[MUSIC] Hello, as far as we consider wind or marine turbine efficiency the measured power curve rarely rely on saved power curves. Why? Does the turbine efficiency is deliberately overestimated, maybe not. Here is an example of a generic power curve calculated for a commercial wind turbine having rated power of 500 kilowatt and a maximum power coefficient of 0.4. And here is the measured power curve. The calculated and the measured curves are very close and the power coefficient could exceed 0.4 and sometimes it reaches 0.5. For this wind turbine, the upstream atmospheric conditions lead to various sources of uncertainties in power performance. We will quantify now the amplitude of the power fluctuations induced by the variation of density. The output power is directly proportional to the density rho, but the atmospheric density depends on both the temperature and pressure fluctuations. At the first order of approximation, we can assume that the atmosphere satisfy the ideal gas law. The density is then proportional to the pressure and is inversely proportional to the temperature. The standard air density used for the calculation of theoretical power curves, is 1.225 kilogram per cubic meter. If the on-site temperature, or the pressure divert from the standard value of 15 degrees celsius and one atmosphere, we may expect some deviations of the output power. If, for instance, the temperature varies by 10 degrees, what will be the relative impact on the density? According to the density curve, we can see that a fluctuation of 10 degrees that may occur between the night and the middle of a sunny day, will induce a relative density fluctuation around 3%. For exactly the same wind speed measure at the hub height, we could have an output power variation of 3 to 4% between the day and the night. If now we consider the atmospheric pressure fluctuations, similar variations could occur during few days from a high to a low pressure system. On this meteorological chart we can see at the mean sea level pressure varies from 1,020 hectopascal to 990 hectopascal across the atmospheric front. If this front moves rapidly we could experience a relative pressure fluctuation of 3% at the turbine location. Hence, in the atmosphere, the variations induced by the temperature or the pressure on the air density should be taken into account, to estimate a currently output power of wind turbines. In marine or river flows, the amplitude of the density fluctuations are different. The state equation of water is nonlinear and depends on temperature, salinity and pressure. Nevertheless, at the first order of approximation the impact of the pressure on density is negligible due to the incompressibility of water. Hence, for all practical use we will consider that the water density mainly depends on the temperature and the salinity. Performing an asymptotic expansion around a reference temperature T0 and a salinity S0, we can write a linear state equation introducing a thermal expansion coefficient, alpha and a high-line contraction coefficient beta. The typical values of this coefficient are weak. And for temperature variation of 10 degrees over salinity variation of 5 gram per liter, we will get a relative density variation of 0.1%. Hence we can neglect the impact of temperature or salinity on the output power of hydrokinetic turbines. To sum up we should take into account the density variations only for wind turbines. The changes in atmospheric temperature and pressure could affect the power production by few percent, hence density normalization is a standard procedure used for wind power assessment. Thank you.