans the the region entire of northern lowest ice part frequency including spans the northeastern the entire CLnorthe coastrn downpart including to the northern the northeastern Neman Delta CL (Figurecoast down9d). to the northern Neman Delta (Figure 9d).

Remote Sens. 2020, 12, 3754 13 of 20 Remote Sens. 2020, 12, x 12 of 19

FigureFigure 9.9. ProbabilityProbability ofof iceice covercover observationsobservations (%)(%) inin thethe CuronianCuronian LagoonLagoon accordingaccording toto spacebornespaceborne SARSAR datadata in:in: ((aa)) 2009–2010;2009–2010; ((bb)) 2010–2011;2010–2011; ((cc)) 2011–2012;2011–2012; ((dd)) 2012–2013.2012–2013. InsertsInserts showshow windwind rosesroses forfor thethe periodperiod fromfrom 2020 NovemberNovember toto 1515 April.April. 4.2. Validation of SAR-Based Ice Thickness Products 4.2. Validation of SAR-Based Ice Thickness Products As already mentioned before, ice thickness (IT) measurements in the Curonian Lagoon were As already mentioned before, ice thickness (IT) measurements in the Curonian Lagoon were performed only at several coastal stations and, therefore, lack the spatial details of IT distribution over performed only at several coastal stations and, therefore, lack the spatial details of IT distribution the entire lagoon. In turn, satellite observations provide a unique possibility to derive spatially detailed over the entire lagoon. In turn, satellite observations provide a unique possibility to derive spatially ice thickness maps for the large ice-covered regions of the Baltic Sea [41]. In this subsection we aim detailed ice thickness maps for the large ice-covered regions of the Baltic Sea [41]. In this subsection to explore the quality of SAR-based IT retrievals over the Curonian Lagoon through the comparison we aim to explore the quality of SAR-based IT retrievals over the Curonian Lagoon through the comparison of satellite-based IT products with in-situ ice thickness measurements. Since the latter

Remote Sens. 2020, 12, 3754 14 of 20

Remote Sens. 2020, 12, x 13 of 19 of satellite-based IT products with in-situ ice thickness measurements. Since the latter were only availablewere only till available the winter till of the 2012–2013 winter of and2012 satellite–2013 and products satellite became products available became fromavailable 2010–onwards, from 2010– ouronwards, analysis isour limited analysis to three is limited winter toseasons three of winter 2010–2013. seasons In winter of 2010 2010–2011,–2013. In wewinter consider 2010– two2011, types we ofconsider ASAR-based two ice types thickness of ASA dataR-based (ASAR ice 0.5-km thickness and data ASAR (ASAR 1-km), 0.5 while-km andin 2011–2012 ASAR 1- ASARkm), while 0.5-km in and2011 RADARSAT-2–2012 ASAR (RS-2) 0.5-km 0.5-km and RADARSAT products are-2 used. (RS-2) Due 0.5-km to failure products of Envisat are used. ASAR Due in to April failure 2012, of weEnvisat consider ASAR only in the April RS-2 2012, 0.5-km we productconsider foronly the the winter RS-2 2012–2013.0.5-km product for the winter 2012–2013. AsAs seen seen from from Figure Figure 10 10a,a the, the ASAR-based ASAR-based ice ice thickness thickness values values nicely nicely follow follow thethe gradual gradual iceice thicknessthickness rise rise from from 5 cm 5 cm to 25to cm25 cm observed observed in the in beginningthe beginning of winter of winter 2010–2011 2010–2011 at Nida at Nida and Ventand e,˙Ventė with, somewith overestimates some overestimates at Otkrytoye at Otkrytoye station. station. The discrepancy The discrepancy between between the two the data two sources data sources grows grows when thewhen true icethe thicknesstrue ice thickness experiences experiences a step-like a step rise-like till around rise till 40around cm in 40 the cm middle in the ofmiddle the season of the around season around 27 January–26 February 2011. After that, the SAR-based estimates grow up and start to 27 January–26 February 2011. After that, the SAR-based estimates grow up and start to reproduce reproduce the observed IT levels at all stations with the ASAR 1-km IT product performing the observed IT levels at all stations with the ASAR 1-km IT product performing somewhat better, somewhat better, while the 0.5-km IT product showed a better performance during the first half of while the 0.5-km IT product showed a better performance during the first half of the season. the season.

Nida Nida Nida

Ventė Ventė Ventė

Otkrytoye Otkrytoye Otkrytoye

(a) (b) (c)

FigureFigure 10. 10.Comparison Comparison of of Envisat Envisat ASAR ASAR and and RADARSAT-2 RADARSAT-2 ice ice thickness thickness products products with with in-situ in-situ measurementsmeasurements for for winter winter seasons seasons of of 2010–2013 2010–2013 for for ( a()a) Envisat Envisat ASAR ASAR 0.5-km 0.5-km andand 1-km1-km productsproducts inin 2010–2011,2010–2011, (b ) Envisat(b) Envisat ASAR ASAR 0.5-km 0.5 and-km RADARSAT-2 and RADARSAT 0.5-km- products2 0.5-km in products2011–2012, in (c ) 2011 RADARSAT-2–2012, (c) 0.5-kmRADARSAT in 2012–2013.-2 0.5-km in 2012–2013.

InIn winter winter 2011–2012, 2011–2012, the ice ice thickness thickness is isgradually gradually rising rising and and peaks peaks at 40 at–42 40–42 cm around cm around 21–23 21–23February February 20122012 (Figure (Figure 10b) .10 Duringb). During this winter this winter season, season, both the both RS the-2 and RS-2 the and ASAR the ASAR products products show showrather rather similar similar results results and, and, in in general, general, reproduce reproduce the the observed observed seasonal seasonal changes changes of of IT IT values. Nevertheless,Nevertheless, satellite satellite products products largely largely underestimate underestimate the the observed observed IT IT values values by by 10–1510–15 cm.cm. OneOne may seesee a cleara clear delay delay of of about about 7–10 7–10 days days between between the the in-situ in-situ and and the the satellite satellite ITITcurves. curves. InIn satellitesatellite data, thethe 40-cm 40-cm ice ice thickness thickness level level is reached is reached only inon thely in end the of end February of February 2012, after 2012, which after the which gap between the gap thebetween satellite the and satellite the in-situ and datathe in is-situ almost data negligible. is almost negligible. The smallest The di ffsmallesterence isdifference observed is for observed Otkrytoye for stationOtkrytoye in the station southern in the CL southern part where CL the part ice where conditions the ice are conditions more stable are more (Figure stable9c). (Figure 9c). TheThe winter winter season season of of 2012–2013 2012–2013 is is characterized characterized by by relatively relatively low low iceice thicknessthickness valuesvalues generallygenerally notnot exceeding exceeding 20–30 20–30 cm cm throughout throughout the the season season (Figure (Figure 10 10c).c). The The performance performance ofof thethe RS-2RS-2 productproduct isis ratherrather adequate, adequate, it it correlates correlates quite quite well well with with the the observed observed ice ice thickness thickness values values withwith onlyonly somesome minorminor underestimatesunderestimates for for Vent Ventėe˙ and and Otkrytoye Otkrytoye stations. stations. The The correlation correlation for for Nida Nida station station is alsois also good good apart apart of twoof short-termtwo short-term episodes episodes of rapid of rapid ice thickness ice thickness growth growth observed observed after after 15 February 15 February 2013. 2013. Figure 11 shows two scatter plots where the in-situ ice thickness measurements are plotted against the satellite products separately for Envisat ASAR and RS-2. The corresponding statistical measures describing the overall performance of the three different SAR-based IT products are given

Remote Sens. 2020, 12, 3754 15 of 20

Figure 11 shows two scatter plots where the in-situ ice thickness measurements are plotted against the satellite products separately for Envisat ASAR and RS-2. The corresponding statistical measures describingRemote Sens. the2020 overall, 12, x performance of the three different SAR-based IT products are given in Table14 of 192. The comparison between the ASAR 0.5-km and 1-km products shows that the 0.5-km product performs in Table 2. The comparison between the ASAR 0.5-km and 1-km products shows that the 0.5-km slightly better with a mean R2 equal to 0.64 (Table2). For this reason, only the ASAR 0.5-km product product performs slightly better with a mean R2 equal to 0.64 (Table 2). For this reason, only the values are plotted in Figure 11a. As seen from Figure 11a and Table2, the best results are obtained ASAR 0.5-km product values are plotted in Figure 11a. As seen from Figure 11a and Table 2, the best for Vente˙ station where R2 is equal to 0.81 and a root mean square error (RMSE) is smallest, 9.7 cm. results are obtained for Ventė station where R2 is equal to 0.81 and a root mean square error (RMSE) The plausible explanation of this fact could be a more stable ice cover conditions near Vente˙ (for is smallest, 9.7 cm. The plausible explanation of this fact could be a more stable ice cover conditions 2 example,near Ventė see (for Figure example,9b,c). The see largestFigure RMSE9b,c). isThe obtained largest forRMSE Nida is stationobtained (14.1 for cm), Nida while station the (14.1 lowest cm), R iswhile obtained the for lowest Otkrytoye R2 is obtained station where for Otkrytoye the ASAR product station where both over- the and ASAR underestimates product both the over observed- and iceunderestimates thickness values. the observed ice thickness values.

FigureFigure 11. 11.Scatter Scatter plotsplots ofof in-situin-situ iceice thicknessthickness measurementsmeasurements made at at Nida, Nida, Ventė Vente˙ and and Otkrytoye Otkrytoye stationsstations versus versus ( a(a)) Envisat Envisat ASAR ASAR 0.5-km 0.5-km andand ((bb)) RADARSAT-2RADARSAT-2 0.5-km0.5-km iceice thicknessthickness products.

TableTable 2. 2.Statistical Statistical measures measures for for SAR-derived SAR-derived ice ice thickness thickness products as compared against against in in-situ-situ measurementsmeasurements in in the the Curonian Curonian Lagoon. Lagoon.

ASARASAR 1-km 1-km ASARASAR 0.5-km 0.5-km RS-2RS-2 0.5-km 0.5-km StationStation NN RR RMSE,RMSE, cm cm NN RR RMSE,RMSE, cmcm NN RR RMSE,RMSE, cmcm Nida 37 0.55 14.9 46 0.61 14.1 35 0.58 8.8 Nida 37 0.55 14.9 46 0.61 14.1 35 0.58 8.8 Ventė 39 0.81 8.6 46 0.81 9.7 33 0.65 9 Vente˙ 39 0.81 8.6 46 0.81 9.7 33 0.65 9 Otkrytoye 40 0.57 13.5 47 0.59 13 41 0.78 6.7 Otkrytoye 40 0.57 13.5 47 0.59 13 41 0.78 6.7 All 116 0.59 12.6 139 0.64 12.4 109 0.68 8.1 All 116 0.59 12.6 139 0.64 12.4 109 0.68 8.1 In contrast to the ASAR performance, the RS-2 ice thickness values have much smaller spread aroundIn contrast 1:1 line to(Figure the ASAR 11b) with performance, RMSE in the RS-2range ice of thickness6.7–9 cm (Table values 2). have Moreover, much smaller RS-2 data spread are aroundfound 1:1to predominantly line (Figure 11 b)underestimate with RMSE inthe the observed range of ice 6.7–9 thickness cm (Table levels2). at Moreover, all stations RS-2 unlike data the are foundASAR to product predominantly that partly underestimate overestimates the observed the observations. ice thickness The levels best at performance all stations unlike is obtained the ASAR for 2 productOtkrytoye that station partly with overestimates R equal to the 0.78 observations. and smallest The RMSE best value performance (6.7 cm). is The obtained correlation for Otkrytoye of RS-2 2 stationdata with with in R-situ2 equal measurements to 0.78 and at smallest Ventė station RMSE valueis R = (6.7 0.65 cm). but Theis characterized correlation ofby RS-2 the relatively data with in-situhigh RMSE measurements (around 9 atcm). Vent e˙ station is R2 = 0.65 but is characterized by the relatively high RMSE (around 9 cm). 5. Discussion The Baltic Sea is a vital waterway connecting densely populated and highly industrialized countries of , and the Port of Klaipėda is one of the busiest in the SE Baltic Sea. Shipping operations in the Port of Klaipėda, as well as navigation and fisheries in the Curonian Lagoon in wintertime depend on accurate now- and forecasting of local weather and ice conditions. Many numerical forecasting models suffer from a lack of calibration data because relevant ice cover properties such as ice extent, concentration, thickness, drift and deformation are difficult and

Remote Sens. 2020, 12, 3754 16 of 20

5. Discussion The Baltic Sea is a vital waterway connecting densely populated and highly industrialized countries of Northern Europe, and the Port of Klaipeda˙ is one of the busiest in the SE Baltic Sea. Shipping operations in the Port of Klaipeda,˙ as well as navigation and fisheries in the Curonian Lagoon in wintertime depend on accurate now- and forecasting of local weather and ice conditions. Many numerical forecasting models suffer from a lack of calibration data because relevant ice cover properties such as ice extent, concentration, thickness, drift and deformation are difficult and expensive to monitor [41]. In turn, the accurate knowledge of ice conditions obtained from satellite data leads to a more adequate modeling of hydrodynamic processes and circulation patterns in the Curonian Lagoon [42]. In this work we use visual interpretation and manual digitization of SAR data to retrieve ice cover extent in the SE Baltic Sea and the Curonian Lagoon. While the human supervised method could have various biases in defining the area of ice-covered regions, it is widely used by ice analysts to produce ice charts at FMI (Finland), Swedish Meteorological and Hydrological Institute (Sweden), and Antarctic Research Institute (Russia), etc. In our case, the main advantage of the method is the ability to preserve the initial high resolution of SAR data at 50–150 m spatial resolution that is critical to study the ice conditions over small coastal areas and semi-enclosed water bodies. Nevertheless, we must acknowledge that more elaborated and fully automated methods of sea ice parameter retrieval from SAR data exist [17–21,28]. We have also attempted to evaluate the quality of several SAR-based IT products produced by FMI and distributed by CMEMS, in the shallow semi-enclosed Curonian Lagoon where the number of in-situ ice thickness measurements is small. To our knowledge, there has been only a limited validation of these FMI IT products versus in-situ IT measurements made from the Finnish ice breakers in the northern Baltic Sea counting only for 27 match-ups in total [22]. In particular, the validation results showed rather large scatter and mismatch up to 50 cm between the SAR-based IT product and in-situ IT measurements for the range of observed IT values of 10–65 cm [22]. In our case, the largest difference between the ASAR 0.5-km IT product and in-situ IT measurements was 34 cm, while for the RS-2 0.5-km IT product it was even less, 22 cm, for the range of observed IT values of 4–55 cm. In this sense, our validation study based on a larger number of match-up points suggests that the considered SAR-based IT products perform better in the shallow Curonian Lagoon than in the northern Baltic Sea. Some information is also available on the performance of FMI multi-sensor IT product for the Barents and Kara Seas. This product is slightly different from the one used in our study. It is based on merging ice thickness derived from AMSR2 radiometer data, TOPAZ-4 coupled ocean-sea ice model and SAR (see details in [43]). The validation of this multi-sensor 1-km IT product based on RADARSAT-2 data versus in-situ IT measurements derived from ship-mounted electromagnetic induction instrument EM31 was done in March 2014 east of Svalbard [44]. The results based on 2758 match-up points showed good accuracy in identification of thin ice class (<30 cm). Only in 10% of cases the multi-sensor IT product underestimated the observed IT values above 30 cm and showed thin ice class (<30 cm) [44]. In our case, the three considered FMI IT products showed thin ice class and underestimated the observed IT values above 30 cm in 30% of cases, suggesting that multi-sensor IT product enriched with model data performs better at least for IT values above 30 cm. However, most of our results and those from the Arctic region were obtained for thin ice class. Yet, the detailed statistics of multi-sensor IT product performance for thin ice class in the Arctic is not available and, hence, we can’t make any solid conclusions on the quality of FMI IT products in the Curonian Lagoon in comparison to the results obtained in the Arctic. We should also note that the contemporary version of the FMI IT product for the Baltic Sea is largely based on Sentinel-1 data (in combination with RADARSAT-2, [39]) that were not considered in this study. Availability of data from two SAR sensors carried by Sentinel-1A and -1B suggests a higher frequency of SAR observations over our study site [3] and a better potential for future validation activities, provided the number of regular in-situ IT measurements would increase. Remote Sens. 2020, 12, 3754 17 of 20

Lastly, L-band SAR data from ALOS-2 PALSAR-2 were recently shown to have better results than conventional C-band SAR data in defining key sea ice parameters in the , including ice thickness [45]. As such, they could be also used in regular sea ice monitoring to complement the data available from operational C-band and X-band instruments for the Baltic Sea and better the quality of existing sea ice products.

6. Conclusions Here we show that a combination of spaceborne SAR and optical imagery with in-situ data from local hydrometeorological stations can significantly improve our knowledge about the ice regime in the seasonally ice-covered regions like the Southeastern Baltic Sea and the Curonian Lagoon. As shown, the ice conditions in the SEB and in the CL are strongly varying from year to year. In the SEB, the ice cover could be observed either only along the coast within 5–15 km band covering only about 3–4% of the study area (January 2013), or spread up to 100 km offshore and cover almost the entire SEB, as in February 2011. In turn, the Curonian Lagoon is almost fully ice-covered during all considered winters apart of its northern part that is subject to sea water inflow and various port activities. The SEB is also characterized by sporadic sea ice growth that may appear and decay several times per season (e.g., 6 episodes in 2010–2011) depending on background weather conditions. In contrast, the ice cover is more stable in the Curonian Lagoon where it usually lasts from the first frosts till the end of winter season. Nevertheless, it also possesses multiple periods of partial melting and re-freezing that becomes more frequent during the last decades. The dates of maximal ice extent in the SEB range from the end of January till the end of February. The mean ice season duration here is equal to 45 days ranging from 23 days in 2011–2012 to 68 days in 2010–2011. The ice season of 2011–2012 was the shortest during 2009–2013 both for the SEB and the CL, lasting for 23 and 71 days, respectively. However, the ice season duration in the sea and in the lagoon do not always correlate with each other. For example, in 2012–2013 it counted only for 24 days in the SEB (second lowest value) versus the longest ice season duration of 129 days in the lagoon. In the Curonian Lagoon, the rapid ice cover formation is often stimulated by the high winds above 10–15 m/s that intensify water mixing and cooling of the shallow brackish lagoon. The ice break-up usually starts in its northern part [3] and could be associated with an intense ice drift sometimes followed by heavy ice accumulation on the coasts under high winds. The highest ice cover frequency (i.e., longest ice cover season) is observed in the southeastern part of the lagoon, while the lowest one is clearly seen in its northern part. Here, the location and area of regions with the shortest ice season differs from year to year and is mainly determined by the wind conditions. Under prevalence of westerly-northwesterly winds, the sea water enters the lagoon and together with the mechanical wind action trigger the ice break up along the north-western CL coast. When the easterly and southeasterly winds dominate, the region of lowest ice cover frequency spans the entire northern part including the northeastern CL coast down to the northern Neman Delta. The comparison and validation of three SAR-based ice thickness products (ASAR 0.5-km, ASAR 1-km and RADARSAT-2 0.5-km) versus in-situ measurements has shown very promising results. All satellite products perform rather well for the periods of gradual ice thickness growth. During the episodes of the rapid ice thickness increase, all products underestimate the ice thickness by 10–20 cm (20–50%) reaching the observed level after about a month period. During such periods, the observed difference could be compensated by simply adding on average 15 cm to the SAR-based ice thickness values. The best results are obtained for the RADARSAT-2 0.5-km ice thickness product, it has the highest station-mean R2 value (0.68) and the lowest mean RMSE (around 8 cm), followed by the ASAR 0.5-km product. The overall performance of satellite products is better near Vente˙ and Otkrytoye stations where the ice conditions are more stable, while the largest RMSE and the lowest R2 values are obtained for Nida station characterized by more dynamic ice cover conditions. All in all, the validation results show that the SAR-based ice thickness products provide rather adequate ice thickness estimates for Remote Sens. 2020, 12, 3754 18 of 20 the shallow brackish coastal lagoon. Though our comparison spans only several winter seasons and doesn’t consider ice thickness products from other contemporary SAR missions (e.g., Sentinel-1), we believe that, upon careful validation, such products can provide useful information for spatial and temporal analysis of ice thickness variations in coastal waters and semi-enclosed shallow water bodies where the number of field observations is insufficient or lacking.

Author Contributions: Conceptualization, I.E.K., E.V.K. and A.G.K.; methodology, E.V.K. and I.E.K.; software, E.V.K. and I.E.K.; validation, I.E.K. and I.D.; formal analysis, I.E.K. and E.V.K.; investigation, all authors; data curation, E.V.K. and I.E.K.; writing—original draft preparation, I.E.K., E.V.K., A.G.K.; writing—review and editing, I.D., A.G.K.; visualization, E.V.K., I.E.K., I.D.; supervision, I.E.K.; funding acquisition, I.E.K. and E.V.K. All authors have read and agreed to the published version of the manuscript. Funding: This research and the APC was funded by the Russian Science Foundation, grant no. 17-77-30019. Collection of satellite data was funded by the state assignment of IO RAS, theme no. 0149-2019-0013. Study of spatial sea ice extent variability was funded by the Russian Foundation for Basic Research, grant no. 19-45-390012. Acknowledgments: The authors would like to thank LUKOIL-KMN, Ltd. (, Russia) for providing satellite radar imagery for the Southeastern Baltic Sea, and the Marine Research Department of the Environment Protection Agency of Lithuania for providing the hydrometeorological and the ice thickness data from coastal stations. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Granskog, M.; Kaartokallio, H.; Kuosa, H.; Thomas, D.N.; Vainio, J. Sea ice in the Baltic Sea—A review. Estuar. Coast. Shelf Sci. 2006, 70, 145–160. [CrossRef] 2. Chubarenko, B.; Chechko, V.; Kileso, A.; Krek, E.; Topchaya, V. Hydrological and sedimentation conditions in a non-tidal lagoon during ice coverage—The example of Vistula Lagoon in the Baltic Sea. Estuar. Coast. Shelf Sci. 2019, 216, 38–53. [CrossRef] 3. Idzelyte,˙ R.; Kozlov, I.E.; Umgiesser, G. Remote sensing of ice phenology and dynamics of Europe’s largest coastal lagoon (The Curonian Lagoon). Remote Sens. 2019, 11, 2059. [CrossRef] 4. Eremina, T.R.; Bugrov, L.Y.; Maksimov, A.A.; Ryabchenko, V.A.; Shilin, M.B. The Baltic Sea. In Second Assessment Report of Roshydromet on Climate Change and Its Concequences on Territory of Russian Federation; Kattsov, V.M., Semenov, S.M., Eds.; Institute of Global Climate and Ecology: Moscow, Russia, 2014; pp. 615–643. (In Russian) 5. Dailidiene,˙ I.; Baudler, H.; Chubarenko, B.; Navrotskaya, S. Long term water level and surface temperature changes in the lagoons of the southern and eastern Baltic. Oceanologia 2011, 53, 293–308. [CrossRef] 6. Omstedt, A.; Elken, J.; Lehmann, A.; Leppäranta, M.; Meier, H.E.M.; Myrberg, K.; Rutgersson, A. Progress in physical oceanography of the Baltic Sea during the 2003–2014 period. Prog. Oceanogr. 2014, 128, 139–171. [CrossRef] 7. Serykh, I.V.; Kostianoy, A.G. About the climatic changes in the temperature of the Baltic Sea. Fundam. Prikl. Gidrofiz. 2019, 12, 5–12. 8. Helsinki Commission (HELCOM). Climate Change in the Baltic Sea Area: HELCOM Thematic Assessment in 2013; Baltic Sea Environment Proceedings No. 137; Helsinki Commission: Helsinki, Finland, 2013; 66p. 9. Elken, J.; Lehmann, A.; Myrberg, K. Recent changes in Baltic Sea oceanographic conditions, with some effects to the status. In Proceedings of the 10th Baltic Sea Science Congress, Riga, , 15–19 June 2015; p. 34. 10. Haapala, J.; Ronkainen, L.; Schmelzer, N.; Sztobryn, M. Recent change—Sea ice. In Second Assessment of Climate Change for the Baltic Sea Basin; The BACC Author Team, Ed.; Springer: Cham, Switzerland, 2008; pp. 145–153. 11. Von Storch, H.; Barkhordarian, A. Detection and attribution of climate change in the Baltic Sea Region. In Proceedings of the 10th Baltic Sea Science Congress, Riga, Latvia, 15–19 June 2015; p. 8. 12. Dailidiene,˙ I.; Davuliene,˙ L.; Kelpšaite,˙ L.; Razinkovas, A. Analysis of the climate change in Lithuanian coastal areas of the Baltic Sea. J. Coast. Res. 2012, 28, 557–569. [CrossRef]

13. Kl, avin, š, M.; Avotniece, Z.; Rodinovs, V. Dynamics and impacting factors of ice regimes in Latvia inland and coastal waters. Proc. Latv. Acad. Sci. 2016, 70, 400–408. [CrossRef] Remote Sens. 2020, 12, 3754 19 of 20

14. Krek, E.V.; Stont, Z.I.; Bukanova, T.V. Ice conditions in the Southeastern Baltic Sea from satellite data (2004–2019). J. Ocean Res. 2020, 48, 18–33. (In Russian) [CrossRef] 15. Dailidiene,˙ I. Kuršiu˛mariu˛hidrologinio režimo pokyˇciai. Geografija 2007, 43, 36–43. 16. Bagdanaviˇciut¯ e,˙ I.; Umgiesser, G.; Vaiˇciut¯ e,˙ D.; Bresciani, M.; Kozlov, I.; Zaiko, A. GIS-based multi-criteria site selection for zebra mussel cultivation: Addressing end-of-pipe remediation of a eutrophic coastal lagoon ecosystem. Sci. Total Environ. 2018, 634, 990–1003. [CrossRef][PubMed] 17. Karvonen, J.; Simila, M.; Heiler, I. Ice thickness estimation using SAR data and ice thickness history. In Proceedings of the IEEE International Geoscience and Remote Sensing Symposium 2003 (IGARSS 03), Toulouse, France, 21–25 July 2003; Volume 1, pp. 74–76. 18. Similä, M.; Karvonen, J.; Haas, C. Inferring the degree of ice field deformation in the Baltic Sea using Envisat ASAR images. In Advances in SAR Oceanography from Envisat and ERS Missions, Proceedings of the SEASAR 2006, Frascati, Italy, 23–26 January 2006; Lacoste, H., Ouwehand, L., Eds.; ESA Publications Division, ESTEC: Noordwijk, The Netherlands, 2006. 19. Askne, J.; Dierking, W. Sea ice monitoring in the Arctic and Baltic Sea using SAR. In Remote Sensing of the European Seas; Barale, V., Gade, M., Eds.; Springer: Dordrecht, The Netherlands, 2008; pp. 383–398. [CrossRef] 20. Karvonen, J. Operational SAR-based sea ice drift monitoring over the Baltic Sea. Ocean Sci. 2012, 8, 473–483. [CrossRef] 21. Karvonen, J. Evaluation of the operational SAR based Baltic Sea ice concentration products. Adv. Space Res. 2015, 56, 119–132. [CrossRef] 22. Eastwood, S.; Karvonen, J.; Dinessen, F.; Fleming, A.; Pedersen, L.T.; Saldo, R.; Buus-Hinkler, J.; Hackett, B.; Ardhuin, F.; Kreiner, M.B.; et al. Quality Information Document for SI TAC Sea Ice Products; EU Copernicus Marine Service, 2020; Issue 2.9; 71p. Available online: https://cmems-resources.cls.fr/documents/QUID/ CMEMS-OSI-QUID-011-001to007-009to012.pdf (accessed on 6 October 2020). 23. Žaromskis, R. Oceans, Seas, Estuaries; Debesija: , Lithuania, 1996. (In Lithuanian) 24. Gasiunait¯ e,˙ Z.R.; Daunys, D.; Olenin, S.; Razinkovas, A. The Curonian Lagoon. In Ecology of Baltic Coastal Waters; Schiewer, U., Ed.; Springer: Berlin/Heidelberg, , 2008; pp. 197–215. [CrossRef] 25. Dailidiene,˙ I.; Davuliene,˙ L. Salinity trend and variation in the Baltic Sea near the Lithuanian coast and in the Curonian Lagoon in 1984–2005. J. Mar. Syst. 2008, 74, 20–29. [CrossRef] 26. Onstott, R.G.; Shuchman, R.A. SAR measurements of sea ice. In Synthetic Aperture Radar Marine User’s Manual; Jackson, C.R., Apel, J.R., Eds.; U.S. Department of Commerce: Washington, DC, USA, 2004; pp. 81–115. 27. Johannessen, O.M.; Alexandrov, V.Y.; Frolov, I.Y.; Bobylev, L.P.; Sandven, S.; Pettersson, L.H.; Kloster, K.; Babich, N.G.; Mironov, Y.U.; Smirnov, V.G. Remote Sensing of Sea Ice in the Northern Sea Route: Studies and Applications; Springer: Chichester, UK, 2007. 28. Dierking, W. Sea ice monitoring by synthetic aperture radar. Oceanography 2013, 26, 100–111. [CrossRef] 29. Muckenhuber, S.; Nilsen, F.; Korosov, A.; Sandven, S. Sea ice cover in Isfjorden and Hornsund, Svalbard (2000–2014) from remote sensing data. Cryosphere 2016, 10, 149–158. [CrossRef] 30. Kozlov, I.; Dailidiene,˙ I.; Korosov, A.; Klemas, V.; Mingelait˙ e,˙ T. MODIS-based sea surface temperature of the Baltic Sea Curonian Lagoon. J. Mar. Syst. 2014, 129, 157–165. [CrossRef] 31. Dabuleviciene, T.; Kozlov, I.E.; Vaiciute, D.; Dailidiene, I. Remote sensing of coastal upwelling in the South-Eastern Baltic Sea: Statistical properties and implications for the coastal environment. Remote Sens. 2018, 10, 1752. [CrossRef] 32. Karimova, S.S.; Gade, M. Improved statistics of sub-mesoscale eddies in the Baltic Sea retrieved from SAR imagery. Int. J. Rem. Sens. 2016, 37, 2394–2414. [CrossRef] 33. Ginzburg, A.I.; Krek, E.V.; Kostianoy, A.G.; Solovyev, D.M. Evolution of mesoscale anticyclonic vortex and vortex dipoles/multipoles on its base in the South-Eastern Baltic (satellite information: May–July 2015). J. Ocean Res. 2017, 45, 10–22. (In Russian) [CrossRef] 34. Ginzburg, A.I.; Krek, E.V.; Kostianoy, A.G.; Solovyev, D.M. On the role of mesoscale vortices in the oil spill transport in the Southeastern Baltic Sea (satellite information). J. Ocean Res. 2019, 47, 8–19. (In Russian) [CrossRef] 35. Kozlov, I.E.; Kudryavtsev, V.N.; Johannessen, J.A.; Chapron, B.; Dailidiene,˙ I.; Myasoedov, A.G. ASAR imaging for coastal upwelling in the Baltic Sea. Adv. Space Res. 2012, 50, 1125–1137. [CrossRef] Remote Sens. 2020, 12, 3754 20 of 20

36. Kozlov, I.E.; Artamonova, A.V.; Manucharyan, G.E.; Kubryakov, A.A. Eddies in the Western Arctic Ocean from spaceborne SAR observations over open ocean and marginal ice zones. J. Geophys. Res. Oceans 2019, 124, 6601–6616. [CrossRef] 37. Kozlov, I.E.; Plotnikov, E.V.; Manucharyan, G.E. Brief Communication: Mesoscale and submesoscale dynamics of marginal ice zone from sequential SAR observations. Cryosphere 2020, 14, 2941–2947. [CrossRef] 38. Stont, Z.; Bulycheva, E.; Bukanova, T. Ice conditions in the South-Eastern Baltic by satellite data and field observations in 2006–2013. In Proceedings of the Baltic Sea Science Congress, Klaipeda,˙ Lithuania, 26–30 August 2013; p. 251. 39. Karvonen, J.; Heiler, I.; Seinä, A.; Hackett, B. Product User Manual for Baltic Sea Ice Products SEAICE_BAL_SEAICE_L4_NRT_OBSERVATIONS_011_004/011; EU Copernicus Marine Service, 2018; Issue 2.7; 28p. Available online: https://resources.marine.copernicus.eu/documents/PUM/CMEMS-OSI-PUM-010-016. pdf (accessed on 6 October 2020). 40. Zemlys, P.; Ferrarin, C.; Umgiesser, G.; Gulbinskas, S.; Bellafiore, D. Investigation of saline water intrusions into the Curonian Lagoon (Lithuania) and two-layer flow in the Klaipeda˙ Strait using finite element hydrodynamic model. Ocean Sci. 2013, 9, 573–584. [CrossRef] 41. Löptien, U.; Axell, L. Ice and AIS: Ship speed data and sea ice forecasts in the Baltic Sea. Cryosphere 2014, 8, 2409–2418. [CrossRef] 42. Idzelyte,˙ R.; Mežin˙ e,˙ J.; Zemlys, P.; Umgiesser, G. Study of ice cover impact on hydrodynamic processes in the Curonian Lagoon through numerical modeling. Oceanologia 2020.[CrossRef] 43. Similä, M.; Mäkynen, M.; Cheng, B.; Rinne, E. Multisensor data and thermodynamic sea-ice model based sea-ice thickness chart with application to the Kara Sea, Arctic Russia. Ann. Glaciol. 2013, 54, 241–252. [CrossRef] 44. Melsheimer, C.; Mäkynen, M.; Rasmussen, T.; Rudjord, Ø.; Simila, M.; Sohlberg, R.; Walker, N. Comparison and validation of four Arctic sea ice thickness products of the EC POLAR ICE project. In Proceedings of the ESA Living Planet Symposium 2016, Prague, Czech Republic, 9–13 May 2016; Volume 740, p. 8. 45. Karvonen, J.; Rinne, E.; Sallila, H.; Mäkynen, M. On suitability of ALOS-2/PALSAR-2 dual-polarized SAR data for Arctic sea ice parameter estimation. IEEE Trans. Geosci. Rem. Sens. 2020, 58, 7969–7981. [CrossRef]

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).