Your Title

5 downloads 870 Views 794KB Size Report
Investigation on the offshore wind energy potential over the North. Western ... opportunities for cost reduction and technical development. Indeed, the offshore wind ..... [2] http://www.thewindpower.net/windfarm_fr_17399_le-carnet.php.
1

Investigation on the offshore wind energy potential over the North Western Mediterranean Sea in a regional climate system model Hiba Omrani Laboratoire de Météorologie Dynamique Ecole Polytechnique, Palaiseau, France Bénédicte Jourdier Laboratoire de Météorologie Dynamique Ecole Polytechnique, Palaiseau, France Karine Béranger Ecole Nationale Supérieure de Techniques Avancées, ParisTech, Palaiseau, France Sophie Bastin Laboratoire Atmosphéres Milieux Observations Spatiales, Guyancourt, France

Abstract—Wind energy is one of the fastest growing renewable energy resources worldwide. The estimation of the wind resources depends on the atmospheric circulation data retrieved from meteorological observations or regional climate models outputs. These models provide weather and climate data over specific domains and over time periods where few or no observations exist. In this work we evaluate the sensitivity of the offshore wind to the horizontal resolution of the regional climate model (WRF) and to the air-sea interactions over the north western Mediterranean sea. Comparisons between different model configurations allowed us to highlight the complexity of the interactions between the atmospheric circulation, the local topography and the sea surface. Results show that the simulated wind potential energy is very sensitive to the model configuration which can modify substantially its space-time variability. Index Terms—wind energy, regional climate model, air-sea interactions, horizontal resolution.

I. I NTRODUCTION HE reduction of greenhouse gas (GHG) emissions to reduce climate change impacts has been a crucial topic in recent years. The European Council adopted, in March 2007, a strategy "for safe, competitive and sustainable energy" known as the "3 x 20 roadmap". It has three major aims for Europe toward 2020: (1) To reduce greenhouse gas emissions by 20%; (2) To improve energy efficiency by 20%; (3) To boost the share of renewable in the total energy consumption to 20% (compared to 10% for Europe today). The wind energy is a key component to achieve this objective, since it represents one of the rare renewable sector with an enough mature technology that can effectively contribute to the energy

T

Philippe Drobinski Laboratoire de Météorologie Dynamique Ecole Polytechnique, Palaiseau, France Cindy Lebeaupin Brossier Centre Nationale de Recherches Météerologiques Météo France, Toulouse, France Sylvain Mailler Laboratoire de Météorologie Dynamique Ecole Polytechnique, Palaiseau, France Thomas Arsouze Ecole Nationale Supérieure de Techniques Avancées, ParisTech, Palaiseau, France

transition. Especially, the offshore wind energy with a rapidly growing industry and a huge potential, can cover seven times the energy demand of Europe [1]. Moreover the offshore wind energy has several advantages and offers a significant opportunities for cost reduction and technical development. Indeed, the offshore wind potential energy is much greater than the onshore wind potential energy with less turbulence and environmental constraints. In the Mediterranean basin as in the rest of the world, the most populated cities are near the coasts. Which make the offshore wind more suitable for large scale development near the major demand centers, avoiding the need for long transmission lines. In France, there are only 6 MW of installed offshore wind power in 2012 [2], compared to 854.2 MW for UK and 184.5 MW for Belgium [1]. This offers a real opportunity for growth and investigation for future site. Energy supply from wind is related to climate, as wind resources are determined by atmospheric circulation. Regional climate models provide weather and climate data and analysis over a specified domain at locations and over time periods where few or no observations exist. With relatively high time and spatial resolution the potential renewable energy resources can be evaluated from this data using well-documented formulations. The aim of this work is to investigate the sensitivity of the offshore wind potential energy simulated with an atmosphere only regional climate model (ARCM) and a atmosphere/ocean coupled regional climate model (AORCM) to the model gridresolution and the ocean/atmosphere feedbacks.

2

A. Experiment set-up This study is a part of two international programs named the Hydrological Cycle in the Mediterranean Experiment (HyMex) [3] and the Coordinated Downscaling Experiment (CORDEX) of the World Climate Research Program (WCRP) [4]. In this framework, an ensemble of 20-years regional climate simulations was produced over the Mediterranean region. Four of them have been used in this study. Three atmosphereonly simulations performed with the Weather Research and Forecasting (WRF) model of the National Center for Atmospheric Research (NCAR) [5]. Initial and lateral conditions are taken from the European Centre for Medium-Range Weather Forecasts (ECMWF) ERA-interim reanalysis [6] provided every 6 h with a 0.75◦ resolution. One coupled simulation with WRF and the ocean model Nucleus for European Modeling of the Ocean (NEMO) [7], used in a regional eddy-resolving Mediterranean configuration MED12. The atmospheric simulations run with the same physics. A detailed description of the model physics can be found in [8]. Only the atmospheric fields were used here. 1) The first atmosphere-only simulation (called hereafter ATM50) is run over the Med-CORDEX domain with a 50 km horizontal resolution (110×69 grid-points). 2) The second atmosphere-only simulation (called hereafter ATM20) has a 20km horizontal resolution (240×130 grid-points). These two simulations have a prescribed sea surface temperature (SST) retrieved from ERAinterim reanalysis. 3) The coupled simulation (named CPL20), with a 20km horizontal resolution, runs with two-way interactive exchanges between the atmospheric and the ocean models. The exchanged variables are the sea surface temperature (SST) and the heat, water and momentum fluxes. The coupling frequency is 3 hours. 4) The third atmosphere-only simulation (named ATMs20) has the same model configuration as the atmospheric component of the CPL20 simulation, except for the SST field. This later was filtered (smoothed) in order to remove the high frequency variability but retain the same climatology and diurnal cycle as the SST of the CPL20 simulation. Twenty-year simulations were performed, starting January, 1st 1989 and ending December, 31st 2008. The three hourly atmospheric fields are stored. Table I summarize the used simulations with the main differences. Table I L IST OF USED SIMULATIONS Name ATM50 ATM20 ATMs20 CPL20

Horizontal resolution 50km 20km 20km 20km

SST ERA-I ERA-I NEMO smoothed NEMO

B. Studied area In this work we focus on the Gulf of Lions area, the most windy region of the entire Mediterranean Sea [9]. The

prevailing winds in this area are from the North and NorthWest , especially in winter. This area is also known by a surrounding complex orography. It is bounded along its coasts from East to the West by the Alps , the Massif central and the Pyrenees separated by deep valleys (Figure 1 ). When cold-air masses flow towards the Mediterranean Sea from the North west they are channeled through the valleys between the Pyrenees, the Massif Central and the Alps. The gulf experiences then a dry, cold, northerly wind known locally as the Mistral and Tramontane [10], [11], [12], [13] as showed by the wind rose computed over the red box (Figure2).

Figure 1. The Gulf of Lions region with the main local winds and the surrounding topography. The area of study is delimited by the dashed red box

Figure 2. The wind rose computed over the studied domain for ATM20 simulation.

II. R ESULTS In this work we investigate, thanks to numerical modeling, the impact of the horizontal resolution and the air-sea interactions on the wind speed and the wind potential energy variability. A. Horizontal resolution In this section we used the two atmosphere-only simulations ATM20 and ATM50 with 20km and 50km horizontal resolution all being otherwise equal. In Figure 3 we plot the 20-year climatology (panel a) for ATM20 (in black) and ATM50 (in red) of the surface wind speed. The quantiles 5, 50 and 95 are represented in the corresponding color by q05, q50 and q95. We note that the wind speed is systematically higher in ATM 20 compared to ATM50 especially during the windy season (January to April and November to December). Indeed, given the geographic localization of the Gulf of Lions, the complex

3

surrounding orography (Massif Central, Alps, and Pyrenees) steers the atmospheric circulation at large scale producing meso-scale flow especially local winds as the Mistral and Tramontane. Therefore, the representation of the topography in the model modulate significantly the atmospheric circulation and thus the local winds.

Figure 4. 3D topography over the Gulf of Lions region with the main mountain as represented in ATM20 (a) and ATM50 (b)

winter (Figure5a and e) where the wind potential energy is important as well as the energy demand for the heating needs. One should note that the average height of wind turbines is between 80 and 100 m with 100 m diameter blades. The wind turbines extract the wind potential energy between 50 and 150 m. A bad estimation of wind potential energy is very critical in terms of cost for wind energy industries. Figure 3. Wind speed climatology over the studied area (panel a) and the PDF (panel b) for ATM20 and ATM50 simulations.

Figure 4 displays the surrounding orography of the Gulf of Lions as represented in ATM20 and ATM50 simulations. We note that the principals mountains are smoothed in ATM50 and the valleys are larger and less deep compared to ATM20. Winds are then less constrained and less channeled by these valleys which explain the weaker wind speed in the ATM50 simulation. The probability density functions (PDF) show that the wind is redistributed between ATM20 and ATM50 simulations with a higher number of strong wind days (≥ 8 m s−1 ) in ATM20 and a higher number of moderate wind days (∼ 5 m s−1 ) in ATM50. Figure 5 shows the vertical profiles of the wind potential energy for ATM20 (in black) and ATM50 (in red) and the differences between the two simulations (ATM20-ATM50) for each season. We computed the wind potential energy using the equation 1: 1 ρAV 3 (1) 2 With P is the wind potential energy in Watt, ρ is the air density (ρ = 1.2 kg m−3 ), A is the swept area of the turbine computed from the length of the turbine blades using the equation for the area of a circle: A = π r2 (here,r = 50m) and finally V is the wind speed (m s−1 ). Regardless the seasons, we note that wind potential energy is larger in ATM20 than ATM50 throughout all the vertical layers (up to 500 m). The maximum difference is mainly between 100 and 300 meters. This is especially true in the

Figure 5. Wind potential energy profile for ATM20 (black line) and ATM50 (red line) (panels a, b, c and d) and the difference ATM20-ATM50 (panel e, f, g and h) for different seasons.

P =

Figure 6. The wind potential energy difference at 100 m between ATM20 and ATM50.

In Figure 6, we plot the spatial difference of the wind po-

4

tential energy at 100 m between ATM20 an ATM50. We note that the maximum difference is located where the two main winds of the region (the Mistral and Tramontane) converge, along the western coasts of the domain. These differences can reach locally 3000 kW which represents 100% the production capacity of an offshore wind turbine. However the minimum seems to be in the leeward side. The spatial distribution of the difference does not change a lot from one season to another, only the magnitude vary depending on the wind speed strength. B. Air-sea interactions The Gulf of Lions is a key region for air-sea interactions. The strong local winds (Mistral, Tramontane) bring continental air over the sea, inducing strong air-sea interactions with intense heat and momentum exchanges [14], [15]. These interactions produce many distinctive flow patterns in both compartments. In this section we focus on the impact of the air-sea interactions on the small scale variability of the wind in the Gulf of Lions. Three simulations were used, two atmosphere-only simulations (ATM20, ATMs20) and one air-sea coupled simulation (CPL20). The ATM20 simulation represents the “classic” configuration of the atmospheric-only models with a prescribed SST. The SST field is generally retrieved from reanalysis products which has a coarse horizontal resolution, typically between 2◦ and 0.75◦ (here 0.75◦ ) and a daily updated outputs. In the coupled simulation (CPL20) a high temporal and spatial resolution SST field is provided by the ocean model. Finally, the ATMs20 uses a smoothed SST where the high-frequency air-sea coupling effects (submonthly variations), are filtered. However, the diurnal cycle, the seasonal variability and the persistent spatial structures that exist in the CPL20 simulation are preserved. In this section we consider the coupled simulation as “reference” because it represent the most realistic configuration. Therefore comparisons will be made with respect to CPL20.

of the near-surface wind for the three simulations. We note that for ATM20 wind speed is systematically stronger than ATMs20 and CPL20. Whereas, the coupled simulation has the weakest wind speed. The differences seem to be larger in summer than in winter. The wind distribution is almost the same. Since the only difference between the three simulations is the SST fields, we will investigate in this section on the relationship between the wind and the SST fields and the corresponding effect on wind potential energy. Many studies have documented this relationship. A negative correlation between surface wind and SST fields is interpreted as evidence since strong winds cool the sea surface through evaporation [16], [17], [18]. This effect has been observed in the northwestern Mediterranean Sea where the Mistral and Tramontane drive the horizontal ocean circulation and the heat and momentum exchanges between the atmosphere and the sea surface. The strong-wind in winter is well known as the major factor in the cooling of the sea surface and thus the formation the deep water in the Gulf of Lions [19], [20]. However, observations over the oceanic fronts and eddies [21], [22], [23] showed a positive correlation between the near-surface wind and the SST field during weak wind episodes. In this study we investigate to what extent the differences in the SST fields (used in the three simulations) can modify the offshore wind.

Figure 8. The correlation coefficients computed between the near surface wind speed and the SST difference for each season between ATMs20 and CPL (panels a, c, e and g) and between ATM20 and CPL20 (panels b, d, f and h) .

Figure 7. Wind speed climatology and the PDF over the studied area for ATM20 (blue line), ATMs20 (red line) and CPL20 (black line) simulations.

In Figure 7, we plot the monthly climatology and the PDF

In Figure 8 we plot the correlation coefficients of the wind speed and the SST difference between ATMs20 and CPL (panels a, c, e and g) and between ATM20 and CPL20 (panels b, d, f and h). We calculate the correlation coefficients for each grid point over the 20 winters (spring, summer, autumn) corresponding to the twenty years of the simulations (19892008 using the equation(2): ¯ E[(X − X)(Y − Y¯ ) (2) CR(X, Y ) = σX σY ¯ and Y¯ the mean of X, where E[.] is the expected value, X Y and σX , σY the standard deviation of X, Y. X represents ∆f1 f = f fAT M 20 − f fCP L20 and ∆f2 f = f fAT M s20 − f fCP L20 and Y represents ∆SST = SSTAT M 20 −SSTCP L20 1 and ∆SST = SST − SST AT M s20 CP L20 . 2

5

The first observation is that the correlation coefficient is positive over all the domain. That means to a warmer SST corresponds a stronger wind speed and vice versa. We also note that the higher correlation is obtained near the coasts and the western part of the domain. This is important since the actual offshore wind turbines are installed between 20 and 80 m water depth and then not far from the coasts. Moreover, the high correlation coefficient in summer (Fig (8)e) for both ATM20-CPL20 and ATMs20-CPL20, show that the wind is more sensitive to SST variations in summer than during the other seasons. The rapid and small scale variations of the SST in summer (not shown) drive the atmospheric response over this area and modify the near surface circulation. In fact, in winter the wind is mainly driven by the large scale atmospheric circulation and the topography forcing which explain the weak correlation coefficient for both ATMs20/CPL20 and ATM20/CPL20. In summer, the large scale forcing is weaker and then the wind it is more sensitive to SST variations especially the SST gradient. Previous studies [24], [25], [26] show that when air blows across an SST gradient, an air-sea temperature difference is generated which modifies the near surface stability and thus the wind profile. It means if we modify the SST gradient we modify the wind field.

Figure 10. The wind potential energy differences at 100 m between ATMs20 and CPL (panels a, c, e and g) and between ATM20 and CPL20 (panels b, d, f and h) .

Figure 10 displays the spatial distribution of the differences of the wind potential difference at 100 m. Overall the difference between ATM20 and CPL20 is larger than the difference between ATMs20 and CPL20 and localized near the coasts. One should note that the offshore wind farms are generally installed at most 50 km from the coast and a difference of 400 kW represents 13% of the production capacity of a classic offshore wind turbine. The spatial variability depends on the season with a maximum a long the northern coast in winter and the western coast in summer and spring. The difference between ATMs20 and CPL20 is small but can reach 300 kW in winter which is considerable for the wind energy industrials. These results show that the high frequency air-sea interactions as well as the horizontal resolution of the simulated fields can modify substantially the space-time wind variability. III. C ONCLUSIONS

Figure 9. Wind speed (panels a, b, c and d) and wind potential energy (panels e, f, g and h) profiles differences between ATMs20 and CPL20 (red lines) and ATM20 and CPL20 (black lines).

In Figure 9, we plot the wind speed and wind potential profiles differences between ATMs20 and CPL20 (in red) and ATM20 and CPL20 (in blue). Compared to Figure5 in section II-A, we note that the difference between the two wind speed profiles is small and decreases rapidly with height to zero. This is not a surprising result since the sea feedback is a surface forcing, however, the high surrounding orography impacts the whole atmospheric column (up to 500 m). We also note that up to 200 m the difference between ATM20 and CPL20 is positive and higher than the difference between ATMs20 and CPL20. In fact the difference in terms of SST fields between ATM20 and CPL20 is larger than the difference between ATMs20 and CPL20 (not shown). These differences impact the near surface thermodynamic fields (temperature and wind) and propagate to the higher levels. Therefore we expect a proportional effect on the wind and thus the wind potential energy. The inversion between the two profile above 200m still to be investigated.

In this study, we investigate the sensitivity of the wind potential energy on different regional climate model configurations. We focus on the impact of the horizontal resolution on the simulated offshore wind. We highlight the interactions between the small scale atmospheric circulation with the local topography which can have a significant impact on the offshore wind and wind potential energy. Indeed, the difference between ATM20 and ATM50 in terms of wind potential energy vary between 1200 kW to 3000kw which represents 60% to 100% of the production capacity of an offshore wind turbine. The assessment of this variability is crucial for wind energy industries. We also investigate the effect of the air-sea interactions on the near-surface wind and the vertical profile of the wind potential energy. The results show that the differences in SST fields can modify the near-surface wind through the SST fronts and the atmospheric boundary layer by modifying the stratification and thus the wind vertical profile. Differences between ATM20 and CPL20 can reach 600 kW near the coasts. The average nameplate capacity of an offshore wind turbine in Europe was about 2/3 MW with a rotor radius of 50/60m and the capacity of future turbines is expected to increase to 6/7 MW with a rotor radius of 75/82m. In this study we

6

computed the wind potential energy available for a 2 MW wind turbine with a rotor radius of 50m. A larger rotor radius means a larger production capacities with a 3 MW to 7 MW wind turbines and a rotor radius. For example, in similar wind condition, one produced 2.3 times more energy than the other since it has a much larger blade diameter, and therefore more collector area. However, it will be more sensitive to a smaller change in wind speed. Indeed, in figure 11, we plot the power curves for a two offshore wind turbines (VESTAS) of 40m and 82m rotor radius. For example, between 10 m/s and 11 m/s wind speed the difference in power production for Vestas-40m is 335 kW against 987 for Vestas-82m.

Figure 11. The power curve of two offshore wind turbines from VESTAS with 40 m (blue line) and 82 m (red line) rotor radius.

Finaly, comparisons with observations will enable as to understand the offshore wind and wind potential energy variability in time and space. Future work will be dedicated to projection of the wind energy in the future climate. We should note that this study is an ongoing work, further investigation on the physical processes still to be done. ACKNOWLEGMENT This research has received funding from ADEME (Agence de l’Environnement et de la Maîtrise de l’Energie) through the MODEOL project (contract #1205C0147). Hiba Omrani was supported by the Ecole Polytechnique/CEA Chaire "Energie Durable" in partnership with EDF (the European Foundation for Tomorrow’s Energies, Institut de France) and “la Fondation de l’École Polytechnique”.It was also supported by the IPSL group for regional climate and environmental studies. This work also contributes to the HyMeX program (HYdrological cycle in The Mediterranean EXperiment) through INSU-MISTRALS support and the Med-CORDEX program (A COordinated Regional climate Downscaling EXperiment -Mediterranean region). R EFERENCES [1] http://www.gwec.net. [2] http://www.thewindpower.net/windfarm_fr_17399_le-carnet.php. [3] P. Drobinski, V. Ducrocq, P. Lionello, and V. Homar, “Hymex, the newest GEWEX regional hydroclimate project,” GEWEX newsletter, vol. 21, pp. 10–11, 2011. [4] F. Giorgi, C. Jones, and G. R. Asrar, “Addressing climate information needs at the regional level: the cordex framework,” WMO Bulletin, vol. 58, pp. 175–5183, 2009. [5] W. C. Skamarock, J. B. Klemp, J. Dudhia, D. O. Gill, D. M. Barker, M. G. Duda, X.-Y. Huang, W. Wang, and J. G. Powers, “A description of the Advanced Research WRF Version 3,” NCAR Tech. Note NCAR/TN475+STR, p. 125, 2008.

[6] A. Simons, S. Uppala, D. Dee, and S. Kobayashi, “Era-interim: New ecmwf reanalysis products from 1989 onwards,” ECMWF Newsletter, vol. 110, pp. 525–535, 2007. [7] G. Madec, “NEMO ocean engine.” Note du Pole de modélisation, Institut Pierre-Simon Laplace (IPSL), France, vol. No 27, pp. ISSN No 1288– 1619, 2008. [8] C. Lebeaupin-Brossier, S. Bastin, K. Béranger, and P. Drobinski, “Regional mesoscale air-sea coupling impacts and extreme meteorological events role on the mediterranean sea water budget,” Clim. Dyn., p. submitted. [9] Ascensio, E. and Bordreuil, C. and Frasse, M. and Orieux, A. and Roux D., “Une approche des conditions météorologiques sur le golfe du lion,” Ann. Inst. Oceanogr., Paris, vol. 53, pp. 155–5169, 1977. [10] P. Drobinski, C. Flamant, J. Dusek, P. H. Flamant, and J. Pelon, “Observational evidence and modeling of an internal hydraulic jump at the atmospheric boundary layer top during a tramontane event,” Bound. Lay. Meteorol., vol. 98, pp. 497–515, 2001. [11] P. Drobinski, S. Bastin, V. Guénard, J.-L. Caccia, A. Dabas, P. Delville, A. Protat, O. Reitebuch, and W. C., “Summer mistral at the exit of the rhone valley,” Q. J. Roy. Meteorol. Soc., vol. 131, pp. 353–375, 2005. [12] V. Guénard, P. Drobinski, J. L. Caccia, B. Campistron, and B. BÃl’nech, “Observational study of the mistral mesoscale dynamics,” Bound Lay Meteorol, vol. 115(2), pp. 263–288, 2005. [13] V. Guénard, P. Drobinski, J. L. Caccia, G. Tedeschi, and C. P, “Dynamics of the map iop15 severe mistral event: observations and high-resolution numerical simulations,” Q J R Meteorol Soc, vol. 132, pp. 757–777, 2006. [14] C. Flamant, “Alpine lee cyclogenesis influence on air-sea heat exchanges and marine atmospheric boundary layer thermodynamics over the western mediterranean during a tramontane/mistral event,” J. Geophys. Res. (Oceans), vol. 108, no. C2, pp. n/a–n/a, 2003. [Online]. Available: http://dx.doi.org/10.1029/2001JC001040 [15] C. Lebeaupin-Brossier and P. Drobinski, “Numerical high-resolution air-sea coupling over the gulf of Lions during two tramontane/mistral events,” J. Geophys. Res., vol. 114, p. doi:10.1029/2008JD011601, 2009. [16] W. T. Liu, A. Zhang, and J. K. B. Bishop, “Evaporation and solar irradiance as regulators of sea surface temperature in annual and interannual changes,” J. Geophys. Res., vol. 99, pp. 12 623–12 637, 1994. [17] N. J. Mantua, S. R. Hare, Y. Zhang, J. M. Wallace, and R. C. Francis, “A pacific interdecadal climate oscillation with impacts on salmon production,” Bull. Am. Meteor. Soc., vol. 78, pp. 1069–1079, 1997. [18] Y. Okumura, S. P. Xie, A. Numaguti, and Y. Tanimoto, “Tropical atlantic air-sea interaction and its influence on the nao,” Geophys. Res. Lett., vol. 28, pp. 1507–1510, 2001. [19] J. Marshall and F. Schott, “Open–ocean convection: observations, theory and models,” Rev. Geophys., vol. 37, pp. 1–64, 1999. [20] C. Lebeaupin-Brossier, K. Béranger, and P. Drobinski, “Sensitivity of the north-western mediterranean coastal and thermohaline circulations as simulated by the 1/12 degree resolution oceanic model NEMOMED12 to the space-time resolution of the atmospheric forcing,” Ocean Modelling, vol. 43–44, pp. 94–107, 2012a. [21] W. T. Liu, X. Xie, P. S. Polito, S. P. Xie, and H. Hashizume, “Atmospheric manifestation of tropical instability wave observed by quikscat and tropical rain measuring mission,” Geophys. Res. Lett., vol. 27, pp. 2545–2548, 2000. [22] D. B. Chelton, S. K. Esbensen, M. G. Schlax, N. Thum, M. H. Freilich, F. J. Wentz, C. L. Gentemann, M. J. McPhaden, and P. S. Schopf, “Observations of coupling between surface wind stress and sea surface temperature in the eastern tropical pacific,” J. Climate, vol. 14, pp. 1479– 1498, 2001. [23] H. Hashizume, S. P. Xie, W. T. Liu, and K. Takeuchi, “Local and remote response to tropical instability waves: a global view from space,” J. Geophys. Res., vol. 106, pp. 10 173–10 185, 2001. [24] W. Sweet, R. Fett, J. Kerling, and P. La Violette, “Air-sea interaction effects in the lower troposphere across the north wall of the gulf stream,” Mon. Wea. Rev., vol. 109, pp. 1042–1052, 1981. [25] J. A. Businger and W. J. Shaw, “The response of the marine boundary layer to mesoscale variations in sea-surface temperature,” Dyn. Atmos. Oceans, vol. 8, pp. 267–281, 1984. [26] S. P. Hayes, M. J. McPhaden, and J. M. Wallace, “The influence of sea surface temperature on surface wind in the eastern equatorial pacific: weekly to monthly variability,” J. Climate, vol. 2.