Definition: Spectral resolution refers to the minimum wavelength difference (Δλ) that a spectrometer can distinguish, typically expressed as the Full Width at Half Maximum (FWHM) at a specific wavelength. Lower values denote higher resolving power, indicating the ability to distinguish finer adjacent spectral features. Physical Significance: It determines an instrument's capacity to resolve closely spaced spectral lines or absorption peaks. High resolution (e.g., 0.1 nm) is vital for fine-structure analysis of gas molecules; mid-resolution (1-5 nm) meets the needs of vegetation red-edge and mineral identification; low resolution (＞10 nm) is mainly used for sorting broad surface feature classes. Practical Application: Choosing resolution requires balancing the Signal-to-Noise Ratio (SNR)—excessively high resolution disperses light energy, resulting in increased noise. Field measurements typically utilize a 3-10 nm resolution, whereas precision laboratory analysis favors 0.5-1 nm. Note that spectral resolution differs from sampling interval (band spacing); the latter must satisfy the Nyquist theorem (sampling interval ≤ resolution/2).

Definition: The Signal-to-Noise Ratio is the ratio of effective signal intensity to noise intensity (typically expressed as a ratio or in dB). In spectroscopy, a higher SNR yields a smoother spectral curve and makes faint absorption/emission features more reliable. Physical Significance: It directly dictates data repeatability and the lower detection limit for trace components. When the SNR falls below 100, reflectance variances under 1% become statistically indistinguishable; when the SNR exceeds 1000, sub-millimeter mineral alteration signals can be resolved. Practical Application: Methods to improve SNR include: increasing integration time (signals scale linearly while noise scales with the square root), averaging multiple scans (for N averages, SNR improves by √N times), and lowering instrument temperature (to suppress dark current). Be cautious, as excessive integration time can saturate the detector (leading to DN value overflow).

Definition: Radiometric calibration is the process of converting the digital numbers (DN) output by an instrument into physically meaningful radiometric quantities (such as radiance L, measured in W·m⁻²·sr⁻¹·μm⁻¹) or apparent reflectance. Physical Significance: It removes sensor-specific response variances and ambient illumination variations, making spectral data captured at different times or by different instruments comparable. Absolute calibration requires a standard radiation source (such as an integrating sphere), while relative calibration utilizes standard reference panels. Practical Application: In field work, the "reference panel correction method" is common: DN_target / DN_white × ρ_white (the known reflectance of the panel) yields target reflectance. Ensure the panel exhibits Lambertian properties and undergoes routine traceable calibration. Absolute radiometric calibration should be completed in a lab using traceable standard lamps, achieving accuracies within ±3%.

Definition: Spectral reflectance is the ratio of reflected radiant flux to the incident radiant flux hitting an object's surface (0~1 or 0%~100%), serving as a function of wavelength λ, denoted as R(λ). Physical Significance: It characterizes the capacity of surface features to reflect electromagnetic waves across different wavelengths, serving as a fundamental physical quantity in remote sensing. Different materials possess characteristic reflectance spectra (e.g., vegetation features a "green peak" near 550 nm and a "red edge" near 700 nm). Practical Application: Measuring reflectance requires simultaneous measurement of a reference white panel (approximating a 100% Lambertian reflector). Distinguish between Directional-Hemispherical Reflectance (measured via integrating spheres) and Bidirectional Reflectance Factor (BRF) (typically measured in the field). Rough surfaces like soil show strong directionality; measurement angles (zenith/azimuth) must be meticulously logged.

Definition: The wavelength interval corresponding to a local minimum in reflectance (or transmittance) along a spectral curve. Driven by internal electronic transitions or molecular vibrations within matter, it contains four quantitative parameters: center wavelength, depth, width, and symmetry. Physical Significance: Absorption features form the core of a substance's "spectral fingerprint." For example, absorption bands around 1400 nm and 1900 nm are caused by H₂O molecules, the 2200 nm feature indicates Al-OH bonds in clay minerals, and absorption near 2340 nm is characteristic of carbonate CO₃²⁻. Practical Application: Extracting absorption features requires normalization via "Continuum Removal" first, scaling features to a shared baseline (0~1) for comparison. Absorption depth increases with substance concentration, though it approaches saturation at high concentrations. Spectral indices like the absorption depth index can be utilized for mineral mapping.

Master Spectral Terms to Enhance Data Interpretation Capabilities

We provide spectral data processing training and instrument application support services

Hagorun Technology Limited  |   Focus on Spectroscopy and Remote Sensing Applications