DART-simulated panchromatic stereo images with RPC camera models over the SERC forest and bare terrain (paired along-track views, 0-70 degrees)

Panchromatic (450-800 nm) images simulated with the DART-Lux Monte Carlo radiative transfer model (DART 5.10.11, build v1432) over a 1000 x 4500 m (450 ha) domain at the Smithsonian Environmental Research Center (SERC), Maryland, USA, each with a Rational Polynomial Coefficient (RPC) camera model fitted to DART's pixel-to-scene correspondences and stored in the GeoTIFF RPC metadata. The images are ready for stereo processing, for example with the NASA Ames Stereo Pipeline. Two scenes with identical terrain and acquisition geometry: (1) forest_images: forest scene built from G-LiHT airborne lidar (31 July 2017) as 1 m turbid plant-area-density voxels over the 1 m DTM; (2) bare_terrain_textured_images: the same terrain without vegetation, with spatially textured Lambertian ground (the forest scene's ground material, loam_gravelly_brown, scaled by 0.6-1.4 in 8 equal-area classes of a Gaussian random field with 1.5 m correlation length, seed 7), so that stereo matching over bare ground has natural-like texture. Views: 68 per scene, view zenith 0, 0.25, 0.5, 0.75, 1-20 (1 degree step) and 25-70 degrees (5 degree step), in both along-track directions (DART view azimuth 0 and 180 degrees); file names Camera_NNN_ZA_DDD_FF_AA_DDD_FF.tiff give the view zenith (ZA) and azimuth (AA). The exact nadir view (ZA 0) was simulated but not used in the associated study. 1 m ground sample distance, 500 km pushbroom platform, no atmosphere, sun zenith 46.5 degrees and azimuth 102.4 degrees (31 July 2017, 14:00 UTC), no sensor noise. Float64, losslessly DEFLATE-compressed; UTM zone 18N. RPC heights: WGS84 ellipsoidal for the forest scene; EGM96 geoid heights for the bare-terrain scene (about 34.4 m above WGS84 ellipsoidal heights at the site). Associated article: Mkaouar et al., Separating acquisition-geometry and canopy controls on spaceborne stereo forest surfaces with paired physics-based simulation (ISPRS Journal of Photogrammetry and Remote Sensing, in preparation). Simulation framework: Yin et al. (2023a, b), Remote Sensing of Environment 298, 113825 and 113824. G-LiHT data: https://gliht.gsfc.nasa.gov/. Version 2 (29 September 2026): the bare-terrain images are replaced by the textured-ground simulation used in the associated article (the uniform-ground bare-terrain images of version 1 remain available in version 1 of this record). The forest images are unchanged.