The Vaisala 2.1 dataset is the second dataset Vaisala created using the new REST2 clear sky algorithm. The main change relative to the V2.0 dataset is the use of aerosol and water vapor data from NASA's MERRA-2 reanalysis.

The REST2 model is a parameterized version of Dr. Gueymard's SMARTS radiative transfer model, as described in C. A. Gueymard, "REST2: High-performance solar radiation model for cloudless-sky irradiance, illuminance, and photosynthetically active radiation – validation with a benchmark dataset," Solar Energy, vol. 82, pp. 272-285, 2008.

We are using version 9.0 of the REST2 clear sky code. Defaults are currently used for ozone, albedo, single scattering albedo, and asymmetry parameter values. Vaisala 2.1 uses NASA's MERRA-2 (Modern Era-retrospective Analysis for Research and Applications) dataset as the source for aerosol optical depth (AOD) and precipitable water.

Whether the Perez or REST2 clear sky model is used, GHI is calculated by modulating the clear sky GHI values by the cloud index values to calculate the cloudy sky GHI value. In the Perez model, DNI is calculated from GHI using Dr. Perez's DIRINT methodology. In the REST2 model, a second modulation function is used to calculate DNI from the cloud index and the clear sky DNI value. Diffuse is then calculated from the GHI and DNI values and solar zenith angle.

Vaisala uses a proprietary cloud index calculation methodology. To derive the cloud index, the satellite radiance measurements are first processed to account for instrument degradation over time. The proprietary Vaisala cloud algorithm then divides the satellite measurements into bins by solar azimuth and zenith angle, and treats each bin independently for the calculation of a cloud index. The fit between cloud index, clear-sky GHI, clear-sky DNI, and observations is done on a regional basis.