Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica
DOI: 10.15517/ri.v33i2.54492
Esta obra está bajo una Licencia de Creative Commons. Reconocimiento - No Comercial - Compartir Igual 4.0 Internacional
Waves Data in Costa Rica: Validating the WAVERYS Reanalysis
using Waves Data
Datos de Oleaje en Costa Rica: Validando el Reanálisis WAVERYS con
Mediciones de Campo
Henry Alfaro-Chavarría
Unidad de Ingeniería Marítima, de Ríos y de Estuarios (IMARES)
University of Costa Rica, San José, Costa Rica
Email: henry.alfaro@ucr.ac.cr
ORCID: 0000-0003-4088-785X
Javier Zumbado-González
Unidad de Ingeniería Marítima, de Ríos y de Estuarios (IMARES)
University of Costa Rica, San José, Costa Rica
Email: javier.zumbado@ucr.ac.cr
ORCID: 0000-0002-6736-199X
Rodney Mora-Escalante
Centro de Investigación en Ciencias del Mar y Limnología (CIMAR)
University of Costa Rica, San José, Costa Rica
Email: rodney.moraescalante@ucr.ac.cr
ORCID: 0000-0001-7484-8299
Felipe Calleja-Apéstegui
Unidad de Ingeniería Marítima, de Ríos y de Estuarios (IMARES)
University of Costa Rica, San José, Costa Rica
Email: felipe.callejaapestegui@ucr.ac.cr
ORCID: 0000-0002-5864-9940
Georges Govaere-Vicarioli
Unidad de Ingeniería Marítima, de Ríos y de Estuarios (IMARES)
University of Costa Rica, San José, Costa Rica
Email: georges.govaere@ucr.ac.cr
ORCID: 0000-0002-0036-4571
Recibido: 14 de marzo de 2023 Aceptado: 16 de abril de 2023
Abstract
Field wave data are a relevant source of information with high impact for marine sciences and engineering.
The present work compiled, in a single database, the dierent wave data records found in Costa Rica.
Such wave data were compared with the WAVERYS reanalysis from the Copernicus Marine Environment
Monitoring Service with the purpose of examining this information and its possible use in future research at
Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica DOI: 10.15517/ri.v33i2.54492 117
both the country and regional levels. The historical wave data compilation considered records documented
in dierent projects carried out on behalf of several governmental institutions. Statistical methods were used
to analyze and compare, spatially and temporally, the information contained both in the eld wave data and
in the WAVERYS reanalysis. Results showed that, in the Caribbean, there are wave records between 2015
and 2017. In the Pacic, there are measurements made during the construction of Puerto Caldera between
1978 and 1985. There are also wave data obtained in dierent sites between 2009 to 2011 and by a recently
established wave gauge network from 2014 onwards. It was veried that the reanalysis database has a high
potential for applications in marine sciences and coastal engineering in this region of the Earth.
Keywords:
Costa Rica, eld measurements, reanalysis, wave data, WAVERYS.
Resumen
Las bases de datos de oleaje, a partir de mediciones de campo, son información relevante y de alto
impacto para las ciencias marinas y la ingeniería. Este trabajo compiló las diferentes bases de datos de oleaje
medidas en campo en Costa Rica con el n de preservar en una única fuente de consulta dicha información;
a su vez, se compararon dichos datos con los ofrecidos por el reanálisis WAVERYS del Copernicus Marine
Enviroment Monitoring Service, con el propósito de analizar esta información y su posible utilización en
futuras investigaciones en el país y en la región. Se recopilaron los datos históricos empleados en distintos
proyectos y registrados por diferentes instituciones del Estado. Los datos fueron analizados espacial y
temporalmente por métodos estadísticos, se realizaron comparaciones entre los distintos registros y la base
de datos de reanálisis de oleaje WAVERYS. Los resultados mostraron que en el Caribe se tienen registros de
oleaje entre los años 2015 y 2017. En el Pacíco, se cuenta con datos que se midieron durante la construcción
de Puerto Caldera (1978-1985); además, existen datos de campañas de campo realizadas en distintos lugares
entre los años 2009 y 2011, y los datos medidos por una red de equipos de oleaje colocados en zonas costeras
desde el 2014 hasta la actualidad. Finalmente, se vericó que los datos del reanálisis tienen un alto potencial
de aplicación en las ciencias marinas y la ingeniería de costas en esta región del globo.
Palabras Clave:
Costa Rica, mediciones de campo, reanálisis, datos de oleaje, WAVERYS
ALFARO, ZUMBADO, MORA, CALLEJA, GOVAERE: Waves Data in Costa Rica...
118
1. INTRODUCTION
Waves aect natural processes and human activities in the coastal zone, such as coastal
erosion and accumulation [1], marine ecosystems and their dynamics [2], marine aquaculture
[3], the design of maritime structures [4], such as breakwaters [5], and the behavior of oating
structures [6].
Wave data measured in situ are one of the most important sources of information to
characterize this variable. However, due to their investment and operation costs, these data are
usually measured in strategic areas and mostly during short periods, which can be representative
of a large spatial extension. Countries with developed coastal and port engineering have networks
of equipment distributed along their coasts on a permanent basis (i.e: Buoy Network of Puertos del
Estado, España; National Data Buoy Center-National Oceanic and Atmospheric Administration's
(NDBC-NOAA), USA; Ocean Data Buoy Observations-Japan Meteorological Agency, Japan).
Nevertheless, in other oceanic regions such as along the Latin American coast, wave data
measured on the eld are usually scarce. Costa Rica is no exception, and the available data
catalogs, until the last decade, were few and present temporal and spatial limitations since they
have solved specic needs for particular projects [7], [8] and [9]. However, as of 2014, Costa
Rica has a network of gauge wave placed in coastal areas, formed by the research groups:
Unidad de Ingeniería Marítima, de Ríos y de Estuarios (IMARES) and Módulo de Información
Oceanográca del Centro de Investigación en Ciencias del Mar y Limnología (MIO-CIMAR),
both from the University of Costa Rica (UCR). The purpose of the network is to measure
waves on both coasts of the country continuously and through equipment distributed in sites of
scientic interest.
Other sources of wave data, which have gained relevance over the last two decades, are
global wave reanalysis based on spectral wave model —i.e: Production Hindcast [10]; GOW2
[11]; WAVERYS [12]. These databases are characterized by having historical information,
homogeneous in space and continuous in time, which is important to understand the dynamic
behavior of waves. Nevertheless, despite the continuous improvement in the models used, the
resolution and quantity of their forcings, they need to be calibrated and validated with information
measured in eld or from satellites [13].
The WAVERYS database of the Copernicus Marine Environment Monitoring Service
(CMEMS) is used in this work due to the high spatial resolution, which is important for our
regional application. WAVERYS is a global reanalysis of wave surface conditions, generated
with the operational model MFWAM (Meteo France WAve Model) version 4 [14]. It is forced
from the ice and wind elds coming from the ERA5 atmospheric reanalysis [15], an initiative
developed by European Centre for Medium-Range Weather Forecasts [16].
WAVERYS has a spatial resolution of 1/5°, currently covers the period between 1993 and
2021, and has information on wave parameters with a resolution of every three hours. The
WAVERYS information was validated for deep water areas with data from satellite altimeter
(HY-2A satellite, not included in the assimilation) and for coastal areas with information from
buoys along the globe between the years 1994 and 2015 [12]. However, the buoys used for
Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica DOI: 10.15517/ri.v33i2.54492 119
validation are mainly located in the high latitudes of the northern hemisphere, because there is
more equipment installed. For the South American region, ve buoys are used for validation,
three in the Pacic Ocean (one in Colombia and two in Chile) and two in the Atlantic (one in
Florianopolis, Brazil and one in Cayenne, French Guiana) [17].
The present work has two main objectives: i) to show the historical wave databases recorded
in the eld on both coasts of Costa Rica and ii) to compare the wave data provided by the
WAVERYS reanalysis with the information measured in the eld at dierent sites and times,
in order to analyze the information provided by this reanalysis and its possible use in future
research in the country and in the region.
2. MATERIALS AND METHODS
This research used historical and available wave data measured by dierent public institutions
in the Pacic and Caribbean coasts of Costa Rica. Moreover, information recently measured in
the eld by IMARES and MIO-CIMAR of the UCR was used. All the wave databases used were
measured by dierent types of wave gauges, placed in dierent places and at dierent times.
In addition, the study included data from the WAVERYS wave reanalysis [12] of the
Copernicus Marine Environment Monitoring Service (CMEMS). The numerical nodes of the
reanalysis were selected close to the sites where the equipment was placed in the eld, in order
to validate the trend over time of the WAVERYS information, based on the measured data.
The wave data measured in eld and those from the reanalysis were standardized by means
of the main wave parameters, such as zero-order moment wave height (Hmo), peak period (Tp)
and mean direction (θ). Both databases were analyzed using descriptive statistical tools; these
were compared using graphical control tools and quality indicators.
2.1 Study zone
The databases and numerical nodes of the WAVERYS model used in this study are located
in the northern and central zone of the Pacic coast of Costa Rica; on Coco Island, located
approximately 500 km from the national territory, between the country and the Galapagos Islands,
and on the Caribbean coast (Fig. 1).
The data sites present natural and social conditions that have favored the measurement of
wave data in their vicinity. For example, they include the most important ports of the country (i.e.,
Caldera and Moín in Fig. 1), natural parks and conservation areas of important ecological interest
(e.g., Coco Island, Cabo Blanco in Fig. 1), and exposed sites that allow accurate characterization
of incident waves (e.g., Cabo Velas, Sámara, Playa Grande and Quepos). In addition, they present
shallow depths where it has been feasible to install measurement equipment (50 m or less).
The study area also includes six numerical nodes of the WAVERYS model (Fig. 1), located
in the vicinity of the wave measurement points, which makes it possible to compare both sources
of information.
ALFARO, ZUMBADO, MORA, CALLEJA, GOVAERE: Waves Data in Costa Rica...
120
Fig. 1. Spatial distribution of gauges and nodes of reanalysis WAVERYS.
2.2 Field measured data and WAVERYS reanalysis
The wave data were collected by public institutions in dierent time periods and were
not centralized in any general database or public access system. Therefore, the collection of
the measured data, as well as the details associated with the recording process (i.e., dates,
coordinates, types of gauges), were done through direct contact with the institutions in charge
of the information. The original data and measurement details were requested; in cases where
the information was printed on paper, it was digitized manually.
Numerical nodes from the WAVERYS reanalysis [12] (Fig 1) had the wave parameters zero-
order moment height, peak period and wave direction (Hmo, Tp and θ) extracted together with
the associated time variable. Subsequently, time periods were selected when both databases
coincided temporally. In this way, the spatial and temporal comparison of both sources of
information was achieved.
2.3 Data analysis
The wave data and the reanalysis were analyzed by means of graphical tools of descriptive
statistics such as time series and wave rose. Also, the H
mo
and T
p
parameters of both data sources
were compared using scatter plots; in addition, statistical descriptors such as BIAS in equation
(1), root mean square error (RMSE) in equation (2), Pearson's correlation coecient (ρ) in
equation (3) and dispersion index (SI) in equation (4) were calculated.
=
−
(1)
Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica DOI: 10.15517/ri.v33i2.54492 121
=1∑(−)2
=1
(2)
=∑(−)( −)
=1
∑(−)2
=1 ∑(−)2
=1
(3)
=
(4)
where xi is the reference data, yi is the reanalysis data and n is the number of observations.
3. RESULTS
3.1 Field measured data
Field data were measured by public institutions such as the Ministerio de Obras Públicas y
Transportes (MOPT), the Comisión de Marinas y Atracaderos Turísticos (CIMAT), the Consejo
Nacional de Concesiones (CNC), and the UCR through the IMARES and MIO-CIMAR research
groups. Data were recorded at eight dierent sites (Fig. 1): Cabo Velas, Cabo Blanco, Sámara,
Playa Grande, Caldera, Quepos, Moín, and Coco Island.
The MOPT wave data were measured as part of eld studies during the design and construction
of Puerto Caldera. The equipment used was an ultrasonic wavemeter model USW-2000A (brand:
Kaijo Denki Corporation/SONIC, Japan), which was placed 1.8 km oshore, at a depth of
15.5 m and measured for a period of 7.3 years, between 1978 and 1985; however, it presented
interruptions, and the eective period of measurement was 3.3 years [7]. The available information
corresponds to the main wave parameters, represented by various statistics of forty-eight storms
with signicant wave height greater than 1.5 m.
The CIMAT wave data were measured as part of a consultancy, whose purpose was to
calibrate the NOAA wave forecast with data measured in eld. The equipment used was a
pressure sensor model TWR-2050 (brand: RBR Ltd., Canada), which was placed in front of
Sámara at 22 m depth and maintained collecting data between 2009 and 2010. Subsequently,
the equipment was moved in front of Playa Grande, where it was placed at 22 m depth during
2011. The sensor sampled at 2 Hz, so that it would take data for 20 minutes every hour; however,
it only took measurements when NOAA predicted that the waves in deep water and in front of
these beaches would exceed approximately 2 m wave height [8].
The CNC wave data were measured during the construction process of the Terminal de
Contenedores de Moín (TCM). The equipment placed was a directional buoy model Waverider
(brand: Datawell, Netherlands), which was located in front of Moín at 14 m depth and measured
for 2 years, between 2015 and 2017 [9].
The MIO-CIMAR group reported wave data during the months between March and November
2020, measured by a directional oceanographic buoy model SB-138P (brand: Tideland, USA),
located about 8 km northwest of Quepos, on the isobath of approximately 50 m and with real-time
ALFARO, ZUMBADO, MORA, CALLEJA, GOVAERE: Waves Data in Costa Rica...
122
transmission. The buoy measured the main wave parameters, current prole and meteorological
variables with a reporting interval of 15 minutes. Directional waves were measured with a
MOTUS sensor (movement in Latin, brand: Aanderaa, Norway), which measures the motion of
the buoy by integrating accelerometers, magnetometers, gyroscopes in its inertial motion unit
(IMU) and in conjunction with an external compass for correction. The sampling rate of the free
surface elevations of the MOTUS sensor was 4 Hz.
The IMARES group has measured waves at dierent sites along the Pacic coast with two
types of equipment: an ADCP (Acoustic Doppler Current Proler) model AWAC (Acoustic
Wave and Current Proler, brand: Nortek, Norway) and a scalar buoy model BARES (brand:
HCTech, Spain). The rst site where IMARES measured waves was in front of Cabo Blanco,
at 15 m depth, in 2014 with an AWAC. In 2015, this equipment was moved approximately 2.5
km to the northwest at 1 m depth and currently remains at that location. The second site was in
Puerto Caldera with the BARES buoy, anchored at 15 m depth, where it recorded data between
2015 and 2018. The third site was at Coco Island, where another AWAC was placed at 20 m
depth, between the months of March and October 2018 and during April 2019. The fourth site
was Cabo Velas, where another AWAC was placed at 18 m depth in October 2019 and continues
measuring nowadays.
The AWAC equipment was set to collect data at 2 Hz for approximately 17 minutes every
3 hours. The data collected free surface records, which were then analyzed in the time and
frequency domain to determine the main wave parameters. The buoy made inertial measurements
in all three-axis using a gyroscope, accelerometer and a compass, and it was controlled by a
low-power microcontroller. The buoy was set up to take data for approximately 17 minutes on
an hourly basis and transmitted in real time. TABLE I summarizes the main information from
the historical and available wave databases.
TABLE I
MAIN INFORMATION OF HISTORICAL AND AVAILABLE WAVES DATA
Location Equipment Recording period Geographic
Coordinates Depth (m) Source
Cabo Velas AWAC 2019-present 10.36°N 85.88°W 18.0 IMARES
Playa Grande TWR-2050 2011 10.20°N 86.00°W 22.0 CIMAT
Sámara TWR-2050 2009-2010 9.85°N 85.50°W 22.0 CIMAT
Cabo Blanco AWAC 2014-present 9.56°N 85.13°W 18.0 IMARES
Puerto
Caldera USW/BARES 1978-1985
2015-2018 9.91°N 84.74°W 15.5 MOPT IMARES
Quepos MOTUS 2020-present 9.39°N 84.23°W 50.0 MIO-CIMAR
Coco Island AWAC 2018-2019 5.50°N 87.06°W 20.0 IMARES
Moín Waverider 2015-2017 10.03°N 83.11°W 14.0 CNC
Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica DOI: 10.15517/ri.v33i2.54492 123
3.2 Comparison of measured and modeled data by WAVERYS
The wave data measured for the Puerto Caldera project do not temporally overlap with
the reanalysis data, so a comparison is not possible; however, this information was included
as supplementary material in order to preserve the only existing wave data record from that
project. The data correspond to the wave parameters of the forty-eight storms greater than 1.5
m signicant wave height, which were recorded during the 7.3 years that the equipment was
in place.
Fig. 2A and C show the time series of the wave parameters Hmo and Tp, respectively,
corresponding to the data measured in front of Moín and the WAVERYS node near that site. Both
sources of information follow the same pattern of behavior over time; in addition, it was found
that the series of measured data shows continuity in the rst year and a half of measurement
and then there are periods of time with missing information. Fig. 2B shows a linear correlation
of the Hmo data with a coecient ρ of 95 %, a BIAS of 17 cm, an SI of 0.14 and a RMSE of 24
cm. Fig. 2D, corresponding to the variable Tp, shows a correlation ρ of 76 %, a SI of 0.11, a
BIAS and RMSE of less than 1 s.
A B
C D
Fig. 2. Moín, A) Time series of Hmo parameter, B) scatter plot of Hmo, C) time series
of Tp parameter and D) scatter plot of Tp.
Fig. 3A and C show the time series of the wave parameters Hmo and Tp, respectively,
corresponding to the data measured in front of Quepos and the WAVERYS node near that site.
The same pattern of behavior is observed over time for the two parameters with respect to each
source; also, it is highlighted that the measured data is composed of a continuous series and
extends from March 2020 to March 2021. Fig. 3B shows a linear correlation of the Hmo data
with a coecient ρ of 87 %, a BIAS of 20 cm, an SI of 0.12 and a RMSE of 25 cm. Fig. 3D,
ALFARO, ZUMBADO, MORA, CALLEJA, GOVAERE: Waves Data in Costa Rica...
124
corresponding to the variable Tp, shows a correlation ρ of 30 %, a SI of 0.24, a BIAS and RMSE
of less than 3 s and 4 s, respectively.
A B
C D
Fig. 3. Quepos, A) Time series of Hmo parameter, B) scatter plot of Hmo, C) time
series of Tp parameter and D) scatter plot of Tp.
Fig. 4A and C show the time series of the wave parameters Hmo and Tp, respectively,
corresponding to the data measured at Cabo Blanco and the WAVERYS reanalysis node near that
site. Both sources of information show the same pattern of behavior over time; Fig. 4A shows
that the Hmo values measured during 2014, recorded at a dierent site, are higher compared to
the rest of the measured and reanalysis data; also, it is highlighted that the measured data are
composed of an almost continuous series that extends to date. Fig. 4B shows a linear correlation
of the Hmo data with a coecient ρ of 88 %, a BIAS of 12 cm, a SI of 0.13 and a RMSE of 21
cm. Fig. 4D, corresponding to the Tp parameter, shows a correlation ρ of 55 %, a BIAS less than
2 s and RMSE less than 3 s.
A B
C D
Fig. 4. Cabo Blanco, A) Time series of Hmo parameter, B) scatter plot of Hmo, C) time
series of Tp parameter and D) scatter plot of Tp.
Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica DOI: 10.15517/ri.v33i2.54492 125
TABLES II and III show the BIAS, RMSE, ρ and SI results for the Hmo and Tp parameters
for each of the sites analyzed; the respective gures were included as supplementary material.
TABLE II shows that there is linear correlation of the Hmo parameter between both sources of
information, with Puerto Caldera being the site with the lowest value of 76 %, but with the
smallest BIAS of 7.7 cm towards the measured data. The rest of the sites show linear correlation
values greater than 80 % and dispersion less than 0.2. TABLE III shows the periods are biased
towards the WAVERYS data, the errors are less than 5 s and the correlation does not exceed 64 %.
TABLE II
STATISTICAL DESCRIPTORS OF Hmo PARAMETER
Location BIAS (cm) RMSE (cm) ρSI
Sámara 11 24 0.81 0.16
Playa Grande 53 55 0.94 0.15
Puerto Caldera -7.7 23 0.76 0.21
Coco Island 32 37 0.83 0.13
Cabo Velas 44 48 0.80 0.19
TABLE III
STATISTICAL DESCRIPTORS OF Tp PARAMETER
Location BIAS (s) RMSE (s) ρSI
Sámara - - - -
Playa Grande 0.45 2.9 0.62 0.22
Puerto Caldera 1.1 2.5 0.64 0.16
Coco Island 2.7 3.7 0.55 0.20
Cabo Velas 2.9 4.5 0.26 0.27
Fig. 5 shows the wave roses corresponding to the Hmo parameter; row 1 corresponds to the
data measured in the eld and row 2 to the reanalysis nodes near each measurement site; columns
A, B, C and D correspond to the measurement sites Cabo Velas, Cabo Blanco, Quepos, and Moín,
respectively. The sites located in the Pacic show that the swell comes from the SW sector; Fig.
5A1 shows that the direction S67.5°W corresponds to 50 % of the time, followed by 25 % of the
time with SW swells. However, Fig. 5B1 and C1 show that the main swell direction is S22.5°W
for about 60 % of the time and 25 % of the time there are S45°W directions. Fig. 5A2, B2 and
C2 coincide with the distributions of the directions reported by the measured data. In the central
part of the Pacic coast, the swell concentrates the directions with a greater southern component
(Fig. 5B1, B2, C1 and C2), while the northern most sector of the coast presents swells with
directions with a greater western component (Fig. 5A1 and A2).
Fig. 5D1 and D2 show the same distribution of directions, with the main direction being
N67.5°E with about 50 % of the time, followed by N45°E with about 30 % of the time; the rest
of the time the swell direction is distributed between N and N22.5°E.
ALFARO, ZUMBADO, MORA, CALLEJA, GOVAERE: Waves Data in Costa Rica...
126
Fig. 5. Hmo parameter roses, 1) eld data, 2) nodes of WAVERYS near to, A) Cabo
Velas, B) Cabo Blanco, C) Quepos y D) Moín.
4. DISCUSSION
The analysis of the eld measured data shows a swell climate that varies in time and space,
and that corresponds, for example, to the Pacic coast, with the weather patterns of other
latitudes. The most extensive wave database, corresponding to Cabo Blanco, shows an oscillatory
annual behavior of the Hmo variable. Between the months of November and March, the Hmo
magnitudes are, on average less than 1 m with a Tp of approximately 15 s; between the months
of April and October, the magnitudes of Hmo and Tp increase and are of the order of 1.5 m and
17 s, respectively; while during the months between June and August, some waves exceed 3 m
height and 20 s of peak period.
This behavior coincides temporally with the climatology of the southern hemisphere,
where the summer and winter months coincide with the months of lowest and highest wave
energy in the Costa Rican Pacic, respectively. Thus, it is found that the swell originates in the
southwestern Pacic Ocean, it travels approximately 9000 km to cross the ocean and reach the
coast of Central America [18]-[19].
On the other hand, wave data measured in the Caribbean show that months with the highest
wave energy occur between December and March, with Hmo and Tp magnitudes of about 2 m
and 10 s, respectively; the lowest energy months occur between September and October, with
Hmo magnitudes of approximately 1m and Tp of 8 s. Between the months of July and August,
wave events occur with high H
mo
magnitudes, but normally lower than those occurring between
months of December and March.
Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica DOI: 10.15517/ri.v33i2.54492 127
This behavior coincides with the climatology of the trade winds, which present their greatest
magnitudes between months of December and March; then their speeds decrease, increase again
between months of July and August, and decrease again between September and October [20].
The wave height roses generated from the data measured in eld and from the reanalysis
conrm that the swell has a greater variability of directions in the northern sector of the Pacic
coast; that is, the swell directions are distributed between south and west. Nevertheless, in the
central sector of the Pacic coast the swell shows less variability with mainly south-southwest
directions. This could be because the Galapagos Islands, located about 1300 km away, alter the
propagation process of the swell coming from the southern hemisphere.
The Hmo parameter measured in Cabo Blanco during 2014 shows higher magnitudes than
those reported in the rest of the record but follows the same annual behavior. This dierence
could be attributed to the fact that the equipment, starting in 2015 and up to the present, was
relocated 2.5 km northwest of the site where it was originally placed in 2014.
In Caldera, the Hmo magnitudes are biased towards the measured data; that is, the data
measured in the eld are higher than those reported by the WAVERYS. It is possibly because
the equipment was placed in front of the breakwater and close to clis that are reective, which
could inuence the measurements and be evidenced as an increase in the energy measured by
the equipment. In the rest of the sites, being coastal sites and directly exposed to waves, the
reanalysis shows higher magnitudes of the Hmo parameter than the data measured in the eld.
As for the Tp parameter, the results at all sites show that there is a temporal correlation
with the database. However, the peak period is an imprecise parameter to compare, as it is the
maximum frequency associated with the spectrum; being the measured spectra of a higher
sampling resolution than the theoretical wave spectra, which generally have programmed wave
reanalysis.
The comparison of the Hmo parameter between the measured and WAVERYS wave data
reveals that there is a temporal and linear correlation between the databases, even though they do
not match spatially. This demonstrates that, despite not being a calibration, WAVERYS reanalysis
is able to adequately follow trends in space and time. The wave parameter Hmo, which, added
to its spatial and temporal homogeneity, gives it a great value and usefulness to undertake, as
a rst approximation, dierent projects of maritime engineering, environmental, mariculture,
wave energy and risk mitigation product of extreme events. Likewise, the magnitudes of the
statistical descriptors related to the period and direction obtained coincide in order of magnitude
with those estimated by [12] in their validations.
5. CONCLUSIONS
This work compiled, in a single source of consultation, the wave information that has been
measured in the country in order to preserve the data, make them available to the scientic
community and the corresponding decision-making authorities, upon request to the authors.
Among them, the wave data measured by IMARES and MIO-CIMAR groups stands out, being
ALFARO, ZUMBADO, MORA, CALLEJA, GOVAERE: Waves Data in Costa Rica...
128
an unpublished source of information of great value for the country and for the region, since it
is an area of the globe where primary information is scarce.
The measured data were used to validate the WAVERYS wave reanalysis information; an
adequate temporal and spatial coincidence was veried in the sites analyzed, so it is concluded
that the reanalysis is suitable for various marine science and marine engineering projects in this
region. However, this source of information is considered complementary to the measured data,
and for more accurate results it is convenient to calibrate with measured data.
ACKNOWLEDGMENTS
We would like to thank the Ministerio de Obras Públicas y Transportes for nancing the wave
measurements at Cabo Blanco, the Comisión de Marinas y Atracaderos Turísticos, the Consejo
Nacional de Concesiones and APM Terminals for providing the wave data. We would also like
to thank the technical sta of IMARES, who made an exceptional contribution to the eldwork.
ROLES
Henry Alfaro-Chavarría: Conceptualización, Análisis formal, Adquisición de fondos,
Investigación, Metodología, Administración del proyecto, Validación, Redacción – borrador
original – Preparación
Javier Zumbado-González: Curación de datos, Análisis formal, Recursos, Software,
Visualización
Rodney Mora-Escalante: Curación de datos, Recursos, Software, Redacción – revisión y
edición – Preparación
Felipe Calleja-Apéstegui: Conceptualización, Redacción – revisión y edición – Preparación
Georges Govaere-Vicarioli: Conceptualización, Adquisición de fondos, Metodología,
Recursos, Supervisión, Redacción – revisión y edición – Preparación
REFERENCES
[1] C. Jaramillo, M. S. Jara, M. González and R. Medina, “A shoreline evolution model considering the
temporal variability of the beach prole sediment volume (sediment gain/loss),” Coast. Eng., vol.
156, pp. 103612, March, 2020, doi: https://doi.org/10.1016/j.coastaleng.2019.103612.
[2] E. Ramos et al, “Coastal waters classication based on physical attributes along the NE Atlantic
region. An approach for rocky macroalgae potential distribution,” Estuar. Coast. Shelf Sci., vol. 112,
pp. 105-114, Oct., 2012, doi: http://dx.doi.org/10.1016/j.ecss.2011.11.041.
[3] F. Calleja, G. J. Chacón and H. Alfaro, “Marine aquaculture in the pacic coast of Costa Rica:
Identifying the optimum areas for a sustainable development,” Ocean. Coast. Manag., vol. 219, pp.
106033, March, 2022, doi: https://doi.org/10.1016/j.ocecoaman.2022.106033.
Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica DOI: 10.15517/ri.v33i2.54492 129
[4] B. Gouldby, F. J. Méndez, Y. Guanche, A. Rueda and R. Mínguez, “A methodology for deriving
extreme nearshore sea conditions for structural design and ood risk analysis,” Coast. Eng., vol. 88,
pp. 15-25, June, 2014, doi: https://doi.org/10.1016/j.coastaleng.2014.01.012
[5] T. Albers and K. Schmitt, “Dyke design, oodplain restoration and mangrove co-management as parts
of an area coastal protection strategy for the mud coasts of the Mekong Delta, Vietnam,” Wetlands Ecol.
Manage., vol. 23, no. 6, pp. 991-1004, Jul., 2015, doi: https://doi.org/10.1007/s11273-015-9441-3.
[6] P. Li, O. M. Faltinsen and M. Greco, “Wave-Induced Accelerations of a Fish-Farm Elastic Floater:
Experimental and Numerical Studies,” J. Oshore Mech. Arct. Eng., vol. 140, no 1, pp. 011201, Feb.,
2018, doi: https://doi.org/10.1115/1.4037488.
[7] JICA, Japan International Cooperation Agency, “Final Report. The Study on the Maintenance Project
of the Port of Caldera in the Republic of Costa Rica,” Ministerio de Obras Públicas y Transportes,
San José, Costa Rica, 1986.
[8] Watermark, “Consultoría para la medición y calibración del oleaje en la costa pacíca de Costa Rica,”
Instituto Costarricense de Turismo, San José, Costa Rica, 2010.
[9] Baird and CH2M HILL, “Meteorological and Oceanographic Report,” APM Terminals, San José,
Costa Rica 2013.
[10] H. Tolman and D. Chalikov, “Source Terms in a Third-Generation Wind Wave Model,” J. Phys. Oceanogr.,
vol. 26, no. 11, pp. 2497-2518, 1996, doi: https://doi.org/10.1175/1520-0485(1996)026<2497:STIAT
G>2.0.CO;2.
[11] J. Pérez, M. Menéndez and I. J. Losada, “GOW2: A global wave hindcast for coastal applications,”
Coast. Eng., vol. 124, pp. 1-11, Jun., 2017, doi: http://doi.org/10.1016/j.coastaleng.2017.03.005.
[12] S. Law-Chune, L. Aouf, A. Dalphinet, B. Levier, Y. Drillet and M. Drevillon, “WAVERYS: a CMEMS
global wave reanalysis during the altimetry period,” Ocean Dyn., vol. 71, pp. 357-378, Jan., 2021,
doi: https://doi.org/10.1007/s10236-020-01433-w.
[13] C. Izaguirr, F. J. Méndez, M. Menéndez and I. J. Losada, “Global extreme wave height variability
based on satellite data,” Geophys. Res. Lett., vol. 38, no. 10, pp. L106007, May, 2011, doi: https://
doi.org/10.1029/2011GL047302.
[14] L. Aouf, D. Hauser, C. Tison and B. Chapron, “On the assimilation of multi-source of directional wave
spectra from Sentinel 1A and 1B, and COSAT in the wave model MFWAM: Toward an operational
use in CMEMS-MFC,” presented at International Geoscience and Remote Sensing Symposium
Conference, Valencia, Spain, Jul. 22-27, 2018.
[15] ERA5-Land hourly data from 1950 to present. Copernicus Climate Change Service (C3S) Climate
Data Store (CDS), Copernicus Climate Change Service (C3S), 2022, doi: 10.24381/cds.e2161bac.
[16] European Centre for Medium-Range Weather Forecasts. Part VII: ECMWF Wave Model. (2013).
Accessed: 2022. [Online]. Available : https://www.ecmwf.int/sites/default/les/elibrary/2013/9248-
part-vii-ecmwf-wave-model.pdf
[17] Copernicus Marine Environment Monitoring Service. GLOBAL_REANALYSIS_WAV_001_032.
(2021). Accessed: 2022. [Online]. Available : https://catalogue.marine.copernicus.eu/documents/
QUID/CMEMS-GLO-QUID-001-032.pdf
[18] Y. Goda, “Analysis of wave grouping and spectra of long-travelled swell,” Rep. Port and Harbour
Res. Inst., vol. 22, no. 1, pp. 3-41, 1983.
[19] R. E. Mora-Escalante and J. P. Ureña-Mora, “Numerical simulation of wave eld around Cocos Island,
Costa Rica,” Rev. Biol. Trop., vol. 68, no. 1, pp. 198-212, Mar., 2020, doi: https://doi.org/10.15517/
rbt.v68iS1.41181.
ALFARO, ZUMBADO, MORA, CALLEJA, GOVAERE: Waves Data in Costa Rica...
130
[20] V. Magaña, J. A. Amador and S. Medina, “The midsummer drought over Mexico and
Central America”, J. Clim., vol. 12, no. 6, pp. 1577-1588, Jun., 1999, doi: https://doi.
org/10.1175/1520-0442(1999)012<1577:TMDOMA>2.0.CO;2.
COMPLEMENTARY MATERIAL
TABLE A1
FORTY-EIGHT HIGHEST WAVES SURVEYED DURING THE CONSTRUCTION OF
PUERTO CALDERA
Ranking Date Hmax
(m)
Tmax
(s)
H1/10
(m)
T1/10
(s)
H1/3
(m)
T1/3
(m)
Hmean
(m)
Tmean
(s)
Year Month Day Time
1 1981 5 21 16 5.44 16.90 4.17 17.80 3.55 17.90 2.19 15.10
21983 7 18 4 4.44 16.80 4.09 17.20 3.47 17.10 2.31 16.10
3 1978 6 18 8 4.10 18.00 3.90 17.80 3.30 17.50 1.90 15.00
4 1985 5 28 17 4.17 16.30 3.74 17.70 2.94 17.80 1.85 15.50
5 1982 3 18 12 3.83 15.70 3.10 16.10 2.86 15.90 1.33 12.30
61985 9 13 17 3.66 19.80 3.47 17.70 2.77 17.60 1.73 13.00
7 1981 56 18 3.98 20.00 3.43 16.70 2.74 17.40 1.58 12.00
8 1983 8 7 24 3.80 16.90 3.27 17.30 2.66 17.50 1.71 16.00
91980 11 58 3.36 17.70 2.95 17.60 2.53 17.10 1.55 12.50
10 1985 10 2 23 3.22 17.60 2.83 17.30 2.49 17.00 1.54 12.30
11 1981 11 29 2 3.14 17.50 2.93 16.40 2.44 16.40 1.50 13.90
12 1980 10 16 22 3.55 18.10 3.11 17.30 2.22 17.10 1.30 12.40
13 1985 10 27 72.92 17.50 2.75 17.10 2.20 17.30 1.33 14.50
14 1982 612 4 2.75 13.90 2.53 14.50 2.13 15.70 1.38 12.80
15 1978 9 18 4 3.20 9.00 2.70 8.60 2.10 8.90 - -
16 1985 5 17 18 3.05 16.90 2.55 17.30 2.10 16.10 1.26 10.70
17 1981 11 20 24 3.10 16.10 2.37 15.50 2.06 15.90 1.41 13.90
18 1985 8 18 8 2.88 14.20 2.56 15.60 2.06 15.60 1.26 12.90
19 1981 7 10 20 2.69 17.60 2.30 14.60 2.02 15.40 1.21 11.00
20 1978 10 3 16 3.00 15.00 2.50 16.00 2.00 16.00 - -
21 1979 8 7 16 2.50 20.00 2.30 19.00 2.00 18.20 1.40 16.20
22 1981 3 21 18 2.48 14.30 2.29 15.50 2.00 15.70 1.21 13.00
23 1981 11 120 2.99 16.30 2.40 16.40 1.99 16.10 1.23 18.30
24 1978 8 6 4 3.20 14.00 2.40 14.50 1.90 14.50 - -
25 1979 5 20 24 2.50 17.00 2.30 16.00 1.90 16.00 1.30 15.00
26 1985 4 17 15 3.00 12.90 2.41 9.60 1.88 12.30 1.08 8.50
Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica DOI: 10.15517/ri.v33i2.54492 131
Ranking Date Hmax
(m)
Tmax
(s)
H1/10
(m)
T1/10
(s)
H1/3
(m)
T1/3
(m)
Hmean
(m)
Tmean
(s)
Year Month Day Time
27 1985 926 92.48 15.70 2.30 16.30 1.85 16.30 1.13 13.40
28 1985 630 13 2.50 14.30 2.18 12.10 1.81 12.70 1.12 8.50
29 1979 9 7 16 2.70 18.00 2.30 16.00 1.80 15.40 1.10 12.20
30 1982 8 7 18 2.62 16.00 2.26 16.20 1.79 16.50 1.06 11.80
31 1985 8511 2.45 17.80 2.17 16.30 1.75 16.50 1.16 13.70
32 1985 3 19 2 2.53 16.00 2.12 17.00 1.74 16.70 1.16 13.70
33 1979 9 25 16 2.50 16.00 2.10 16.00 1.70 15.00 1.20 12.00
34 1985 9 8 15 2.46 15.00 2.04 14.30 1.70 14.50 1.07 12.20
35 1981 1 17 6 2.26 13.80 1.98 13.90 1.67 14.20 1.01 13.00
36 1982 3 9 14 2.35 14.30 2.03 13.70 1.61 13.50 0.96 9.70
37 1978 9 12 20 2.70 14.00 2.00 15.40 1.60 15.00 - -
38 1983 9 9 6 2.45 15.60 1.98 14.90 1.60 15.00 1.01 11.40
39 1978 7 9 16 2.00 16.00 1.90 15.90 1.60 16.00 - -
40 1985 7 17 41.93 16.00 1.80 15.70 1.58 16.10 1.06 14.00
41 1982 5 19 10 2.42 14.90 1.93 14.90 1.57 14.80 1.00 12.50
42 1981 2722 2.40 14.30 1.96 14.20 1.57 14.30 0.97 13.10
43 1982 815 62.55 15.80 1.99 15.20 1.56 15.50 1.00 11.90
44 1984 2 26 22 1.89 13.40 1.59 14.70 1.56 11.30 0.90 9.50
45 1985 615 72.29 13.10 2.04 13.30 1.51 13.30 0.91 11.60
46 1983 8 15 22 2.12 12.80 1.90 13.70 1.51 13.70 0.93 9.90
47 1980 911 22 2.06 11.50 1.82 11.80 1.51 11.70 0.97 9.70
48 1978 10 15 4 1.90 16.00 1.70 15.00 1.50 14.00 - -
A B
Fig. B1. Sámara Beach, A) Time series of Hmo parameter and B) scatter plot of Hmo.
ALFARO, ZUMBADO, MORA, CALLEJA, GOVAERE: Waves Data in Costa Rica...
132
A B
C D
Fig. B2. Playa Grande, A) Time series of Hmo parameter, B) scatter plot of Hmo, C)
time series of Tp parameter and D) scatter plot of Tp.
A B
C D
Fig. B3. Puerto Caldera, A) Time series of Hmo parameter, B) scatter plot of Hmo, C)
time series of Tp parameter and D) scatter plot of Tp.
Ingeniería 33(2): 116-133, Julio-Diciembre, 2023. ISSN: 2215-2652. San José, Costa Rica DOI: 10.15517/ri.v33i2.54492 133
A B
C D
Fig. B4. Coco Island, A) Time series of Hmo parameter, B) scatter plot of Hmo, C)
time series of Tp parameter and D) scatter plot of Tp.
A B
C D
Fig. B5. Cabo Velas, A) Time series of Hmo parameter, B) scatter plot of Hmo, C) time
series of Tp parameter and D) scatter plot of Tp.
