Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251

Contents lists available at ScienceDirect

Renewable and Sustainable Energy Reviews

journal homepage: www.elsevier.com/locate/rser

An overview of global ocean wind energy resource evaluations

Chong Wei Zheng a,b,c,n, Chong Yin Li a,b, Jing Pan b, Ming Yang Liu a, Lin Lin Xia a
a
College of Meteorology and Oceanography, People's Liberation Army University of Science and Technology, Nanjing 211101, China
b
National Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics (LASG), Institute of Atmospheric Physics, the
Chinese Academy of Sciences, Beijing 100029, China
c
Dalian Naval Academy, Dalian 116018, China

art ic l e i nf o a b s t r a c t

Article history: With the rapid development of human society, the demand for energy has accordingly increased, and
Received 13 January 2015 along with this increasingly serious energy and environmental crises have developed. Many countries
Received in revised form have been focusing on new energy resources to combat these crises, and offshore wind energy resources
29 May 2015
are especially attractive; they are safe, non-polluting, renewable, and widely distributed with large
Accepted 17 September 2015
Available online 10 November 2015
reserves, which has made them become the focus of developed countries. However, the distribution of
wind energy has strong regional and seasonal differences, which determines the success and efficiency of
Keywords: wind energy developments. Therefore, there is a clear need for “resource evaluation and planning in
Offshore wind energy advance” in the wind energy development. Previous research has made a great contribution to the
Resource evaluation
evaluation of offshore wind energy resources, mostly through analysis of the climatic characteristics of
Short-term forecasting
wind energy. In the actual development process of wind energy resources, these analyses of the climatic
Medium- to long-term prediction
Wind energy development index characteristics of wind energy provide a reference for site selection. However, after constructing wind
Integrated application system farms, to aid their operation, there needs to be a more comprehensive understanding of other factors,
such as the short-term forecasting and medium- to long-term predictions of wind energy. This paper
reviews the research progress of the wind energy resource evaluations, and then considers where future
research needs to focus, for the evaluation of wind energy resources. This mainly includes further ana-
lyses of the climatic characteristics of wind energy, short-term forecasting, medium- to long-term pre-
dictions, early disaster warning systems, the establishment of a wind energy development index (WEDI)
and an integrated application system, in hope of providing a reference for offshore wind power gen-
eration, seawater desalination and other wind energy resource developments, and accelerating the
industrialization and utilization of offshore wind energy. Doing this will alleviate the energy and
environmental crises, and promote the sustainable development of human society.
& 2015 Elsevier Ltd. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1241
2. Comparison between offshore and onshore wind energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1241
2.1. Common advantages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1241
2.2. Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1242
3. Research progress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1242
3.1. Offshore wind energy resource evaluations based on surface observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1242
3.2. Application of satellite observational data into offshore wind energy resource evaluations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1244
3.3. Application of numerical simulation methods into offshore wind energy resource evaluations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1245
3.4. Application of reanalysis data into offshore wind energy resource evaluations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1245
4. Prospects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
4.1. Further research into climatic characteristics of wind energy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1246
4.2. Short-term forecasting of wind energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1247
4.3. Medium- and long-term predictions of wind energy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1248

n
Correspondence to: No. 667, Jiefang Road, Zhongshan District of Dalian City, Dalian 116041, China. Tel.: þ86 18640814027.
E-mail address: [email protected] (C.W. Zheng).

http://dx.doi.org/10.1016/j.rser.2015.09.063
1364-0321/& 2015 Elsevier Ltd. All rights reserved.
 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251 1241

4.4. Early warning systems for the prevention and reduction of natural disasters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1249
4.5. WEDI and integrated application systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1249
5. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1249
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1249
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1249

1. Introduction Coastal areas around the world tend to have advanced economies
and, as such, are often also centers of high demand for electric
The environmental crises have been attracting much attention power; therefore, dependent on the local conditions, exploiting
in recent years. With shortages of conventional energy, such as and making full use of the offshore wind energy in these areas,
coal and oil, predicted in the near future, humans have been could effectively alleviate the energy crisis and promote sustain-
focusing on new energy resources, researching which resource is able development. Remote islands and deep sea areas have a
the best for coping with climatic change and the shortages of particularly urgent demand for electricity. Taking advantage of the
conventional energy; this has been a common strategy, adopted by marine wind resources and developing offshore wind energy in
many countries. There has been a gradual move towards the these areas could not only solve the power dilemma, but will also
industrialization and large-scale utilization of solar and onshore protect the environment of these ecologically sensitive islands,
wind energy resources, but their use is restricted severely by their and avoid destruction brought about by diesel power generation
uneven spatial distributions. Nuclear energy can provide sig-
(e.g. pollution).
nificant amounts of energy, but also poses a potentially big threat
Onshore wind power generation technologies have been
to human life.
established worldwide, but the technology for offshore wind
Many developed countries have encouraged the development
power generation has only really been developed in a few Eur-
of new energy resources by legislating, as well as reducing or
opean countries [15,16]. In 2014, the cumulative capacity of global
exempting tax, amongst other measures [1]. For example, in order
offshore wind power rose to 8759 MW; the added capacity in that
to support the production and use of renewable energy, Turkey
year was 1713 MW. More than 91% of all the offshore wind
implemented a new law in 2010 [2], within which the principles of
installations worldwide are in European waters, particularly in the
“Wind Power Plant Supporting Mechanisms”, such as price, time
and payments were determined. Only 1% of the licensing costs North Sea (5094 MW: 63.3%) [7]. Thus, the offshore wind po-
during the application process are paid by corporate entities in wer generation potential in many countries is still basically
Turkey, and these entities do not pay annual licensing costs for the undeveloped.
first 8 years [3]. Past renewable energy policies in Japan have had a The distribution of wind energy resources has strong regional
weak market focus on wind resources, so there has been no and seasonal differences. Therefore, for the large-scale develop-
increase in the wind energy share in Japan. As a result, only 1% of ment of wind energy resources, there needs to be clear resource
the global total installation of wind energy has occurred in Japan, evaluation and planning in advance. To realize the ordered and
which is far behind the Unites States, China, Germany, Spain, and efficient exploitation of wind energy resources, there is a need for
many other countries [4]. Within the European Union (EU), each comprehensive evaluations of wind energy resources, along with
member state is supposed to reach a mandatory 20% share of the creation of strategic plans for wind power development and
renewable resources in their total energy consumption by 2020 the construction of associated power networks, through detailed
[5]. In the United Kingdom, wind power is recognized to be the investigations of the wind energy resources. This paper reviews
main renewable energy resource that will enable the country to the research progress and presents some prospects for offshore
achieve the European Union 2020 renewable energy targets. Cur- wind energy resource evaluations, in the hope of making a con-
rently, over 50% of renewable power in the United Kingdom is tribution to alleviate the energy and environmental crises.
generated from onshore wind, with a large number of offshore
wind projects in development [6].
The advantages of offshore wind energy have made it become
2. Comparison between offshore and onshore wind energy
extremely attractive as a potential resource. These advantages
include that it is a safe, non-polluting, renewable resource, with
2.1. Common advantages
large reserves, across a wide distribution, and that its development
will not take up land resources. The biggest challenge for utilizing
There are some common advantages between offshore and
offshore wind resources is the relatively high cost of development.
onshore wind energy resources. Firstly compared with conven-
However, according to a study commissioned by the European
tional energy resources (coal, oil, natural gas, etc.), both onshore
Wind Energy Association (EWEA) in 2015, with the rapid devel-
and offshore wind energy resources are clean, non-polluting,
opment of science and technology, offshore wind costs could be
reduced to EUR 100/MW h by 2020 and EUR 90/MW h by 2030 [7]. renewable and widely distributed. The utilization of wind energy
Junginger and Faaij [8] and Blanco [9] have shown that despite the does not produce toxic or greenhouse gases, which benefits the
recent increase in the capital costs of wind power generation, the environment. Secondly, compared with nuclear energy, wind
long-term trends have indicated a substantial reduction, due to the energy is safer. Although it has the advantages of high efficiency,
rapid development of technology. Considering the costs of power long functional time-scales and high energy outputs, nuclear
generation, environmental pollution [10], wind energy conversion, energy comes with a high risk and can result in environmental and
and other external costs synthetically [11,12], wind power will humanitarian disasters, such as the 1986 Soviet Chernobyl nuclear
become a significant competitor in the energy market. disaster, caused by operator error, and the 2011 Japanese
Wind power generation is the main way to develop and utilize Fukushima nuclear leakage, caused by a tsunami. Finally, solar
wind energy resources, which are also used widely in desalination, energy can only be generated during the daytime, but wind energy
navigation, irrigation, wind-heating and other projects [13,14]. can be generated round-the-clock.
1242 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251

Nomenclature JJA June, July, August

MCP Measure-Correlate-Predict
ANN Artificial Neural Network NASA National Aeronautics and Space Administration
ARIMA autoregressive integrated moving average NREL National Renewable Energy Laboratory
ARPS Advanced Regional Prediction System PO.DAAC Physical Oceanography Data Active Archive Center
CCMP Cross-Calibrated, Multi-Platform RAN Resource Allocating Network
CLLJ Caribbean low-level jet SAR Synthetic Aperture Radar
CMIP5 fifth phase of Coupled Model Intercomparison Project SVM support vector machine
DJF December, January, February WAsP Wind Atlas Analysis and Application Program
DOE Department of Energy of the United State WEDI Wind Energy Development Index
EEMD Ensemble Empirical Mode Decomposition WEST Wind Energy Simulating Toolkit
EWEA European Wind Energy Association WPD wind power density
EU European Union WPPT Wind Power Prediction Tool
GIS Geographic Information System WRF Weather Research and Forecasting Model

2.2. Differences In 1980, the United States generated a distribution diagram of its
wind energy resources using surface wind data from 975
Compared with offshore wind energy, onshore wind energy has meteorology stations, and in 1986 a further 270 meteorology
several advantages. It is easier to install wind energy generators stations were added to the data set, to generate a more compre-
onshore, which inherently means that the construction costs are hensive distribution diagram of the onshore wind energy resour-
much lower; the cost of constructing a wind farm offshore is 1.5–2 ces. The Risoe National Laboratory in Denmark collected weather
times greater than that of constructing one onshore. In addition, observational data from 220 meteorology stations across 12 Eur-
the maintenance and repair of offshore wind farms are more opean countries, and, by incorporating the impacts of the build-
challenging, due to their location [17]. The application of onshore ings, the terrain surface conditions, and surface roughness at all
wind energy technology is well established, and it is easy to the weather stations, a distribution diagram of European onshore
incorporate the electricity generated into local networks. It is also wind energy densities was generated. In India, observational data
easy to obtain observational data onshore, which is beneficial for from 570 anemometer towers, installed since 1987, were used by
evaluations of the wind energy resource, and can improve the Indian Energy Consultants Ltd. to generate a distribution diagram
precision of simulated data, to optimize the results of resource of Indian local onshore wind energy resources. The China
evaluations. Meteorological Administration conducted wind energy resource
There are also several advantages to the development of off- censuses in the 1980s and 1990s, respectively, which analyzed the
shore, compared with onshore, wind energy. As a resource, offshore historical wind data using statistical methods, and from these they
wind energy is attractive, because offshore winds are higher in generated a distribution diagram of Chinese onshore wind energy
velocity, more reliable and more consistent [18]; the sea surface resources.
wind speeds 10 km off the coast are usually 25% greater than those Due to the difficulties of collecting ocean observational data,
on land, and offshore wind energy resources can generally be uti- there is a great shortage of data, so there are relatively fewer
lized to generate electricity for 2–3 times longer, during the same evaluations of offshore wind energy resources and the research
time period, than those onshore [19,20]. The fact that offshore wind scope is relatively small. Offshore wind energy resource evalua-
energy developments are offshore means less use of land resources, tions have gone through the following stages: (1) offshore wind
no human migration as a result of development, and a lower public energy resource evaluations based on limited surface observa-
visual impact. The offshore wind turbines are inherently far away tional data; (2) application of satellite observational data into the
from the public, so that the issue of noise, which generally results in offshore wind energy resource evaluations; (3) application of
public complaints with onshore wind developments, can be ignored numerical simulation methods into the offshore wind energy
[17,21]. Due to the low roughness and small friction on the sea resource evaluations; and (4) application of reanalysis data into
surface, the variation of wind speed with height is small, which offshore wind energy resource evaluations. These stages are dis-
reduces the cost of wind turbines because they do not need to be as
cussed in detail below.
high as those on land. In addition, with the low turbulence intensity
and friction on the sea surface, the fatigue loads of the wind tur-
3.1. Offshore wind energy resource evaluations based on surface
bines, caused by changing winds is reduced, which extends the life
observations
of the wind turbines; the foundations are also reusable, with a
design life that can be up to 50 years [22]. Finally, with the large
Towards the end of the last century, scientific researchers
area of sea adjacent to land masses around the world, offshore wind
began to focus on offshore wind energy; despite the great shortage
energy resources have an effectively limitless development space.
of oceanic data, their research made a great contribution to cur-
Although it has not been developed as much as onshore wind
rent knowledge of offshore wind energy, through exploring the
energy, offshore wind power generation has been established in
certain European countries, so the technology is available. Obser- methods of evaluating offshore wind energy resources. Research-
vational data at sea has also become increasingly abundant, and ers utilized the extremely limited site observational data and ship
simulations have improved, which is beneficial for the evaluations reported data to evaluate the offshore wind energy resources at
of wind energy resources across large-scale marine areas. single stations or small-scale coastal sea areas. From these studies,
near shore wind energy resource development started in Eur-
3. Research progress opean and American developed countries.
Using surface observational data, Youm et al. [23] analyzed the
Benefiting from abundant observational data, there have been distribution characteristics of wind energy resources in the
numerous research projects into onshore wind energy resources. northern coastal sea area of Senegal, and found that the annual
 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251 1243

average wind speed and wind power density (WPD) in the area anemometer towers (Fig. 2). Their results showed that wind
was, 3.8 m/s and 158 W/m2, respectively. In 2004, Musial and energy in the coastal area of Fujian Province was abundant, and
Butterfield [24] evaluated the development state of the United the maximum energy resources were located in the coastal area,
States offshore wind energy resources, and, based on their study, from central and southern Fujian to the south of Quanzhou. The
Michael [25] examined the wind energy resources in the seas annual effective WPD was calculated to be 517–930 W/m2, and
surrounding the United States; Michael [25] presented a detailed more than 930 W/m2 in Pingtan Island. The wind energy resources
analysis of the wind power around the coastline, which provides a were also found to be very rich in the town of Chichu, in Zhangpu
scientific basis for the development of offshore wind energy county, in southern Fujian, where the annual effective WPD was
resources in the United States (Fig. 1). Kucukali and Dinçkal [26] more than 510 W/m2. In the coastal areas of Fujian, there were
analyzed the wind energy resources at Izmit, located in the Wes- 7837.3 annual average effective wind energy hours, that is, 89.4%
tern Black Sea (a coastal region of Turkey), based on data from of the time wind energy power generation would be effective. The
measurement masts 50 m high, covering the period 06/2008–06/
development of wind energy resources for power generation in
2009. González-Longatt et al. [27] created a wind resource atlas of
the coastal area from central and southern Fujian to the south of
Venezuela, based on 32 weather stations. Their results showed
Quanzhou, was classified as “best”, with the remaining areas
that the best wind energy resources were located in the northern
classified as between "better" and "good".
coastal area of Venezuela. In 2005, Archer and Jacobson [28] were
Surface observational data remains important for wind energy
the first to perform an evaluation of global onshore and coastal
evaluations in current research. Based on observational wind data,
wind power, at heights of 10 m and 80 m above sea level.
Oh et al. [21] analyzed the wind energy potential at a demon-
China and other Asian countries started researching into their
stration offshore wind farm in the sea to the southwest of the
offshore wind energy resources after the European and American
countries, but are making rapid progress on their evaluations Korean Peninsula. They found that the wind energy potential of
[29,,30]. In China, Chen et al. [31] analyzed the distribution char- the demonstration wind farm was Wind Class 3, in terms of the
acteristics of wind resources in Lianyungang and its coastal sea energy density. Since the major wind direction was distinct (the
area, using 30 years of meteorological data. Their study showed prevailing wind was northwest), they showed that a wind farm
that the area had abundant and relatively stable wind energy layout that takes into account the prevailing wind direction would
resources, with broad prospects for wind energy resource be the most efficient. Luong [33] have given an overview of wind
development. energy potential, its current application and future development
A comprehensive study was performed by Wen et al. [32], who in Vietnam. They identified the major barriers that need to be
analyzed the wind energy resources in the coastal areas of Fujian addressed for the future development of wind energy in Vietnam;
Province, using 70 m height observational data from 18 these include technical and technological issues, under developed

Fig. 1. Wind resource assessment map of the state of Massachusetts, including offshore areas. Wind speed and wind power density are shown at a 50 m height. This map
was created by the National Renewable Energy Laboratory for the Department of Energy. Source: Windpoweramerica.gov.
1244 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251

infrastructure, a lack of financial resources and services, and wind energy resource evaluations. The final results showed that it
institutional issues. was feasible to evaluate the offshore wind energy resources using
There is a large deficiency in surface observational data for the satellite-derived wind speed distributions. To provide further
wind energy resource evaluations. Due to the relatively backward verification, using the ocean surface wind speed data derived from
measuring methods and the shortage of data, scientific researchers SAR, Charlotte et al. [34] studied wind energy resources over the
can only evaluate the offshore wind energy resources at a single Baltic Sea. They compared the wind speed data derived from SAR
station or small-scale coastal sea areas. However, extensive wind and surface observational wind speed data, and found that the
energy resource evaluations over large-scale areas are more useful former had high accuracy in the Baltic Sea, in which the wind
for selecting the regions with the best energy resources available. energy density was 300–800 W/m2.
For the construction of wind energy developments, the height The National Aeronautics and Space Administration (NASA)
of the wind turbines is directly related to the input costs. Conse- [35] used wind speed data derived from QuikSCAT satellite data to
quently, comparing the results of wind energy resource evalua- generate the distribution of WPDs in the oceans globally, during
tions at different height levels is also useful, to reduce the cost of JJA (June, July, August) and DJF (December, January, February)
the development of wind farms. Wind energy resources can gen- (Fig. 3). During JJA, the WPDs in the southern ocean were sig-
erally only be studied at a fixed level, when based on the limited nificantly greater than those in the northern oceans. In the oceans
surface observational data; for instance, ship reported data is of the Northern Hemisphere, the sea around Somalia had sig-
generally located 10 m above the ocean surface, and buoy data is nificantly high WPDs, which were probably caused by the strong
also generally at a fixed altitude. southwest monsoon. During DJF, due to the powerful cold air in
A further issue is that the installment time of anemometer the Northern Hemisphere, the WPDs in the oceans of the Northern
towers is often different, which leads to the disunity of the period Hemisphere were significantly greater than in the Southern
of wind data, and this in turn affects the accuracy of the wind Hemisphere oceans. Thus, they found that, in general, the WPDs in
energy resource evaluations. Installing anemometer towers also the winter hemisphere were greater than those in the summer
requires manpower and material resources; this makes it impos- hemisphere.
sible to establish an intensive observation network on a large-scale Liu et al. [36] also analyzed the global ocean WPDs in winter
and conduct regular observations, like with onshore weather sta- and summer, based on the QuikSCAT Level-2 data at a 12.5 km
tions. In accordance with all the issues highlighted, it is infeasible resolution, which were obtained from the Physical Oceanography
to evaluate regional wind energy resources solely on the basis of Data Active Archive Center (PO.DAAC). A good agreement between
surface observational data. the results of Liu et al. [36] and NASA [35] was found. Capps and
Zender [37,38] evaluated the global ocean wind power potential
3.2. Application of satellite observational data into offshore wind and global ocean 80 m wind energy, accounting for surface layer
energy resource evaluations stability. They demonstrated that during 2000–2006, available
80 m high WPDs, between 100 and 500 W/m2, existed over
With the progress of ocean measuring methods, more and approximately 50% of the ice-free ocean surface area.
more satellite data have been used to study marine resources. Satellite observational data has also been used widely to assess
During 2004–2006, the Risoe National Laboratory in Denmark and local wind energy. Jiang et al., [39] presented a distribution study of
offshore wind power in China using QuikSCAT Level-2 satellite
several other research institutions conducted the SAT-WIND
measurements, at a 0.5° horizontal resolution, over the past 9 years.
research program and verified the possibility of applying
They found that the coastal region along Fujian Province had the best
satellite-derived data, including surface wind distribution data
wind resources, compared with other offshore regions of China; this
derived from passive microwave remote sensors, altimeters,
Scatterometers and Synthetic Aperture Radars (SAR), into offshore

Fig. 2. Wind power densities at a height of 70 m along the coastal areas of Fujian Fig. 3. Distribution of wind power density (W/m2) in June–August (a) and
Province, China. December–February (b), derived by NASA.
 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251 1245

agreed well with the results presented by Wen et al. [32]. Based on a mesoscale meteorological numerical model, usually the Weather
more than 1000 satellite SAR images, Hasager et al. [40] found a WPD Research and Forecasting Model (WRF) or Mesoscale Model 5, and
range from 300 to 800 W/m2 for the 14 existing and 42 planned a complex terrain dynamical diagnosis model, usually the Cali-
wind farms in the Baltic Sea. Based on the synergetic use of Envisat fornia Meteorological Model or The Advanced Regional Prediction
ASAR, ASCAT and QuikSCAT data, Hasager et al. [41] identified that System (ARPS). Carvalho et al. [46] forced the WRF model with
significant coastal wind speed gradients were identified with SAR different initial and boundary conditions (NCEP-R2, ERA-Interim,
data. In regard of eight new offshore wind farm areas in Denmark, NCEP-CFSR, NASA-MERRA, NCEP-FNL and NCEP-GFS) and con-
they showed that the spatial variability of the mean energy density, ducted ocean surface wind simulations to assess which one of the
based on the SAR data, ranged from 347 W/m2 in Sejerøbugten to data sets provided the most accurate ocean surface wind simula-
514 W/m2 at Horns Rev 3. tion and offshore wind energy estimates. The simulation results of
Compared with the conventional surface observations, satellite the models driven by the NCEP-FNL and NCEP-GFS analyses were
data can cover a wider spatial range and allows more compre- better than those produced using the NCEP-CFSR and NASA-
hensive evaluations of the wind energy resources offshore, but the MERRA reanalyses.
data also come with some deficiencies. Due to the limited number It is feasible to apply numerical models to simulate wind in a
of satellites and orbits, satellite data have some deficiencies in region, which has a number of advantages. Firstly, simulation
time synchronization and spatial resolution; namely they cannot results can be used as auxiliary information to wind energy
cover large-scale areas at the same observation time point. In resource surveys, and so can contribute to evaluations of wind
addition, while there is a clear need to compare evaluations of energy resources in a particular region for which there are no
wind energy resources at different levels (to reduce the develop- wind observational data. Simulated results can especially make up
ment costs of wind energy resources, as already discussed), for the deficiency of oceanic surface observational data, which can
satellite-derived data cannot reflect variations in the wind field at guide the site selection of wind farms. The numerical simulation
different levels. Therefore, the data cannot fully meet the needs of methods can also generate wind energy data of a higher spatial
assessments for wind power development. resolution, which enables the accurate determination of the area
of available wind energy resources and the available wind energy
3.3. Application of numerical simulation methods into offshore wind reserves at the height of wind turbines. A key drawback of
energy resource evaluations observational data (both surface and satellite) is that the altitude is
usually fixed, which limits the wind energy resource evaluations
With the rapid development of computer technology, more and to some extent (as discussed above); by applying numerical
more numerical models have been used for wind energy resource simulation methods in wind energy resource evaluations, the wind
evaluations; such models have enabled the gradual realization of energy parameters for all grid points on the three-dimensional
fine-scale evaluations of wind energy resources in local sea areas, space within the research area can be obtained, and so a more
as well as the study of regions for which observational data is comprehensive evaluation of the wind energy resources can take
lacking. In general, the numerical simulations have been found to place [47]. However, there are some deficiencies of numerical
produce reliable data [42,43]. A number of advanced wind energy modeling. Namely, the influences of topography, among other
simulation software programs have been developed in American factors, mean that the simulation data are not very good in some
and in European countries. For example, in the 1980s and 1990s, complex regions, which needs further improvement.
the Risoe National Laboratory in Denmark developed resource
analysis software for the micro-site selection of wind farms, called 3.4. Application of reanalysis data into offshore wind energy
the ‘Wind Atlas Analysis and Application Program’ (WAsP). The resource evaluations
United States TrueWind Solutions Company [44] developed the
MesoMap and SiteWind wind resource evaluation systems, which Thanks to the rapid development of technology to produce
have been applied in wind energy resource evaluations across observational data and model simulations, more and more rea-
more than 20 countries and regions. The SiteWind system can nalysis data have been applied widely for wind energy resource
correct wind maps using real wind field data, which can greatly evaluations. For example, through statistical analysis of 10 m
reduce model errors. The Canadian Meteorological Bureau estab- NCEP/DOE reanalysis wind data, during 1979–2010, Chadee and
lished the Wind Energy Simulating Toolkit (WEST) numerical Clarke [48] derived a regional annual wind resource map, which
model [45], by combining the mesoscale model MC2 and the showed that the Caribbean low-level jet (CLLJ) region was an area
small-scale model Ms-micro. with superb WPDs, of 400–600 W/m2. In addition, they identified
Currently, researchers mostly apply integrated model systems the eastern Caribbean and the Netherland Antilles as locations
to evaluate wind energy resources; such systems are composed of where there were excellent wind energy resources (300–400

Fig. 4. Wind power density at 10 m (a) and 100 m (b) height over the South China Sea, unit: W/m2.
1246 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251

W/m2), and the Greater Antilles and the Bahamas as areas with between 3 and 25 m/s are regarded as ‘available’), stability of the
very good wind resources (200–300 W/m2). WPD (including the coefficient of variation, monthly variability
Based on multiple satellite data (SAR and Scatterometer ASCAT index and seasonal variability index), long-term trends of the
images) and WRF modeling data, Chang et al. [49] analyzed the WPD, and resource storage (including total storage per unit area,
offshore wind resources over the South China Sea (Fig. 4). Their effective storage, and exploitable storage), amongst other aspects.
results showed that the South China Sea was rich in wind energy, According to the technical standards for the development of wind
at both 10 m and 100 m heights. The areas with the best WPDs energy resources, created by the Department of Energy of the
were located in the Beibu Gulf and waters to the east of Hainan United States [54] the and National Development and Reform
Island. Over a 30-year period (1980–2010), daily averaged high Commission of China [55], Zheng and Pan [53] produced a class
altitude wind data was extracted using the NCEP/DOE AMIP-II division of global ocean wind energy resources (Table 1; Fig. 6).
Reanalysis (Reanalysis-2) by Ban et al. [50], who analyzed the high
altitude wind energy in the southeast of Europe. Their results
showed the WPDs at an altitude of 2.5 km, to identify the best 4. Prospects
places to position ground stations in the southeast of Europe.
However, the highest WPDs were found offshore, with increasing Previous research has made a great contribution to scientific
tendency towards the south, between Italy and Greece. understanding of the offshore wind energy resources, but the
Previous research has made a great contribution to the scientific research has mostly concentrated on analyzing the climatic char-
understanding of WPDs and the stability of wind energy, but there acteristics of wind energy. In terms of the actual development of
has been little research on the class division of ocean wind energy the wind energy resources, these advanced analyses of the climatic
resources globally. In October 2005, the United States Department characteristics of wind energy can provide a reference for site
of Energy, National Renewable Energy Laboratory (NREL) [51] pre- selection. However, for the operational working of wind farms,
sented a wind power class map of the global oceans, based on different types of information are needed and the focus of wind
QuilSCAT wind data (Fig. 5). The study showed that satellite-derived energy resource evaluations in the future should not only include
estimates of wind resources in near shore, coastal, and island areas further research into the climatic characteristics of wind energy,
do not always agree with observational data. but also consider: (1) short-term forecasting of wind energy;
In 2014, using Cross-Calibrated, Multi-Platform (CCMP) wind (2) medium- to long-term predictions of wind energy; (3) early
data, Zheng and Pan [53] analyzed the global ocean wind energy warning systems for the prevention and reduction of natural dis-
resources. The CCMP data possess the characteristics of a long- asters to wind farms; and (4) development of a Wind Energy
term series, high spatial resolution and high precision [52]. Zheng Development Index (WEDI) and integrated application system.
and Pan [53] comprehensively considered the magnitude of the
WPD, energy level occurrence (wind power can generally be uti- 4.1. Further research into climatic characteristics of wind energy
lized when the WPD is above 100 W/m2, and energy-rich regions
have WPDs above 200 W/m2; these levels are classified according Systematic analysis of the climatic characteristics of wind energy
to their ‘energy level occurrence’), gale occurrence (strong wind includes: the magnitude of the WPD, energy level occurrence, gale
occurrence), occurrence of the effective wind speed (wind speeds occurrence, the occurrence of effective wind speeds, the stability

Fig. 5. Wind power class in the global ocean.

 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251 1247

and long-term trends of WPD, resource storage, and other aspects, this needs to be especially considered in the future. Assessments of
hereafter collectively referred to as the “wind energy climatic fac- the resource grade division at the large-scale ocean can provide
tors”. In this study, a “wind energy rose” (as opposed to a wind rose) reference for large-scale site selection, but a high-resolution grade
was created, which summarizes the wind energy climatic factors at division for small-scale ocean areas is also required, to provide
a specific site (rather than just wind speed and direction; Fig. 7). reference for microscopic site selection. This work could be com-
From the wind energy rose, it can clearly be seen that the gross pleted by the integration of observation, satellite and simulation
energy at station 42001 (25.89°N, 89.66°W) depended on the wind data. These ideas need to be realized during future research
direction and magnitude of the WPD. In February, the wind energy endeavors.
source mainly came from the SE–SSE direction, followed by a
northerly direction, but the large levels of wind energy came from 4.2. Short-term forecasting of wind energy
the NNW direction. In May, the wind energy mostly came from the
E–SE direction, but was largest when the wind came from the SE or Short-term forecasting of wind energy can provide reference
NW direction. In August, the wind energy mainly came from the for wind farm operators, which can then improve the collection
NE–SE direction, especially the ESE direction. In November, the and conversion efficiency of wind energy and provide an accurate
wind energy mainly came from the NE–E direction, and when basis for regulating short-term electricity generation. The typical
coming from the NE and NNE directions, it was strongest. And this methods for predicting wind energy include numerical weather
“wind energy rose” can also be used in the short term forecasting of prediction and wind forecasting, ensemble forecasting, physical
wind energy resource. methods, statistical and learning approach methods, benchmark-
In terms of the wind energy grade division, on the basis of the ing and uncertainty analysis, and hybrid methods, amongst others
traditional macroscopic annual resource grade division, in the [56,57]. Shi et al. [58] compared hybrid forecasting methods using
above analysis, this study emphasizes the grade division in dif- three major single prediction models, that is, the autoregressive
ferent months and the microscopic grade division. Although Zheng integrated moving average (ARIMA) model, the artificial neural
and Pan [53] assessed the wind energy grade division in oceans network (ANN) models, and the support vector machine (SVM), to
globally, by comprehensively considering the wind energy climatic predict hourly wind speeds and power generation. Their results
factors, their results mostly concentrated on the annual resource showed that the hybrid methodology was always superior to the
grade division. However, in the actual site selection process, the single models.
resource grade division during different months is important, and Han et al. [59] proposed a forecasting model for wind power
based on the Resource Allocating Network (RAN). Compared with
Table 1 observational data, their results showed that the self-
Standards for wind power classification [54,55]. reconstruction structure of the RAN made it possible to forecast
with a look-ahead horizon of up to more than one day, for different
Wind Annual Annual average wind Significant Wind
power average power density (W/m2) interval (h) energy
types of wind turbines at different locations. Tascikaraoglu and
class wind speed division Uzunoglu [60] outlined the combined forecasting approaches and
(m/s) Method 1 Method 2 presented an up-to date annotated bibliography of the wind
forecasting literature; they also proposed combined approaches
1 0.0–4.4 o100 o 50 o 2000 Indigent
composed of several spatial correlation models, which have not
area
2 4.4–5.1 100–150 50–150 2000–3000 Available been included in this study. The evaluation and comparison of
area these models should be considered in a future study, because of
3 5.1–5.6 150–200 150–200 3000–5000 Subrich their successful implementation.
area
Wind energy forecasting software has also developed rapidly in
4 5.6–6.0 200–250 200–250 45000 Rich area
5 6.0–6.4 250–300 250–300 recent decades [61]. In the early 1990s, a number of European
6 6.4–7.0 300–400 300–400 countries started to develop wind energy forecasting systems,
7 7.0–9.4 400–1000 400–1000 which were used for their forecasting services [62]. Forecasting
techniques mostly use medium-term prediction models, which

Fig. 6. Grade classification map of the global ocean wind energy resources. Note: a higher grade means a richer resource.
1248 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251

N N
N
NNW NNE NNW NNE

NW NE NW
NW NE
15% 30%
30%

WNW 10% ENE WNW

WNW 75%
20% ENE
ENE
50%
5% 10%
25%
W E W
W
W E
E 1800 - 2000
1600 - 1800
W/m2 W/m
1400 - 1600 2
>=1200 1200
>=1200 - 1400
WSW ESE 1000 - 1200 WSW
WSW ESE
ESE 1000
1500 -- 1200
1500
>=1000
1000 1800
1800
1200
800 - 1000 800
1500
1200
>=800
1200
800 -- 1000
--1000
1800
1500
1500
600 - 800 600
>=1200
1200
900
>=1200
900
600 -- 800
-1200
1500
1200
800
SW SE 400 - 600 400
>=3
900
800 -- 1200
600
600
400 600
900
1200
900
600
SW
SW SE
200 - 400 200
>=2
2
600
300- 3-- 900
400
300
200 400
600
800
600
400
0 - 200 00 -- 200
>=1
<2
<600 300
400
300
200
S SW S SE S SW S SE
S S
S

N N
NNW NNE NNW NNE

NW NE NW NE
15%
20% 15%
20%
30%

15% 15%
WNW 10% ENE WNW 20%
10% ENE
10% 10%
5% 10%
5%
5% 5%
W E W E

W/m2 W/m2
>=1200
1200 - 1400 >=1200
1200 - 1400
WSW ESE 1000 - 1200 WSW ESE 1000 - 1200
800 - 1000 800 - 1000
600 - 800 600 - 800
SW SE 400 - 600 400 - 600
200 - 400 SW SE
200 - 400
0 - 200 0 - 200
S SW S SE S SW S SE
S S

Fig. 7. Wind energy rose at station 42001 (NDBC buoy) in February, May, August and November 2013.

nest high-resolution effective area models (or nest higher resolu- Spain), and HIRPOM (developed in Ireland and Denmark), are
tion regional models), and generation capacity models, to forecast proposed combination models that could also be applied for wind
the generation capacity of wind farms. For instance, the Denmark energy forecasting [69].
Predictor forecasting system has been applied for short-term wind
energy business operations in Denmark, Spain, Ireland and Ger-
many, while the Wind Power Prediction Tool (WPPT) is also used 4.3. Medium- and long-term predictions of wind energy
for wind energy forecasting business operations in some
European areas. The medium- to long-term trends of wind energy resources
In the mid-1990s, the United States True Wind Solutions are important for the long-term planning of wind energy devel-
Company began their commercial wind energy forecasting service, opment and the power market [70,71]. Three methods are
through which they developed the wind energy forecast software usually applied for medium and long-term predictions of wind
‘eWind’, which consists of high-resolution mesoscale meteor- energy. The first method employed is the use of the long-term
ological numerical models and statistical models that have been forecasting wind field, such as the fifth phase of the Coupled
used to forecast the wind field and generation capacities. eWind Model Intercomparison Project (CMIP5), to produce long-term
and Predictor are currently both used for the forecasting service of wind energy resource predictions. Although not widely used
two large wind farms in California [63,64]. In October 2002, the
currently, this method will be popular in the future. Wang et al.
European Commission started the program ANEMOS, which aims
[72] analyzed the changes in global ocean wave heights, pro-
to develop advanced forecasting models that are better than the
jected with multi-model CMIP5 simulations. The second method
existing ones, with emphasis on the condition of complex terrains
is to use some regular and strong astronomical and earth factors,
and extreme weather in the forecasts, as well as the development
to assist the medium and long-term prediction of wind energy.
of nearshore wind energy forecasting [65,66]. By combining MC2
and WAsP, the Canadian wind energy resource numerical evalua- The third method uses the ANN, SVM, ARIMA, RAN [73],
tion and forecasting software, WEST, generates an atlas of wind Measure-Correlate-Predict (MCP), or Ensemble Empirical Mode
energy, with a resolution of 100–200 m [67,68]. In addition, Pre- Decomposition (EEMD) [74] methods to produce medium and
viento (developed in Germany), LocalPred and RegioPred (from long-term predictions of wind energy.
 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251 1249

4.4. Early warning systems for the prevention and reduction of The combined exploitation of wave and offshore wind energies
natural disasters has also become an attractive research topic recently [80]. The
ocean is the cradle of human life, and a huge treasure of resources.
For the development of offshore wind power, apart from wind Completing the resource evaluations, along with the rational
energy evaluations, it is necessary to accurately analyze the development of marine energy resources, will effectively alleviate
extreme wind speeds [75]. Disaster reduction and provision is the energy and environmental crises that we face today.
especially important in remote islands, because they are far away
from the continent. In the engineering design of wind farms, it is
necessary to calculate the gale occurrence (occurrence of strong 5. Conclusion
wind speeds) and extreme wind speeds, to enable the provision of
equipment that will cope with such high wind speeds [76,77]. In modern society, humans are being significantly affected by
With the temporary provision of equipment, it is highly possible to energy and environmental crises. It is undoubtedly an ideal choice
cause dangers which could have been avoided. to develop renewable resources, to cope with the energy and
Calculating the annual gale occurrence and extreme wind environment crises; this needs to be done according to local
speeds provides a reference for engineering design and can extend conditions, thus promoting the sustainable development of human
the lifespan of wind turbines. Accurate calculations of gale society. Offshore wind energy is attractive as a renewable non-
occurrence and extreme wind speeds in different months must be polluting, resource with large reserves, and a wide distribution.
made, because these factors will affect the actual construction and However, its instability makes it a complex resource to exploit and
operational working processes. For example, during medium- to utilize. In order to develop offshore wind energy resources effi-
long-term planning, it is necessary to use the extreme values as a ciently and properly, there need to be comprehensive evaluations
reference to determine whether to reinforce the equipment of of the resources, such as analyses of the climatic characteristics
wind farms, in case of adverse sea conditions, for certain months. and grade divisions for different months of the year, short-term
The annual largest gale occurrence and largest extreme values can forecasting and medium- to long-term predictions of wind energy,
be used, however, their use will cause a great amount of avoidable early warning systems to prevent and reduce damage from natural
waste, due to the high cost of offshore construction, if construction disasters, WEDI and integrated application systems.
could occur in a month when the extreme events are very unlikely Just as a coin has two sides, offshore wind energy has both
to occur. advantages and disadvantages. Leung and Yang [17] found that
offshore wind turbines impact on marine animals, like dab, sal-
4.5. WEDI and integrated application systems mon, and migratory birds, and can cause changes to the climate.
Tabassum-Abbasi et al. [81] showed that, with the rapidly
The construction of offshore wind farms have many constraints, increasing utilization of wind energy, environmental concerns also
such as the water depth, the distance from the coast, sea-route, rise, such as the visual, noise and wildlife impacts. Saidur et al. [82]
fishing areas, the distance from the nearest port, the seabed con- have also shown that wind turbines have significant impacts on
ditions, the impacts of waves, sea level, the occurrence of wildlife, such as birds, bats, and raptors. In the development of
typhoons or earthquakes, and military facilities, amongst others. offshore wind energy, the impacts of utilizing the wind energy on
Lee [78] described eight key factors, which were the wind power, the environment, in addition to the wind energy evaluations
earthquake occurrence, the land use, shore type, typhoon occur- themselves, should be carefully considered, before any wind farm
rence, the distance to shoreline, flight safety and the water depth, is constructed, or a decision is made. Only the rational and sci-
that should be considered during the process of offshore wind entific exploitation of offshore wind energy resources will pro-
energy resource evaluations. On the basis of the distribution dia- mote the sustainable and harmonious development between
gram of offshore wind energy resources, and by combining these humans and the environment.
restraining factors, the offshore areas suitable for wind farm
developments in the future can be identified [79].
This study proposes the establishment of a WEDI, which Acknowledgments
includes the factors of WPD, occurrence of effective wind speed,
occurrence of WPD 4 100 W/m2, occurrence of WPD 4200 W/m2, This work was supported by the National Key Basic Research
gale occurrence, and water depth, amongst other factors. Each Development Program Astronomy and Earth Factor on the Impact
factors should be weighted differently, according to its importance. of Climate Change (Grant nos. 2015CB453200, 2013CB956200,
For example, the magnitude of the WPD is closely related to the Nos. 2012CB957803, 2010CB950400) and National Nature Science
richness of wind energy, which should occupy a key positive Foundation of China (Grant nos. 41490642, 41275086, 41475070).
weight. In addition, a WPD 4100 W/m2 and a WPD 4200 W/m2,
are needed for the resource to be considered ‘available’ and ‘rich’,
respectively, so the occurrences of these two factors should be References
positively weighted. In contrast, increases in the gale occurrence
and water depth will cause difficulties in the development and [1] Saidur R, Islam MR, Rahim NA, Solangi KH. A review on global wind energy
utilization of the wind resource, so these two factors should be policy. Renew Sustain Energy Rev 2010;14:1744–62.
negatively weighted. [2] Dursun B, Gokcol C. Impacts of the renewable energy law on the develop-
ments of wind energy in Turkey. Renew Sustain Energy Rev 2014;40:318–25.
Building an integrated application system for the development [3] Kaplan YA. Overview of wind energy in the world and assessment of current
of wind energy is also very important, much more important than wind energy policies in Turkey. Renew Sustain Energy Rev 2015;43:562–8.
a platform for wind farm site selection. For example, in order to [4] Mizuno E. Overview of wind energy policy and development in Japan. Renew
Sustain Energy Rev 2014;40:999–1018.
support the decision making process effectively and efficiently, an
[5] Hadžić N, Kozmar H, Tomić M. Offshore renewable energy in the Adriatic Sea
application system should be established, which includes the with respect to the Croatian 2020 energy strategy. Renew Sustain Energy Rev
WEDI, wind energy climatic characteristics, location queries, 2014;40:597–607.
short-term forecasting, medium- to long-term predictions, and [6] Higgins P, Foley A. The evolution of offshore wind power in the United
Kingdom. Renew Sustain Energy Rev 2014;37:599–612.
automated risk early warning systems, amongst other factors, [7] GWEC (Global Wind Energy Council). Global wind report: annual market
based on a Geographic Information System (GIS). update 2014; March 2015.
1250 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251

[8] Junginger M, Faaij A, Turkenburg WC. Cost reduction prospects for offshore [44] Ayotte KW, Davy Robert J, Coppin PA. A simple temporal and spatial analysis
wind farms. Wind Eng 2004;28:97–118. of flow in complex terrain in the context of wind energy modeling. Bound-
[9] Blanco MI. The economics of wind energy. Renew Sustain Energy Rev Layer Meteorol 2001;98:275–95.
2009;13:1372–82. [45] Yu W, Benoit R, Girard C. Wind Energy Simulation Toolkit (WEST): a wind
[10] Varun G, Prakash R, Bhat IK. Energy, economics and environmental impacts of mapping system for use by the wind-energy industry. Wind Eng 2006;30
renewable energy systems. Renew Sustain Energy Rev 2009;13(9):2716–21. (1):15–33.
[11] Diaf S, Notton G. Evaluation of electricity generation and energy cost of wind [46] Carvalho D, Rocha A, Gómez-Gesteira M, Santos CS. Offshore wind energy
energy conversion systems in southern Algeria. Renew Sustain Energy Rev resource simulation forced by different reanalyses: comparison with observed
2013;23:379–90. data in the Iberian Peninsula. Appl Energy 2014;134:57–64.
[12] McKenna R, Hollnaicher S, Fichtner W. Cost-potential curves for onshore wind [47] Lee ME, Kim G, Jeong ST, Ko DH, Kang KS. Assessment of offshore wind energy
energy: a high-resolution analysis for Germany. Appl Energy 2014;115:103–15. at Younggwang in Korea. Renew Sustain Energy Rev 2013;21:131–41.
[13] Markus F, Fredrik M, Fernando D. Feasibility study on wind-powered desali- [48] Chadee XT, Clarke RM. Large-scale wind energy potential of the Caribbean
nation. Desalination 2007;203:463–70. region using near-surface reanalysis data. Renew Sustain Energy Rev
[14] Zheng CW, Pan J, Li JX. Assessing the China Sea wind energy and wave energy 2014;30:45–58.
resources from 1988 to 2009. Ocean Eng 2013;65:39–48. [49] Chang R, Zhu R, Badger M, Hasager CB, Xing XH, Jiang YR. Offshore wind
[15] Ackermann T, Soder L. An overview of wind energy-status 2002. Renew Sus- resources assessment from multiple satellite data and WRF modeling over
tain Energy Rev 2002;6(1–2):67–127. South China Sea. Remote Sens 2014;6:1–21.
[16] Thompson S, Duggirala B. The feasibility of renewable energies at an off-grid [50] Ban M, Perković L, Duić N, Penedo R. Estimating the spatial distribution of high
community in Canada. Renew Sustain Energy Rev 2009;13(9):2740–5. altitude wind energy potential in Southeast Europe. Energy 2013;57:24–9.
[17] Leung DYC, Yang Y. Wind energy development and its environmental impact: [51] NREL. QuikSCAT Annual wind power density at 10m. From/http://en.
a review. Renew Sustain Energy Rev 2012;16:1031–9. openei.org/w/index.php?title=File:QuikSCAT-_Annual_Wind_Power_Den
[18] Kaplan YA. Overview of wind energy in the world and assessment of current sity_at_10m.pdf&page=1S; 2005.
wind energy policies in Turkey. Renew Sustain Energy Rev 2015;43:562–8. [52] Atlas R, Hoffman RN, Ardizzone J, Leidner SM, Jusem JC, Smith DK, Gombos D.
[19] Tambke J, Lange M, Focken U. Forecasting offshore wind speeds above the A cross-calibrated, multiplatform ocean surface wind velocity product for
North Sea. Wind Energy 2005;8:3–16. meteorological and oceanographic applications. Bull Am Meteorol Soc
[20] Wang JZ, Qin SS, Jin SQ, Wu J. Estimation methods review and analysis of 2011;92:157–74.
offshore extreme wind speeds and wind energy resources. Renew Sustain [53] Zheng CW, Pan J. Assessment of the global ocean wind energy resource.
Energy Rev 2015;42:26–42. Renew Sustain Energy Rev 2014;33:382–91.
[21] Oh KY, Kim JY, Lee JK, Ryu MS, Lee JS. An assessment of wind energy potential [54] Wind Energy Resource Atlas of the United States (WERA). the National
at the demonstration offshore wind farm in Korea. Energy 2012;46:555–63. Renewable Energy Laboratory; 1986. Available from.http://rredc.nrel.gov/
[22] Li XY, Yu Z. Developments of offshore wind power. Acta Energiae Sol Sin wind/pubs/atlas.
2004;25(1):78–84. [55] National Development and Reform Commission. Technical requirements for
[23] Youm I, Sarr J, Sall M. Analysis of wind data and wind energy potential along wind energy resource assessment. Beijing: Chinese Standard Press; 2004 GB/T
the northern coast of Senegal. Renew Energy 2005;8:95–108. 18710-2002.
[24] Musial W, Butterfield S. Future for offshore wind energy in the United States. [56] Foley AM, Leahy PG, Marvuglia A, McKeogh EJ. Current methods and advances
In: Proceedings of the energyocean proceedings NREL/CP-500-36313. Palm in forecasting of wind power generation. Renew Energy 2012;37:1–8.
Beach, Florida; 28–29 June 2004. [57] Xiao L, Wang JZ, Dong Y, Wu J. Combined forecasting models for wind energy
[25] Michael CR. Renewable energy technologies for use on the outer continental forecasting A case study in China. Renew Sustain Energy Rev 2015;44:271–88.
shelf. Colorado: National Renewable Energy Lab; 2006. [58] Shi J, Guo JM, Zheng ST. Evaluation of hybrid forecasting approaches for wind
[26] Kucukali S, Dinçkal Ç. Wind energy resource assessment of Izmit in the West speed and power generation time series. Renew Sustain Energy Rev
Black Sea Coastal Region of Turkey. Renew Sustain Energy Rev 2014;30:790–5. 2012;16:3471–80.
[27] González-Longatt F, González JS, Payán MB, Santos JMR. Wind-resource atlas [59] Han L, Romero CE, Yao Z. Wind power forecasting based on principle com-
of Venezuela based on on-site anemometry observation. Renew Sustain ponent phase space reconstruction. Renew Energy 2015;81:737–44.
Energy Rev 2014;39:898–911. [60] Tascikaraoglu NA, Uzunoglu M. A review of combined approaches for pre-
[28] Archer CL, Jacobson MZ. Evaluation of global wind power. J Geophys Res diction of short-term wind speed and power. Renew Sustain Energy Rev
2005;110:5462–81. http://dx.doi.org/10.1029/2004JD005462 D12110. 2014;34:243–54.
[29] Chang LX, Zhan FS. Wind energy in China: current scenario and future per- [61] Mederos ACM, Padrón JFM, Lorenzo AEF. An offshore wind atlas for the Canary
spectives. Renew Sustain Energy Rev 2009;13(8):1966–74. Islands. Renew Sustain Energy Rev 2011;15:612–20.
[30] Zhang D, Zhang XL, He JK, Chai QM. Offshore wind energy development in [62] Lars L. Short-term prediction of the power production from wind farms. J
China: current status and future perspective. Renew Sustain Energy Rev Wind Eng Ind Aerodyn 1999;90:207–20.
2011;15(9):4673–84. [63] Bailey B, Brower MC, Zack J. Short-term wind forecasting. In: Proceedings of
[31] Chen F, Ban X, Qi X. Evaluation of wind energy resources on the coastland and the European wind energy conference. (ISBN l902916X) [c]//. Nice, Frace;
adjacent sea of Lianyungang. Sci Meteorol Sin 2008;28:101–6. March 1–5, 1999. p. 1062–65.
[32] Wen MZ, Wu B, Lin XF. Distribution characteristics and assessment of wind [64] Milligan M, Schwartz M, Wan Y. Statistical wind power forecasting for U.S.
energy resources at 70 m height over Fujian coastal areas. Resour Sci 2011;33 wind farms. In: Proceedings of the conference WINDPOWER 2003. Austin;
(7):1346–52. May 18–21, 2003.
[33] Luong ND. A critical review on potential and current status of wind energy in [65] Focken U, Lange M, Waldl HP. Previento–a wind power prediction system with
Vietnam. Renew Sustain Energy Rev 2015;43:440–8. an innovative upscaling alogrithm. In: Proceedings of the European wind
[34] Charlotte BH, Metrete B, Alfredo P. SAR-Based wind resource statistics in the energy conference. Copenhagen, Denmark; June 2–6, 2001. p. 826–29.
Baltic Sea. Remote Sens 2011;3:117–44. [66] Kariniotakis G, Moassafir J, Usaola J. ANEMOS: development of a next gen-
[35] NASA. Global Ocean Wind energy potential. From http://earthobservatory. eration wind resource forecasting system for the large-scale integration of
nasa.gov/IOTD/view.php?id ¼8916; 2008. onshore and offshore wind farms. In: Proceedings of the European wind
[36] Liu WT, Tang WQ, Xie XS. Wind power distribution over the ocean. Geophys energy conference and exhibition, EWEC 2003. Madrid. Spain; 2003.
Res Lett 2008;35:L13808. http://dx.doi.org/10.1029/2008GL034172. [67] Pinard JP, Benoit R, Yu WA. West wind climate simulation of the mountainous
[37] Capps SB, Zender CS. Global ocean wind power sensitivity to surface layer Yukon. Atmos-Ocean 2005;43:259–82.
stability. Geophys Res Lett 2009. http://dx.doi.org/10.1029/2008GL037063 [68] Nielsen TS, Madsen H, Nielsen HA. Short-tem wind power forecasting using
L09801. advanced statistical models. In: Proceedings of the European wind energy
[38] Capps SB, Zender CS. Estimated global ocean wind power potential from conference; 2006.
QuikSCAT observations, accounting for turbine characteristics and siting. J [69] Xiao L, Wang JZ, Dong Y, Wu J. Combined forecasting models for wind energy
Geophys Res 2010;115:12679–91. http://dx.doi.org/10.1029/2009JD012679 forecasting: a case study in China. Renew Sustain Energy Rev 2015;44:271–88.
D09101. [70] Tascikaraoglu A, Uzunoglu M. A review of combined approaches for prediction
[39] Jiang D, Zhuang DF, Huang YH, Wang JH, Fu JY. Evaluating the spatio-temporal of short-term wind speed and power. Renew Sustain Energy Rev
variation of China's offshore wind resources based on remotely sensed wind 2014;34:243–54.
field data. Renew Sustain Energy Rev 2013;24:142–8. [71] Zheng CW, Li CY. Variation of the wave energy and significant wave height in
[40] Hasager CB, Badger M, Peña A, Larsén XG, Bingöl F. SAR-based wind resource the China Sea and adjacent waters. Renew Sustain Energy Rev 2015;43:381–7.
statistics in the Baltic Sea. Remote Sens 2011;3:117–44. [72] Wang XL, Feng Y, Swail VR. Changes in global ocean wave heights as projected
[41] Hasager CB, Mouche A, Badger M, Bingöl F, Karagali I, Driesenaar T, Stoffelen A, using multimodel CMIP5 simulations. Geophys Res Lett 2014;41:1026–34.
Peña A, Longépé N. Offshore wind climatology based on synergetic use of http://dx.doi.org/10.1002/2013GL058650.
Envisat ASAR,ASCAT and QuikSCAT. Remote Sens Environ 2015;156:247–63. [73] Beccali M, Cellura M, Brano VL, Marvuglia A. Short-term prediction of
[42] Dvorak MJ, Archer CL, Jacobson MZ. California offshore wind energy potential. household electricity consumption: assessing weather sensitivity in a Medi-
Renew Energy 2009;1(1–11). http://dx.doi.org/10.1016/j.renene.2009.11.022. terranean area. Renew Sustain Energy Rev 2008;12(8):2040–65.
[43] Al-Yahyai S, Charabi Y, Gastli A. Review of the use of numerical weather [74] Carta JA, Velázquez S, Cabrera P. A review of measure-correlate-predict (MCP)
prediction (NWP) Models for wind energy assessment. Renew Sustain Energy methods used to estimate long-term wind characteristics at a target site.
Rev 2010;14(9):3192–8. Renew Sustain Energy Rev 2013;27:362–400.
 C.W. Zheng et al. / Renewable and Sustainable Energy Reviews 53 (2016) 1240–1251 1251

[75] Wang JZ, Qin SS, Jin SQ, Wu J. Estimation methods review and analysis of [80] Pérez-Collazo C, Greaves D, Iglesias G. A review of combined wave and off-
offshore extreme wind speeds and wind energy resources. Renew Sustain shore wind energy. Renew Sustain Energy Rev 2015;42:141–53.
Energy Rev 2015;42:26–42. [81] Tabassum-Abbasi Premalatha M, Abbasi T, Abbasi SA. Wind energy: increasing
[76] Wu J, Wang ZX, Wang GQ. The key technologies and development of offshore deployment, rising environmental concerns. Renew Sustain Energy Rev
wind farm in China. Renew Sustain Energy Rev 2014;34:453–62. 2014;31:270–88.
[77] Perveen R, Kishor N, Mohanty SR. Off-shore wind farm development: present [82] Saidur R, Rahim NA, Islam MR, Solangi KH. Environmental impact of wind
status and challenges. Renew Sustain Energy Rev 2014;29:780–92. energy. Renew Sustain Energy Rev 2011;15:2423–30.
[78] Lee TL. Assessment of the potential of offshore wind energy in Taiwan using
fuzzy analytic hierarchy process. Open Civil Eng J 2010;4:96–104.
[79] Dong J, Feng TT, Yang YS, Ma Y. Macro-site selection of wind/solar hybrid
power station based on ELECTRE-II. Renew Sustain Energy Rev 2014;35:194–
204.