ISSN 1991-346X Series physico-mathematical. 3.

2020

NEW S
OF T H EN A TIO N A L A C A D EM Y OF SCIENCES OF TH E REPU BLIC OF K A ZA K H STA N
P H Y S IC O -M A T H E M A T IC A L S E R IE S
ISSN 1991-346Х https://doi.org/10.32014/2020.2518-1726.52
Volum e 3, N um ber 331 (2020), 183 - 190

UDC 528.85/.87(15)

Z h . Sh. Z h a n ta y e v , D. V. C hepashev,
А. А. K a ld y b a ev , S. M . N u ra k y n o v , K . M . U n g a ro v

Institute of Ionosphere JSC "NCSRT", Almaty, Kazakhstan.

E-mail: [email protected], [email protected], [email protected],
[email protected], [email protected]

MAPPING FREQUENCY OF OIL SPILLS

IN THE CASPIAN SEA USING SENTINEL-2 OPTICAL IMAGES

A bstract. The main goal of this work is to evaluate the capabilities of optical images in the problem of
detecting oil spills, testing them in the task of mapping zones with a high frequency of oil spills during one year.
Using SAR imagery, sometimes it is difficult to separate anthropogenic oil spills and false targets, usually called
“look-alikes”. Using optical images for oil spill detection, we rely on the spectral characteristics of objects, and
although the reflectivity varies depending on the thickness and composition of the slick, in most cases, oil spills can
be distinguished from “look-alikes”. The region near Absheron peninsula was selected as a study area. Optical
images of the Sentinel-2 (A, B) satellites were used to detect oil spills and calculate the zones of their most frequent
occurrence. Object based approach was chosen for segmentation and classification images. The map of hot spots
with high frequencies of oil spills during 2019 year was presented in this article. Conclusions about the advisability
of using optical images along with radar data for monitoring oil spills were made.
Key w ords: oil spills detection, Caspian Sea, Sentinel-2, optical imagery.

In tro d u c tio n . The environm ental situation in the Caspian Sea, w hich is literally saturated w ith
products o f the oil industry, is one o f the acute issues for the littoral states. The Caspian Sea is subject to
increasing anthropogenic impact, there is an increase in w ater pollution by oil products, heavy m etals, and
chem ical products. A t a rapid pace, there is a decrease in m arine bioresources. The m ain sources o f sea
pollution are: river runoff; industrial and m unicipal effluents; oil production at sea and onshore;
transportation o f oil products by sea or through pipelines; flooding/drainage o f the coastal zone as a result
o f the rise/fall o f the Caspian level [1].
Oil spills at sea can be o f natural origin (natural oil em issions due to fractures o f geological structures,
griffins, m ud volcanoes), but in m ost cases they are o f anthropogenic nature: they arise as a result o f ships
and tankers discharging ballast and bilge w ater, due to the rem oval o f oil stains or pollution w ith river
flows, industrial effluents, in the event o f accidents o f ships and tankers, in the breakdow n o f offshore oil
pipelines, in the discharge o f drill cuttings and in accidents due to exploratory drilling and industrial
production [2,3]. This is especially felt in offshore zones w here oil and gas fields are being developed, oil
is extracted and transported. Oil pollution in coastal zones has a detrim ental effect on flora and fauna,
leads to pollution o f coastlines, w hich entails large m aterial losses associated w ith the cleaning o f these
territories. The m arine ecosystem o f the Caspian Sea, a closed reservoir, w hich self-cleaning ability is very
small, is especially sensitive. A lm ost every year, there are reports o f the death o f sturgeon, seals, birds, in
the carcasses o f w hich oil products are found [4].
International experience shows that space m onitoring m ethods are the m ost effective w ay to detect
and evaluate the developm ent dynam ics o f m ost natural and a num ber o f m an-m ade em ergencies, in
particular oil spills. M odern system s o f remote sensing m ake it possible to obtain overview and detailed
inform ation about em ergencies o f various sizes in large areas. A t the same tim e, space m onitoring

183
News o f the National Academy o f sciences o f the Republic o f Kazakhstan

m ethods are m uch more econom ical than traditional ground-based data collection m ethods. It is practically
im possible to control oil spills in the Caspian Sea scattered over a gigantic area (over 377,155 thousand
square kilom eters) by traditional land-based m ethods, and the use o f aerial observations is expensive and
can only be carried out in lim ited local areas. Therefore, remote sensing is the m ost effective m ethod for
tim ely detection o f oil spills in the Caspian Sea w ith its large territory.
There are tw o types o f sensors used in remote sensing - active (radar) and passive (optical).
B oth types o f sensors have their advantages and disadvantages. Optical sensors can detect oil spills
only in the daytime w hen there is no cloud, fog or sm oke above the area o f interest, that is, w hen nothing
obscures the object from the sensor. Radar im ages can detect oil spills despite the cloudiness and tim e o f
day. A lso , a large coverage area o f radar im ages is an advantage. N evertheless, there are weather
conditions under w hich radar im ages are "powerless" for this task, nam ely, rain, and w ind force above the
sea surface beyond the interval o f 2.5-12.5 m /s [5-8]. Due to m any o f the advantages listed above, radar is
n ow the m ost com m on m eans o f o il spill remote sensing.
H ow ever, using SA R im agery, som etim es it is difficult to separate anthropogenic oil spills and false
targets, usually called “look -alik es”, w hich include lo w w ind areas, areas, sheltered by land, rain cells,
organic film s, grease ice, w ind fronts, up-w elling zon es, oceanic fronts, algae bloom s, current shear zones,
etc. One group o f researchers show ed that even w ith extensive processing o f SA R (synthetic aperture
radar) im agery, 20% o f the im ages reported as “oil sp ills” were still “look-alikes” [9]. This m eans that
using on ly radar im ages to m onitor o il spills, about a fifth o f all detected objects w ill be false targets, and
w hen calculating their total area or analyzing m aps built on such data, the error w ill be large.
U sin g optical im ages for o il spill detection, w e rely on the spectral characteristics o f objects, and
although the reflectivity varies depending on the thickness and com position o f the slick, in m ost cases, oil
spills can be distinguished from “look -alik es”. O f course, lim itations o f optical im agery are a serious
obstacle to m onitoring efficien cy. But in cases w hen m apping accuracy is m ore important, optical im ages
are needed.
So, purpose o f this work w as to test opportunities o f optical im agery in oil spill detection task by
trying m apping hot spots w ith high frequency o f oil spills during year.
S tu d y area. The region near A bsheron peninsula w as chosen for a reason. Oil spills detected more
often in this region than in other regions o f the Caspian Sea. Accurate oil spill frequency map o f this
region, built by using optical sensors, w ill be useful for future oil spill m onitoring. There are 3 oil fields in
this region: Oil R ocks Settlem ent, C hilov and Pirallahi Islands. Oil R ocks Settlem ent is located 35 km
from the coast o f the Caspian Sea. Its developm ent began in 1949. Oil R ocks is an industrial settlem ent in
Baku, Azerbaijan. A full tow n on the sea, it w as the first oil platform in Azerbaijan, and the first operating
offshore o il platform in the world, incorporating num erous drilling platforms [10]. Pirallahi Island is
located 1.6 km from the coast o f the Caspian Sea. C hilov Island is located 16 km from the coast o f the
Caspian Sea [11]. The map o f the Oil Rocks Settlement, Chilov and Pirallahi Islands is presented in figure 1(b).

a) b)
Figure 1 - a) Study area; b) Map of the Oil Rocks Settlement, Pirallahi and Chilov Islands

184
ISSN 1991-346X Series physico-mathematical. 3. 2020

M eth o d s a n d d a ta p ro cessin g . Optical im ages o f the Sentinel-2 (A , B ) satellites w ere used to detect
oil spills and calculate the zones o f their m ost frequent occurrence during 2019. Sentinel-2 is an Earth
observation m ission from the Copernicus Programme that system atically acquires optical im agery at high
spatial resolution (10 m to 60 m ) over land and coastal waters. The m ission is a constellation w ith tw o
tw in satellites, Sentinel-2A and Sentinel-2B [11].
E ighteen L1C im ages w ith less than 20% cloud cover were dow nloaded from the U SG S
Earthexplorer w ebsite. D ow nloaded im ages w ere pre-processed to obtain L 2A im agery from L1C.
Sen2C or Processor w as used for that purpose. Sen2Cor is a processor for Sentinel-2 L evel 2A product
generation and formatting that performs the atm ospheric, terrain and cirrus correction o f Top-O f-
A tm osphere L evel 1C input data. Sen2C or creates B ottom -O f-A tm osphere, optionally terrain and cirrus
corrected reflectance im ages. In addition, Sen2C or generates A erosol Optical Thickness, W ater Vapor,
Scene C lassification M aps and Q uality Indicators for cloud and snow probabilities. Its output product
format is equivalent to the L evel 1C U ser Product: JPEG 2000 im ages, three different resolutions, 60, 20
and 10 m [12, 13].
O bject based segm entation and classification w as used for oil spill detection. In contrast to pixel-
based classification that classify pixels one-by-one, object-based approach first groups im age p ixels into
spectrally h om ogenous im age objects using an im age segm entation and then classifies the individual
objects [14, 15]. Then detected o il spills objects w ere checked through visual interpretation. A fter all, GIS
overlay analysis w as perform ed for com putation o f oil spill frequency, and determination o f
spatiotem poral distribution. The w orkflow for the detection o f oil spill frequency presented in figure 2.

Sentinel-2
L1C im a g e ry

SenZC or

S e g m e n ta tio n

M ultire so lu tio n Spectral Difference

Segm entation C=> Segm entation

C lassification

Select Samples fo r Nearest Neighbor

Class Hierarchy Classification

Й .
V isual In te rp re ta tio n and GIS O v e rla y Analysis

Figure 2 - Workflow for the detection of oil spill frequency

185
News o f the National Academy o f sciences o f the Republic o f Kazakhstan

R esu lts an d d iscu ssio n . Oil spills in the study area for the m ost cases are easily detected visually in
the visib le range o f the electrom agnetic spectrum, they usually have a silver-blueish color (figure 3 a, b),
w hich indicates the thinness o f the slicks [16].

Figure 3 - Well visible oil spills (a-The Oil Rocks Settlement, b - Pirallahi Island)

B ut it should be noted that in som e cases in the zones o f accum ulation o f biogenic material such as
surface seaw eeds or sunken kelp beds oil detection w as difficult. One o f that cases presented in figure 4.
Under m any conditions, oil is not visib le on the surface o f the water [17]. Oil spills on shorelines is
difficult to identify p o sitively because seaw eeds look sim ilar to oil and oil cannot be detected on darker
shorelines [18].

Figure 4 - One of difficult cases

In addition, there remain lim itations such as cloudiness in the im ages. Indeed, m any pictures for 2019
w ere cloudy over 20%. This is com pletely unsuitable for operational monitoring.
Therefore, it is better to take an integrated approach, nam ely the use o f both radar and optical im ages.
This approach also partially solv es the issue o f validation in those days w hen the sensing days coincides o f
both radar and optical sensors. This allow s us to compare the obtained oil spill detection results. Exam ple
o f that com paring presented in figure 5 (a, b), where w e see the correspondence.

a) b)
Figure 5 - a) Radar image on 26thof April; b) Optical image on same date

186
ISSN 1991-346X Series physico-mathematical. 3. 2020

The detection o f oil spill from Sentinel-2 im ages and m ulti-tem poral GIS overlay analysis resulted in
the map o f hot spots w ith high frequencies o f oil spills during 2019 year, presented in figure 6. On the map
w e see that the m ain hotspot o f oil spills is the Oil R ocks Settlem ent, what w as to be expected. The
m edium class o f oil leak sources w ere observed around the Oil R ocks Settlem ents and C hilov Islands and
the sm all and occasional leak sources w ere observed around Pirallahi Island.
This map w ill allow to increase the accuracy and reliability o f detected oil spills in future monitoring.
For exam ple, the oil spill areas w ith no overlaps and far aw ay from the oil spill hot spots w ill not be
m apped because o f the low er reliability o f occurrence.
50‘ 200"E 50‘ 400'E 51 W E Z
1 1--------------------------------------- 1--------------------------------
---------------------------------------------------------------- ----------------------------------------------------------------------

10 4 i km N
14

A
Mxdakyon

Z l fyo

8-

O il S p ill F re q u e n cy 2019

High

I
Sources: Esri. HERE, Garmin, Intermap, increment P Corp..
GEBCO, USGS, FAO, NPS, NRCAN, GeoBase, IGN, Kadaster
l-o w NL, Ordnance Survey. Esri Japan. METI. Esri China (Hong Kong).
(c) OpenStreetMap contributors, and the GIS User Community
--------------------- 1--------------------------------------- 1----------------------------------------1-------------------
50'20’0‘E 50*4(T0"E 51W E

Figure 6 - Map of oil spill frequency in 2019

C o n clu sio n . R esearch in this direction w ill be continued w ith the use o f more remote sensing data,
their statistical analysis, w ith the aim o f increasing the likelihood o f detecting oil pollution and its
practical application. Preliminary results are considered successful and consistent, w ith a high degree o f
applicability to other Sentinel-2 satellite im agery. Further testing and proper tuning o f the proposed
object-oriented m ethodology should be carried out using SA R im ages.
T hese w orks are carried out w ith financial support by the CS M ES RK, in the framework o f the
scientific and technical program 0 .0 7 8 2 “D evelopm ent o f a m ulti-purpose aerospace forecast monitoring
system (M A K SM ), as w e ll as the creation on its basis o f services for the com prehensive presentation o f
em ergency and natural-generated em ergency information in conjunction w ith sem antic and geospatial
data”.

187
News o f the National Academy o f sciences o f the Republic o f Kazakhstan

Ж. Ш. Ж антаев, Д. В. Чепашев, А. А. ^алды баев, С. М. Нуракынов, К. М. Унгаров

ЕЖШС «Ионосфера институты» АК «¥лттык гар^1штык зерттеулер мен

технологиялар орталыгы», Алматы, Казакстан

SENTINEL-2 ОПТИКАЛЬЩ СУРЕТТЕР1Н П АЙДАЛ АНА ОТЫ РЫ П,

КАСПИЙ Т Е Щ ЗЩ Д Е П МУНАЙДЫН Т 0Г1Л У Ж Ш Л1ГШ КАРТАГА T Y C IPy

Аннотация. Мунай eнеркэсiбiнiн eнiмдерiмен ласташан Каспий генiзiндегi экологиялык жагдай -

жагалау манындагы мемлекеттер Yшiн мацызды проблемалардын бiрi. Каспий тенiзi куннен кYнге
антропогендiк эрекетке ушырауда. Жыл сайын судын мунай eнiмдерiмен ластануы артып келедi. Тенiз
биоресурстарынын кыскаруы тез каркын алуда. Халыкаралык тэж1рибе керсеткендей, гар^1штык мониторинг
эдiстерi кептеген табиги жэне техногендiк тетенше жагдайлардьщ, атап айтканда, мунайдын тeгiлуiнiн даму
динамикасын аныктау мен багалаудын негурлым тиiмдi тэсiлi болып табылады. Ка^рп замангы
кашыктыктан зондылау жYЙелерi Yлкен аумактардагы эртyрлi кeлемдегi тетенше жагдайлар туралы шолу
жэне толык акпарат алуга мYмкiндiк бередi. Сонымен катар, гарыштык мониторинг эдiстерi деректердi
жинаудын дэстYрлi жерYCтi эдiстерiнен элдекайда Yнемдi. Кашыктыктан зондылауда сенсорлардын екi т д а
колданылады - активтi (радарлык) жэне пассивт (оптикалык). Сенсорлардын екi туршщ де eз
артыкшылыктары мен к е м ш ш г а ^ бар. Кeптеген артыкшылыктардын аркасында радарлык суреттер казiргi
уакытта мунайдын тeгiлуiн кашыктыктан зондылаудын ен кeп таралган куралы болып табылады. Алайда,
радарлык суретгердi пайдалана отырып, кейде антропогендш мунай тeгiлуiн жэне оларга уксас жалган
объектiлердi бeлiп алу киын болады жэне эдетте оларга "look-alikes" деп аталатын, желсiз аймактарды,
табиги текп органикалык пленкаларды, балдырлардын гулденуш жэне т. б. камтиды. Бул жумыстын н епзп
максаты - бiр жыл iшiнде мунай тeгiлуiнiн жогары жиiлiктерi пайда болатын аймактарды картага тYсiру
жэне мунай тeгiлуiн аныктау максатында оптикалык суретгердiн мYмкiндiктерiн багалау. Апшерон тYбегiнiн
манындагы Каспий тещзi ауданы зерттеу ауданы ретiнде тандалды. Мунайдын тeгiлулерi Каспий тенiзiнiн
баска аймактарына караганда осы eнiрде жш аныкталады, сондыктан оптикалык сенсорларды пайдалана
отырып салынган осы eнiрдегi «Мунайдын жш тeгiлу аймактары картасы» болашакта мунай тeгiлуiнiн
мониторингi Yшiн пайдалы болады. Sentinel-2 (A, B) жерсерштершщ оптикалык бейнелерi мунайдын тeгiлуiн
аныктау жэне олардын жш пайда болу аймактарын есептеу Yшiн пайдаланылды. ЖYктелген суреттер Ь1С-ден
L2A суретiн жасайтын Sen2Cor утилитасы аркылы алдын ала eнделдi. Пикселдердi бiр-бiрден жiктейтiн
пикселдер негiзiндегi жжтемеге караганда, объектiлi-баFытгалFан тэсш алдымен бейненiн пикселдерiн
бейненiн сегменттеуш пайдалана отырып, спектрлiк бiртектi объектшерге топтастырады, содан кейiн
жекелеген объектiлердi жiктейдi. Мунай тeгiлуi ретiнде жжтелген объектiлер кeзбен шолу аркылы
тексершдг 2019 жыл iшiнде мунай тeгiлуiнiн жоFары жиiлiктерi бар аймактардын картасы усынылды.
СалынFан карта бойынша 2019 жылы мунай тeгiлуiнiн ен жш кeздерi аныкталды, атап айтканда "Мунай
тастары", "Чилов" жэне "Пираллахи" кен орындары. Сондай-ак, макалада кейбiр тeгiлулердi кeрнекi
тYсiндiру кезiнде туындаFан киындыктардын мысалдары келтiрiлген. Нэтижесiнде, «мунайдын тeгiлуiн
мониторингiлеу Yшiн радиолокациялык деректермен катар оптикалык бейнелердi пайдалану да орынды», -
деген корытынды жасалды.
TYЙiн сездер: мунай тeгiлуiн аныктау, Каспий тенiзi, Sentinel-2, оптикалык суреттер.

Ж. Ш. Ж антаев, Д. В. Чепашев, А. А. Калдыбаев, С. М. Нуракынов, К. М. Унгаров

ДТОО «Институт ионосферы» АО «Национальный центр космических исследований

и технологий», Алматы, Казахстан

КАРТИРОВАНИЕ ЧАСТОТЫ РАЗЛИВОВ НЕФТИ В КАСПИЙСКОМ МОРЕ

С ИСПОЛЬЗОВАНИЕМ ОПТИЧЕСКИХ ИЗОБРАЖ ЕНИЙ SENTINEL-2

Аннотация. Экологическая ситуация на Каспийском море, которое буквально насыщено продуктами

нефтяной промышленности, является одной из острых проблем для прибрежных государств. Каспийское
море подвержено возрастающему антропогенному воздействию. С каждым годом увеличивается загрязнение
воды нефтепродуктами. В быстром темпе происходит сокращение морских биоресурсов. Международный
опыт показывает, что методы космического мониторинга являются наиболее эффективным способом
выявления и оценки динамики развития большинства природных и техногенных чрезвычайных ситуаций, в
частности разливов нефти. Современные системы дистанционного зондирования позволяют получать

188
ISSN 1991-346X Series physico-mathematical. 3. 2020

обзорную и подробную информацию о чрезвычайных ситуациях различных размеров на больших

территориях. В то же время методы космического мониторинга намного экономичнее традиционных
наземных методов сбора данных. В дистанционном зондировании используются два типа сенсоров -
активный (радарный) и пассивный (оптический). Оба типа сенсоров имеют свои преимущества и недостатки.
Благодаря многим преимуществам, радарные снимки в настоящее время являются наиболее
распространенным средством дистанционного зондирования разливов нефти. Однако, используя радарные
снимки, иногда бывает трудно отделить антропогенные разливы нефти и ложные объекты, похожие на них, и
обычно называемые «look-alikes», которые включают безветренные области, органические пленки
естественного происхождения, цветение водорослей и т. д. Используя же оптические изображения для
обнаружения разливов нефти, мы полагаемся на спектральные характеристики объектов, и хотя
отражательная способность варьируется в зависимости от толщины и состава пятна, в большинстве случаев
разливы нефти можно отличить от похожих ложных объектов. Основная цель данной работы - оценить
возможности оптических снимков в задаче обнаружения разливов нефти, опробовав их в задаче
картирования зон с высокой частотой появления разливов нефти в течение одного года. Район Каспийского
моря возле Апшеронского полуострова был выбран в качестве района исследований. Разливы нефти
обнаруживаются в этом регионе чаще, чем в других регионах Каспийского моря, поэтому карта зон с
частыми разливами нефти в этом регионе, построенная с использованием оптических сенсоров, будет
полезна для будущего мониторинга разливов нефти. Оптические изображения спутников Sentinel-2 (A, B)
использовались для обнаружения разливов нефти и расчета зон их наиболее частого появления. Скачанные
снимки были предварительно обработаны с помощью утилиты Sen2Cor, генерирующей изображения L2A из
L1C. Объектно-ориентированный подход был выбран для сегментации и классификации изображений. В
отличие от классификации на основе пикселей, которая классифицирует пиксели по одному, объектно­
ориентированный подход сначала группирует пиксели изображения в спектрально однородные объекты
изображения с использованием сегментации изображения, а затем классифицирует отдельные объекты.
Объекты, классифицированные как разливы нефти, проверялись путем визуальной интерпретации. Был
проведен анализ наложения и представлена карта зон с высокими частотами появления разливов нефти в
течение 2019 года. По построенной карте, были определены наиболее частые источники разливов нефти за
2019 год, а именно месторождения «Нефтяные камни», «Чилов» и «Пираллахи». Также в статье приведены
примеры сложностей, возникших при визуальной интерпретации некоторых разливов, и сделаны выводы о
целесообразности использования оптических изображений наряду с радиолокационными данными для
мониторинга разливов нефти.
Ключевые слова: обнаружение разливов нефти, Каспийское море, Sentinel-2, оптические снимки.

Information about authors:

Zhantayev Zhu. Sha., Institute of Ionosphere, JSC NCSRT, Almaty, Kazakhstan, Director, Doctor of Geological and
Mineralogical Sciences, Corresponding member of NAS RK; [email protected]; https://orcid.org/0000-0002-9189-9546;
Chepashev D. V., Institute of Ionosphere, JSC NCSRT, Almaty, Kazakhstan, Researcer, PhD student;
[email protected]; https://orcid.org/0000-0002-8417-3990;
Kaldybaev А. А., Institute of Ionosphere, JSC NCSRT, Almaty, Kazakhstan, Deputy Director, PhD;
[email protected], https://orcid.org/0000-0002-0563-282X;
Nurakynov S. M., Institute of Ionosphere, JSC NCSRT, Almaty, Kazakhstan, head of the Cartography and GIS Laboratory,
PhD student; [email protected]; https://orcid.org/0000-0001-9735-7820;
Ungarov K. M., “Institute of Ionosphere”, JSC “National Center of Space Research and Technology”, Almaty, Kazakhstan;
[email protected]; https://orcid.org/0000-0001-9692-1346

REFERENCES

[1] Lavrova O.U., Mityagyn M.I. Satellite monitoring of coastal anthropogenic pollution [Sputnikovyi monitoring
antropogennyh zagryaznenyi pribrezhnoi zony] // Zemlya i Vselennaya, (2008) №1. pp. 26-34. ISSN: 0044-3948 (in Russian).
[2] Nesterova M.P. Methods and means of combating oil pollution of the World Ocean waters. Problems of chemical
pollution of the World Ocean waters [Metody i sredstva borby s neftyanym zagryazneniem vod Mirovogo okeana. Problema
himicheskogo zagryazneniya vod Mirovogo okeana], Т. 8., L.: Gidrometeoizdat, (1989), p. 208 (in Russian).
[3] Izrael И.А., Ciban’ А ^ . Anthropogenic Ocean Ecology [Antropogennaya ekologiya okeana], Moscow: Flinta Nauka,
(2009). p. 529 ISBN: 978-5-9765-0879-8 (in Russian).
[4] [Mass death of fish occurs in the Caspian Sea. Available at: https://neftegaz.ru/news/incidental/323916-v-kaspiyskom-
more-proiskhodit-massovaya-gibel-ryby/ (accessed on 28 March 2020)]
[5] Stark J.S., Riddle M.J., Snape I, Scouller R.C. Human impacts in Antarctic marine soft sediment assemblages:
correlations between multivariate biological patterns and environmental variables at Casey Station. Estuarine, Coastal and Shelf
Science (2003);56:717. DOI: 10.1016/S0272-7714(02)00291-3
[6] Stark J.S, Riddle M.J, Smith S.D.A. Influence of an Antarctic waste dump on recruitment to nearshore marine soft-
sediment assemblages. Marine Ecology Progress Series (2004); 276:53. DOI: 10.3354/meps276053

189
News o f the National Academy o f sciences o f the Republic o f Kazakhstan

[7] Raymond T.C, Snape I. Using triage for environmental remediation in Antarctica: Triage for environmental remediation
in Antarctica. Restoration Ecology (2016); 25(1). DOI: 10.1111/rec.12383
[8] Northcott K.A, Snape I, Connor M.A, Stevens G.W. Water treatment design for site remediation at Casey Station,
Antarctica: site characterization and particle separation. Cold Regions Science and Technology (2003); 37:169. DOI:
10.1016/S0165-232X(03)00039-9
[9] Nedwed T, Coolbaugh T, Tidwell A. Subsea dispersant use during the Deepwater Horizon incident. AMOP (2012);506.
[10] [Oil Rocks field in Azerbaijan provides up to 70% of SOCAR’s oil gas extraction volume. Available at:
https://www.oilandgas360.com/oil-rocks-field-in-azerbaijan-provides-up-to-70-of-socars-oil-gas-extraction-volume/ (accessed on
03 April 2020)]
[11] Bayramov E., Buchroithner M., Bayramov R. Detection of oil pollution hotspots and leak sources through the
quantitative assessment of the persistence and temporal repetition of regular oil spills in the Caspian Sea using remote sensing and
Gi S. ISPRS Annals of Photogrammetry, Remote Sensing and Spatial Information Sciences (2015). II-3/W5. 451-457. DOI:
10.5194/isprsannals-II-3-W5-451-2015
[12] [Copernicus Open Access Hub. Available at: https://scihub.copernicus.eu/ (accessed on 25 March 2020)]
[13] [Sentinel-2 Toolbox. Available at: https://sentinel.esa.int/web/sentinel/missions/sentinel-2/ (accessed on 27 March
2020)]
[14] Main-Knorn M., Pflug B., Louis J., Debaecker V., Muller-Wilm U., Gascon F. Sen2Cor for Sentinel-2. In: Conference
on Image and Signal Processing for Remote Sensing, 23., (2017), Warsaw, Poland. Proceedings. Bellingham: SPIE. DOI:
10.1117/12.2278218
[15] Darwish A., Leukert K., Reinhardt W. Image segmentation for the purpose of object-based classification. In:
Geoscience and Remote Sensing Symposium, (2003). IGARSS'03. Proceedings. 2003 IEEE International. IEEE, pp. 2039-2041
DOI: 10.1109/IGARSS.2003.1294332
[16] Karathanassi, V., Topouzelis K., Pavlakis P., Rokos D. An object-oriented methodology to detect oil spills. Interna­
tional Journal of Remote Sensing 27, no. 23 (2006): 5235-5251. DOI: 10.1080/01431160600693575
[17] Smith T.G, Geraci J.R, St Aubin D.J. Reaction of bottlenose dolphins, Tursiops truncatus, to a controlled oil spill.
Canadian Journal of Fisheries and Aquatic Sciences (1983);40(9):1522e5. Do I: 10.1139/f83-175
[18] Fingas M., Brown C.E. A Review of Oil Spill Remote Sensing. Sensors, (2017); 18(1):91. DOI: 10.3390/s18010091

190