The New Hyperspectral Satellite PRISMA: Imagery for Forest Types Discrimination

Area of Science:

• Earth Observation
• Remote Sensing
• Forestry

Background:

• Traditional multispectral sensors struggle to differentiate forest types with similar spectral signatures.
• Hyperspectral imagery offers a more detailed spectral representation for accurate classification.
• The Italian Space Agency's PRISMA sensor provides continuous spectral data across 240 bands.

Purpose of the Study:

• To compare the forest type discrimination capabilities of the PRISMA hyperspectral sensor against the Sentinel-2 MSI.
• To evaluate sensor performance using pairwise separability analysis across different forest nomenclature systems.
• To assess the impact of spectral resolution on forest classification accuracy.

Main Methods:

• Utilized pairwise separability analysis in two Italian study areas.
• Employed two distinct forest nomenclature systems and four separability metrics.
• Compared PRISMA hyperspectral data with Sentinel-2 Multi-Spectral Instrument (MSI) data.

Main Results:

• PRISMA demonstrated superior forest type discrimination compared to Sentinel-2 MSI across all tested scenarios.
• Classification performance improved with PRISMA as the complexity of the forest nomenclature increased.
• PRISMA achieved an average improvement of 40% for broadleaf vs. coniferous discrimination and 102% for five main tree species groups.

Conclusions:

• Hyperspectral sensors like PRISMA are highly effective for detailed forest type mapping.
• PRISMA's advanced spectral capabilities surpass traditional multispectral instruments for forestry applications.
• The study validates PRISMA's potential for diverse remote sensing applications, including precise forest monitoring.