137

1 INTRODUCTION

Currently, the second generation intact stability

criteria are being discussed at the International

Maritime Organization (IMO) (IMO 2015, 2018). The

criteria include a new concept, “operational measures

(OMs)” (IMO 2019), for ensuring safety of ships at sea

during ship operations. The implementation of the

OMs has never been addressed in the history of intact

stability criteria. Therefore, the rationality and

practicality of the OMs in terms of implementation

during actual ship operation must be carefully

investigated. To facilitate discussion of the OMs,

Hashimoto et al. presented a set of pioneering case

studies on the OMs using a voyage simulation for an

ocean-going container ship for a variety of scenarios

(Hashimoto et al. 2017). This voyage simulation was

developed based on a weather routing model

(Kobayashi et al. 2011, Kobayashi et al. 2015).

Regarding the decision making for voyage routing

during an actual voyage, captains consider fuel

consumption, voyage distance, and safety factors,

such as rolling and pitching effects on cargos

(Koshimizu & Ishizuka 1994). A weather routing

service, which considers above points, is commonly

used in actual navigation (Fujii et al. 2017). This

means that it is reasonable to develop a simulation

tool for the investigation of OMs, which simulates

practical navigation routes correctly, based on a

weather routing model. However, captains decide a

voyage route not only with the reference to a route

recommended by weather routing service but also the

safety margin especially when the rough weather is

expected. Reliable voyage simulations need to

simulate ship navigation routes with sufficient

similarity to actual routes decided, by accounting for

the preferred safety margin based on the weather

conditions. Therefore, the route decision-making

criterion of captains taking into account of the safety

margin needs to be clarified.

Comparison of Master’s Route Selection Criteria of

ehicle Carriers in North Pacific and North Atlantic

sing Satellite AIS and Ocean Wave Data

M. Fujii

Marine Technical College, Japan agency of Maritime Education and Training for Seafarers, Ashiya, Japan

. Hashimoto

Kobe Ocean

-Bottom Exploration Center, Kobe University, Kobe, Japan

. Taniguchi

Graduate School of Maritime Sciences, Kobe University, Kobe, Japan

ABSTRACT: The operational measures in which a ship needs to avoid specified areas to escape ship stability

failures were discussed at the International Maritime Organization as a part of the second generation intact

stability criteria. It is necessary that the rationality and practicality of the operational measures are carefully

investigated. In this study, master’s route decision-making criteria of trans-ocean vehicle carriers have been

clarified by comparing the Pacific and the Atlantic data, derived from Satellite AIS and ocean wave data.

Features of voyage routes of each ocean were discussed, followed by analysis of the encountered wave

direction and height during a voyage. The master’s route selection criteria were defined by comparing the

probability densities of the wave heights that occurred in the navigable area and that of the actual encountered

waves. The navigation hours in a stormy area were also studied.

http://www.transn

av.eu

the

International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 14

Number 1

March 2020

DOI:

10.12716/1001.14.01.16

138

The wave criterion of navigators has been defined

by questionnaires and route planning experiments by

seafarers in the past (Hayashi & Ishida 2004).

However, it was derived from a limited amount of

data covering only the North Pacific and could be

significantly different from the universal criterion.

Another research discussed the relationship between

route selection and encountered wave height,

combined with satellite AIS data and ocean wave data

supplied by the National Centres for Environmental

Prediction (NCEP) (available at http://www.

cdc.noaa.gov/) (Fujii et al. 2019, Fujii et al. 2017). This

research demonstrated that the encountered wave

height during the voyage can be obtained as objective

data by combining the ship’s position obtained from

the satellite AIS data and the weather information at

the corresponding position and time. However, these

researches were focused on the container carrier or a

limited number of vehicle carriers only trans-Pacific.

In this study, the master’s route decision-making

criteria aimed at developing a reliable voyage

simulation for discussion of OMs has been clarified

by comparing the actual voyages of the trans-Pacific

and the Atlantic. Firstly, the features of voyage routes

of each ocean were discussed by using the ship’s

position from the Satellite AIS data. Secondly, the

encountered wave direction and height during a

voyage were analysed by data combined with satellite

AIS data and ocean wave data. Thirdly, the master’s

route selection criteria were obtained by comparing

the probability densities of wave heights that

occurred in the navigable area with that of the actual

encountered waves. Finally, the calculation result of

navigation time, which is duration hours between

entering and exiting a stormy area, was determined.

2 DETAILS OF THE ANALYSIS DATA

Currently, the automatic identification system (AIS)

equipment is required to be installed on all ships,

including the vehicle carriers over 300 GT, on

international voyages. The signal is continuously

transmitting during a voyage in the ocean. In this

study, the ship’s position data from the AIS, for a

large number of vehicle carriers, were collected and

analysed. The AIS data were purchased from

exactEarth (https://www.exactearth.com), and

collected by several satellites from December 2015 to

February 2016. Figure 1 shows the received position

of the AIS signal by the satellites in the purchased

data. Nowadays, with the improvement in the AIS

service, and increase in the number of satellites, the

quality of data is becoming better year by year. The

purchased data included a large number of received

AIS data, as shown in Figure 1.

Figure 1. Received position of AIS signal by the satellites

This study is focused on the master’s judgment in

a rough sea, especially the Pacific Ocean and the

Atlantic Ocean in winter. Therefore, the AIS data for

analysis was picked up from the purchased data

which contained the worldwide data. Analysis data

for the Pacific Ocean was picked up from the received

data between latitude 0 °N to 70 °N and longitude 100

°E to 100 °W, and for the Atlantic Ocean between

latitude 10 °N to 70 °N and longitude 0 °E to 80 °W.

For exclusion of error data in the AIS ship position,

the speed between neighbouring positions was

calculated from the position and received time. If the

speed was greater than the speed limit (24.7 kn), the

corresponding data were omitted as error data (Fujii

et al. 2017). In addition, voyages that were not

received within 24 hours were excluded from the

analysis due to poor reliability. Secondly, the trans-

ocean voyages were picked up from the data in the

designated area. The definitional lines were set on

both the eastern and western sides of the North

Pacific for the definition of the term “trans-Pacific”, as

shown in Figure 2. Similarly, the lines were set for the

North Atlantic in reference to the major routes (Vettor

& Soares 2015). Here, east-bound voyage means that a

ship departs from westerly of the west side

definitional line to easterly of the east side definitional

line, west-bond voyage is opposite.

Figure 2. Definitional line for trans-ocean in the (a) Pacific

and the (b) Atlantic

From these steps, the analysis voyages were

picked up, the details of which are shown in Table 1.

The number of ships was 174 in the Pacific Ocean and

124 in the Atlantic Ocean. The number of voyages was

198 in the Atlantic Ocean and 257(1.3 times more) in

the Pacific Ocean. However, this number is more

significant than the previous research (Fujii et al.

2017) and sufficient to analyse.

Table 1. Number of analysed ships and voyages

_______________________________________________

Number Number of Voyages

of ships East-bound West-bound Total

_______________________________________________

Pacific Ocean 174 151 106 257

Atlantic Ocean 124 108 90 198

_______________________________________________

The AIS data includes a ship’s position, the time of

transition, and more. However, the wave height at

that point is not included. Therefore, weather data has

to be imported from a different source while

analysing the height and direction of the encountered

wave. In this study, the oceanic data corresponding to

the ship’s position during the navigation was

obtained from the ocean wave data supplied by the

National Centers for Environmental Prediction

(NCEP). The encountered wave height and direction

are defined by combining the received ship’s position

data and the oceanic data. The ship’s position every 3

139

hours based on 00 UTC was used in this study, which

was obtained from the position information of the AIS

data. This time interval is the same as that of the

NCEP weather forecast data. The mesh of the weather

data has a longitudinal interval of 1.25° and a

latitudinal interval of 1.0°. To determine the weather

conditions at a ship’s position at a given time, a

simple linear interpolation was used. The accuracy of

analysis for the encountered wave and direction

depends on the reliability of the weather data.

3 COMPARISON OF ANALYSIS DATA WITH THE

PACIFIC OCEAN AND ATLANTIC OCEAN

3.1 Wave height in each ocean

The average of the top 10% wave heights were

calculated for the Pacific and Atlantic from December

2015 to February 2016, and are shown in Figure 3. The

Americas are depicted in the centre of the map. The

area with wave height above 7 m is widely

distributed in both the North Pacific Ocean and the

North Atlantic Ocean. The wave height in the centre

of the North Pacific Ocean was above 10 m. In the

Atlantic, the area with wave height above 8 m is

distributed in the Northern part. The past research

showed that the PCC, which means vehicle carrier in

this study, could navigate in area wave height above 4

m (Fujii et al. 2017). It means that both oceans were

stormy at the timing of analysis.

Figure 3. The top 10% wave heights between December 2015

to February 2016

3.2 Voyage Routes

Voyage routes in the Pacific Ocean are shown in

Figure 4. The east-bound voyages mainly used Great-

circle sailing route between latitude 30 °N to 40 °N. In

addition, the vessels tend to navigate around the

centre of the North Pacific Ocean. On the other hand,

the west-bound voyages seem to have two major

routes: one through the Bering Sea and the other

around the south of latitude 30 °N. Most west-bound

voyages used Mercator sailing, while Great-circle

sailing was only used in the case of the eastern

departure point in latitude 50 °N or more.

In the case of the Atlantic, it is shown that the east-

bound voyages navigated almost on the Great-circle

route, not deviating too far from the shortest route.

However, the west-bound voyages show that routes

in the southern part of the Great-circle, were taken

especially in the area latitude 45 °N and longitude 30

°W. This area had higher wave height than the other

areas, as shown in Figure 3. Hence, it seems that ships

avoided the stormy sea. However, the routes of the

east-bound and west-bound vessels are similar to

each other, unlike the Pacific cases.

Figure 4. Voyage routes of vehicle carrier in the Pacific

Ocean

Figure 5. Voyage routes of vehicle carrier in the Atlantic

Ocean

3.3 Direction and height of the encountered waves

The trends of east-bound and west-bound trans-

Pacific voyages are different. However, the trans-

Atlantic routes of both east-bound and west-bound

vessels are similar to each other, as mentioned before.

The cause of the route trend difference from wave

direction and height point of view, is discussed in this

section.

Figure 6 shows the probability density of the

directions of encountered waves for the east-bound

and west-bound voyages. Here, 0° on the horizontal

axis represents head waves, 180° represents following

waves, a positive angle means that the wave is

incoming from the starboard side, and a negative

angle means the opposite. The figures show that the

east-bound voyages received following waves and the

west-bound received head waves from the starboard.

This trend is quite similar to the Pacific voyages and

Atlantic voyages.

Figure 7 shows the histograms of the encountered

wave heights in the Pacific and the Atlantic which is

not categorised in voyage direction. The vertical axis

of these figures represents the probability density,

and the horizontal axis represents the encountered

wave height. The number of data (n) in each ocean is

over 10,000. It means that the value is enough for the

analysis. Here, the cause of difference of the number

between the Pacific and the Atlantic is expected

voyage time, in addition to the differences of a

number of navigated vessels. Because the voyage

distance between the western and the eastern in the

Pacific is much farther than the Atlantic one. The

mean encountered wave height in the Pacific is lower

than the Atlantic; the mean height is 3.17 m in the

Pacific and 3.45 m in the Atlantic. In the Pacific, a

peak of the encountered wave height is between 2.0 m

140

to 3.5 m. However, the Atlantic one is distributed

more widely between 2.0 m to 4.0 m. Besides, a

changing rate of the histogram is gentler in the

Atlantic.

Figure 6. Probability density of encountered wave direction

corresponding to the (a) Pacific east-bound, (b) Pacific west-

bound, (c) Atlantic east-bound, and (d) Atlantic west-bound

voyages

Figure 7. Probability density of encountered wave heights

corresponding to the (a) Pacific, and (b) Atlantic voyages

Figure 8. Assumed navigable area in the (a) Pacific and (b)

Atlantic

To compare the occurred wave height in each

ocean, the wave height in the assumed navigable area

where was shown in figure 8 was calculated. It was

set based on the AIS position data, and the area

between 10 °N to 57 °N and 138 °E to 120 °W in the

North Pacific, and 18 °N to 56 °N and 70 °W to 19 °W

in the Atlantic. As a result, the average wave height of

the area in the Pacific was found to be higher than in

the Atlantic; 3.42 m in the Pacific and 3.29 m in the

Atlantic. However, the encountered wave height in

the Pacific is lower than the Atlantic. Therefore,

opposite results were achieved in occurred wave

height and encountered wave height. As a result, it

can be said that the trans-Pacific voyage can be

selected avoiding a route through heavy weather

area, while a trans-Atlantic voyage cannot choose a

route through a non-stormy area.

4 LIMITATION WAVE HEIGHT AND

NAVIGATION TIME IN THE ROUGH SEA

Table 2 is indicating the average wave height

categorised by the head wave and the following

wave. In both the Pacific and the Atlantic, the head

wave is lower than the following wave. It means that

the Captain’s judgment for selection of route have

different standards between the case of the head wave

and the following wave.

Table 2. Average wave height categorised by encountered

direction

_______________________________________________

Head Wave (m) Following Wave (m)

Mean SD Mean SD

_______________________________________________

Pacific 3.00 1.01 3.28 1.14

Atlantic 3.35 1.18 3.53 1.28

_______________________________________________

Figure 9. Comparison of probability densities of waves

observed in navigable areas and encountered waves in the

Pacific Ocean

Figure 10. Comparison of probability densities of waves

observed in navigable areas and encountered waves in the

Atlantic Ocean

The probability densities of wave heights that

occurred in the navigable area and actual encountered

waves are shown in Figure 9 for the Pacific and in

Figure 10 for the Atlantic. If Captains did not avoid

the stormy area intentionally, the two histograms

would be the same. However, the two histograms do

not match. From this fact, it is expected that captains

avoid rough sea area if the wave height is beyond the

141

threshold. It can be said that the threshold would be

the wave height where the two histograms were

reversed. Therefore, the thresholds are 4.5 m in head

wave and 5.0 m in following wave in the Pacific, as

shown in Figure 9. Similarly, 5.5 m in head wave and

6.0 m in following wave in the Atlantic as shown in

Figure 10.

Figure 9 and Figure 10 show probability densities

indicated even when the wave height is beyond the

threshold. It means that a few ships still navigate on

the area beyond the threshold wave height. It is

considered that navigation time in stormy weather

affects the Captain’s judgment. Hence, the calculation

results of navigation time in a stormy area beyond the

threshold are shown in Table 3. This result has been

measured in the time between entering and exiting

the area. In this study, the ship’s position is aligned

with the weather data which means every 3 hours. So,

the calculation time is also 3 hours unit. The

navigation time during a stormy area in the Pacific is

approximately twice as much as the Atlantic one.

From the results, the threshold in the Atlantic is

higher than the Pacific, however, it can be said that

once entered in the heavy weather area, the voyages

in the Pacific need to stay for a long time.

Table 3. Navigation time in area beyond the threshold

_______________________________________________

Mean Median

_______________________________________________

Pacific Ocean 34.6 hours 30 hours

Atlantic Ocean 18.4 hours 15 hours

_______________________________________________

5 CONCLUSION

This study compared the master’s route selection

pattern, especially encountered wave direction and

wave height for vehicle carriers derived by Satellite

AIS and ocean wave data. The summary of this study

is as follows:

− Comparing the voyage routes in the Pacific and

the Atlantic derived by the Satellite AIS position

data of vehicle carriers, the trans-Pacific east-

bound routes were in the middle of the North

Pacific Ocean and west-bound routes were divided

to two major routes to avoid the middle. However,

in the Atlantic cases, east-bound and west-bound

routes are not different from the Pacific cases.

− The average wave height of the area in the Pacific

is higher than the Atlantic; 3.42 m in the Pacific

and 3.29 m in the Atlantic. However, the ship’s

encountered wave height in the Pacific is lower

than the Atlantic. Therefore, it can be said that the

trans-Pacific voyage can be selected avoiding route

to the heavy weather area while a ship under

trans-Atlantic voyage could not choose a way to a

non-stormy area in comparison to the Pacific

voyage.

− Based on the result of the comparison of the

probability densities of wave heights that occurred

in the navigable area and encountered waves, it is

expected that Captains avoid rough sea area. The

wave height in the area was 4.5 m in the head

wave and 5.0 m in the following wave, in the

Pacific; and 5.5 m in in head wave and 6.0 m in

following wave, in the Atlantic.

− The calculation result of navigation time, which is

the duration (hours) between entering and exiting

a stormy area beyond the threshold, shows that the

trans-Pacific voyages need to stay for a long time

than the trans-Atlantic, if a ship enters a heavy

weather area once. The navigation time is

approximately twice the Atlantic one. The median

time of the Pacific is 30 hours.

In the past research, the analysis is focused on the

container ships or analysis with a few data. These

results can supplement the previous research.

Additionally, the figures derived from satellite AIS

data and ocean wave data will help when the master’s

route selection pattern is put into the voyage

simulation.

ACKNOWLEDGEMENTS

This work was supported by the research activity of the

Goal-Based Stability Criterion Project of Japan Ship

Technology Research Association in the fiscal year of 2018,

funded by the Nippon Foundation. This work was also

supported by JSPS KAKENHI Grant number 17H03493.

REFERENCES

International Maritime Organization. 2015. Finalization of

second-generation intact stability criteria. SDC 3/6/7

International Maritime Organization. 2018. Draft report to

the maritime safety committee. SDC 5/WP.1

International Maritime Organization. 2019. Report of the

Experts' Group on Intact Stability. SDC 6/WP.6 Annex2

Hashimoto, H., Taniguchi, Y., Fujii, M. 2017. A case study

on operational limitations by means of navigation

simulation. In: Proc. of the 16th international ship

stability workshop, Belgrade, pp 41–48.

Fujii, M., Hashimoto, H., Taniguchi, Y. 2017. Analysis of

satellite AIS data to derive weather judging criteria for

voyage route selection. TransNav Intern J Mar Navig Saf

Sea Transp 11(2):271–277.

Fujii, M., Hashimoto, H., Taniguchi, Y. 2019. Statistical

validation of a voyage simulation model for ocean-going

ships using satellite AIS data. J Mar Sci Technol:

https://doi.org/10.1007/s00773-019-00626-3.

Hayashi, M., Ishida, H, 2004. Weather routing simulation of

ocean-going ship by practical navigators and

encountered wind and wave conditions on simulated

ship’s routes. J Jpn Inst Navig 110:27–35 (in Japanese).

Kobayashi, E., Asajima, T., Sueyoshi, N. 2011. Advanced

navigation route optimization for an oceangoing vessel.

TransNav Intern J Mar Navig Saf Sea Transp 5(3):377–383.

Kobayashi, E., Hashimoto, H., Taniguchi, Y., Yoneda, S.

2015. Advanced optimized weather routing for an

ocean-going vessel. In: Proceedings of the 2015

international association of institutes of navigation

world congress, Prague, pp 1–8.

Koshimizu, Y., Ishizuka, M. 1994. Weight of considerable

factors of course selecting using the A.H.P. method. J Jpn

Inst Navig 90:307–319 (in Japanese).

Vettor, R., Soares, C.G. 2015. Detection and analysis of the

main routes of voluntary observing ships in the North

Atlantic. J Navig 68(2):397–410.