Impact of sea breeze on wind-seas off Goa, west coast of India

Impact of sea breeze on wind-seas off Goa, west coast of India
of 6
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Impact of sea breeze on wind-seas off Goa,west coast of India S Neetu 1 , ∗ ,  Satish Shetye 1 and  P Chandramohan 21 Physical Oceanography Division, National Institute of Oceanography, Dona Paula, Goa 403 004, India. 2 Indomer Coastal Hydraulics (P) Ltd, Ragamalika, Kumran Colony Main Road, Vadapalani,Chennai 600 026, India. ∗ e-mail:  After withdrawal of the Indian Summer Monsoon and until onset of the next monsoon, i.e., roughlyduring November–May, winds in the coastal regions of India are dominated by sea breeze. It has animpact on the daily cycle of the sea state near the coast. The impact is quite significant when largescale winds are weak. During one such event, 1–15 April 1997, a Datawell directional waveriderbuoy was deployed in 23m water depth off Goa, west coast of India. Twenty-minute averagedspectra, collected once every three hours, show that the spectrum of sea-breeze-related ‘wind-seas’peaked at 0 . 23 ± 0 . 05Hz. These wind-seas were well separated from swells of frequencies less than0.15Hz. The TMA spectrum (Bouws  et al   1985) matched the observed seas spectra very well whenthe sea-breeze was active and the fetch corresponding to equilibrium spectrum was found to be77 ± 43km during such occasions. We emphasize on the diurnal cycle of sea-breeze-related seaoff the coast of Goa and write an equation for the energy of the seas as a function of the localwind. 1. Introduction While the importance of seasonally reversing mon-soon winds to ocean processes in the Indian coastalregion has been well appreciated, virtually noattention has been paid in the available literatureto diurnal winds along the coast. After withdrawalof the Indian Summer Monsoon and until onset of the next monsoon, i.e., roughly during November–May, winds in the coastal region of India are dom-inated by sea breeze. Studies carried out elsewherehave shown that sea breeze can have a significantimpact on coastal processes. For example, it hasbeen shown that sea breeze has important impli-cations to alongshore sediment transport in thecoastal areas off Perth, Australia (Masselink  et al  1998a, b; Pattiaratchi  et al   1997). Most often theimpact of sea breeze is felt because of the wave fieldthat is excited by the breeze in the shallow areasnear the coast. This paper examines the impact of the breeze on near-shore waves off the west coastof India using wave data collected with a Datawelldirectional waverider buoy in 23m of water, andwinds from a location close to the coastlineand nearer the buoy position. Simultaneouswind and wave measurements were made duringJune 1996 to May 1997 off Goa (Mormugao), cen-tral west coast of India (figure 1) to record the seabreeze effect as most of the Indian coast experi-ence it. Here we focus on the first two weeks of April 1997, when the winds showed a distinct sea-breeze cycle and its impact was seen on the prevail-ing waves. During April, the large-scale circulationover India and the north Indian Ocean is gener-ally weak, this being the transition time betweenthe northeast monsoon and the southwest monsoon(see Shetye and Gouveia 1998, for a description of the environment along the coast). It is thereforeexpected that the swells at the location of obser-vation would be minimal at this time. Keywords.  Wind-wave; sea breeze; coastal processes; energy density spectra; Arabian Sea. J. Earth Syst. Sci.  115 , No. 2, April 2006, pp. 229–234 © Printed in India.  229  230  S Neetu, Satish Shetye and P Chandramohan  Figure 1. Location of the wind and wave observationsreported in the paper (bottom topography contours are inkms). 2. Observations and analysis The wind observations were made using ananemometer located on the terrace of the mainbuilding of the National Institute of Oceanography,Goa. The building is on the coast and is approx-imately 28km from the wave rider buoy location.The height of the anemometer was about 50mabove sea level. The recorded data were ten-minutevector averages of wind speed and direction. Dur-ing the period of study the winds usually pickedup at 1000 hours (IST) and peaked around 1500hours and the peak wind speed remained around5m/s (figure 2a upper panel). By 2000 hours thewinds dropped below 3 m/s and decreased slowlythereafter. The mean daily hodograph traced bythe wind vector during the period of study isshown in figure 3. The hodograph is elliptical withwind towards the land from about 1030hrs. As thewind picked up, the wind vector turned towardsthe right. By 1800hrs the wind was approximatelyalongshore. The vector continued to turn towardsits right. Subsequently wind magnitude decreased.During early morning hours the wind had a com-ponent oriented towards the sea: this is the landbreeze. Its magnitude was much weaker than thatof the sea breeze. In fact, minimal wind speed of about 1.5m/s was observed around 0600 hours.Before commencement of the sea breeze, the windspeed generally remained less than 2m/s. Theonset of sea breeze was marked by an abruptchange in wind direction. The wind was from about90 ◦ before 1000hrs and after onset of sea breeze thedirection changed to about 300 ◦ (figure 2a lowerpanel). During 1000–2000 hours the wind directionchanged slowly from 300 ◦ to about 360 ◦ . The coast-line in the vicinity of Goa is oriented along approx-imately 340 ◦ (figures 1 and 3). Hodographs similarto the one shown in figure 3 have been observed atKinloss and Aberdeen, Scotland (Simpson 1994).Hence, the sea breeze at a height of about 50mabove sea level recorded by the anemometer isalmost oriented along the coastline. At sea level,the breeze is expected to be at an angle somewhatlower than the angle at the height of the anemo-meter owing to veering from frictional effects in theatmospheric boundary layer.The observed 20-minute averaged wave spec-tra during the first two weeks of April 1997,was computed once every three hours over the0.03–0.6Hz frequency band. Wind-wave spectrum1–30sec could be separated invariably into twodistinct parts (figure 4). The lower frequency partcorresponds to swell waves of frequencies lesserthan 0.15Hz. The higher frequency part which cor-responds to the sea waves, reveals a distinct diurnalcycle. This part of the spectrum was isolated andanalyzed in greater detail. The spectrum peaked at0 . 23 ± 0 . 05Hz. The mean direction during 1000–2000 hours was around 300 ◦ when sea breeze wasusually active (figure 2b). The mean wave direc-tion deviates, by about 30 ◦ than the average angleof sea breeze measured by the anemometer (330 ◦ ).This difference may be attributed to the veering of wind angle in the boundary layer. However, thesetwo directions were close enough to support theidea that the high frequency part of the spectrumwas primarily due to the effect of local winds.To investigate this possibility further, we identi-fied the peak frequency of the wind-sea componentof the observed spectrum, and fitted a theoreticalequilibrium spectrum. The theoretical spectrumchosen was the TMA spectrum (Bouws  et al   1985),formulated as an extension of the JONSWAP spec-trum (Hasselmann  et al   1973) for wind-generatedseas in a finite water depth. The depth was takento be 23m. The JONSWAP spectrum is: S  ( f  ) =  αg 2 (2 π ) 4 f  5  exp  − 1 . 25  f  m f   4  γ  τ  ,τ   = exp  − ( f   − f  m ) 2 2 σ 2 f  2 m  , where  γ   = peak-shape parameter (average value3.30),  α  = 0 . 076¯ x − 0 . 22 ,σ  = 0 . 07 ,f   ≤ f  m  and 0.09, f > f  m ,f  m  = 3 . 5( g/U  )¯ x − 0 . 33 ,  x  is the fetch length,¯ x  =  gx/U  2 is the dimensionless fetch,  U   is the  Impact of sea breeze on wind-seas off Goa   231 Figure 2.  (a)  Upper panel: wind speed (m/s) during 1–15 April 1997. Lower panel: same as upper panel except winddirection (degree). The horizontal axis covers the time span from 0000hrs on 1 April 1997 to 0000hrs on 16 April 1997. ws:wind speed; wd: wind-direction.  (b)  Daily variation of wind and wave direction. Solid circles indicate average wind directionduring 0000–1000hrs; wind direction during 1000–2000 hrs is shown by solid squares; stars give average wave direction. (c)  Comparison of total wave energy calculated from wave-rider buoy data (solid line) and calculated by equation (dashedline). mean wind speed and  g  is the acceleration due togravity.The spectrum represents wind-generated seaswith fetch limitation. Wind speed and fetch lengthare inputs for the above spectrum while the peakfrequency was identified from the observed spec-trum. Bouws  et al   (1985) applied a transformationfactor to the JONSWAP spectrum and defined theTMA spectrum as follows: S  ( f  ) =  S  ( f  ) · φ ( ω h ) , where φ ( w h ) = ( k ( w,h )) − 3  ∂ ∂w k ( w,h )( k ( w, ∞ )) − 3  ∂ ∂w k ( w, ∞ ) . In the above transformed formulation  w h  is adimensionless frequency defined by 2 πf    ( h/g )where  h  is the water depth and  k ( w,h ) is the wavenumber associated with the dispersion relationshipfor waves in finite water depth.The TMA spectrum that could provide the bestfit for the sea-related observed spectrum off Goa,was determined by choosing the equivalent fetchand wind speed that minimized the sum of squaresof departure between the observed and the theoret-ical spectra. The results of this exercise on a typ-ical day is summarized in figure 5. The spectrumshowed distinct diurnal variability. The energy dueto the spectrum peaked around 1800–2000 hours.The wind speed determined from the best-fit theoretical spectrum and that measured bythe anemometer for the corresponding 20-minuteperiod were well correlated when the computedwind speed was in excess of 2m/s (figure 6). Thecorrelation between the observed and estimated  232  S Neetu, Satish Shetye and P Chandramohan  Figure 3. Sea breeze hodograph for the average of 1–15April 1997. The wind vectors at 0600, 1200, 1800 and 0000hours are identified. Though there is day-to-day variationin the sea breeze as seen in figure 2(a), on an average thesea breeze at about 1200 hours is approximately from thenorthwest. At this time the wind has a component perpen-dicular to the coastline and oriented towards the coast (seeorientation of the coastline marked). By 1800 hours the windvector turns towards the right to a direction that is alongthe coast, and continues to turn further while the magnitudedecreases.Figure 4. Spectral density (m 2 /Hz) as a function of frequency (Hz) and time (3 hourly interval) during 1–15 April 1997.Along the time axis, 0000 hours on 1 April 1997 is given by zero. wind speeds was 0.64 (for 39 points), which is wellabove the 99% significance level ( ∼ 0 . 4) and sup-ports the argument that the observed spectrumwas mostly in equilibrium with the local wind whenthe sea breeze was active with speed in excess of 2m/s. The fetch corresponding to the equilibriumspectrum was found to be 77 ± 43km during suchoccasions. The least-square fit through the pointsmarked in figure 6 was  y  = 0 . 5426 x +1 . 2264, where y  is the observed wind and  x  is the wind equivalentof the wave spectrum. The wave spectral energypeaked around 1800–2000 hours though the onsetof the sea breeze was around 1000 hours; maximumwind speed was observed at 1500 hours.It is known from empirical data that there isalways a lag between onset of wind and the time atwhich maximum wave energy is attained. Empir-ical formulae for wave growth, which have beenderived from large data sets (WMO-No. 702, 1988),state that for the wind speed of   ∼ 5m/s and fetch ∼ 77km, the time at which the maximum wavegrowth takes place is approximately 9 hours. Afterthis period, wave growth becomes saturated. Thisis the reason for the time lag between the maxi-mum wind speed and the resultant wave spectrahaving maximum energy.Our analysis of the wind and wave data indi-cates that, the higher frequency peak of the wave  Impact of sea breeze on wind-seas off Goa   233 Figure 5. Computed best-fit theoretical spectrum (dashed line) and the observed spectrum (solid line) on 11 April 1997.Figure 6. Each point gives wind speed (m/s) calculatedfrom TMA spectra versus observed wind speed (m/s). Theline is the least-squares fit to all the points. spectrum (figure 4) was forced by local winds alone.This implies that the energy associated with thiswind-sea spectrum should be a function of the localwind speed. More specifically, the diurnal variabil-ity of the energy under the spectrum should beassociated with diurnal variability of the winds.This may be expressed as, dE dt  =  α 1 U  3 − α 2 E, where  E   is the wave energy,  U   is the mean windspeed, and  α 1  and  α 2  are constants. The first termon the right hand side represents the generationterm. The energy transmitted to the water columnby wind has been taken to be proportional to  U  3 following Denman (1973) who used a similar para-meterization in connection with energy enteringthe ocean surface mixed-layer. The second term onthe right hand side is the dissipation term. The val-ues of   α 1  and  α 2  were chosen so as to get the bestfit to the observed variation. The observed energyand the one computed from the above equation isgiven in figure 2(c). Values of the constants usedin the computed spectrum were  α 1  = 1 . 93 × 10 − 4 ,and  α 2  = 3 . 0 × 10 − 5 . Comparison of the observedand computed energy shown in figure 2(c) encou-rages us to conclude that the high-frequency peakof the spectrum represents the wind-seas forced bylocal wind activity or sea breeze. 3. Summary and conclusions The analysis in the present study was restricted tothe period 1–15 April 1997; the correlation betweenwind speed inferred from the TMA spectrum andthe observed wind speed (figure 6) was quite sig-nificant for this period. This appears to be dueto the contribution to the high frequency peaksarising from the wind-seas. Contribution of theswell waves was insignificant for the period stud-ied here. March–April are the months during whichtransition from northeast monsoon (November–February) to the much stronger winds of the south-west monsoon occurs. Along the west coast of Indiathe winds due to the latter generally start blowingfrom the west in May. They strengthen once themonsoon sets in. The period we have analyzed istherefore rather special: the large-scale winds wereparticularly weak and hence the diurnal cycle dueto sea breeze could be identified.
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks