Solar-induced chlorophyll fluorescence (SIF) is a valuable metric for assessing photosynthesis and vegetation stress. However, as SIF radiance constitutes less than 3% of the reflected canopy radiance, the spectral resolution of a spaceborne SIF detector should be below 0.3 nm. To ensure adequate signal-to-noise ratio (SNR), current spaceborne SIF imagers typically achieve spatial resolutions above 1 km. European Space Agency’ FLEX (Fluorescence explorer) mission recommends spatial resolution below 300 m, particularly for monitoring field and forest areas in Europe. The complex terrain and vegetation types in China, however, demand even higher spatial resolutions. In this paper, we propose a mid-resolution ultraspectral imager (MIRUS), designed for satellite-based SIF detection at a spatial resolution of 100 m. To evaluate the SIF retrieval performance of MIRUS, we develop a model that leverages SIF imaging spectrometer (SIFIS) data to calculate SIF retrieval accuracy.

Given the weak SIF radiance relative to canopy reflectance, the design specifications for MIRUS are shown in Table 1. MIRUS employs a low F# optic and a Littrow-Offner spectrometer to improve irradiance on the focal plane array (FPA) and reduce chromatic aberration, as shown in Fig. 2. The modulation transfer function (MTF) is optimized to be greater than 0.9, as shown in Fig. 3, while smile and keystone distortions are controlled to below 1.0% pixels and 3.3% pixels, respectively, as shown in Table 3. The spectral resolution is set at 0.3 nm, with a convex grating designed using rigorous coupled-wave analysis (RWCA) (Fig. 5), achieving an average diffraction efficiency of 0.7 (Fig. 7). The mechanism is designed as shown in Fig. 8. A prototype of MIRUS is produced, combining a telescope, a spectrometer, and an FPA. The prototype’s performance, including instrument line shape (ILS), spectral resolution, smile, keystone, and SNR, is tested, with results shown in Figs. 9?13.

To assess MIRUS’s SIF detection performance, we build a model relating SIF retrieval accuracy to the SNR of the spectral imager. The spectral range and resolution of the SIFIS on the Goumang satellite are comparable to the designed performance of MIRUS, with SIFIS achieving a maximum spatial resolution of 0.375 km×0.800 km in non-binning mode. In addition, we analyze 48 SIFIS image orbits from January to October in 2023, covering diverse environments such as tropical rainforests, savannas, deserts, and polar regions. Using singular value decomposition (SVD) in the 743?758 nm range, we measure retrieval errors across different radiance levels (Fig. 17) and develop an SNR model for SIFIS data (Fig. 16). The radiance of MIRUS within the same wavelength range is simulated using the MODTRAN 6.0 model, based on MIRUS’s orbital parameters and typical atmospheric, aerosol, and ground albedo parameters, as shown in Table 4. Subsequently, the SNR for MIRUS is calculated from this radiance, as shown in Fig. 16. Finally, the SIF retrieval accuracy for MIRUS is evaluated using a polynomial function of SNR, with results shown in Fig. 18. The relative errors for SIF retrieval are 1.22%?1.38% for 100 m GSD mode and 0.69%?0.86% for the 200 m GSD mode.

SIF serves as a “probe” for photosynthetic activity. “The remote sensing of chlorophyll fluorescence is a rapidly advancing front in terrestrial vegetation science, with emerging capability in space-based methodologies and prospects for diverse applications,” as noted by G. H. Mohammed. Due to the weak SIF radiance, it must be captured with an ultraspectral imager. Considering the imaging SNR, the spatial resolution of SIF radiance retrieved by spaceborne instruments typically exceeds 1 km. In this paper, we propose a mid-resolution spaceborne SIF detector featuring a small F# TMA and a Littrow-Offner spectrometer. The high-groove-density convex grating is designed using rigorous coupled-wave analysis (RCWA), resulting in significantly greater irradiance on the FPA compared to SIFIS. A prototype is produced and tested, achieving a full width at half maximum (FWHM) of 0.3 nm, a spectral sampling interval (SSI) of 0.1 nm, a smile distortion of less than 0.0035 nm, a keystone of under 0.06 pixel, and an SNR exceeding 206 at 10 mW·m^-2·sr^-1·nm^-1. To evaluate the SIF retrieval accuracy of MIRUS, we develop a method to estimate accuracy based on the instrument’s SNR. The designed spectral performance of SIFIS matches that of MIRUS, and the spatial resolution of SIFIS is comparable to MIRUS. A polynomial relationship between SNR and SIF retrieval accuracy is established using radiance data from SIFIS. After calculating the typical radiance received by MIRUS, we use the SNR model to determine its typical SNR. Finally, the SIF retrieval accuracy is calculated using the polynomial, yielding relative errors of 1.22%?1.38% for the 100 m GSD mode and 0.69%?0.86% for the 200 m GSD mode, comparable to SIFIS performance.