Search for people | The University of Manchester - 03/09/2019Β Β· 2019-09-03Β Β· 03/09/2019 2...

Post on 20-Jul-2020

1 views 0 download

Transcript of Search for people | The University of Manchester - 03/09/2019Β Β· 2019-09-03Β Β· 03/09/2019 2...

13/09/2020

1

Waves

3. Statistics and Irregular Waves

Statistics and Irregular Waves

Statistics and Irregular Waves

● Real wave fields are not regular

● Combination of many periods, heights, directions

● Design and simulation require realistic wave statistics:

β€’ probability distribution of heights

β€’ energy spectrum of frequencies (and directions)

1

2

3

13/09/2020

2

3. STATISTICS AND IRREGULAR WAVES

3.1 Measures of height and period

3.2 Probability distribution of wave heights

3.3 Wave spectra

3.4 Reconstructing a wave field

3.5 Prediction of wave climate

Statistics and Irregular Waves

Measures of Wave Height

π»π‘šπ‘Žπ‘₯ Largest wave height in the sample π»π‘šπ‘Žπ‘₯ = max𝑖(𝐻𝑖)

π»π‘Žπ‘£π‘’ Mean wave height π»π‘Žπ‘£π‘’ =1

𝑁𝐻𝑖

π»π‘Ÿπ‘šπ‘  Root-mean-square wave height π»π‘Ÿπ‘šπ‘  =1

𝑁𝐻𝑖

2

𝐻 Ξ€1 3 Average of the highest 𝑁/3 waves 𝐻 Ξ€1 3 =1

𝑁/3

1

𝑁/3

𝐻𝑖

π»π‘š0 Estimate based on the rms surface elevation π»π‘š0 = 4 Ξ·2Ξ€1 2

𝐻𝑠 Significant wave height Either 𝐻 Ξ€1 3 or π»π‘š0

Measures of Wave Period

𝑇𝑠 Significant wave period (average of highest 𝑁/3waves)

𝑇𝑝 Peak period (from peak frequency of energy spectrum)

𝑇𝑒Energy period (period of a regular wave with same significant wave height andpower density; used in wave-energy prediction; derived from energy spectrum)

𝑇𝑧 Mean zero up-crossing period

4

5

6

13/09/2020

3

3. STATISTICS AND IRREGULAR WAVES

3.1 Measures of height and period

3.2 Probability distribution of wave heights

3.3 Wave spectra

3.4 Reconstructing a wave field

3.5 Prediction of wave climate

Statistics and Irregular Waves

Probability Distribution of Wave Heights

For a narrow-banded frequency spectrum the Rayleigh probability distribution is appropriate:

𝑃 height > 𝐻 = exp βˆ’ Ξ€(𝐻 )π»π‘Ÿπ‘šπ‘ 2

Cumulative distribution function:

𝐹 𝐻 = 𝑃 height < 𝐻 = 1 βˆ’ eβˆ’ ΀𝐻 π»π‘Ÿπ‘šπ‘ 2

Probability density function:

𝑓 𝐻 =d𝐹

d𝐻= 2

𝐻

π»π‘Ÿπ‘šπ‘ 2 eβˆ’ ΀𝐻 π»π‘Ÿπ‘šπ‘ 

2

Rayleigh Distribution

π»π‘Žπ‘£π‘’ ≑ 𝐸 𝐻 = ΰΆ±0

∞

𝐻 𝑓 𝐻 d𝐻

π»π‘Ÿπ‘šπ‘ 2 ≑ 𝐸 𝐻2 = ΰΆ±

0

∞

𝐻2 𝑓 𝐻 d𝐻

π»π‘Žπ‘£π‘’ =Ο€

2π»π‘Ÿπ‘šπ‘  = 0.886 π»π‘Ÿπ‘šπ‘ 

𝐻 Ξ€1 3 = 1.416 π»π‘Ÿπ‘šπ‘ 

𝐻 Ξ€1 10 = 1.800 π»π‘Ÿπ‘šπ‘ 

𝐻 Ξ€1 100 = 2.359 π»π‘Ÿπ‘šπ‘ 

𝑃 height > 𝐻 = exp βˆ’ Ξ€(𝐻 )π»π‘Ÿπ‘šπ‘ 2

Single parameter: π»π‘Ÿπ‘šπ‘ 

7

8

9

13/09/2020

4

Example

Near a pier, 400 consecutive wave heights are measured.Assume that the sea state is narrow-banded.

(a) How many waves are expected to exceed 2π»π‘Ÿπ‘šπ‘ ?

(b) If the significant wave height is 2.5 m, what isπ»π‘Ÿπ‘šπ‘ ?

(c) Estimate the wave height exceeded by 80 waves.

(d) Estimate the number of waves with a height between1.0 m and 3.0 m.

Near a pier, 400 consecutive wave heights are measured. Assume that the sea state isnarrow-banded.(a) How many waves are expected to exceed 2π»π‘Ÿπ‘šπ‘ ?(b) If the significant wave height is 2.5 m, what is π»π‘Ÿπ‘šπ‘ ?(c) Estimate the wave height exceeded by 80 waves.(d) Estimate the number of waves with a height between 1.0 m and 3.0 m.

𝑃 height > 𝐻 = eβˆ’ ΀𝐻 π»π‘Ÿπ‘šπ‘ 2

𝑃 height > 2π»π‘Ÿπ‘šπ‘  = eβˆ’4(a)

= 0.01832

In 400 waves, 𝑛 = 400 Γ— 0.01832

𝐻 Ξ€1 3 = 1.416 π»π‘Ÿπ‘šπ‘ (b)

2.5 = 1.416 π»π‘Ÿπ‘šπ‘ 

π‘―π’“π’Žπ’” = 𝟏. πŸ•πŸ”πŸ” 𝐦

(c) 𝑃 height > 𝐻 = eβˆ’ ΀𝐻 π»π‘Ÿπ‘šπ‘ 2=

80

400= 0.2

𝐻/π»π‘Ÿπ‘šπ‘ 2 = βˆ’ ln 0.2

𝐻 = π»π‘Ÿπ‘šπ‘  Γ— βˆ’ ln0.2 = 𝟐. πŸπŸ’πŸŽ 𝐦

𝑃 1.0 < height < 3.0 = 𝑃 height > 1.0 βˆ’ 𝑃 height > 3.0

= eβˆ’ 1.0/π»π‘Ÿπ‘šπ‘ 2βˆ’ eβˆ’ 3.0/π»π‘Ÿπ‘šπ‘ 

2

= 0.6699In 400 waves, 𝑛 = 400 Γ— 0.6699

(d)

= πŸ•. πŸ‘πŸπŸ–

= πŸπŸ”πŸ–. 𝟎

3. STATISTICS AND IRREGULAR WAVES

3.1 Measures of height and period

3.2 Probability distribution of wave heights

3.3 Wave spectra

3.4 Reconstructing a wave field

3.5 Prediction of wave climate

Statistics and Irregular Waves

10

11

12

13/09/2020

5

Regular vs Irregular Waves

Regular wave:- single frequency

Irregular wave:- many frequencies

Wave energy depends on Ξ·2

Energy Spectrum

● β€œSpectral” means β€œby frequency”

● A spectrum is usually determined by a Fourier transform

● This splits a signal up into its component frequencies

● Energy in a wave is proportional to Ξ·2, where Ξ· is surface displacement

● The energy spectrum, or power spectrum, is the Fourier transform of Ξ·2

Energy Spectrum

For regular waves (single frequency):

𝐸 =1

2ρ𝑔𝐴2 = ρ𝑔 )Ξ·2(𝑑 =

1

8ρ𝑔𝐻2

For irregular waves (many frequencies):

𝑆 𝑓 d𝑓

ࢱ𝑓1

𝑓2

𝑆 𝑓 d𝑓

is the β€œenergy” in a small interval d𝑓 near frequency 𝑓

is the β€œenergy” between frequencies 𝑓1 and 𝑓2

(strictly: energy/ρ𝑔)

The energy spectrum 𝑆(𝑓) is determined by Fourier transforming Ξ·2

Ξ· = 𝐴 cos π‘˜π‘₯ βˆ’ ω𝑑

13

14

15

13/09/2020

6

Model SpectraOpen ocean: Bretschneider spectrum

Fetch-limited seas: JONSWAP spectrum

Key parameters: peak frequency 𝑓𝑝 ( =1/(peak period, 𝑇𝑝) )

significant wave height π»π‘š0

Model Spectra

Bretschneider spectrum:

JONSWAP spectrum:

𝑆 𝑓 =5

16π»π‘š02

𝑓𝑝4

𝑓5exp βˆ’

5

4

𝑓𝑝4

𝑓4

𝑆 𝑓 = 𝐢 π»π‘š02

𝑓𝑝4

𝑓5exp βˆ’

5

4

𝑓𝑝4

𝑓4γ𝑏

𝜎 = ࡝0.07 𝑓 < 𝑓𝑝0.09 𝑓 > 𝑓𝑝

𝑏 = exp βˆ’1

2

΀𝑓 𝑓𝑝 βˆ’ 1

Οƒ

2

Ξ³ = 3.3

Significant Wave Height, π‘―π’ŽπŸŽ

Bretschneider spectrum: 𝑆 𝑓 =5

16π»π‘š02

𝑓𝑝4

𝑓5exp βˆ’

5

4

𝑓𝑝4

𝑓4

Total energy:𝐸

ρ𝑔≑ ΰΆ±

0

∞

𝑆 𝑓 d𝑓 =1

16π»π‘š02

π»π‘š0 = 4 ΀𝐸 ρ 𝑔 = 4 )Ξ·2(𝑑

Total energy for a regular wave:𝐸

ρ𝑔=1

8π»π‘Ÿπ‘šπ‘ 2

Same energy if π»π‘š0 = 2π»π‘Ÿπ‘šπ‘  = 1.414π»π‘Ÿπ‘šπ‘ 

Rayleigh distribution: 𝐻𝑠 = 𝐻 Ξ€1 3 = 1.416π»π‘Ÿπ‘šπ‘ 

π‘―π’ŽπŸŽ and 𝑯 ΀𝟏 πŸ‘ can be used synonymously for 𝑯𝒔 ...

… but π»π‘š0 is easier to measure!

16

17

18

13/09/2020

7

Finding Wave Parameters From a Spectrum

● Surface displacement Ξ·(𝑑) is measured (e.g. wave buoy)

● Ξ·2(𝑑) is Fourier-transformed to get energy spectrum 𝑆(𝑓)

● Height and period parameters are deduced from the peak and the momentsof the spectrum:

π‘šπ‘› = ΰΆ±0

∞

𝑓𝑛𝑆 𝑓 d𝑓

Peak period: 𝑇𝑝 =1

𝑓𝑝

Significant wave height: 𝐻𝑠 = π»π‘š0 = 4 π‘š0

Energy period: 𝑇𝑒 =π‘šβˆ’1

π‘š0

Multi-Modal Spectra

frequency f (Hz)

spect

ral d

ensi

ty S

(m

^2 s

)

swell

wind

3. STATISTICS AND IRREGULAR WAVES

3.1 Measures of height and period

3.2 Probability distribution of wave heights

3.3 Wave spectra

3.4 Reconstructing a wave field

3.5 Prediction of wave climate

Statistics and Irregular Waves

19

20

21

13/09/2020

8

Using a Spectrum To Generate a Wave Field

Ξ· 𝑑 = )π‘Žπ‘– cos(π‘˜π‘–π‘₯ βˆ’ ω𝑖𝑑 βˆ’ πœ™π‘–

amplitude

𝑆 𝑓𝑖 d𝑓 = 𝐸𝑖 =1

2π‘Žπ‘–2 π‘Žπ‘– = 2𝑆 𝑓𝑖 d𝑓

(angular) frequency

ω𝑖 = 2π𝑓𝑖

wavenumber

πœ”π‘–2 = π‘”π‘˜π‘– tanhπ‘˜π‘–β„Ž

random phase

Simulated Wave Field𝑇𝑝 = 8 s

𝐻𝑠 = 1.0 m

β„Ž = 30 mRegular waves:

Irregular waves:

Focused waves:

Example

An irregular wavefield at a deep-water location is characterisedby peak period of 8.7 s and significant wave height of 1.5 m.

(a) Provide a sketch of a Bretschneider spectrum, labelling bothaxes with variables and units and indicating the frequenciescorresponding to both the peak period and the energy period.

Note: Calculations are not needed for this part.

(b) Determine the power density (in kW m–1) of a regular wavecomponent with frequency 0.125 Hz that represents thefrequency range 0.12 to 0.13 Hz of the irregular wave field.

22

23

24

13/09/2020

9

An irregular wavefield at a deep-water location is characterised by peak period of 8.7 s andsignificant wave height of 1.5 m.

(a) Provide a sketch of a Bretschneider spectrum, labelling both axes with variables and unitsand indicating the frequencies corresponding to both the peak period and the energyperiod.

frequency f (Hz)

spect

ral d

ensi

ty S

(m

^2 s

)

fp fe

An irregular wavefield at a deep-water location is characterised by peak period of 8.7 s andsignificant wave height of 1.5 m.

(b) Determine the power density (in kW m–1) of a regular wave component with frequency0.125 Hz that represents the frequency range 0.12 to 0.13 Hz of the irregular wave field.

𝑓 = 0.125 Hz

Δ𝑓 = 0.01 Hz

𝑇𝑝 = 8.7 s 𝐻𝑠 = 1.5 m

𝑓𝑝 =1

𝑇𝑝

𝑆 𝑓 = 1.645 m2 s

𝐸 = ρ𝑔 Γ— 𝑆 𝑓 Δ𝑓

𝑃 = 𝐸𝑐𝑔 = 𝐸(𝑛𝑐) Deep water: 𝑛 =1

2

𝑐 =𝑔𝑇

2Ο€

𝑇 =1

𝑓= 8 s

𝑷 = 𝟏. πŸŽπŸ‘πŸ‘ π€π–π¦βˆ’πŸ

= 0.1149 Hz

= 165.4 J mβˆ’2

= 12.49 m sβˆ’1

𝑆 𝑓 =5

16𝐻𝑠2𝑓𝑝4

𝑓5ex p( βˆ’

5

4

𝑓𝑝4

𝑓4)

3. STATISTICS AND IRREGULAR WAVES

3.1 Measures of height and period

3.2 Probability distribution of wave heights

3.3 Wave spectra

3.4 Reconstructing a wave field

3.5 Prediction of wave climate

Statistics and Irregular Waves

25

26

27

13/09/2020

10

Terminology

A wave climate or sea state is a model wave spectrum, usually defined by a representative height and period, for use in:

● forecasting (from a weather forecast in advance)

● nowcasting (from ongoing weather)

● hindcasting (reconstruction from previous event)

Prediction of Sea State

Require: significant wave height, 𝐻𝑠significant wave period, 𝑇𝑠 (or peak period, 𝑇𝑝)

Depend on: wind speed, π‘ˆfetch, 𝐹duration, 𝑑gravity, 𝑔

Dimensional analysis:

Empirical functions: JONSWAP or SMB curves

ΰ΅°π‘”π»π‘ π‘ˆ2 = π‘“π‘’π‘›π‘π‘‘π‘–π‘œπ‘›(

𝑔𝐹

π‘ˆ2 ,𝑔𝑑

π‘ˆ

α‰‡π‘”π‘‡π‘π‘ˆ

= π‘“π‘’π‘›π‘π‘‘π‘–π‘œπ‘›(𝑔𝐹

π‘ˆ2 ,𝑔𝑑

π‘ˆ

Fetch-Limited vs Duration-Limited

distance travelled by wave energy greater than fetch F

F

distance travelled by wave energy less than fetch F

F

Feff

FETCH-LIMITED

DURATION-LIMITED

28

29

30

13/09/2020

11

JONSWAP CurvesFor fetch-limited waves:

π‘”π»π‘ π‘ˆ2

= 0.0016𝑔𝐹

π‘ˆ2

Ξ€1 2

(up to maximum 0.2433)

π‘”π‘‡π‘π‘ˆ

= 0.286𝑔𝐹

π‘ˆ2

Ξ€1 3

(up to maximum 8.134)

Waves are fetch-limited provided the storm has blown for a minimum timeπ‘‘π‘šπ‘–π‘› given by

𝑔𝑑

π‘ˆ π‘šπ‘–π‘›= 68.8

𝑔𝐹

π‘ˆ2

Ξ€2 3

up to maximum 7.15 Γ— 104

Otherwise the waves are duration-limited, and the fetch used to determineheight and period is an effective fetch determined by inversion using the actualstorm duration 𝑑:

𝑔𝐹

π‘ˆ2𝑒𝑓𝑓

=1

68.8

𝑔𝑑

π‘ˆ

Ξ€3 2

JONSWAP Curves (reprise)

𝐻𝑠 = 0.0016 𝐹 Ξ€1 2

𝑇𝑝 = 0.286 𝐹 Ξ€1 3

ΖΈπ‘‘π‘šπ‘–π‘› = 68.8 𝐹 Ξ€2 3

𝐹 ≑𝑔𝐹

π‘ˆ2 , Ƹ𝑑 ≑𝑔𝑑

π‘ˆ, 𝐻𝑠 ≑

π‘”π»π‘ π‘ˆ2 , 𝑇𝑝 ≑

π‘”π‘‡π‘π‘ˆ

JONSWAP and SMB Curves

Solid lines: JONSWAP

Dashed lines: SMB

31

32

33

13/09/2020

12

Example

(a) Wind has blown at a consistent π‘ˆ = 20m sβˆ’1

over a fetch 𝐹 = 100 km for 𝑑 = 6 hrs .Determine𝐻𝑠 and 𝑇𝑝 using the JONSWAP curves.

(b) If the wind blows steadily for another 4 hourswhat are 𝐻𝑠 and 𝑇𝑝?

(a) Wind has blown at a consistent π‘ˆ = 20 m sβˆ’1 over a fetch 𝐹 = 100 km for 𝑑 = 6 hrs.Determine 𝐻𝑠 and 𝑇𝑝 using the JONSWAP curves.

π‘ˆ = 20 m sβˆ’1

𝐹 = 105 m

𝑑 = 6 Γ— 3600 = 21600 s

𝐹 ≑𝑔𝐹

π‘ˆ2

Ƹ𝑑 =𝑔𝑑

π‘ˆ

ΖΈπ‘‘π‘šπ‘–π‘› = 68.8 𝐹 Ξ€2 3

ΖΈπ‘‘π‘šπ‘–π‘› = 68.8 𝐹 Ξ€2 3 = 12510

𝐻𝑠 = 0.0016 𝐹1/2

𝑇𝑝 = 0.286 𝐹1/3

𝐹 ≑𝑔𝐹

π‘ˆ2Ƹ𝑑 ≑

𝑔𝑑

π‘ˆπ»π‘  ≑

π‘”π»π‘ π‘ˆ2

Duration-limited

𝐹𝑒𝑓𝑓 =Ƹ𝑑

68.8

Ξ€3 2

𝐻𝑠 = 0.0016 𝐹𝑒𝑓𝑓΀1 2

𝑇𝑝 = 0.286 𝐹𝑒𝑓𝑓΀1 3

𝐻𝑠 = 𝐻𝑠 Γ—π‘ˆ2

𝑔

𝑇𝑝 = 𝑇𝑝 Γ—π‘ˆ

𝑔

= 1910

= 0.06993

= 3.548

= 𝟐. πŸ–πŸ“πŸ 𝐦

= πŸ•. πŸπŸ‘πŸ‘ 𝐬

= 2453

= 10590

(b) If the wind blows steadily for another 4 hours what are 𝐻𝑠 and 𝑇𝑝?

π‘ˆ = 20 m sβˆ’1

𝐹 = 105 m

𝑑 = 10 Γ— 3600 = 36000 s

𝐹 = 2453

Ƹ𝑑 =𝑔𝑑

π‘ˆ

ΖΈπ‘‘π‘šπ‘–π‘› = 68.8 𝐹 Ξ€2 3

ΖΈπ‘‘π‘šπ‘–π‘› = 12510

𝐻𝑠 = 0.0016 𝐹1/2

𝑇𝑝 = 0.286 𝐹1/3

𝐹 ≑𝑔𝐹

π‘ˆ2Ƹ𝑑 ≑

𝑔𝑑

π‘ˆπ»π‘  ≑

π‘”π»π‘ π‘ˆ2

Fetch-limited

𝐻𝑠 = 0.0016 𝐹1/2

𝑇𝑝 = 0.286 𝐹1/3

𝐻𝑠 = 𝐻𝑠 Γ—π‘ˆ2

𝑔

𝑇𝑝 = 𝑇𝑝 Γ—π‘ˆ

𝑔

= 0.07924

= 3.857

= πŸ‘. πŸπŸ‘πŸ 𝐦

= πŸ•. πŸ–πŸ”πŸ‘ 𝐬

= 17660

34

35

36

13/09/2020

13

Examples on Blackboard

Try examples 10-13

37