Collecting Data from Oklahoma Wind Sites
(rev. 01/16)

If a wind energy plant is going to be built in Oklahoma, the most important factor to consider is obvious: location. The purpose of this assignment is to collect data on the wind conditions at the Mesonet sites across Oklahoma and project which site has the greatest potential for a wind farm. To ensure the quality of our choices for possible wind farm sites, each must collect the accurate data for which they are responsible.

To project which site will be optimal for constructing a wind farm, we will analyze average wind speeds over the course of one year for a particular site. By doing so, we will gain information on which sites seem to have more wind, and therefore more energy. On top of this, we will also begin to analyze correlations in data, and how we can know that two variables are linked. To do this we will look at the relationship between solar radiation and temperature, and use a linear regression to see how closely related they truly are.

To collect data, go to Here you can select the month, year, and location, and a corresponding table will appear. Each student is responsible for collecting data from three sites for the year 2014. A list of the sites and student assignments can be found in an Excel file on the ?Course Content? page of Harvey.

Data Collection:
Create a table in Excel that contains the following information that looks similar to the image below. Copy and paste this table three times to fill in data for each of the three sites that you?ve been assigned.

Mesonet Site
Month # Day w/Data Ave. Wind Speed (mph) Days * Ave. Wind Speed

Total Days with Data Sum Of Days * Ave. Wind
Average Monthly Air
Average Daily Wind Speed

When you select a month, year, and site, a table at the URL above, a table similar to the following will appear:

To determine the # of days with data, count the number of NAs in the SPEED AVG column (highlighted in red) and subtract it from the number of days in the month. Next, record the average wind speed for the month (highlighted in purple). The two other boxed numbers will be important in the later questions, but for now, ignore them.

For the last column in the table, multiply the average wind speed for the month by the number of days that data was collected. To get a rough estimate of how windy the site is, sum up the last column and divide by the total days that data was present. Report your calculations in the table similar to the example figure on the first page.

This assignment will require the examination of 36 files! The work must be completed in a careful and diligent manner. You might want to spread data collection over a couple of days to prevent making errors from being tired.

Questions and Exercises:
Submit both a Word Document and an Excel Spreadsheet for this lab that answers the following:

1. Set up a single Excel Spreadsheet with a formatted table for each of the assigned Mesonet sites. Collect the data and perform the calculations as described above.

2. Which, if any, of your sites have months that lack more than three days? List the site, number of missing days, and month for each. How does missing data affect this lab?

3. Give an example of extra data that you could collect that would help in determining how windy a site is. (For example, you could collect dew temperature data, but that won?t have an affect on wind speeds?)

4. It is important to look for correlations, or trends in data. To determine if two variables influence one another, we can construct a linear plot comparing these two variables. The two variables we are interested here are the Average Temperature and the Average Solar Radiation (the teal and the green box on the monthly summaries, respectively). Pick one of your sites and for each month record these two numbers. After you have made a chart in excel with this data, we are going to plot the two on a scatter plot. Arrange your chart so that the independent variable is the left column, and the dependent variable is the right. Highlight both columns and click Insert > Chart > Scatter. Right-click the data points and click Add Trend line. Choose a linear trend and display the R2.

In statistics, the R2 term is called the coefficient of determination. This value lies somewhere between 0 and 1 and indicates how accurately the model reflects the data. Thus, an R2 of 1 means there is a perfect correlation while an R2 of 0 means there is no correlation between to the two variables. Interpret your R2 value. What does this say about the relatedness between temperature and solar radiation?

5. In the question above, do your R2 values make sense? In two or three sentences, explain why or why not.