Effects of land cover type and greenness on advanced very high resolution radiometer bidirectional reflectances: Analysis and removal

Abstract

The objectives of the study are (1) to examine the effects of land cover type, green biomass, and solar zenith angle (SZA) on the bidirectional reflectance distributions (BRDs) of the visible and near‐infrared advanced very high resolution radiometer (AVHRR) data from NOAA 11 over terrestrial surfaces; and (2) to correct for these effects by developing BRD functions (BRDF). Four land cover types are examined: barren, grassland, forest, and cropland. The data used consist of 1‐km daily AVHRR measurements for three growing seasons collected over 19 homogeneous land sites (20×20 km2 for each site) in the conterminous United States and parts of Canada. BRD is found to be strongly land cover dependent. For the same cover type, BRD is altered significantly by the green biomass present (represented by the normalized difference vegetation index (NDVI)). The effects of SZA on BRDs are also observed over all surface types under investigation. Semiempirical BRDFs were developed to account for these effects that are the functions of the SZA, satellite viewing zenith angle, relative azimuth angle, and NDVI. Good agreements were found between the observed and the modeled bidirectional dependencies for wide ranges of NDVI and SZA. A single BRDF appears to be sufficient for bidirectional correction of the clear‐sky AVHRR measurements made over a specific land cover type throughout a season. Finally, the developed BRDFs were used to normalize AVHRR reflectance data to a common geometry and to infer the hemispherical albedos for monitoring the seasonal variations of land surfaces.

Seasonal Separation of African Savanna Components Using Worldview-2 Imagery: A Comparison of Pixel- and Object-Based Approaches and Selected Classification Algorithms

“…Therefore it is important to note, that given the varying viewing angles the BRDF might have an effect on the classification accuracy results presented in this study, as for instance showed by Sue et al (2009) [69], Vanonckelen et al (2013) [70] and Wu and Cihlar (1995) [71].…”

CLARA-SAL: a global 28 yr timeseries of Earth's black-sky surface albedo

“…The narrow-to-broadband conversion algorithms have reported uncertainties of 5-10 % (Liang, 2000;Xiong et al, 2002). The BRDF correction and spectral albedo calculation for vegetation (Roujean et al, 1992;Wu et al, 1995) is estimated to have an uncertainty in the same order of magnitude. Cloud mask misclassifications cause the largest retrieval errors, but they occur only sporadically so that their effect on the temporal means is generally low.…”

“…Next, anisotropic reflectance (over vegetated/barren surfaces) is corrected and the surface reflectances are converted into spectral albedo after Wu et al (1995) and Roujean et al (1992). Broadband albedo is retrieved according to land cover type.…”

“…For non-snow covered land surfaces, we correct for anisotropy in the surface reflectance prior to the conversion into spectral albedo. The process uses the kernel-based BRDF/albedo model by Roujean et al (1992), using land cover-specific BRDF coefficients from Wu et al (1995). Landcover information is obtained with a nearest-neighbour retrieval from a 1 km USGS land cover dataset for each GAC overpass.…”

“…The preprocessing by the PPS module (Dybbroe et al, 2005) also creates a cloud mask and provides CLARA-SAL with Sun/satellite geometry data and a land cover classification for each pixel in the image swath. AVHRR GAC reflectances from channels 1 and 2 are used in CLARA-SAL Atmospheric correction with SMAC (Rahman & Dedieu), 1994 Compute BRDF correction for vegetation (Wu et al, 1995), NDVI as main parameter VEGT.…”