2.2.5. Albedo Validation

Using the BRDF parameters at HLS resolution, we derive the surface albedo which is validated against field measurements. The surface albedo is estimated following the same methodology as the official MCD43 MODIS product [7] by integrals of the BRDF model through the black-sky albedo and the white-sky albedo [7]. Surface albedo is typically estimated in the spectral range 280–2800 nm and is therefore comparable with the broadband MODIS albedo (300–5000 nm). Therefore, we estimate the broadband albedo using the narrow to broadband equation proposed by Liang [23].

For validation purposes, assuming that every pyranometer has an effective field of view of 81◦, the downward pyranometer covers an area equal to tan(81◦)\*height. According to this equation and

considering that the tower height in every SURFRAD is 10 m, the area covered by the pyranometer is 63 m in radius (3 by 3 HLS pixels). Then, the weighted average of the nine HLS pixels is estimated considering the angle of the instrument and the center of each pixel. The same approach is applied to OzFlux sites to account for the field of view as function of the height of the instrument.

SURFRAD instruments provide radiation measurements every minute, and we just consider the observations that match the time of Landsat 8 and Sentinel 2 overpass time. On the other hand, since OzFlux instruments provide radiation measurements every 30 min, a linear interpolation is done to estimate the value at the Landsat 8 and Sentinel 2 overpass time. A careful analysis of cloud-free conditions around the tower was performed to discard observations that may include residual effects.
