633
1 INTRODUCTION
The awareness of environmental protection,
especially on air pollution issue, has been rising
rapidly in the past few years. The International
Maritime Organization(IMO) consecutively released
series of Greenhouse Gas Research Report, which
indicated that ship exhaust gas has become an
important source of air pollution in certain regions
such as big port cities. GHGs (CO
2, CO, CH4 and
N
2O), as a dominant composition of ship emissions,
areassumedtoincreaseby50%‐250%ifno actionis
taken (Buhaug et al.,2009; Smith et al.,2014). Such
pollutants, together with other ingredients of ship
emissions such as SO
X, Particular Matter(PM),
Hydrocarbon(HC) and Diesel Particular
Matter(DPM),poseseriousthreattothequalityofthe
environment.
To investigate ship emissions and their influence
on the environment, many researches have been
conducted on estimating the quantity of emissions.
Eyring et al. (2005) estimated the amount of global
emissions of CO
2 and NOX from 1950 to 2001 using
fuel consumption and fleet numbers. Hulskotte and
Gon (2009) proposed a method to estimate the
emissions of ships at berth based on the actual fuel
consumption and the fuel quality. Schrooten et al.
Ship Emission Inventories in Estuary of the Yangtze
River Using Terrestrial AIS Data
X.Yao,J.Mou,P.Chen&X.Zhang
WuhanUniversityofTechnology,Wuhan,China
ABSTRACT:Estuaryformsatransitionzonebetweeninlandriverandopensea.InChina,theestuaryofthe
Yangtze Riverplays a vitalrole in connecting the inland andoversea shipping,and witnessesheavy vessel
trafficintherecentdecades.Nowadays,moreattentionshavebeendirectedtotheissueofshippollut
ionin
busy waterways. In order to investigate the ship emission inventory, this paper presents an Automatic
Identification System(AIS) based method. AIS data is the realistic data of vessel traffic including dynamic
information(position,speed,course,etc.)andstaticinformation(shiptype,dimensions,name,etc.).According
toshipdimensions,thepowerofenginesisesti
matedfordifferentshiptypes.ByusingAISbasedbottom‐up
approach,shipemissioninventoriesandsharesofairpollutantsandGHGs(Greenhousegases)aredeveloped.
Spatial distribution of ship emissions is illustrated in the form of heat map. As a case study, the emission
invent
oriesareanalyzedusingAISdataof2010intheestuary,andfollowingresultsaremade:(1)sharesofthe
emission are cruise ships 6.59%, bulk carriers 5.16%, container ships 52.96%, tankers 15.16%, fishing ships
9.16%,otherships10.97%;(2)CO2isthedominantpartoftheemission.(3)Areasofhighestemissionint
ensity
aregenerallyclusteredaroundtheSouthChannel,theNorthChannelandportsinthevicinity.Theproposed
methodispromising becauseitisderivedfromtheAISdatawhichcontainsnotonlyrealdataofindividual
shipbut also vesseltrafficsituation in thestudy area.It can serveras a reference for other researchers and
policyma
kersworkinginthisfield.
http://www.transnav.eu
the International Journal
on Marine Navigation
and Safety of Sea Transportation
Volume 10
Number 4
December 2016
DOI:10.12716/1001.10.04.13
634
(2009) developed the total emissions using
comprehensive maritime transport database of
activity data, specific energy consumption, emission
factors. These studies demonstrated useful methods
to quantify ship emissions and their composition.
However,thedataformodellingtheshipemissionin
certain area is very much limited. Therefore,
calculatingtheshipemission
inventoryischallenging.
Since2002newshipsandlateralllargersea‐going
vessels (>300 GT) and all passenger vessels are
requiredtocarryanAutomaticIdentificationSystem
(AIS)onboard.ThroughdedicatedVHFfrequencies,
AISinformationistransmittedbetweenvessels,from
vesselstoshore,orviceversa.Insimple
terms,AISis
a technology to make ships “visible” to each other.
AISdataistherealisticdataoftrafficdataincluding
dynamic information (position, speed, course, etc.)
andstaticinformation(shiptype,dimensions,name,
etc.). Such development provides profound data
foundation of individual ship activity information
and regional traffic situation
to study on ship
emissions.Therefore,severalresearchersappliedsuch
perspectiveontheirstudyasnewapproachtoobtain
shipemissioninventoriesaccurately.
The Ship Traffic Emission Assessment
Model(STEAM) was established by Jalkanen et al.
(2007)toestimatetheemissionsofSO
X,NOX,CO2in
the Baltic sea. Laurie and Brett (2014) established a
model to calculate ship engine exhaust emissions in
ports and extensive coastal waters using terrestrial
Automatic Identification System data for ship
movements and operating modes. Winther et al.
(2014) calculated a detailed Black Carbon(BC), NO
X
and SO
2 emissions inventories in the Arctic in 2012
and under two shipping scenarios (with or without
arctic diversion routes). Meanwhile, they forecasted
theemissionfortheyears2020,2030and2050.Since
ships emission inventory is closely linked with the
activity of every ship, and AIS data can explicitly
reflectindividual
shipactivityandmacroscopictraffic
situationofcertainregion,methodsbasedonAISdata
can estimate emissions more accurately and become
promising.
However, the relevant research and management
of ship emission are backward in China to some
extent.ThedrasticdevelopmentofChineseeconomy
has boosted the increment of ship number
and the
intensityofshiptrafficincongestedwaterareassuch
as estuaries, which causes serious environment
pollution due to ship emissions. The estuary of
Yangtze River plays a vital role in connecting the
inland and oversea shipping, and witnesses heavy
vesseltraffic.Suchproblemsareobviousinthisarea.
With improvement and implementation of ship
pollution prevention and control requirements,
government of China has been speeding up the
establishment of the ship Emission Control Areas
(ECA) all over the country. Three ECAs are
establishedinthePearlRiverDelta,theYangtzeRiver
Delta and the Bohai Rim (Beijing City, Tianjin
City
andHebeiProvince)respectively.TheYangtzeRiver
Delta has already started the implementation of
emission reduction measures from April 1, 2016.
Those port authorities encourage ships to use low
sulfur fuel (less than 0.5%m/m) when they are
mooring in the hub port areas along the Yangtze
River delta ECA.
To control the emission in ECAs,
shipemissioninventoryshouldbedeveloped.JINet
al.(2009)presented the emissionsinventory ofNO
X,
HC,COandPM10fortheTianjinportin2006based
fuel consumption data. FU et al. (2012) applied a
bottom‐up dynamic approach to calculate the ship
emissions for Shanghai port. YE et al. (2014)
presented the ship emissions inventories for
Guangdong province in 2010 using two different
methods
for emission factors, and analyzed the
temporalandspatialcharacteristicsofemissionsfrom
different ship type. However, it is very difficult to
build the bottom‐up approaches with satisfactory
accuracy for ship emission on the basis of fuel
consumptionorturnovervolumeofgoods.
In this context, this paper presents
a bottom‐up
AIS‐based method to calculate the ship emission
inventories in the estuary of Yangtze River. We
establish emission calculation models for different
ship types to present a detail CH
4, CO2, CO, DPM,
HC, N
2O, NOX, PM10, PM2.5 and SOX emission
inventories for ships in the estuary of the Yangtze
Riverin2010.Theemissioncontributionwhichcomes
from each ship type to the whole‐year emissions
inventories are reckoned. From emission spatial
allocation, the highest emission period of each ship
typeis identified,and theshipping routecomprised
the region
with the highest emissions can be
distinguished.
Hopefully, the proposed method can serve as a
reference for other researchers and policy makers
workinginthisfield.
2 METHODOLOGY
Emissions from the shipping industry will be
calculatedusinga bottom‐upAIS‐basedmethodand
AIS data to derive vessel activity.
The ship length
interval, sailing speed, sailing time and position are
abstracted from AIS data to calculate the ship
emissions. The ships within the study area will be
divided into several differenttypes. Before
developing the ship emissions inventory, the length
between perpendiculars (Lpp) should be calculated
usinga formula.Lpp
canbeusedasinputparameter
for calculation of the propulsion power. Then, a
method will be developed based on the emissions
calculation formula to enable the calculation of
emissionsusingenginepower,engineoperatingtime,
emissionsfactorsandloadfactor.Fig.1illustratesthe
frameworktocalculatetheshipsemissions.
2.1 Emissionsunderdifferentloadconditions
Theship emissions isclosely relatedtospeed ofthe
ship because the load factor of main engine is
different at different speed. Therefore, ship sailing
state is divided into three statuses: in port or at
anchor,maneuveringandcruising.
When ship speed is less
than 1knot, we believe
thattheshipisinportoratanchor,atthistime,the
auxiliaryengineisconsideredtobethesoleworking
engine (WEN et al., 2016). The installed auxiliary
enginepowercanbeusedtocalculatetheemissions
635
with load factor and emission factors according to
equation(1):

ij i j ij
E
PTLF EF
(1)
where E is the emission in kilogram, P is engine
power in kilowatt, T is working time of auxiliary
engineinhour,LFisloadfactorofauxiliaryenginein
dimentionless,EFisemissionfactorindimentionless,
istandsfortheshiptype,jstandsforthecategoryof
thepollutant.
However,theloadfactorofmainengineisrelated
to ship design maximum speed, which can be
calculatedas:
3
V（ /V）
js
LF
(2)
where
V
is the mean velocity for a considerable
period of time and
s
V
is the ship design maximum
velocity.Theunitofthemarem/s.
Whenashipismaneuvering,webelievethatboth
mainengineandauxiliaryengineareworking.Ifload
ofmainengineismorethan20%,theemissionfactor
of main engine can be regarded as constant. And if
the
loadislessthan20%,emissionfactorisnegatively
correlated with load of main engine, so revision is
needed,theformulaas:
0
EF EF AF
(3)
where
A
F isrevisionfactorwhenthemainengineis
underlowload.
Figure1.Frameworktocalculatetheshipsemissions
Whenashipisinmaneuvering, theshipemission
iscalculatedas:
,,,,,,

ij m i m m j m ij a i a a j a ij
EPTLF EF PTLFEF
(4)
where
m
P
stands for power of main engine,
a
P
standsforauxiliaryengine.
When a vessel keeps her speed for a
considerable long period of time, the state of this
ship can be regarded as cruising, and both main
engineandauxiliaryengineparticipateinthework,
sotheformula tocalculate shipemissions issame
asequation(4).
2.2 propulsivepower
Propulsivepowerofshipistheessentialparameterto
calculateshipemission.Mainenginepower(MP)and
auxiliary engine power (AP) for each type of ship
should be estimated before calculating the emission
because different type of ship with different engine
power and the data ofengine power
islimited. The
propulsive power of cruise ships, bulk carriers,
containerships,tankersfishingshipsandotherships
were calculated by using the methodology of
Kristensen (2012) and Kristensen and Lützen (2012).
The estimation model for propulsive power of
differentshiptypeisnotthesame.Heretheywillnot
be detailed because ofthe limited space. Taking the
process of calculating the propulsive power of
containershipsasanexample.Thelengthofwaterline
of a certain container ship should be obtained by
analyzing AIS data, then length between
perpendiculars(Lpp)canbecalculatedas:
wl
/1.01
Lpp L
(5)
whereL
wlisthelengthofthecontainership.
We could calculate the deadweight (DW) of this
container ship by using equation (6) and (7), and
equation (8) and (9) are used to calculate the
propulsivepowerofmainengine.

32
286.93
DW = 0.00591 3.44776
341.6925 10265
Lpp m
L
pp Lpp
Lpp

(6)

2
286.93m
3.66734 1383.89 175999
Lpp
DW Lpp Lpp

(7)



15 9
43
64000
5
2
8.446 10 1.0035 10
3.745 10 1.24 2503
DW t
MP DW DW
DW DW



(8)

6
2
64000t
3.092 10 1.11 14816
DW
MP DW DW

(9)
Finally, thepropulsive power of auxiliary engine
iscalculatedaccordingtofunctionsdevelopedbythe
IMOMEPC(IMO,2010)withassumingaseamarginof
85%:

10000
0.05 / 0.85
MP kW
AP MP (10)
636

10000
0.025 / 0.85 250
MP kW
AP MP
 (11)
The validation of power estimation was worked
outbyusingadatasetof10containershipsfromHIS
Fairplay databases to ensure the accuracy of the
emissioninventoryestimationtoa higherlevel.
Table1 shows that the absoluteerrors of Lpp are
less than 10%, and absolute errors
of propulsive
power of main engine are less than 15%, both of
which are reasonable. But the absolute errors of
auxiliary engine power are relatively large and the
fluctuation of which is intense, probably due to
collected data of the auxiliary engine power do not
matchmainenginepower.
2.3 Emission
factors
Emission factor is also essential for calculating the
ship emissions. In this ma nuscript, the emission
factors are classified according to engine type, fuel
type, and are referred from Shanghai emission
inventory(FUetal.,2012)andthereportofPortofLos
Angeles Inventory of air emission (Agrawal et
al.,2012).
Thoseemissionfactors,expresseding/kWh,
are wildly accepted by most of the scholars, see
Table2andTable3.
3 CASESTUDY
3.1 Studyareaanddatasource
TheestuaryoftheYangtzeRiverplaysavitalrolein
connecting the inland and oversea shipping and
witnessheavyvesseltraffic.
Withthedevelopmentof
inlandandopenseashippingofChina,trafficvolume
of inbound and outbound ship witnesses a rapid
incrementinthepastfewyears.Shanghaiport,whose
containerthroughputisthelargestamongtheworld,
hasacceleratesuchtrend.Therefore,airpollutiondue
toshipemissionsin
thisareaneeds moreattentionto
study. This paper chose a rectangular area with
coveragebetweenlatitudes 31°00′Nand31°50′
Nandlongitudes121°05′Eand122°40′E.South
channel, North channel and other main shipping
routesareincluded.
Asfordatasource,AISdataof
2010inthisareaare
applied. In 2010, the world expo was held in
Shanghai, China. Data quality of that year are very
good thanks to the comprehensive surveillance and
management on vessel traffic. Based on the whole
year’s AIS data, 14,845 cruise ships, 11,425 bulk
carriers, 174,314 container ships,
50,060 tankers,
17,260fishingshipsand38,971otherships(pilotboat,
tugboat,engineeringshipsandshipsdonotbelongto
thesefivespecifictypes),aretakenintostudy.
4 RESULT
4.1.1 Shipemissioninventory
Parameters of the model to calculate emission
inventoriesareobtainedbyanalyzingtheAISdataof
the
Yangtze River Estuary in 2010. According to
equation(1)‐(4),theemissioninventoryofstudyarea
in 2010 estimated (see Table 4). The quantity of
emissions ofsome main pollutants arePM
10 18,862t,
PM
2.5 15,408t, SOX 146,876t and NOX 244,578t
respectively. The largest emission of pollutant is
alwaysCO
2,whichisinconsistentwithothersimilar
study.
Table1.Powerestimationandverificationofcontainership
__________________________________________________________________________________________________
MMSI real estimated error
L
wl(m) Lpp(m) MP(kW)AP(kW)Lpp MP APLpp MP AP
__________________________________________________________________________________________________
212***000 220.23 210.26 35715 1320 218.0 313301171.5 3.7% 12.3% 11.3%
212***000 304 288.8 68520 3500 301.0 609732043.3 4.2% 11.0% 41.6%
304***000 147.87 140.3 13229 600 146.4 11707594.3 4.4% 11.5% 0.9%
538***590 161.3 149.6 12640 1720 159.7 14575678.7 6.8% 15.3% 60.5%
308***000 145.12 134 10860 940 143.7 11147
577.9 7.2% 2.6% 38.5%
352***000 168.05 158 16253 1080 166.4 16127724.3 5.3% 0.8% 32.9%
356***000 182 172.5 19670 1230 180.2 19642827.7 4.5% 0.1% 32.7%
356***000 144.73 135.63 10860 940 143.3 11068575.5 5.7% 1.9% 38.8%
412***430 121.2 112 7342 430 120.0 6493.8 382.0 7.1% 11.6% 11.2%
412***880 99.9
93.9 2991‐98.9 2681.6 157.7 5.3% 10.3%‐
__________________________________________________________________________________________________
Table2.Emissionfactorsoftwotypeofmainengine
_____________________________________________________________________________
engineemissionfactors
____________________________________________________________________
type CH4 CO2 CO DPM HC N2O NOX PM10 PM2.5 SOX
_____________________________________________________________________________
SSD
1
 0.012 620.00 1.40 1.50 0.60 0.031 17.00 1.50 1.20 10.50
MSD
2
0.010 683.00 1.10 1.50 0.50 0.031 13.00 1.50 1.20 11.50
_____________________________________________________________________________
1
SSDstandsforslow‐speeddieselengine,themaximumrotationspeedofwhichislessthan130r/min.
2
MSDstandsformedium‐speeddieselengine,therotationspeedofwhichismorethan130r/min.
Table3.Emissionfactorsofauxiliaryengine
_____________________________________________________________________________
emissionfactors
_____________________________________________________________________________
CH4 CO2 CO DPM HC N2O NOX PM10 PM2.5 SOX
0.005 683 0.21.50.40.031 13 1.51.212.3
_____________________________________________________________________________
637
Table4.Emissionsperpollutantofeachshiptypein2010
_______________________________________________
Cruise Bulk Container TankersFishing Other
ships carriers shipsships ships
(t)  (t) (t)(t)  (t) (t)
_______________________________________________
CH4 248 10832589 23811219936
CO
2 165,839 115,815 1180,348361,515 199,211 238,382
CO 1149 174011,583 414335,2583510
DPM2856981 12,502 258536043591
HC 411 593 4115 141812441244
N
2O 32 39 305 95 31 90
NO
X 12,49816,260125,53638,64313,91537,727
PM
101452188214,627 446722324362
PM
2.511651506117,048365519273500
SO
X 11,67314,664 117,15835,02015,50135,000
_______________________________________________
Table 4 indicates that the emission of each
pollutant in the Yangtze River Estuary is relatively
large,whichismuchlargerthanthatofShanghaiport
(FUetal.,2012).Forexample,theemissionofSO
Xof
Shanghai port in 2010 is 35400t, while that in the
Yangtze River is 229,015t, the difference is 193,615t.
The reason for this difference may be that the ship
trafficflowintheYangtzeRiverEstuaryislargerthan
that of the Shanghai port, and ships in the Yangtze
River
Estuary are always cruising, while ships in
Shanghai port waters are always maneuvering or at
anchor.
4.1.2 Emissionvalidation
Althoughthemethodologyofthisstudyiswidely
accepted around the world and the parameters fits
our study, it is essential to validate the emission
results.Theemissionofindividualshipcan
beeasily
measured by device while it is challenging to test
thousandsshipsintheestuaryarea.Inaddition,itis
hard to compare our emission result with other
researches due to quite different traffic volume,
researchareaandtimeperiod.Therefore,theEnergy
EfficiencyOperationalIndicator(EEOI)willbeused
to
validateouremissionresults.EEOIisdefinedasthe
ratio of mass of CO
2 emitted per unit of transport
work.Thebasicexpressionisshownasequation(12).
When the EEOI is similar to other researches, we
believethattheemissionestimationisreasonable.
arg
kFk
k
co
F
CC
EEOI
mD
(12)
where
k
is the fuel type,
k
F
C
is the mass of
consumed fuel in tons,
C
is the fuel mass to
2
CO
massconversionfactor,
argco
m
isthe massoftheship
intons,
D isthesailingdistanceinnauticalmiles.
Table5showstheEEOIinourstudycomparingto
other researches. The EEOI is calculated based on
numbers of vessels within study area, while
MEPC_684 and SUN et al. (2013) studied the single
vessel.Thecomparisonshowsthatthenumbersarein
same magnitude and because of some different
condition, small difference exists which can be
accepted.
Table5.ComparisonofEEOI
_______________________________________________
This MEPC_684 SecondIMO SUNet
study casestudy GHGStudy al.(2013)
general/bulk
_______________________________________________
EEOI10
‐6
12.6 13.5 14‐27/5.5‐46.3 78‐409
_______________________________________________
4.1.3 sharesofshipsemissions
Thetotalemissionsperpollutantinthestudyarea
isCH
48457t,CO22261,109t,CO57,384t,DPM26,118t,
HC9025t,N
2O592t,NOX244,579t,PM1029,022t,PM2.5
128,801t,andSO
X229,015t.Fig.1showsthesharesper
pollutant of each ship type. The emissions from
container ships is largest due to the number of
container ships accounts for the largest proportion
and the speed is relatively faster. Fishing ships
contribute the largest share of CO emission because
engineconditionof
fishingshipsisalwaysnotupto
standard, and the combustion of inferior fuel in the
engineisnotcomplete.
Large‐sized container ship is not only the
mainstream ship type of the current shipping
industry, but also the future development of ship
type,soitisreasonabletocontrolthe
dischargeofthe
containership.

Figure1.Sharesofeachshiptype
4.1.4 Patternsdistributionofshipemissions
The emission result is calculated based on AIS
data,andarepresentedastheformofheatmaps.Fig.
2 shows the spatial distribution of CH
4, CO2, CO,
DPM,HC,N
2O,NOX,PM10, PM2.5,SOXemissionson
January3,2010.Thereasonforshowingtheresultof
this day isthat thenumber of ships on January 3 is
same as the mean value and the proportion of all
kinds of ships is consistent. The distributions of
emissionare verysimilar. Areas of highestemission
intensity are generally clustered around the South
channel, the North channel and some main jetties,
followedbytheareanorthoftheHengshaislandfor
the shallow water depth. There are few small‐sized
fishingshipswouldsailinthewaterareanorthofthe
Hengshaisland.
638
a.CH4
b.CO2
c.CO
d.DPM
e.HC
f.N2O
g.NOX
h.PM10
639
i.PM2.5
j.SOX
Figure3.(a‐j):SpatialdistributionperpollutantonJanuary
3,2010
5 DISCUSSIONS
CalculatingshipemissionsusingaAISbasedmethod
can obtain detail estimation of emissions. Emission
inventories show that thepollution of theestuary is
veryseriousandtheamountofeachpollutantinthis
areaislargerthanthatofotherportsarea.Whilethe
shares of each pollutant
is consistent with other
researches.TheamountofCO
2islargestfollowedby
NO
X and SOX.The number of container ships is the
largest which contributed the largest amount of
pollution.Therefore,moreattentionshouldbepaidto
thecontrolandreductionofemissionsfromcontainer
ships. Areas of highest emission intensity are
generally clustered around the South Channel, the
NorthChannelandportsinthe
vicinity.
Itisdifficulttomakeapreciseinventoryforsome
reasons:
1 AIS base station cannot receive the definitely
accurateandcompleteAISinformationfromships
duetothetechnicalerror.
2 The static characteristics of a ship in AIS
information are set by the ship operators, thus
existing
someman‐madeerrors.
3 The typeof ship obtainedby analyzing AIS data
doesnotmatchtheshipitselfinsomecases.
4 In order to facilitate the calculation, in the
classification of ship type, only select the more
common fivetypes, the relativelyrare or a small
numberof
shipsareclassifiedasothertypes.
6 DISCUSSION
The drastic development of Chinese economy has
boosted the increment of ship number and the
intensityofshiptrafficincongestedwaterareassuch
as estuaries, which causes serious environment
pollution due to ship emissions. The estuary of the
Yangtze River plays
a vital role in connecting the
inland and oversea shipping and witnesses heavy
vesseltraffic.Suchproblemsareobviousinthisarea.
This paper established emission calculation
models for different ship types based on AIS data,
shipenginepowerfunctionsandtechnologystratified
emission factors to present a detail emission
inventory
forshipsintheestuaryoftheYangtzeRiver
in 2010. In this study, we calculated the propulsive
power through several functions. Then, ship
emissions in the study area in 2010 were estimated
andemissionsresults are showninthe formof heat
maps toanalyze the spatial characteristics. The
total
emission per pollutant in the study area between
latitudes 31°00′N and 31°50′N and longitudes
121 ° 05 ′ E and 122 ° 40 ′ E is CH
4 8457t, CO2
4054,004t, CO 57,384t, DPM 26,118t, HC 9025t, N
2O
592t, NO
X 244,579t, PM10 18,863t, PM2.5 15,408t and
SO
X146,877t.
After analyzing the AIS data broadcasted within
the study area, ships are divided into cruise ships,
bulk carriers, container ships, tankers, fishing ships
andotherships.In2010,thesharesoftotalemissions
arecruiseships6.59%,bulkcarriers5.16%,container
ships52.96%,tankers15.16%,fishingships9.16%and
other ships 10.97%. Obviously, container ships
accounts for the biggest proportion because the
numberofcontainershipsisthelargest.
Itshouldbenotedthattherearesomedeficiencies
in this study. We will solve them in a follow‐up
study. But thisstudy also can be a pilot research to
calculatetheshipemissioninbusywaterways.
ACKNOWLEDGEMENT
The work presented in this paper is financially
supported by the National Natural Science Fund of
China(GrantNo.51579201)andself‐determinedand
innovative research funds of Wuhan University of
Technology(GrantNo.2015‐zy‐109).
REFERENCES
AgrawalA,AldreteG,AndersonB,etal. 2012.PortofLos
Angeles Air Emissions Inventory. Starcrest consulting
group,LLC,LosAngeles,USA.
Buhaug Ø, Corbett J J, Endresen Ø, et al. 2009. Second
imoghg study 2009. International Maritime
Organization(IMO),London,UK,24.
DesignationofanemissioncontrolareaforNitrogen
Oxides
Sulphuroxidesandparticulatematter.
IMO, MARINE ENVIRONMENT PROTECTION
COMMITTEE,2010.InterpretationsofandAmendments
toMARPOLandrelatedinstruments61stsession.
FU Qing‐yan, SHENYin, ZHANG Jian. 2012,On the ship
pollutant emission inventory in Shanghai port. Journal
ofSafetyandEnvironment,12(2012);57‐64.
640
KristensenHO.DeterminationofRegressionFormulasfor
MainDimensionsofTankersandBulkCarriersBasedon
IHS Fairplay Data.Technical University of Denmark
(2012) Project no. 2010‐56,
Emissionsbeslutningsstøttesystem. Work Package 2,
ReportNo.02.
Kristensen H O. Statistical Analysis and Determination of
RegressionFormulasforMainDimensionsofContainers
Ships
BasedonIHSFairplayData.TechnicalUniversity
of Denmark (2012) Project no. 2010‐56,
Emissionsbeslutningsstøttesystem. Work Package 2,
Reportno.03.
Kristensen H O, Lützen M. Prediction of Resistance and
Propulsion Power of Ships. Technical University of
Denmark and University of Southern Denmark (2012)
Project no. 2010‐56, Emissionsbeslutningsstøttesystem.
Work
Package2,Reportno.04.
JalkanenJP,BrinkA,KalliJ,PetterssonH,KukkonenJ,Stip
T. A Modelling System for the Exhaust Emissions of
MarineTrafficanditsApplicationintheBalticSeaarea.
AtmosChemPhys,9(2009),pp.9209–9223.
JIN Tao‐sheng, YIN Xiao‐ge, XU Jia,
et al. 2009. Air
pollutantsemissioninventoryfromcommercialshipsof
Tianjin Harbor. MARINE ENVIRONMENTAL
SCIENCE,28(2009),623‐625.
Hulskotte J H J, Gon H A C, Denier van der. 2010. Fuel
consumption and associated emissions from seagoing
ships at berth derived from an on‐board survey.
AtmosphericEnvironment,44(2014),
pp.1229–1236.
Laurie Goldsworthy and Brett Goldsworthy. 2015.
Modellingofshipengineexhaustemissionsinportsand
extensivecoastalwatersbasedonterrestrialAISdata–
An Australian case study.Environmental Modelling&
Software,63(2015),pp.45‐60.
MortenWinther,JesperHChristensen,MarleneSPlejdrup,
Erik S Ravn, Ómar F
Eriksson, Hans Otto Kristensen.
Emission inventories for ships in the arctic based on
satellite sampled AIS data. Atmospheric
Environment,91(2014),pp.1‐14.
SmithTWP,JalkanenJP,AndersonBA,CorbettJJ,Faber
J, Hanayama S, OʹKeefe E, Parker S, Johansson L,
AldousL,RaucciC,
TrautM,EttingerS,NelissenD,Lee
D S, Ng S, Agrawal A, Winebrake J J, Hoen M,
Chesworth S, Pandey A, 2014. Third IMO GHG Study
2014.InternationalMaritimeOrganization,London,UK.
Schrooten Liesbeth, De Vlieger Ina, Panis Luc Int, et al.
2009. Emissions of maritime transport: A European
reference
system. Science of The Total Environment,
408(2009),pp.318‐323.
EyringV,KoehlerHW,AardenneJvan,LauerA.Emission
from international shipping: the last 50 years.JGR, 110
(D17305)(2005),pp.2171–2194.
WEN Yuan‐qiao, REN Jun‐yong, LI Zhao‐ran. 2016.
Research of Ship Optimal Speed with Minimum
EmissionsintheProcessofShipsEnteringandLeaving
Port. Journal of Wuhan University of Technology
(TransportationScience & Engineering), 40(2016), 2095‐
3844.
YESi‐qi,ZHENGJun‐yu,PANYue‐yun,etal.2014.Marine
emission inventory and its temporal and spatial
characteristics in Guangdong Province. Acta Scientiae
Circumstantiae,34(2014),
537‐547.