Weather Based Irradiance Modeling of the Solar Strand at the University at Buffalo

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Tuesday, 6 January 2015: 11:15 AM
224B (Phoenix Convention Center - West and North Buildings)
John McGreevy, SUNY, Buffalo, NY; and T. Singh and M. Majji

A method of predicting the solar irradiance at the Solar Strand is developed and presented. The parameters of time of day, cloud cover, and relative humidity are used in a polynomial model to estimate the transmittance of the atmosphere, which in turn provides the estimated irradiance at the Solar Strand.

It is easy to measure the solar irradiance that reaches the top of the Earth's atmosphere as it is just a geometric problem of finding the angle and distance between the Sun and the surface of the Earth at that specific location and time. The difficulty arises in modeling the fraction of the irradiance which makes its way through the atmosphere to the surface. Because the irradiance at the top of the atmosphere is a known quantity, it can be used to normalize the measured irradiance at the surface at the Solar Strand to determine the transmittance of the atmosphere. This provides a sort of “atmospheric efficiency” measure of the fraction of the solar radiation which is transmitted through the atmosphere to the solar panels. In theory, it should be easier to model the transmittance instead of the total irradiance, as the weather has a direct impact on the atmosphere's transmittance as opposed to the total irradiance.

To test this method, the needed weather data from 2013 as recorded at the Buffalo airport by NOAA is downloaded from NOAA's National Climatic Data Center. This dataset has recorings of the cloud cover, visibility, relative humidity, and other weather data at three hour intervals. This data is interpolated to align with the the irradiance data recorded at the Solar Strand location in 15 minute intervals. A least squares fit was performed on the data for the entire year of 2013. To test the accuracy of the model, the values of the model based on the data in 2013, is used to see if it can accurately forecast the first three months of 2014, which is the most up-to-date weather archive from NOAA's National Climatic Data Center. The forecast results are very encouraging and are being further verified on the latest set of data.