Home Random Page


CATEGORIES:

BiologyChemistryConstructionCultureEcologyEconomyElectronicsFinanceGeographyHistoryInformaticsLawMathematicsMechanicsMedicineOtherPedagogyPhilosophyPhysicsPolicyPsychologySociologySportTourism






Satellite oceanography: New results 5 page

The conducted sub-satellite experiments indicate that the airborne microwave surveys, at wavelengths of 5.0 cm, allow identification of three types of ice: new ice, nilas, and first-year ice (the number of gradations depends on the content of water fraction in the sea ice body, sea ice current state, and aggregations). Microwave data on the size of floes are also important. Indeed, based on such data it is possible to assess the ice concentration inside the ice area. At the same time, the location of rookeries during the initial stage of whelping is invariably correspondent with the definite gradation of ice concentration (7-8, 8-9 tenths).


 

The airborne validation carried out according to the multi-level scheme

aircraft-satellite-ground-based in situ observations , suggested the possibility for routine remote monitoring of ice drift parameters (speed and direction), as well as of documenting the processes of breaking and destruction of individual floes of drifting ice (using a time series of SAR images). Multi-level SAR investigations of specific features of ice state and dynamics show that the location of whelping rookeries are relevant to the ice floes of special type and origin (Figure 6.29(a, b), see color section), namely the most freshened and, consequently, the strongest ice, which is formed in places where quasi-stationary fronts and spiraling structures and eddies originate (Melentyev and Chernook, 2002; Melentyev et al., 2003a).

A RADARSAT SAR image (Figure 6.30) shows a situation corresponding to mild, lightly frosty winter conditions. A quasi-stationary cyclonic circulation at the border between the Bassein and the Gorlo is clearly visible. In the lower left-hand corner of this SAR image, a demarcation of water masses and formation of a meander in the frontal zone (the dark signature) have been identified. The RADARSAT SAR image of the ice also reveals the initial stage of formation of another cyclonic structure in the coastal zone near the Terskiy coast.

Analysis of the SAR data displayed in Figure 6.30 shows that the ice in the zone of mass settlement of harp seals for whelping has a characteristic radar signal unlike that arising from surrounding water areas. Multi-level airborne investigations were conducted in various regions located in the Gorlo and the Bassein that are con- ventionally used by Greenland seals for whelping. Ample material from the satellite SAR imagery of the ice-seal system are directly relevant to Dvinskiy and Kanda- lakshskiy Bays, where there are no mass settlements of sea animals. Analogous data are also available for the Voronka and Mezenskiy Bay. This makes it possible to assess the specific regional features of ice dynamics as well as the changes occurring in ice parameters. It is shown (Chernook et al., 1998a) that the ice regime in the Voronka is dangerous for a newborn generation: various processes of ice deforma- tion and breaking of ice floes could provoke a mass destruction of harp seals.



Satellite SAR monitoring of ice cover parameters allows us to establish some indications of substantial differences in migration patterns and population numbers of Greenland seals, which are due to large-scale climatic factors, such as winter severity, ice coverage, and intensity of ice and water exchange between the White and Barents Seas (Chernook et al., 2002). Based on the analyses of SAR images, we obtained documented confirmation of an essential influence of the regional atmo- spheric processes (wind speed and direction) upon the routes of ice drift northwards (i.e., upon the process of migrations of seals toward the feeding regions in the Barents Sea). Interannual variations in the thermal regime and moistening processes within the catchment of the Severnaya Dvina River in summer and autumn are shown to determine the location of zones of whelping. The relationship between these phenomena may be due to interannual changes in the position of quasi-stationary hydrodynamic structures.

In some cases, SAR data help to reveal both the process of ice blocking inside the Bassein, and a reverse ice drift in the southern direction. It is shown that a return drift of ice inside the White Sea may occur during winters of different severity. Not


6.3

 

Figure 6.30. RADARSAT SAR image of the White Sea taken on 21 December, 2001. Development of cyclonic movements are visible in the southern part of the Gorlo and the north-western part of the Bassein.

 

 

infrequently, the drift becomes the cause of a subsequent ecological catastrophe consisting of the mass mortality of a newborn generation owing to ice destruction and contact of un-molted pups of harp seals with water and hyperthermia.

Figures 6.31 and 6.32 exemplify ice situations that occurred during the migra- tion of whelps of the harp seal as they are documented by ERS-2 SAR images taken on 4 March, 1999 and 20 March, 2000, respectively. The satellite swath is 100 km. Analysis of Figure 6.31 and other SAR images relevant to the winter of 1998-1999 allowed the assumption that there was a real possibility of mass mortality of harp


 

 

Figure 6.31. RS-2 SAR image of the White Sea and adjacent regions taken on 4 March, 1999.


6.3

 

Figure 6.32. RS-2 SAR image of the White Sea and the adjacent regions taken on 20 March, 2000.


 

seals due to the specific wind regime and the severity of this winter. As was revealed from satellite SAR data, the initial period of whelping was favorable for seals but the further development of ice conditions was anomalously severe; extensive ice coverage and a harmful influence of the prevailing wind could lead to a situation of high mortality of seals - a prediction that was supported by actual observations. The passive-microwave satellite data that was also used for the analysis have confirmed anomalously extensive ice coverage of the White Sea, and prevailing north-easterly and easterly winds: such weather conditions generally cause the blocking of ice areas (and ice-driven rookeries) in the Bassein and Gorlo. This forecast of a possible reduction of the population number and change of direction of conventional migration routes has been confirmed by subsequent observations during the following spring and summer periods. The results of thematic interpretation are as follows:

 

• A zone of mass settlement of seals for whelping is found in the north-western part of the Bassein.

• Another seal settlement is located in the central part of the Gorlo.

• Due to strong east/south-easterly winds, the areas of settlement of seals in the Gorlo are displaced towards the Terskiy coast.

• The texture and spatial distribution of SAR signatures of ice suggests that water temperature was predominantly low enough to initiate the freezing of open areas in the studied region during the satellite overpass. These conclusions are based on the SAR data that document intense formation of new ice in the low-salinity water areas in Dvinskiy Bay, around the Zimnegorsky Cape and within the zone influenced by the Belomorskiy stable current.

 

Results ensuing from the above corollaries are as follows:

 

• Areas of whelping rookeries of harp seals are situated in the Gorlo and the Voronka. A significant spatial variability of the radar signal suggests the poss- ibility of a high dispersion of seals over the entire area.

• Due to strong southerly and south-easterly winds, settlements of seals found are on the large and medium ice floes that were safely moved from the Gorlo into the Voronka.

• The formation of the initial types of ice, viz. slush and shuga ice, takes place under conditions of low air temperatures in the open parts of the sea.

• In general, the meteorological, hydrological and ice conditions in the 1999/2000 winter during the completion of migration of whelps are more favorable for seals than they were in the 1998/1999 winter: seal populations probably managed to successfully arrive at the main feeding grounds located along the southern border of the Barents Sea.

 

In practice, wider use of satellite SAR data is rather limited by its high cost. Therefore, an optimal methodology of airborne counting of mammal population numbers suggests a combined use of SAR and passive microwave data, as well as


6.3

optical satellite data. Figure 6.33 (see color section) presents a NOAA satellite image of the White Sea obtained on 21 March, 2000, which reveals a number of character- istic features relating to the distribution and state of ice. At the same time, the limitations of the optical method residing in insufficient sky illumination and cloudi- ness conditions in the Arctic are obvious.

The results of thematic interpretation are as follows:

 

• The NOAA data confirm the results of thematic interpretation of satellite SAR images; ice cover conditions are relatively favorable for the migration of harp seals. Triggered by strong southerly and south-westerly winds, a strong transfer of ice floes and whelping rookeries from the White Sea into the area of the Kanin Nos Cape is clearly visible.

• Patterns of distribution of drifting ice observed during late March suggest that molting rookeries of adults of the Greenland seal population in the White Sea in the spring of 2000 should have been found close to the north-western limits of the Bassein and off the area of the Kanin Nos Cape.

 

Satellite SAR mapping of ice and ice-associated seal migration routes became a reality with the application of RADARSAT and ENVISAT SAR imagery. These sensors allow monitoring of the sea surface within an image area measuring 500 x 500 km (on horizontal and vertical polarizations, respectively). In some field experiments, we combined RADARSAT, ENVISAT, and ERS data to attain a more comprehensive analysis of the White Sea.

Figure 6.34 illustrates the feasibility of satellite SAR data for mapping the spatial and temporal variability of ice cover parameters and ice drift features (white arrows on both the SAR images indicate the direction and daily averaged speed of ice floe displacement). These SAR images taken over Dvinskiy Bay and the adjacent part of the Bassein were obtained on 27 and 28 February, 1998, respec- tively. Combined use of RADARSAT and ERS SAR data makes it possible to better investigate ice cover parameters, taking advantage of polarization effects as addi- tional diagnostic options.

It is useful to return to Figure 6.27 and consider it in more detail. It presents an ERS-2 SAR image taken over the White Sea (12 March, 1997; 08:39 GMT) and a photograph of a whelping rookery located at the nadir (sub-satellite point). The satellite survey encompasses the Gorlo, the south-western part of the Voronka, and a part of Mezenskiy Bay (including Morzhovets Island). The Greenland seal conventionally chooses these parts of the White Sea for establishing whelping and molting rookeries.

The typical ice conditions and specific features of animal welfare in the initial stage of whelping are illustrated in Figure 6.27(b). In this particular case, the ice cover can be characterized as thin, first-year ice 50-70 cm thick. The ice surface is level and snow-free. From indirect characteristics (the color of the ice floe and color tints of the hummock fractures), it can be assessed as brackish-water ice. The female Greenland seals have a gray-white coat with large dark spots on their backs; the calf is white. Ice aerial photography, ice reconnaissance, and detection of not only indi-


 

 

 

Figure 6.34. RADARSAT and ERS SAR images of the White Sea and adjacent regions taken on (a) 27 March and (b) 28 March, 1998, respectively.


6.3

vidual animals but also their considerable assemblages on drifting ice is a difficult, costly, labor and time-consuming problem. Aerial counting of sea animals - parti- tioning them into numbers of females and pups - is one of the most important goals of marine biological and nature-protective studies in the Arctic. This is an essentially difficult task because of the slight contrast between the white coat of pups, white color of snow, and ice in the Arctic. Therefore, the possibility of effective detection of ice zones potentially suitable for organizing whelping of the Greenland seal is an essential achievement of SAR-based satellite surveillance.

The accomplished investigations of ice as an abiotic factor of the Greenland seal ecology make it possible to study the modification of ice states and relevant trans- formations of radar signatures with the ice growth, drift, and changing of state (ridging and rafting, stratification, hummocking, breaking, ice and snow moistening, pollution, as well other factors). Figure 6.27 shows that the ice area used for whelping is characteristic of a deep dark signature (zone A), contrasting with the areas of open water and broken, hummocky or rafted ice whose signatures are various tones of gray (zone B). The bright white signal corresponds to regions of ice-free wind-roughened sea areas (zone C). The radio signal of land also has a dark tone, but nevertheless differs in texture from the signal coming from a level ice zone (zone D).

Analysis of a consecutive 3-day series of satellite SAR images reveals the dis- placement, destruction, and melting of the above-mentioned dark areas of ice- breccia floes of the level first-year ice, where whelping rookeries are generally situated. Such data allow one to follow the dynamics of the floe drift vector and monitor animal migrations within and outside the White Sea.

Figure 6.27 provides grounds to suppose that the sites of settlements of seals for whelping are located in the central, eastern, and north-eastern parts of the Gorlo. The camping zone A are broken into two parts by the previous wind direction reversals: influenced by the Belomorsky stable current, they are extended along the Zimniy coast. Part of the rookeries is pressed to the Zimniy coast by strong (up to 10-12 m s-1) north- westerly winds that were still active at the moment of the ERS-2 satellite overpass. Detailed analyses of these satellite ice data indicate that SAR images can be used for the solution of various applied problems, including those associated with the health state and the preserved population numbers of the newborn generation. SAR images can also be helpful in terms of developing eco-tourism and preparing recommendations to locate icebreaker and transport ship navigation routes off the zones of mass settlements of seals during whelping and molting.

The aforementioned issue of eco-tourism is highly relevant for the White Sea region. Special satellite monitoring of the arrangement and spatial distribution of drift and fast ice in the White Sea are necessary to determine the most appropriate and secure areas eligible for eco-tourism. As has been shown by Melentyev and Chernook (2002), that the possibility of sub-surface sounding of ice, by using spectral SAR signatures, makes it possible to detect animals on the ice not only with high precision but also to predict their appearance in specific areas of the ice- covered White Sea. This can optimize a proper choice of the strongest and most rapidly floating ice floes.


 

Counting of the population numbers and tracing migration features can also be achieved using IR spectrometry. Some results of comprehensive aerial studies, carried out in March, 2000, are presented in Figures 6.35 and 6.36(a, b) (see color section). Analysis of these data revealed that the 1999/2000 winter was, on the whole, favorable for the Greenland seal population in the White Sea: the air temperature and wind conditions contributed to a speedy movement of rookeries from the Bassein and the Gorlo towards the feeding regions situated in the Barents Sea.

In conclusion, we need to mention that at present, comprehensive studies of the Greenland seal using both airborne surveillance and SAR remote sensing are conducted regularly. The SAR-based studies of sea ice cover and seals as a single system have been supported by marine biologists. The next step is the development of methods of remote monitoring of behavioral patterns of some other sea animals. In particular, during the 2003/2004 winter season, in cooperation with some Swedish colleagues, pilot satellite studies of migration routes of gray seals across the Gulf of Bothnia in the Baltic Sea were conducted (Melentyev et al., 2003b).

A further improvement in the methodology of remotely sounding the White Sea ice cover parameters as a habitat of ice-associated seals (harp and ringed seals) is conditioned by optimization of observation tools as well as by perfection of tech- niques of thematic interpretation of images taken in spectral regions ranging from the visible to microwave. Also, to achieve a higher efficiency of airborne investigation and broaden the number of meteorological and oceanographic parameters retriev- able from satellite data, a combined use of the advantages inherent in multispectral airborne surveillance and satellite images should be made routine.

 

 

6.4 STUDIES OF THE WHITE SEA ICE COVER USING SATELLITE PASSIVE MICROWAVE SENSORS

 

6.4.1 Introduction

Data from satellite passive-microwave sensors are at the moment among the longest continuous and globally geophysical of observational records, extending over more than two decades (Johannessen et al., 1999). The Nimbus-7 Scanning Multichannel Microwave Radiometer (SMMR) had been in use from 1978 up to 1987 and was replaced by follow-up SSM/I sensors, which remain operational to the present time. Recent data analyses using the NASA Nimbus-7 satellite microwave sensors has been undertaken to study changes in sea ice concentration in the White Sea during the period November 1978-December 1999.

The data on multi-frequency brightness temperature (Tb) from these passive microwave sensors are used to calculate the total sea ice concentration, providing therefore the information about the variability of total ice extent and sea ice con- centration. Since different radiative properties of open water are not infrequently similar to those of sea ice, a wealth of algorithms using multi-channel microwave data have been established in order to estimate the relative extent of open water and sea ice area (NORSEX, 1983; Gloersen and Cavalieri, 1986; Comiso, 1990;


Sec. 6.4]


 

6.4
Studies of the White Sea ice cover using satellite passive microwave sensors 235


 

Grenfell, 1992). The surface spatial resolution of these data is 25km. The satellite data were used only during wintertime since the White Sea is completely ice-free during summer. During winter, the microwave signature of the sea surface is rela- tively stable compared with the rest of the year.

The Norwegian Remote Sensing Experiment (NORSEX) algorithm (Svendsen et al., 1983) has been used to numerically assess the concentration of sea ice from the SMMR data using values of Tb in the 18 and 37-Ghz channels and SSM/I Tbin the

19 and 37-Ghz channels. The data on total ice concentration was iteratively estimated and adjusted for the overlap period when SMMR and SSM/I sensors were concurrently operational (i.e., from July to August, 1987). This was done in order to eliminate inevitable differences in the sea ice mapping by the two kinds of sensors (Bj0rgo et al., 1997). No adjustments were made to processing SSM/I data from different satellites, viz. F8, F11, and F13. More details regarding the sensor calibration are given elsewhere (Bj0rgo et al., 1997; Johannessen et al., 1999).

 

 

6.4.2 Data analysis

In order to estimate and compare the seasonal and interannual sea ice concentration variability in different regions of the White Sea, the marine basin has been divided into five regions (Figure 6.37). In order to avoid any interference with land pixels, only pixels devoid of any contamination with signals inherent in the land surface emission were carefully selected and spatially averaged. Figure 6.38 (see color section) shows the areas in the White Sea Basin left for further analyses; dashed lines demarcate the border of each region.

Figures 6.39 and 6.40 (see color section) display mean monthly values of ice concentration for January and February of each year. There is a surprisingly high correlation between all these areas. All regions show some decrease in the total ice content with the least values in 1992. This minimum, however, was then followed by a slight but remarkable rise observed for all selected regions. These results are ex- plicitly indicative of being in remarkable compliance with previously published results (Johannessen et al., 1999).

Figure 6.41 (see color section) illustrates the retrieval results for all winter months (from January to April). The retrieval results appear more scattered. This can perhaps be explained by a lower stability of the signatures of the brightness temperature in March and April when early melting of sea ice cover begins turning the entire sea ice system into a transitional state. The signals could have been affected by the presence of meltwater layers on the ice cover.

The results of the analyses are summarized in Table 6.14 and reveal a significant decrease in the ice concentration for all areas in the White Sea varying from -20% (Onezhskiy Bay) to -40% (the Voronka) during the period 1978-1992. Further on, all regions showed an opposite tendency (i.e., an increasing trend in the sea ice concentration ranging from 8% to 22% during the 1992-1999 period). In the Voronka, the sea ice concentration attained its highest increase (up to 22%) with respect to the minimum in ice concentration registered in 1992.


 

 

Figure 6.37. The White Sea Basin as partitioned into five regions: 1. Onezhskiy Bay; 2. Dvinskiy Bay; 3. the Bassein; 4. Mezenskiy Bay; and 5. the Voronka.

 

 

Table 6.14. A summary of trend analyses of changes in the ice concentration in the White Sea. Changes (!i) in % are assessed based on mean monthly values. The observation period has been divided into two sub-periods: from November 1978 to January 1992 and from January 1992 to December 1999. The total length of the combined times series is 21 years.

 

Summary of trend analyses for two sub-periods

 


 

Approximately


11/1978-01/1992 01/1992-12/1999 Total change


Aaa in pixels Slope Slope Slope Region (each 625km2) !i (%) (% yr-1) !i (%) (% yr-1) !i (%) (% yr-1)

 

Mezenskiy Bay -22 -1.67 +2 +0.25 -20 -0.94
Voronka -40 -3.03 +32 +4.0 -8 -0.38
Bassein -27 -2.04 +7 +0.88 -20 -0.94
Onezhskiy Bay -20 -1.52 +9 +1.13 -11 -0.52
Dvinskiy Bay -22 -1.67 -22 -1.04

6.5

Analyzing the entire observational period extending from 1978 to 1999, it can be stated there was a 16.2% reduction in the ice cover over the 21-year period, or "' 8.1% per decade. The total surface area of the White Sea that has been investi- gated amounts to 30,000 km2.

In conclusion, it is important to note that the present work provides evidence that the sea ice cover in the White Sea is in a transitional state. This might lead to changes in atmospheric conditions over the White Sea Basin. However, it is clear that 21 years of observations are insufficient in order to confidently establish any long-term variations in the sea ice cover. Instead, it could be that these fluctuations are driven by a variability in the North Atlantic Oscillation (NAO) (cf. Chapter 3). Indeed, several studies (e.g., Mysak and Venegas, 1998; Deser et al., 2000; Dickson et al., 2000) indicate that the NAO is strongly coupled with the regional sea ice fluctua- tions on regional scales.

 

 

6.5 CONCLUSIONS

 

Large amounts of satellite remote sensing data from AVHRR, SeaWiFS, and microwave sensors were collected and processed for the White Sea. Data from SeaWiFS were collected for cloudless/low-cloudiness days. The data processing consisted of a number of steps: the raw satellite data were subjected to the MUMM atmospheric correction code. The atmospherically corrected images were processed with the LMMVOA (Levenberg-Marquardt Multivariate Optimization Algorithm), at the base of which the NERSC hydro-optical model has been utilized. The results of retrieval (pertaining to the water quality proxies: concentra- tion of phytoplankton chlorophyll, suspended minerals, and dissolved organic matter) allowed mapping of the spatial and temporal distributions of the above proxies. The data were collected for the time period 1997-2002. Some periods were investigated in more detail (e.g., the entire phytoplankton vegetation period in 2001 was studied closely and compared with numerical simulation results). Moreover, a dedicated field campaign was conducted in the summer of this year, which allowed direct comparison of the retrieval results with in situ measurements. The results of water quality proxies and SST retrievals provided the spatial and temporal variations of these parameters. Seasonal as well as interannual variations in the White Sea algal bloom development and trophy dynamics were distinctly revealed from SeaWiFS data. Three periods of algal bloom development were detected: early April (algae development in the ice, which, due to somatic mechan- isms, is discernible on the melting ice/snow surface), May-June (a most pronounced algal bloom), and September-October. Dissolved organics and suspended minerals exhibit distinct seasonal patterns. The first of them is mostly controlled by algae lifecycles, whereas the second is governed by seasonal variations of river discharge. Apart from the visible image data, NOAA AVHRR IR data on the White Sea were collected with the purpose of retrieving water surface temperature. The data from the available remote sensing archives were collected for the period 1983-2001. The McClain/NASA standard algorithm was used for retrieving the SST from


 

atmospherically corrected AVHHR data. The results of water quality proxies and SST retrievals from SeaWiFS and AVHRR data respectively, compared very well with the quasi-synchronous in situ validation data.

The satellite-retrieved SST data provide information on the location and temporal variations of upwelling phenomena and fronts in the White Sea, as well as seasonal cooling/warming of littoral and pelagic waters. A permanent front is distinctly detectable in the area separating the central part of the Bassein and the Gorlo connecting the White Sea with the boundary between the waters inherent in the White and Barents Seas. A permanent upwelling is clearly distinguishable in the area of the Solovetskiy Archipelago. This satellite remote sensing documentation of the water temperature dynamics in the littoral zone within the major bays of the White Sea (Kandalakhskiy, Onezhskiy, Dvinskiy, and Mezenskiy) proved to be very important for the interpretation of both thermo-hydrodynamic and biological changes recorded by ground-based observations and obtained from numerical simulations.


Date: 2016-03-03; view: 610


<== previous page | next page ==>
Satellite oceanography: New results 4 page | Satellite oceanography: New results 6 page
doclecture.net - lectures - 2014-2024 year. Copyright infringement or personal data (0.013 sec.)