Analysis of the 21/22 October 2014 Storm Experienced by the

Sailboat ECO40 in the Gulf of Lion

P. De Girolamo

, A. Romano

, G. Bellotti

, A. Pezzoli

, A. Boscolo

, M. Crespi

, A. Mazzoni

M. Di Risio

, D. Pasquali

, L. Franco

and P. Sammarco

Sapienza Università di Roma, DICEA, Via Eudossiana, 18, 00184, Roma, Italy

Università di Roma Tre, Dip. di Ingegneria, Via Vito Volterra, 62, 00146, Roma, Italy

Politecnico di Torino & Università di Torino, DIST, L.R. MeteoSport, Viale Mattioli 39, 10125, Torino, Italy

Università dell’Aquila, DICEAA, Laboratorio di Idraulica Ambientale e Marittima LIAM,

P.le Pontieri, 1, 67040 Monteluco di Roio, L’Aquila, Italy

Università di Roma Tor Vergata, DICII, Via del Politecnico, 1, 00133, Roma, Italy

Keywords: Sailboat, Waves, Wind, Forecast.

Abstract: On October 19, 2014 Matteo Miceli, a known Italian oceanic sailor, left the Port of Riva di Traiano (Rome,

IT) with the sailboat ECO40, an Italian vessel, for the Roma Ocean World Project. This ambitious challenge

consists in a non-stop sailing alone around the World in energy and food self-sufficiency. ECO40 is a Class

40 oceanic vessel (LOA of 12,0 m) that has been equipped with a data acquisition system for both the met-

ocean parameters recorded on-board (i.e. apparent and real wind speed and wind direction, atmospheric

pressure, current velocity, air and sea temperature, etc.) and the kinematic characteristics of the boat itself

(i.e. speed and course over ground). Furthermore, the boat has been equipped with a three high precision

GPS receivers, provided by Leica Geosystem, for measuring the movements of the boat and with an inertial

platform. Due to these high-precision instruments it has been possible to fully measure and characterize the

six degrees of freedom of the boat, and accordingly to use the boat as a “sailing wave buoy”. In this paper

we present the first analysis of the met-ocean data measured by the boat during the storm occurred in the

Gulf of Lion on October 21-22, 2014 that ECO40 faced just few days after its departure.

1 INTRODUCTION AND AIM OF

THE RESEARCH

On October 19 , 2014 Matteo Miceli, a famous

Italian oceanic sailor, left the Port of Riva di Traiano

located close to Rome (Italy) with the Italian

sailboat ECO40 for the Roma Ocean World Project.

This ambitious challenge consists in a non-stop

sailing alone around the World in energy and food

self-sufficiency. The planned route was the classic

clipper route which runs from west to east through

the Southern Ocean, in order to make use of the

strong westerly winds. Namely the route, very

similar to that of several prominent yacht races as

Around Alone and Vendèe Globe, consists in

passing the Gibraltar Strait, then in descending the

Atlantic Ocean and sailing around the Antarctic, at a

mean latitude of 50

S, from west to east rounding

the most famous capes of the world: Cape of Good

Hope, Cape Leeuwin and Cape Horn. Finally,

ascending the Atlantic Ocean and passing again the

Strait of Gibraltar coming back to the homeport. The

total distance to be covered by the sailboat was

estimated in about 28,000 nautical miles, while the

duration was estimated in about five months.

When Matteo was on the way back to Italy, after

rounding the three capes and sailing for 25.000

nautical miles, ECO40 capsized at the equator.

Matteo was about 600 miles offshore the Brazilian

coasts. Matteo was saved by a cargo. When he came

back to Italy, he organized a first expedition with

four friends to try to recover ECO40, which was not

successful. After one month Matteo and his friends

tried again and found ECO40 300 miles offshore the

Brazilian coasts. Now ECO40 is again in Italy.

ECO40 is a Class 40 oceanic vessel (LOA of 12,0

m) that has been equipped with a data acquisition

system for both the met-ocean parameters recorded

290

Girolamo, P., Romano, A., Bellotti, G., Pezzoli, A., Boscolo, A., Crespi, M., Mazzoni, A., Risio, M., Pasquali, D., Franco, L. and Sammarco, P..

Analysis of the 21/22 October 2014 Storm Experienced by the Sailboat ECO40 in the Gulf of Lion.

In Proceedings of the 3rd International Congress on Sport Sciences Research and Technology Support (icSPORTS 2015), pages 290-298

ISBN: 978-989-758-159-5

on-board (i.e. apparent and real wind speed and

wind direction, atmospheric pressure, current

velocity, air and sea temperature, etc.) and the

kinematic characteristics of the boat itself (i.e. speed

and course over ground). Furthermore, the boat has

been equipped with a three high precision GPS

receivers, provided by Leica Geosystem, for

measuring the movements of the boat and with an

inertial platform. Thanks to these instruments it is

possible to fully measure and characterize the six

degrees of freedom of the boat.

In order to reduce the significant cost of the data

transfer just a small part of the measured data were

sent to the land team on daily basis; it is worth to

note that the transfer of the data is achieved by using

a satellite modem. The sampling frequency for the

met-ocean data acquisition is 2 Hz, nevertheless

only the data averaged over a time window which

had a duration of 10 minutes were sent to shore. It is

important to stress that these data, measured by the

boat and transmitted almost in real time, helped

significantly the team in charge of the safety of

ECO40; in fact the knowledge of the actual weather

conditions that the boat is really facing during its

navigation can improve the route strategy and

increase the boat safety. Furthermore, the boat

performance data (i.e. speed and course over

ground) can allow, after a certain amount of time

that is required for considering the database

statistically meaningful, to estimate the real polar

velocity curves of the boat. Indeed these curves were

used for the prediction of the optimal route (routage)

made by the land team by means of a route

optimization software, which was sent daily to

ECO40.

It is worth noticing that the actual polar velocity

curves of the boat differ from the theoretical ones

estimated by the designer of the boat. This is due to

several causes, among which plays an important role

the ability of the crew to "push" the boat to the

maximum of its performance and the presence of

waves that normally is not taken into account when

calculating the curves themselves.

The remaining data were supposed to be analyzed

once the boat had come back. Fortunately the data

were saved together with the boat.

The measurements of the boat movements, obtained

from the three GPS receivers, if properly analyzed

can provide the characteristics of waves that ECO40

encountered during its navigation. In other words it

can be stated that ECO40 can be used as a “moving

wave buoy” for measuring the waves characteristics

(i.e. significant wave height H

, peak period T

and

mean direction θ) experienced during navigation.

These data are used mainly for two technical and

scientific purposes. The first purpose is the

calibration of the numerical models output that are

commonly used for the wind and wave forecast

and/or analysis into the oceans, and the calibration

of the remote sensing data (e.g. satellite wind and

wave measurements). The second purpose is directly

related to the vessel design. Indeed the knowledge of

both the movements and the loads that these kind of

vessel can deal with, together with the response of

the materials to the fatigue stresses, can improve

significantly the designing methods.

The aim of this paper consists in analyzing the first

storm experienced by ECO40 during its navigation

that occurred in the Gulf of Lion on October 21-22,

2014. In the following sections, first we present the

analysis of the Gulf of Lion weather event, and the

description of the available wind and waves data.

Then we show the results of the comparison between

the wind measurements carried out on board and

those reconstructed in forecasting and analysis, by

using the numerical data of the GFS (Global

Forecast System) and ECMWF (European Center

for Medium-range Weather Forecast). Finally, we

show a comparison of the present data with those

related to the storm that occurred in the Gulf of Lion

on 2 November 1995, which caused the sinking of

the Italian racing boat Parsifal, killing six of its nine

crew members. The weather conditions that caused

the sinking of the boat have been rebuilt by (Bertotti

et al., 1988).

2 THE ECO40’S ROUTE DURING

THE GULF OF LION EVENT

On 19 October 2014, when ECO40 left from the

Italian Port Riva di Traiano, the weather conditions

appeared to be clear: within the next 24/48 hours the

first seasonal front of cold air, expected to cause

mistral winds having wind speed exceeding 40

knots, would come from the Gulf of Lion.

Fortunately, ECO40 was able to reach the Asinara

Island and began to follow the route towards the

Balearic Islands before the arrival of the main storm:

the boat has faced the storm running on the quarter.

The route between the Asinara Island and the

Balearic Islands that the boat ECO40 has followed is

represented in Figure 1. The figure shows also

information of the travel times.

Eco 40 has covered more or less 200 nautical miles

in 24 hours, with a mean speed of almost 8.3 knots.

Figure 2 shows the plots of the speed over ground

(SOG) and of the course over ground (COG) during

Analysis of the 21/22 October 2014 Storm Experienced by the Sailboat ECO40 in the Gulf of Lion

291

the Gulf of Lion Event. We recall that these values

are not the instantaneous ones but are the values

averaged over a time interval of 10 minutes.

Figure 1: The route between the Asinara Island and the

Balearic Islands that the boat ECO40 has followed.

Figure 2: Time series of the speed over ground (upper

panel) and of the course over ground (lower panel) during

the Gulf of Lion Event.

3 WEATHER ANALYSIS OF THE

GULF OF LION STORM

In the days 20 and 21 October 2014, the atmospheric

circulation was characterized by a zonal flux from

West to East in the Northern of Europe. A high

pressure was centred on the Mediterranean Sea

(Figure 3 and Figure 4).

Figure 3: Surface Analysis for 20 October 2014 hour

00.00UTC (source: MetOffice).

Figure 4: Surface Analysis for 21 October 2014 hour

00.00UTC (source: MetOffice).

Figure 5: Upper Atmosphere Analysis at 500hPa for 21

October 2014 hour 12.00UTC (source: MetOffice). Color

filled area: temperature at 500hPa [°C]. Red solid line:

height level at 500hPa [mx10].

WPPDSports 2015 - Special Session/Symposium on Weather, Position and Performance data in Outdoor Sports

292

At the same time a frontal system, positioned

between a low pressure centred on the Atlantic

Ocean and a low pressure centred in the Poland’s

plan, started to move slowly toward South. The

movement of this frontal system was generated by

the faster change of direction of the zonal flux in the

UK island, in turn generated by the movement

towards East of the tropical low pressure “Gonzalo”.

The trough in upper atmosphere (see Figure 5)

associated with the cold air movement on the day of

October 21 and in the night of the October 22 (see

Figure 6), generated the movement to South of the

cold front (see Figure 7).

Figure 6: Upper Atmosphere Analysis at 500hPa for 22

October 2014 hour 00.00UTC (source: MetOffice). Color

filled area: temperature at 500hPa [°C]. Red solid line:

height level at 500hPa [mx10].

Figure 7: Surface Analysis for 22 October 2014 hour

00.00UTC (source: MetOffice).

Figure 8: Upper Atmosphere Analysis at 500hPa for 22

October 2014 hour 12.00UTC (source: MetOffice). Color

filled area: temperature at 500hPa [°C]. Red solid line:

height level at 500hPa [mx10].

Figure 9: Surface Analysis for 23 October 2014 hour

00.00UTC (source: MetOffice).

When the cold front encountered the Alps, it

developed the classical low pressure down-wind to

the mountain. Then a low pressure with the center at

998hPa was positioned on the Venice Gulf. The

opposition between this low pressure and the high

pressure with center at 1033hPa positioned on the

Biscay Gulf, jointly with the movement of the cold

air in upper atmosphere (see Figure 8), generated an

atmospheric situation characterized by an elevated

instability. This weather pattern was active on the

Mediterranean Sea for the entire day on October 22,

2014 and for a part of the October 23, 2014 (see

Figure 9)

This particular baric configuration generated

strong wind from N on the Mediterranean Sea in

Analysis of the 21/22 October 2014 Storm Experienced by the Sailboat ECO40 in the Gulf of Lion

293

particular on the Gulf of Lion, on the West of

Mediterranean Sea, on the Sicily Channel and on the

Sardinia Channel. Moreover, the slow movement of

the cold front and of the low pressure made an

adverse meteorological situation for a time as long

as 48h.

4 WIND DATA ANALYSIS

In this section the wind data measured on board

during the the Gulf of Lion event are presented and

analyzed.

The true wind speed V

and the true wind

direction θ

during the Gulf of Lion event, obtained

from the apparent wind and from the boat kinematic

characteristics, are shown in Figure 10. The upper

panel of Figure 10 shows the time series of the mean

values (i.e. averaged over 10 minutes) of the true

wind speed (black line) and the true wind gust (red

line) that represents the highest values of true wind

speed measured over 10 minutes. It is worth to note

that the maximum value of the averaged wind speed

reached 45 knots at the peak of the event, while the

values of the gusts were greater than 50 knots,

reaching a maximum value of 56 knots. The lower

panel of Figure 10 shows the time series of the true

wind directions; also in this plot both the quantities

that refer to the averaged values and the ones that

refer to the gust values are represented.

The upper panel of Figure 11 represents the

scatter plot between the averaged true wind speed

and the true wind gust. The plot shows that the wind

gust values are generally larger of about 15-20%

Figure 10: Upper panel: time series of the true wind speed

averaged over 10 minutes (black line) and of the true wind

gust (red line). Lower panel: time series of the true wind

direction averaged over 10 minutes (black line) and of the

true wind gust direction (red line); note: the dotted black

line refers to the mean direction of the true wind (θ

w-m

during the event is 313.6°).

Figure 11: Upper panel: scatter plot between the averaged

true wind speed and the true wind gust; the red line

represents the functional relationship, as obtained by

applying a linear regression to the measured data. Lower

panel: scatter plot between the averaged true wind

direction and the true wind gust direction.

than the averaged true wind speed; also, the

difference between the averaged wind speed and the

gusts tends to increase as the wind speed increases.

In order to highlight this feature, the functional

relationship, as obtained by applying a linear

regression to the measured data, has been

represented (red line) in the upper panel of Figure

11.

These wind data, measured by ECO40 during the

Gulf of Lion event, have been compared with two

different sources of data. The first comparison has

been carried out with the ECMWF (European Centre

for Medium-Range Weather Forecasts) data. Data

are provided with a spatial resolution of 0.125° and

with a time resolution of 6 hours.

A second comparison has been carried out by

using the forecast data provided by the numerical

model GFS (Global Forecast System). It has to be

stressed that the output of this model, that has a

WPPDSports 2015 - Special Session/Symposium on Weather, Position and Performance data in Outdoor Sports

294

spatial resolution of 1.0° and a time resolution of 3

hours, are the ones which have been used for the

routing of ECO40.

The results of these comparisons are shown in

Figure 12. The upper panel of Figure 12 represents

the time series of (i) the averaged wind speed (black

line) measured by ECO40 by the anemometer placed

on the top of the mast (as shown in Figure 10), and

(ii) the time series of the wind speed obtained in

analysis and by the GFS, linearly interpolated (in

time and space) along the route of the boat between

Sardinia Island and the Balearic Islands. The results

of the GFS model (i.e. forecast data) are identified

by a red dashed line, while the results of the

ECMWF model (i.e. analysis data) are identified by

a continuous blue line. Now, the forecast data have

been sent to the boat on October 10, 2014; therefore

we can evince that they have a higher accuracy in

the first 24-48 hours; obviously, as shown in Figure

12 as the time increases (over 24-48) the agreement

between the forecasted data and the measured ones

deteriorate.

The upper panel of Figure 12 shows that the

wind speed values measured on board are larger than

those estimated by the numerical models. Indeed,

this comparison is not completely correct given that

the results of the numerical models are provided at

the conventional height of +10 m above the mean

sea level, while the wind data measured on board

have been registered at the actual position of the

anemometer that is placed on the top of the mast:

when the boat is at its rest position (i.e. the boat is

not heeled), then the height of the anemometer is

about +19 m above the mean sea level (i.e. the

height of the top of the mast). Thus, in order to

properly compare the measured data with the

numerical results, it is therefore necessary to reduce

the measured wind data to the height of +10 m

above the mean sea level. This has been carried out

by using the vertical profile of the wind speed

proposed by (Pierson, 1955). This relationship

describes a logarithmic vertical profile of the wind

velocity, that has the following functional form:

(1)

where U

is the wind speed at the height z, U

represents the wind speed at +10 m above the mean

sea level, C

is a coefficient that is function of the

flow regime of the wind speed and of the surface

roughness that theoretically depends on the waves

conditions. In order to estimate C

the functional

form proposed by (Wu, 1969) has been used.

Figure 12: Upper panel: comparison between the true wind

speed (averaged over 10 minutes) measured at the top of

the mast (black line) and the numerical model output of

the GFS (dashed red line) and of the ECMWF data in

analysis (continuous blue line). Middle and lower panel:

comparison between the true wind speed (averaged over

10 minutes), evaluated at the height z = +10 m above the

mean sea level (black line) l.m.m as a function of two

heeling angles (α = 35°, middle panel; α = 45°, lower

panel), ) and the numerical model output of the GFS

(dashed red line) and of the ECMWF (continuous blue

line).

As shown in the following relationship the

coefficient C

depends just on the wind velocity,

while it does not take into account the wave field

Analysis of the 21/22 October 2014 Storm Experienced by the Sailboat ECO40 in the Gulf of Lion

295

(2)

represents the Von Karman coefficient, equal

to 0.4. Thus, eq. (1) has been used to evaluate the

measured wind speed at the conventional height of

+10 m above the mean sea level, once the actual

wind speed U

and the actual measurement height z

are known.

As already pointed out, the anemometer is

placed on the top of the mast, thus its height (and of

course the measurement height z) can change over

time. This is related to the changes in the boat

heeling, which for certain sailing trim, can be

characterized by very pronounced heel angles.

Consequently, in order to carry out a proper

comparison between the measured wind data and

computed ones, a parametric analysis, by varying the

measurement height z, has been performed.

As qualitatively shown in Figure 13, the heeling

angle α modifies the measurement height to be used

in the equation (1). The middle and the lower panel

of the Figure 12 show the results of the comparison

between the measured wind data and the numerical

ones. In these case equation (1) has been used by

varying the heeling angle α (equal to 35° and 45°),

and the measurement height z (equal to 15.56 m and

13.43 m respectively). It appears that as the heeling

angle increases (e.g. α > 35°) then the measured

wind data and the numerical ones (both those

obtained by using the GFS and the ECMWF) are in

agreement.

Figure 13: Sketch of the measurement height z as a

function of the heeling angle α.

Ongoing research is focused on improving the

comparison between the measured wind data and the

calculated ones by using the trim boat

measurements, as obtained from GPS measurements,

and by using a more accurate relationship for the

vertical profile of the wind speed, that takes into

account the sea surface roughness in a real sea state

(i.e. presence of waves).

5 WAVE DATA ANALYSIS

This section describes briefly the analysis carried out

on the wave data collected during the Gulf of Lions

event. The available wave data sources are: the buoy

of Alghero (Italy), that is part of the RON (Italian

National Network for the wave measurements; see

Piscopia et al., 2002) and the data provided by the

numerical model of the ECMWF. Figure 14 shows

the time series of the significant wave height H

(left

panels) in few points of interest placed along the

route of the boat (right panels). The upper panel

shows the comparison between the significant wave

height H

measured by the Alghero buoy and the H

obtained by linearly interpolating the data provided

in analysis by ECMWF (dashed line). It is worth to

note that the two set of data are in good agreement:

indeed both the shape and the maximum values

appear quite similar. The remaining panels (second,

Figure 14: First panel (i.e. upper panel): comparison

between the significant wave height Hs measured by the

Alghero buoy and the same quantity obtained from the

analysis data of the ECMWF. Remaining panels (i.e.

second, third and fourth): evolution over time of the

significant wave height Hs, as obtained from the

numerical model of the ECMWF, evaluated in three points

of interest placed along the route of the boat. Note: the

vertical dashed red lines identify the time at which the

boat has passed in that point.

WPPDSports 2015 - Special Session/Symposium on Weather, Position and Performance data in Outdoor Sports

296

third and fourth) show the evolution over time of the

significant wave height H

, as obtained from the

numerical model of the ECMWF, evaluated in three

points of interest placed along the route of the boat.

The vertical dashed red lines identify the time at

which the boat has passed in that point and therefore

show the values of the significant wave height

experienced by the boat. Note that, in the future,

such numerical wave data will be compared with the

ones obtained by the signals measured by the three

high-precision GPS.

Finally, Figure 14 shows the significant wave

height time series during the Gulf of Lions storm of

November 1995 that has been reconstructed by

(Bertotti et al., 1998). We recall that this storm

caused the sinking of the Italian sailboat Parsifal and

the death of 6 crew members. Although a

quantitative comparison is not actually possible here,

we can qualitatively assess that the magnitude of the

sea state that ECO40 has faced is quite comparable

with that suffered by the Parsifal.

Figure 15: Significant wave height time series during the

Gulf of Lions storm of 2 November 1995 that has caused

the sinking of the Parsifal (Figures 5 and 6 of the paper of

Bertotti et al., 1998).

6 CONCLUDING REMARKS AND

ONGOING RESEARCH

This paper analyses the storm of October 21/22,

2014 faced by the boat ECO40 during the sailing in

the Gulf of Lion. The storm has been analysed on

the basis of the wind data, averaged over ten

minutes, measured and transmitted almost in real

time from the boat. These wind data have compared

with the ones obtained by using two numerical

models: the forecast data as from the GFS and the

analysis data as from the ECMWF. This comparison

has shown that the wind measurement height can

play an important role and therefore has to be

considered by knowing the arrangement of the boat

(i.e. heeling angle, etc.).

The measured values of the wind, averaged over ten

minutes, are comparable with the results obtained

from the numerical models: a good agreement is

noticeable within the first 24-48 hours. As the time

increases, then the numerical results tend to

deteriorate.

Furthermore, the forecast data seem to well predict

not only the magnitude of the event, in terms of

wind speed, but also the exact time of occurrence of

the storm peak, a parameter often prone to error.

A direct comparison between the forecast and the

measured wind showed relevant discrepancies, that

can be up to 50%. This difference is due to: (i) the

differences in altitude (i.e. the height at which the

wind measurements are carried out) that is estimated

to yield almost a 35% error and (ii) the gust

(approximately a 15% error).

Finally, we presented qualitative comparison

between the present wave data (obtained from the

RON buoy of Alghero and from the ECMWF

analysis data) and those reconstructed by (Bertotti et

al., 1998) that refer to the storm of November 1995

that caused the sinking of Parsifal. We have found a

substantial similarity in terms of sea-state magnitude

between the two events.

Currently, the ongoing research activities are

focused on improving the comparison between the

measured wind data and the calculated ones by using

the trim boat measurements, as obtained from GPS

measurements, and by using a more accurate

relationship for the vertical profile of the wind

speed, that takes into account the sea surface

roughness in a real sea state (i.e. presence of waves).

Furthermore, it is worth noticing that in the future a

comparison between the measured wave data, as

obtained by the signals measured by the three high-

precision GPS placed on the sailboat, and the ones

obtained from the available source of wave data (e.g.

numerical models, remote sensing techniques, wave

buoys, etc.) will be performed.

ACKNOWLEDGEMENTS

The authors would like to acknowledge Dr. Pier

Paolo Pecoraro (IDROGEOTEC) for his expertise

and help during the installation of the instruments.

The skipper Matteo Miceli is also acknowledged for

his contribution. A special acknowledgement is due

to Leica Geosystem for providing the high-precision

GPS reveivers.

REFERENCES

Bertotti, L., Cavaleri L., De Girolamo, P., Franco, L.,

Magnaldi S., 1998. Hindcast and forecast of the

Analysis of the 21/22 October 2014 Storm Experienced by the Sailboat ECO40 in the Gulf of Lion

297

Parsifal storm. Il Nuovo Cimento Vol 21, N. 3, pp.

281- 298.

Pierson, W. J., 1955. Wind generated gravity waves.

Advances in geophysics, 2, 93-178.

Piscopia, R., Inghilesi, R., Panizzo, A., Corsini, S.,

Franco, L., 2002, Analysis of 12-year wave

measurements by the Italian wave network.

Proceedings of the Coastal Engineering Conference.

Wu, J., 1969. Froude number scaling of wind-stress

coefficients. Journal of the Atmospheric Sciences,

26(3), 408-413.

WPPDSports 2015 - Special Session/Symposium on Weather, Position and Performance data in Outdoor Sports

298