User:Pitpied/sandbox

Hyperspectral bare soil index (HBSI)

The Hyperspectral Bare Soil Index (HBSI) is a remote‑sensing spectral index designed to detect and map bare soil patches more accurately within complex agricultural and urban landscapes. The index was developed by Dr. Eric Ariel L. Salas and Dr. Sakthi S. Kumaran and first introduced in 2023 in the journal Land.^[1]

Bare soil plays a critical role in ecosystem monitoring, agricultural management, and land degradation assessment. It influences hydrological processes, carbon cycling, and vegetation dynamics, making it a fundamental component of Earth system functioning. Accurate identification of bare soil areas is essential for understanding land surface conditions, evaluating erosion risks, and supporting precision farming. However, conventional remote sensing indices often struggle to distinguish bare soil from spectrally similar surfaces such as built-up areas or sparsely vegetated lands. The HBSI addresses this challenge by utilizing the enhanced spectral resolution of hyperspectral and multispectral imagery. By combining blue, near-infrared (NIR), and shortwave infrared (SWIR2) bands, HBSI improves the classification accuracy of bare soil pixels, especially in heterogeneous and mixed land cover environments.

Development and Spectral Basis

Spectral bands: HBSI uses blue, near-infrared (NIR), and shortwave infrared 2 (SWIR2) bands—particularly effective in differentiating soil from other surfaces.

Hyperspectral enhancement: The incorporation of the blue band from the Airborne Visible Infrared-Imaging Spectrometer-Next-Generation (AVIRIS-NG) hyperspectral imagery is a key innovation, boosting the index's power compared to earlier bare-soil indices.

HBSI Equation

$HBSI={\frac {(SWIR2+Green)-(NIR+Blue)}{(SWIR2+Green)+(NIR+Blue)}}$

The spectral bands are averaged to represent the regions of green (500–600 nm), blue (400–500 nm), NIR (750–850 nm), and SWIR2 (~2080–2350 nm).

Application

HBSI showed to be effective in separating muddy surfaces from water, a common challenge in intertidal zone mapping^[2]. It showed strong Spectral Discrimination Index (SDI) performance especially in distinguishing mudflats from water bodies in two separate study areas:

Yellow River Delta: HBSI = 1.26
Qinzhou Bay: HBSI = 1.27

References

^ Salas, E.A.L., Kumaran, S.S. (2023) 'Hyperspectral Bare Soil Index (HBSI): Mapping Soil Using an Ensemble of Spectral Indices in Machine Learning Environment.' Land, vol. 12(7), 1375. https://doi.org/10.3390/land12071375
^ Yang, G., Shao, C., Zuo, Y., Sun, W., Huang, K., Wang, L., Chen, B., Meng, X., Ge, Y. (2024) 'MFI: A mudflat index based on hyperspectral satellite images for mapping coastal mudflats.' International Journal of Applied Earth Observation and Geoinformation, vol. 133, 104140. https://doi.org/10.1016/j.jag.2024.104140

Pitpied (talk) 23:57, 7 June 2025 (UTC)

[1] Salas, E.A.L., Kumaran, S.S. (2023) 'Hyperspectral Bare Soil Index (HBSI): Mapping Soil Using an Ensemble of Spectral Indices in Machine Learning Environment.' Land, vol. 12(7), 1375. https://doi.org/10.3390/land12071375

[2] Yang, G., Shao, C., Zuo, Y., Sun, W., Huang, K., Wang, L., Chen, B., Meng, X., Ge, Y. (2024) 'MFI: A mudflat index based on hyperspectral satellite images for mapping coastal mudflats.' International Journal of Applied Earth Observation and Geoinformation, vol. 133, 104140. https://doi.org/10.1016/j.jag.2024.104140

[1]

[2]