Include measurement station weighting in the turbine wind resource assessment

[jonssonchristian]

The turbine wind resource assessment data model currently does not include any information on how the predictions based on different measurement stations were weighed at each turbine to arrive at final predictions. This is a key piece of information that I would argue is important to include. It should be relatively easy to accommodate. I think we essentially just need weights binned by turbine ID and measurement stations ID, where the sum of the weights for each turbine must equal unity.

It is possible that different weights are used for different variables, such as distance and ground level altitude weighted for wind speed and just using the closest station for wind direction. Or it can be that only the long-term wind speed is weighted and the wind speed and direction frequency distribution is taken from the measurement station considered most representative for each turbine. To avoid a complex data model, I propose that we only include the effective weights relating to the long-term wind speed. That would mean that the final predicted long-term wind speed at a turbine should be equal to the sum of the predicted long-term wind speed at that turbine based on each measurement station multiplied with the weight of that station for the turbine in question. The weights on long-term wind speed should also correspond to the weights applicable to wind speed uncertainty.

I would propose including this data not under results but under a separate section for results weighting.

[jonssonchristian]

Pull request #57 includes a proposed extension to address this.