Retrieval Of Solar-induced Chlorophyll Fluorescence Based On Domestic Satellites

Solar-induced chlorophyll fluorescence(SIF)is an optical signal re-emitted by the excited chlorophyll molecules of vegetation under natural illumination.It occurs in the light reactions of photosynthesis and is directly related to the linear electron transport rate.Compared to the traditional vegetation indices(VIs)and reflectance spectra,SIF can directly reflect the physiological status of vegetation.Thus,it has been regarded as the ideal proxy for monitoring the actual vegetation photosynthesis.In recent decade,SIF remote sensing has stimulated a flurry of research among researchers in the whole world.More and more works have shown that space-borne SIF products can provide a new tool for global vegetation gross primary production(GPP)estimation.With the rapid advances of optical sensors and satellite remote sensing,demostic satellites or sensors have the potential to detect SIF signals yet.However,the research of SIF retrieval making use of space-borne observations measured by these satellites is still at early stage.Against the requirements of SIF retrieval based on domestic satellite observations and validation of SIF remote sensing products,this paper carried out a series of research on remote sensing of SIF,including the ground-based and tower-based SIF observations,the investigations of relationships between GPP and SIF at different bands,the evaluation and optimal of specifications of SIFIS imaging spectrometer onboard TECIS-1 satellite and remote sensing retrieval of SIF using remote sensing measurements by Tan Sat satellite.The major works and conclusions are included as follows:(1)In order to provide the ground-based SIF measurements for validating remote sensing SIF products derived from domestic space-borne observations and optimization of specifications of SIF payload,focusing on the issue on long-term in-situ SIF measurements,this paper adopted a novel automated SIF measurement system(SIFSpec)for long-term SIF observations in the field and presented the details of the system components,radiometric calibration as well as data collection.In addition,the details of data processing flow for ground-based and tower-based measurements were also included.4 automated SIF measurement systems developed in this paper play important role in the network of ground-based continuous SIF observations,China Spec,and have been run for about 3 years since 2017.The results show that the canopy SIF can track the variation of GPP at both daily and seasonal scales.These results will be helpful to make use of domestic satellite SIF products to estimate GPP at global scale.(2)Against the great requirements of development on satellite-based SIF remote sensing in China,this paper conducted the evaluation and optimization of specifications of SIF payload onboard TECIS-1 satellite.Firstly,the high-spectral-resolution simulated dataset for SIFIS payload was obtained by combing the Soil Canopy Observation Photosynthesis and Energy(SCOPE)and Moderate resolution atmospheric TRANsmissin(MODTRAN)models before TECIS-1 satellite's launch.Then in order to evaluate and optimize the specifications of the SIFIS,this paper conducted the SIF retrievals for different specifications of spectral resolution(SR)and signal-to-noise ratio(SNR)making use of the PCA-based data driven algorithm.The results show that the best optimal combination is an SR of 0.1 nm and SNR of 127,with corrected root?mean?squared difference(RMS diff*)value of 0.15 m W m^-2 sr^-1 nm^-1at the far-red band and 0.43 m W m^-2 sr^-1 nm^-1 at red band.Given an SR of 0.3 nm,which is the pre-configured for the SIFIS,the RMS diff*will be 0.17 m W m^-2 sr^-1 nm^-1 at far-red band and 0.47 m W m^-2 sr^-1 nm^-1 at red band if the SNR is improved to 450.(3)Against the requirements on remote sensing SIF retrieval using observations of Tan Sat satellite,this paper developed the SVD-based data-driven approach for detecting space-based SIF signals.The orthogonal singular vectors,denoted as fundamental spectra,of the optical characteristics of atmosphere are obtained making use of SVD algorithm based on the measurements over vegetation-free surfaces observed by Tan Sat satellite.It was found that the atmospheric radiative transfer characteristics could be greatly reconstructed with high precision using several fundamental spectra.In addition,this paper developed the SIF product production system for obtaining Tan Sat SIF products.And the first global SIF map was produced based on the remote sensing observations measured by domestic Tan Sat satellite.The spatial and temporal pattern of the retrieved SIF agrees well with the true terrestrial vegetation pattern.The results of comparisons between Tan Sat SIF products and other remote sensing datasets show that,Tan Sat SIF is highly linear with OCO-2 SIF,MODIS NDVI,EVI and GPP with R² of 0.86,0.78,0.80 and 0.68.And the difference between Tan Sat and OCO-2 SIF products is about 0.168 m W m^-2 sr^-1 nm^-1.These results all indicate that Tan Sat SIF product is reliable.The main innovative contributions of this paper are summarized below:(1)Against the requirements of SIF retrieval based on Tan Sat domestic satellite,this paper developed the data-driven method for retrieving SIF signals using Tan Sat observations,especially for the optimization of selection method of training dataset,which dealing with the problems on SIF retrieval with high precision using space-borne observations.And this paper developed the production system for retrieving Tan Sat SIF products and firstly produced the SIF products derived from domestic remote sensing observations.The averaged residual between the measured and fitted radiance calculated based on Tan Sat SIF is an order of magnitude smaller compared to that of OCO-2 and GOSAT.(2)In order to meet the requirements of developing remote sensing of SIF based on the domestic SIF satellites,this paper conducted variety of simulations for different specifications and evaluated and optimized the key spectral specifications of SIFIS imaging spectrometer onboard TECIS-1 satellite.And the systematic biases derived by the PCA data-driven method were analyzed and evaluated.The simulated dataset and retrieval algorithms before TECIS-1's launch will be beneficial to improve the characteristic of the first Chinese space-borne SIF spectrometer.