Dear Omid, 

Although the Weibull distribution is adopted also in the international standards, in academic field some other methods have been investigated and compared to Weibull distribution to describe the wind speed in a given site. Some other methods or techniques are based on Machine Learning theory but they are still at the early stages.

Here below you find some of the papers I collected in the past:

http://www.sciencedirect.com/science/article/pii/S0196890407003536

http://www.sciencedirect.com/science/article/pii/S0360544211000867

http://www.sciencedirect.com/science/article/pii/S0960148112005976

http://www.sciencedirect.com/science/article/pii/S0196890408000228

http://www.ijrer.org/index.php/ijrer/article/view/92

http://www.sciencedirect.com/science/article/pii/S1364032112002584

http://www.sciencedirect.com/science/article/pii/S0960148104004033

and others available...

I hope it helps.

Best,

Stefano

Dear Omid, 

Although the Weibull distribution is adopted also in the international standards, in academic field some other methods have been investigated and compared to Weibull distribution to describe the wind speed in a given site. Some other methods or techniques are based on Machine Learning theory but they are still at the early stages.

Here below you find some of the papers I collected in the past:

http://www.sciencedirect.com/science/article/pii/S0196890407003536

http://www.sciencedirect.com/science/article/pii/S0360544211000867

http://www.sciencedirect.com/science/article/pii/S0960148112005976

http://www.sciencedirect.com/science/article/pii/S0196890408000228

http://www.ijrer.org/index.php/ijrer/article/view/92

http://www.sciencedirect.com/science/article/pii/S1364032112002584

http://www.sciencedirect.com/science/article/pii/S0960148104004033

and others available...

I hope it helps.

Best,

Stefano

Thank you Mr. Stefano,

Dear Stefano,

Is it possible to synthesize hourly wind speed data (8760 hr) by using only Average Monthly Wind Speed data (12 months) provided by NASA-SSE at  50m above surface of the earth.

Dear Tarun,

I guess you are talking about spatial data of wind speed.

This type of "downscaling" is quite unlikely to be done. Deriving a 12 data sets into a 8760 data sets is not really feasible unless you try to use other hourly global data that affect the wind speed flow and you try to find some correlation between the monthly and the hourly data. You can do such work for 6h data down to hourly data using other partial hourly data measurement. But Monthly data down to Hourly data is quite hard both in term modeling and accuracy.

wind blows slowly at low altitude and then increases at higher altitude,for calculating k,c weibull parameters in other height you can use this :

kh = k0 [1-1.088*ln(h/50)]

ch = c0 [(h/50)^n] 

n = [0.37 - 0.088 * ln(c0)]

desired height = h

n= the power law exponent (coefficient)

k0 = shape parameter in 50m

c0 = scale parameter in 50m

If you're estimating weibull parameters in order to generate synthetic wind speed time series for conversion to wind power time series, I'd recommend reading this article which allows for the generation of synthetic, correlated time series that preserves (a) the autocorrelation function of each time series, (b) cross correlation between simulated time series, and (c) the pdfs of the individual time series. This can allow you to synthesize time series from multiple wind turbines (or solar power, system load, etc) that have non-zero temporal correlation, which is the case in most power systems.

http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=4453993

Article MCMC for wind power simulation