GC022-0011
Estimation of quartz content in soils using the VIS–NIR–SWIR spectral region
Estimation of quartz content in soils using the VIS–NIR–SWIR spectral region
Tuesday, 8 December 2020
Poster
Abstract:
Quartz is a silicate mineral that makes up around 20% of the earth's crust and is the most abundant surface mineral on the planet. Quartz is spectrally active in the LWIR region and has two stable absorption features at 8.21 µm and 8.85 µm. On the other hand, in the optical region, i.e., 400–2500 nm (VIS–NIR–SWIR), quartz has no significant spectral features and hence cannot be directly detected. Current attempts to mount optical image spectrometers in space are gaining much attention at several space agencies, such as NASA (EMIT), ESA (CHIME), DLR (EnMAP and DESIS) and ASI (PRISMA), among others. One of the missions assigned to these sensors is to map raw materials. However, the spectral recognition of quartz mineral across the optical region, where these sensors are active, remains problematic. In this study, we demonstrate that indirect relationships between the optical and LWIR spectral ranges can be used to assess the presence of quartz using solely the optical region. To that end, we used the Israeli soil spectral library (SSL) which characterizes arid and semi-arid soils. The soils underwent comprehensive chemical and mineral analyses along with spectral measurements across the VIS–NIR–SWIR region (reflectance) and LWIR region (emissivity), and mineral determination using XRD. Recently, the soil quartz–clay mineral index (SQCMI) was developed using the emissivity extracted from the LWIR region to account for quartz relative to clay minerals. Whereas a direct estimation of the quartz content (from XRD) using the gradient-boosting algorithm against the VIS–NIR–SWIR region provided poor results (R2 = 0.45, RMSE = 15.63, RPD = 1.32), the SQCMI index showed a high and significant correlation with the VIS–NIR–SWIR spectral region (R2 = 0.82, RMSE = 0.01, RPD = 2.34). These results suggest that the reflectance from the optical hyperspectral sensors across the 400–2500 nm spectral region can be used to estimate quartz content relative to clay content quite well.

