A single-channel vegetation hyperspectral high-frequency observation system highly coupled with eddy flux

Abstract

The present invention relates to a single-channel vegetation hyperspectral high-frequency observation system highly coupled with eddy flux, comprising: a spectrometer, an optical fiber and a cosine corrector, wherein the tail end of the optical fiber is connected to the spectrometer, the head end of the optical fiber is connected to the optical path of the cosine corrector, and the system further comprises a motor for directly or indirectly flipping the cosine corrector so that the light-receiving surface of the cosine corrector switches between two states of vertically facing upward and vertically facing downward. The present invention has the following advantages: 1. The influence of wavelength drift is reduced by observing with a single spectrometer; 2. The optical path switching switch is abandoned and the motor is used to rotate a single optical fiber, which significantly improves the integrity of the optical path, increases the transmittance of the optical path, and significantly shortens the time interval of the upper and lower channel observations, effectively reducing the influence of the time mismatch of the upper and lower channels on the accuracy of SIF inversion; 3. The cosine observation method is used to expand the SIF observation footprint, thereby improving the coupling with flux data.

Description

Technical Field

[0001] The invention relates to a single-channel vegetation hyperspectral observation system and an observation method highly coupled with eddy flux, and belongs to the technical field of spectrometer observation. Background Art

[0002] Gross Primary Production (GPP) represents the rate of organic carbon fixed by terrestrial plants through photosynthesis at the ecosystem scale, and is the largest component in the study of the global carbon cycle. GPP determines the input of material and energy cycles in terrestrial ecosystems, provides oxygen and food for higher life forms to survive, making higher life forms possible. At the same time, long-term carbon fixation and synthesis of O2 molecules also provide fossil fuels for the initialization of human industrial society. Therefore, GPP-related research has always been at the forefront of academic research in the fields of global carbon cycle and terrestrial ecosystem ecology.

[0003] At present, the universal and accurate GPP estimation method is the eddy covariance (EC) method. This method can accurately measure the net CO2 exchange (NEE) between terrestrial ecosystems and the atmospheric environment within a certain range (~1km), and then split NEE into GPP and soil respiration (Re) according to the respiration equation. However, due to the relatively small measurement range, although more than 1,000 flux stations have been established worldwide, it is still difficult to estimate regional or global GPP.

[0004] In recent years, significant progress has been made in the study of plant-light interactions and photosynthesis processes. In particular, remote sensing observations of solar-induced chlorophyll fluorescence (SIF) have opened up a new perspective for studying photosynthesis. SIF is a type of light with a wavelength of 650-800nm ​​released by green plants during photosynthesis under natural conditions. It can be regarded as a direct substitute for describing the mechanism and state of plant photosynthesis. SIF is a byproduct of photosynthesis and is closely related to photosynthesis. It can provide a more direct and effective method to diagnose the physiological and chemical conditions of vegetation. Therefore, in-depth research on chlorophyll fluorescence will help humans understand the mechanism of photosynthesis. Therefore, SIF observation technology and how to improve the accuracy of SIF observations have gradually become a hot spot for development, but also a difficult point.

[0005] At present, the main SIF observation scales are satellite, aviation, near-ground (e.g., ground-based) and leaves. Although satellite and aviation-scale SIF observations have an important impact on global and regional photosynthesis-related research, research on the mechanism of SIF coupling with photosynthesis is still limited, mainly due to the low spatiotemporal resolution of SIF and the difficulty in matching it with the flux site scale (footprint). Although the leaf scale has a high spatiotemporal resolution, the observation footprint is only ~1 cm2, making it difficult to couple flux observation data for related analysis. Near-ground SIF observations have gradually become an important topic in photosynthesis research due to their ultra-high spectral acquisition frequency and a footprint that is relatively consistent with flux.

[0006] The irradiance spectrum of vegetation reflection measured under natural lighting conditions includes both the emission spectrum of SIF and the reflection spectrum of the leaves to the incident light. Since the SIF emission spectrum and the vegetation canopy reflection spectrum are mixed together, accurate observation is first required to extract the SIF spectrum from the canopy spectrum. According to the idea of ​​SIF filling dark lines in the FLD principle, in order to extract weak SIF signals, the observation needs to achieve high spectral resolution (sub-nanometer) and nearly synchronous observation of solar incident light and canopy reflected light. At the same time, in order to achieve coupled analysis with flux data, the SIF observation footprint should also be consistent with the flux footprint. Therefore, a SIF observation system with reliable data should meet the following requirements:

[0007] 1. Spectral channel optical path integrity;

[0008] 2. The solar incident radiation channel and the ground object reflected radiation channel should be time-matched;

[0009] 3. The footprint of SIF observations matches that of EC fluxes;

[0010] 4. Integration time setting and dark current correction.

[0011] In order to achieve the above goals, a series of SIF observation system modes have been developed (such as dual spectrometers, single spectrometer + optical path switching switch, etc.), but they still cannot fully meet the above requirements and have the following problems:

[0012] 1) Wavelength drift. The spectral response characteristics of the two spectrometers in the dual-spectrometer system are not completely consistent. The solar incident spectrum and the canopy reflectance spectrum may have spectral drift problems and are difficult to correct, which may lead to large SIF inversion uncertainties.

[0013] 2) Time matching. The use of a single spectrometer can effectively solve the wavelength drift problem, but a single spectrometer combined with an optical path switching method can only obtain the solar incident spectrum and canopy reflectance spectrum in a sequential measurement mode, and the optical path loss is large, the optical path integrity is insufficient, and there is a large time interval between the upper and lower channels. In cloudy and other weather conditions with rapid changes in light conditions, obtaining solar incident spectra and canopy reflectance spectra at different times may increase the SIF inversion error.

[0014] 3) Small footprint, difficult to couple with flux data for analysis. Existing systems mainly rely on bare fiber observations, and the SIF observation footprint is much smaller than the flux observation footprint. Summary of the invention

[0015] The technical problem to be solved by the present invention is to overcome the above-mentioned shortcomings of the prior art and provide a single-channel hyperspectral observation system that is highly coupled with the eddy flux, which effectively solves the problems of spectral drift, time matching and small footprint.

[0016] The technical solution adopted by the present invention to solve its technical problems is: a single-channel vegetation hyperspectral high-frequency observation system that is highly coupled with eddy flux, comprising: a vegetation fluorescence spectrometer and a first optical fiber, the tail end of the first optical fiber is connected to the optical signal input end optical path of the vegetation fluorescence spectrometer, characterized in that: it also includes a cosine corrector connected to the optical path of the first optical fiber head end, and a motor for directly or indirectly flipping the cosine corrector so that the light-receiving surface of the cosine corrector is switched between two states: vertically upward and vertically downward.

[0017] Furthermore, the observation system also has a vegetation index spectrometer and a forked optical fiber. The tail end of the forked optical fiber is connected to the optical path of the vegetation index spectrometer, and its head end is a forked end. The two ends of the fork are respectively connected to two optical fibers through an optical switch. The head ends of the two optical fibers are respectively vertically upward and vertically downward for receiving light.

[0018] The observation method of the single-channel vegetation hyperspectral high-frequency observation system highly coupled with the eddy flux is characterized by comprising the following steps:

[0019] Step A: performing calibration processing before measurement to obtain a calibration coefficient corresponding to the first optical fiber;

[0020] Step B, calculate the current solar zenith angle according to the local longitude and latitude and the current time, observe when the solar zenith angle is ≤90°, and execute step C, otherwise do not observe;

[0021] Step C: rotating the motor so that the light-receiving surface of the cosine corrector faces vertically upward, and measuring the solar irradiance according to the calibration coefficient corresponding to the first optical fiber; blocking the internal switch of the vegetation fluorescence spectrometer, measuring the corresponding dark noise, and obtaining the solar spectrum information;

[0022] Step D: rotate the motor to make the cosine corrector point vertically downward, measure the radiance of the ground object according to the calibration coefficient corresponding to the first optical fiber; block the internal switch of the vegetation fluorescence spectrometer, measure the corresponding dark noise, and obtain the spectral information of the ground object.

[0023] The single-channel hyperspectral observation system highly coupled with eddy flux of the present invention can be used for vegetation SIF observation and vegetation index calculation. The system effectively solves the above three major problems: 1) The influence of wavelength drift is reduced by observing with a single spectrometer; 2) The optical path switching switch is abandoned and a single optical fiber is rotated by a motor, which significantly improves the integrity of the optical path, increases the transmittance of the optical path, and significantly shortens the time interval for observation of the upper and lower channels, effectively reducing the influence of the time mismatch of the upper and lower channels on the accuracy of SIF inversion; 3) The cosine observation method is used to expand the SIF observation footprint, thereby improving the coupling with flux data. Therefore, compared with other SIF observation systems, the system of the present invention significantly improves the SIF observation frequency and inversion accuracy. More importantly, this integrated system achieves long-term, automatic maximum matching with the flux footprint, ensuring the coupling of SIF-flux data. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The drawings in the specification, which constitute a part of this application, are used to provide a further understanding of the present invention. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0025] Figure 1 It is a schematic diagram of a single-channel hyperspectral observation system of the present invention.

[0026] Figure 2 It is a structural schematic diagram of a single-channel hyperspectral observation system of the present invention. DETAILED DESCRIPTION

[0027] To make the purpose, technical solution and advantages of the embodiments of the present invention more clear, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0028] System composition

[0029] like Figure 1As shown, the single-channel vegetation hyperspectral high-frequency observation system highly coupled with eddy flux of an embodiment of the present invention comprises: a QEPRO spectrometer (vegetation fluorescence spectrometer) 6, a QEPRO optical fiber 3 and a cosine corrector 5 connected to the optical path of the QEPRO optical fiber head end, the tail end of the QEPRO optical fiber 3 is connected to the optical path of the optical signal input end of the QEPRO spectrometer 6, an optical fiber rotation system for directly or indirectly flipping the cosine corrector 5 so that the light receiving surface of the cosine corrector 5 is switched between two states of vertically facing upward and vertically facing downward, and the optical fiber rotation system comprises a motor 4 and a controller for controlling the rotation of the motor.

[0030] like Figure 2 As shown, the single-channel vegetation hyperspectral high-frequency observation system of this embodiment also includes a bracket 1 and a support arm 2 fixed to the bracket 1, a motor 4 is fixed to the support arm 3, and a cosine corrector 5 is directly or indirectly driven by the output head of the motor. When the light-receiving surface of the cosine corrector is vertically facing upward, the observation system measures the solar irradiance; when the light-receiving surface of the cosine corrector is vertically facing downward, the observation system measures the radiance of the ground object.

[0031] In order to observe the vegetation index, as an improved solution, the observation system also has a Flame spectrometer (vegetation index spectrometer) and a Flame bifurcated optical fiber. The tail end of the Flame bifurcated optical fiber is connected to the optical path of the vegetation index spectrometer, and its head end is a bifurcated end. The two ends of the bifurcation are respectively connected to two optical fibers through an optical switch. The head ends of the two optical fibers are respectively vertically upward and vertically downward for receiving light. The bifurcated optical fiber and the optical switch can be replaced by a splitter.

[0032] The advantages of this system are as follows:

[0033] 1. Small impact of wavelength drift

[0034] Compared with dual spectrometers, the single spectrometer system significantly reduces the SIF inversion uncertainty caused by wavelength drift. Previous studies have shown that even very small spectral differences between the solar incident radiation channel and the canopy reflected radiation channel may lead to large errors in the inverted SIF. In particular, for the method of using two spectrometers to observe solar incident radiation and canopy reflected radiation respectively, the wavelength correction is more complicated due to the different spectral response characteristics (such as spectral response center wavelength, spectral resolution, etc.) between different spectrometers, which may increase the SIF inversion error.

[0035] 2. High optical path integrity and time matching between upper and lower channels

[0036] The single fiber combined with the motor rotation method designed in this scheme greatly ensures the optical path integrity of the SIF observation system. Compared with the single fiber method, the solutions adopted by other systems (such as the forked fiber combined optical path switching and opening method, fiber combined prism, etc.) have poor optical path integrity, and there is obvious optical signal loss in the upper and lower optical path channels, resulting in a longer integration time under the same radiation conditions, which greatly reduces the time matching and spectral acquisition frequency of the upper and lower channels. The single fiber method designed in this scheme only has a small amount of optical signal loss at the fiber bend, which can still be solved by ensuring that the fiber bending radius is not too small. The system designed in this scheme adopts a single fiber method to ensure a highly complete optical path. The integration time is significantly shortened under the same radiation conditions, and the spectral acquisition frequency is effectively increased. About 2,000 spectra can be collected every day.

[0037] 3. High coupling between SIF and flux

[0038] This scheme design system adopts a single fiber in series with a cosine corrector to ensure that the SIF matches the flux footprint. If the bare fiber observation method is used, the observed SIF signal will hardly cover the main contribution area of ​​the flux signal, resulting in a mismatch with the flux footprint. For example, the bare fiber field of view (FOV) is about 25°. Assuming vertical downward observation, the bare fiber observation radius is QUOTE (h is the height of the bare fiber probe from the top of the canopy), and the main contribution area of ​​the flux footprint is about 50 times this length. The bare fiber series cosine corrector can expand the FOV to about 140°, which can effectively improve the SIF observation footprint. Studies have shown that about 75% of the 90% total radiance comes from the range of flux tower observation, which significantly improves the matching degree between the SIF observation footprint and the flux footprint.

[0039] Software System

[0040] For this observation system, the present invention has developed corresponding software, which integrates visualization function, parameter setting function, motor and inline control function, data acquisition function, data recording function, and data preprocessing function. Compared with the previous observation system, the software design of the present invention has made certain optimization and improvement in the motor control module and the data preprocessing module. The following will explain the motor control part and the fluorescence data processing part.

[0041] Spectrum acquisition control system

[0042] The present invention designs a set of special hyperspectral acquisition control software system, which greatly optimizes the integration time setting and the time matching problem of the upper and lower channels. The software system controls the spectrum acquisition steps as follows:

[0043] Step A: perform calibration processing before measurement to obtain the calibration coefficient corresponding to the QEPRO optical fiber.

[0044] In this step, a halogen lamp is used for calibration processing to obtain the calibration coefficient corresponding to the optical fiber, and calibration is performed according to the following formula:

[0045] cof= Rad / ((DN-DC) / IT)

[0046] Among them, cof represents the calibration coefficient, Rad represents the known brightness of the light source, DN represents the recorded digital signal corresponding to the light source collected by the spectrometer, IT represents the integration time, and DC represents the corresponding dark current when collecting the DN value.

[0047] Step B, calculate the current solar zenith angle according to the local longitude and latitude and the current time, observe when the solar zenith angle is ≤90°, and execute step C, otherwise do not observe;

[0048] Step C, rotating the motor so that the light-receiving surface of the cosine corrector faces vertically upward, and measuring the solar irradiance according to the calibration coefficient corresponding to the QEPRO optical fiber; blocking the internal switch of the QEPRO spectrometer, measuring the corresponding dark noise, and realizing the acquisition of solar spectrum information;

[0049] Step D: Rotate the motor to make the cosine corrector point vertically downward, and measure the radiance of the ground object according to the calibration coefficient corresponding to the QEPRO optical fiber; block the internal switch of the QEPRO spectrometer, measure the corresponding dark noise, and obtain the spectral information of the ground object.

[0050] Before measuring the solar irradiance in step C and the ground object radiance in step D, it is necessary to determine the optimal integration time optinttime. The optimal integration time optinttime is calculated according to the following formula:

[0051] optinttime=inttime*(targetDN / max)

[0052] Among them, inttime represents the preset integration time, targetDN represents the size of the preset collected spectrum value DN, and max represents the maximum spectrum value in a spectrum collected according to the preset integration time inttime.

[0053] In addition, a maximum integration time T is set in the system. If the optimal integration time calculated in step C or step D is greater than the maximum integration time T, the maximum integration time T is used for measurement.

[0054] As a further improvement, this observation system also has a Flame spectrometer (vegetation index spectrometer) and a Flame dual-channel optical fiber. The tail end of the Flame dual-channel optical fiber is connected to the optical path of the vegetation index spectrometer, and its head end is a forked end, which is connected to two optical fibers through an optical switch. The head ends of these two optical fibers are vertically upward and vertically downward, respectively, for receiving light. Among them, the forked optical fiber and the two optical switches can be replaced by a splitter.

[0055] In the calibration process of step A, the calibration coefficient corresponding to the second optical fiber is also obtained; in step C, the solar irradiance is measured according to the calibration coefficient corresponding to the second optical fiber; the internal switch of the vegetation index spectrometer optical path is blocked, the corresponding dark noise is measured, and the solar spectrum information is obtained; in step D, the vegetation radiance is measured according to the calibration coefficient corresponding to the second optical fiber; the internal switch of the vegetation fluorescence spectrometer is blocked, the corresponding dark noise is measured, and the vegetation index is obtained.

[0056] The motor spectrum high-frequency observation system and spectrum observation method designed in the present invention provide a solution for continuously and efficiently acquiring ground object spectra and solar spectra. By switching the motor control channel, the solar incident spectrum and the ground object reflection spectrum are acquired alternately, thereby ensuring the consistency of the two channels and the acquisition of the solar spectrum and the ground object spectrum under the same lighting conditions, avoiding the error in calculating the ground object reflectivity caused by the channel error and the switching of the TTL switch, and automatically and continuously acquiring the ground object spectrum and the corresponding reflectivity product with high quality.

[0057] In the present embodiment, the motor is connected to the control computer through the 485 interface. In the control program, the motor is first connected through the serial port. The connection status of the motor will be displayed on the interface of the software. Success and failure can have simple prompts on the interface of the software. After the computer is successfully connected to the motor, the initialization command is first sent to the motor. The specific initialization process is to rotate the motor 180 degrees twice. At the end of the initialization, the motor can reach the initialized vertical posture by induction. After the motor is initialized, the program sends the command to rotate to the motor through the serial port, so that the motor is controlled to the spectrometer after the rotation is completed, and the spectrum is collected. It is worth mentioning that in order to ensure the stability of data acquisition, after each rotation of the motor, it is necessary to pause for 5 seconds, so that each time the data is collected, it is ensured that the motor is in a stable state.

[0058] This observation system includes both the QEPRO spectrometer observation module and the Flame spectrometer observation module. The Flame observation module can obtain rich ground object information due to its wide band range from visible light to near infrared. The following are three important vegetation indices, namely NIRv (near-infrared reflectance of vegetation, vegetation near-infrared reflectance), NDVI (normalized difference vegetation index), and PRI (Photochemical Reflectance Index). Among them, r750, r685, r530, ​​and r570 represent the reflectance in the corresponding bands, respectively, and Rnir represents the reflectance in the near infrared. The following three vegetation indices are obtained from the observation results of Flame, which can reflect rich vegetation information. NIRv has been proven to be highly correlated with the observation of eddy flux, can be used to monitor the total primary productivity of ecosystems, and has the potential to convert canopy SIF to light system scale SIF. NDVI is a reflection of vegetation greenness and canopy structure. NDVI is often used to monitor vegetation growth status. When vegetation grows well, the NDVI value is high, and when vegetation is stressed or growing poorly, NDVI can be observed more accurately. Photochemical reflectance index is mainly used to monitor vegetation stress, because when vegetation is under environmental stress, the concentration of pigments in the central chloroplasts of photosynthesis will change, which directly leads to changes in the reflectance of the vegetation's apparent reflectance in the corresponding band of pigment absorption and reflection. When vegetation is under stress, photosynthesis weakens, and the corresponding energy used for heat dissipation and photosynthesis regulation will increase, that is, the increase of NPQ (non-photochemical quenching), and PRI is widely used to monitor NPQ. In summary, the Flame module can provide rich vegetation information, including the total primary productivity of vegetation, the greenness and structure of vegetation, the growth status of vegetation and the stress status. The corresponding information can be obtained through the observation results of this observation system.

[0059] In addition to the above embodiments, the present invention may also have other implementation modes. Any technical solution formed by equivalent replacement or equivalent transformation falls within the protection scope required by the present invention.

Claims

1. An observation method for a single-channel vegetation hyperspectral high-frequency observation system highly coupled with eddy flux, the observation system comprising: a vegetation fluorescence spectrometer, a first optical fiber, a cosine corrector connected to the optical path of the first optical fiber head end, and a motor for directly or indirectly flipping the cosine corrector so that the light-receiving surface of the cosine corrector switches between two states of vertically facing upward and vertically facing downward, the tail end of the first optical fiber is connected to the optical path of the optical signal input end of the vegetation fluorescence spectrometer, characterized in that The observation method comprises the following steps: Step A: performing calibration processing before measurement to obtain a calibration coefficient corresponding to the first optical fiber; Step B, calculate the current solar zenith angle according to the local longitude and latitude and the current time, observe when the solar zenith angle is ≤90°, and execute step C, otherwise do not observe; Step C: rotating the motor so that the light-receiving surface of the cosine corrector faces vertically upward, and measuring the solar irradiance according to the calibration coefficient corresponding to the first optical fiber; blocking the internal switch of the vegetation fluorescence spectrometer, measuring the corresponding dark noise, and obtaining the solar spectrum information; Step D: rotating the motor so that the cosine corrector is vertically downward, and measuring the radiance of the ground object according to the calibration coefficient corresponding to the first optical fiber; blocking the internal switch of the vegetation fluorescence spectrometer, measuring the corresponding dark noise, and obtaining the spectrum information of the ground object; In step A, a halogen lamp is used for calibration processing to obtain the calibration coefficient corresponding to the optical fiber, and calibration is performed according to the following formula: cof= Rad / ((DN-DC) / IT) Among them, cof represents the calibration coefficient, Rad represents the known brightness of the light source, DN represents the recorded digital signal corresponding to the light source collected by the spectrometer, IT represents the integration time, and DC represents the corresponding dark current when collecting the DN value.

2. The observation method of the single-channel vegetation hyperspectral high-frequency observation system highly coupled with eddy flux according to claim 1, characterized in that: Before measuring the solar irradiance in step C and the ground object radiance in step D, it is necessary to determine the optimal integration time optinttime. The optimal integration time optinttime is calculated according to the following formula: optinttime=inttime*(targetDN / max) Among them, inttime represents the preset integration time, targetDN represents the size of the preset collected spectrum value DN, and max represents the maximum spectrum value in a spectrum collected according to the preset integration time inttime.

3. The observation method of the single-channel vegetation hyperspectral high-frequency observation system highly coupled with eddy flux according to claim 1, characterized in that: The maximum integration time T is set. If the optimal integration time calculated in step C or step D is greater than the maximum integration time T, the maximum integration time T is used for measurement.

4. The observation method of the single-channel vegetation hyperspectral high-frequency observation system highly coupled with eddy flux according to claim 1 is characterized in that: The observation system also has a vegetation index spectrometer and a bifurcated optical fiber. The tail end of the bifurcated optical fiber is connected to the optical path of the vegetation index spectrometer, and its head end is a bifurcated end. The two ends of the bifurcation are respectively connected to two optical fibers through an optical switch. The head ends of the two optical fibers are respectively vertically upward and vertically downward for receiving light.

5. The observation method of the single-channel vegetation hyperspectral high-frequency observation system highly coupled with eddy flux according to claim 3 is characterized by: In the calibration process of step A, the calibration coefficient corresponding to the second optical fiber is also obtained; in step C, the solar irradiance is measured according to the calibration coefficient corresponding to the second optical fiber; the internal switch of the vegetation index spectrometer optical path is blocked, the corresponding dark noise is measured, and the solar spectrum information is obtained; in step D, the vegetation radiance is measured according to the calibration coefficient corresponding to the second optical fiber; the internal switch of the vegetation fluorescence spectrometer is blocked, the corresponding dark noise is measured, and the vegetation index is obtained.

6. The observation method of the single-channel vegetation hyperspectral high-frequency observation system with highly coupled eddy flux according to claim 1 is characterized by: The observation system comprises a bracket, a support arm fixed to the bracket and a controller for controlling the rotation of a motor, wherein the motor is fixed to the support arm, and the cosine corrector is driven directly or indirectly by an output head of the motor.

Images