This paper introduces an ontology-driven model aligned with the Semantic Sensor Network Ontology (SSN) to formally describe spectral observations from remote sensing instruments. The model addresses key limitations in existing frameworks by explicitly representing the internal structure of spectral data—such as the relationship between wavelength intervals and measured intensities—using description logic-based constructs. The proposed ontology integrates principles from the Quantities, Units, Dimensions and Data Types (QUDT) ontology and introduces new classes to describe the basis of spectral measurements, supporting multispectral and hyperspectral data. By enabling machine-readable representations of spectra, this framework enhances semantic interoperability, facilitates advanced queries, and supports use cases such as spectral data fusion, validation, and transformation across sensors. Implementation examples include real-world instruments such as the ASD FieldSpec 4 HiRes spectroradiometer and Landsat 8 OLI sensor, demonstrating the utility of the model. The framework contributes to the development of FAIR-aligned, semantically rich infrastructures for Earth Observation data.