Quantification method and system for time-lag response of extreme precipitation-vegetation regulation and storage

Abstract

The present application relates to the technical field of hydrology and meteorology, in particular to a method and system for quantifying the time-delayed response of extreme precipitation-vegetation regulation and storage, which comprises the following steps: obtaining near-infrared, mid-infrared and Raman multi-band molecular spectrum data of a vegetation sample, constructing a water state spectrum diagram containing free water, bound water and hydrogen bond network index, extracting molecular hydrodynamic time-delay characteristics such as water response, balance and release time, combining with microstructure parameters of vegetation tissues such as epidermis, mesophyll and vascular bundle to determine the regulation and storage efficiency index at the cell and tissue scales, and finally generating regional vegetation regulation and storage response prediction results and regulation and storage efficiency spatial distribution diagram. Through multi-band molecular spectrum analysis, water molecules in different bound states in the vegetation tissue can be distinguished, the molecular mechanism of vegetation water regulation and control is revealed in depth, and the characterization accuracy is improved by more than 90% compared with traditional methods.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydrology and meteorology, in particular to a method and system for quantifying the time-lag response of extreme precipitation-vegetation regulation and storage, and more particularly to a technical scheme for studying the regulation mechanism of vegetation to extreme precipitation by using molecular spectral analysis technology, quantifying the time-lag response characteristics, and providing a scientific basis for sponge city planning, watershed management, and flood control and disaster reduction. BACKGROUND

[0002] With the intensification of global climate change, the frequency and intensity of extreme precipitation events are increasing, and the resulting flood disasters have become one of the major natural disasters threatening human life and property safety. As an important component of the terrestrial ecosystem, vegetation has a significant regulation effect on the hydrological process of the watershed through processes such as interception of precipitation, regulation of infiltration, and evapotranspiration, which can effectively reduce flood peak flow and reduce the risk of flood disasters.

[0003] However, the existing technology has many shortcomings in quantifying the regulation effect of vegetation. Traditional hydrological analysis methods are mostly based on simple linear correlation models, which are difficult to capture the complex nonlinear time-lag relationship between vegetation and hydrological system under extreme precipitation conditions. Such methods usually regard vegetation as a static factor, ignoring its dynamic regulation effect, which makes it difficult to accurately reflect the actual situation. At the same time, existing researches mostly focus on single time scale analysis, lacking systematic investigation of multi-scale time-lag characteristics. The regulation effect of vegetation varies at different time scales, from the interception effect at the hourly level to the soil moisture regulation at the monthly scale, and single-scale analysis cannot fully reflect this complex effect.

[0004] Although the existing spectral analysis technology (such as CN110118742A disclosed "Land surface vegetation canopy ecological water content remote sensing inversion method based on spectral analysis") can retrieve vegetation water content through spectral index, it only focuses on the static measurement of macroscopic water content, and cannot distinguish the water distribution in different binding states of vegetation organization, nor can it monitor and analyze the dynamic change process of vegetation water content before and after extreme precipitation. In addition, the existing technology fails to establish a time relationship model between precipitation and vegetation water response, and cannot quantify the time-lag effect, nor can it consider the influence of different vegetation types, structures and physiological characteristics on water absorption, transmission and release processes.

[0005] Therefore, it is urgent to develop a method that can systematically quantify the time-lag response characteristics of extreme precipitation-vegetation regulation from the perspective of molecular hydrodynamics, and provide technical support for scientific disaster prevention and reduction and ecosystem management. SUMMARY

[0006] The application aims to provide an extreme precipitation-vegetation regulation and storage hysteresis response quantification method and system, to realize accurate quantification of the hysteresis response characteristics of vegetation regulation and storage by introducing molecular spectrum analysis technology to study the regulation and storage mechanism of vegetation to extreme precipitation from the perspective of molecular hydrodynamics, and to provide a scientific basis for sponge city planning, watershed management and flood control and disaster reduction.

[0007] The application discloses an extreme precipitation-vegetation regulation and storage hysteresis response quantification method, comprising:

[0008] Obtaining multi-band molecular spectrum data of the vegetation sample, wherein the multi-band molecular spectrum data comprises near-infrared spectrum data, mid-infrared spectrum data and Raman spectrum data;

[0009] Based on the multi-band molecular spectrum data, a water state spectrum graph representing the state of water molecules in the vegetation tissue is constructed, wherein the water state spectrum graph comprises a free water index, a bound water index and a hydrogen bond network index;

[0010] According to the water state spectrum graph, molecular hydrodynamic hysteresis characteristics are extracted, wherein the molecular hydrodynamic hysteresis characteristics comprise a water response time, a water balance time and a water release time;

[0011] Obtaining vegetation tissue microstructure parameters, wherein the vegetation tissue microstructure parameters comprise epidermis structure parameters, mesophyll structure parameters and vascular bundle structure parameters;

[0012] Based on the molecular hydrodynamic hysteresis characteristics and the vegetation tissue microstructure parameters, multi-scale vegetation regulation and storage efficiency indexes are determined, wherein the multi-scale vegetation regulation and storage efficiency indexes comprise cell-scale regulation and storage indexes and tissue-scale regulation and storage indexes;

[0013] According to the multi-scale vegetation regulation and storage efficiency indexes, a regional vegetation regulation and storage response prediction result and a regulation and storage efficiency spatial distribution graph are generated.

[0014] Preferably, the multi-band molecular spectrum data of the vegetation sample comprises:

[0015] Obtaining near-infrared spectrum data with a wavelength range of 780-2500 nm, wherein the near-infrared spectrum data contains combined band and multiple band vibration characteristics of water molecules;

[0016] Obtaining mid-infrared spectrum data with a wavelength range of 2500-25000 nm, wherein the mid-infrared spectrum data contains fundamental frequency vibration characteristics of water molecules;

[0017] Obtaining Raman spectrum data with a wave number range of 200-4000 cm-1, wherein the Raman spectrum data contains scattering characteristics of intermolecular hydrogen bond structures of water molecules;

[0018] The multi-band molecular spectroscopy data is pre-processed, and the pre-processing includes spectral calibration, noise elimination, baseline correction, spectral standardization and spectral differential transformation.

[0019] As preferred, the water state spectrum representing the state of water molecules in the vegetation tissue is constructed based on the multi-band molecular spectroscopy data, including:

[0020] The water absorption peak characteristics near 1450 nm and 1940 nm are extracted from the near-infrared spectroscopy data;

[0021] The -OH stretching vibration characteristics near 3400 and the H-O-H bending vibration characteristics near 1640 are extracted from the mid-infrared spectroscopy data;

[0022] The -OH stretching vibration scattering characteristics in the range of 3200-3600 are extracted from the Raman spectroscopy data;

[0023] Based on the extracted spectral characteristics, the free water index representing the relative content of free water, the bound water index representing the relative content of bound water, and the hydrogen bond network index representing the strength of the hydrogen bond network between water molecules are calculated;

[0024] A water state dynamic distribution diagram is constructed with time as the horizontal axis and the proportion of water in different states as the vertical axis.

[0025] As preferred, the molecular water dynamics lag time characteristics are extracted according to the water state spectrum, including:

[0026] Based on the water state spectrum, the extreme precipitation response process is divided into absorption stage, equilibrium stage and release stage;

[0027] The state conversion path and conversion rate of water molecules in the vegetation tissue in each stage are analyzed;

[0028] The time interval from the peak value of extreme precipitation intensity to the peak value of vegetation water state parameters is determined as the water response time;

[0029] The time required for the change rate of water state parameters to drop below a preset threshold is determined as the water balance time;

[0030] The time required for the water state parameters to drop from the peak value to the half peak value is determined as the water release time.

[0031] As preferred, the multi-scale vegetation regulation and storage efficiency index is determined based on the molecular water dynamics lag time characteristics and the microstructure parameters of the vegetation tissue, including:

[0032] analyze the relationship between the epidermis structure parameters and the water response time to determine the influence of the epidermis on the initial water absorption;

[0033] analyze the relationship between the mesophyll structure parameters and the water balance time to determine the influence of the mesophyll on the water storage;

[0034] analyze the relationship between the vascular bundle structure parameters and the water conversion rate to determine the influence of the vascular bundle on the water transport;

[0035] determine the cell water capacity based on the total amount of water that can be stored in a unit volume of cells;

[0036] determine the cell water retention time based on the sum of the water balance time and the water release time;

[0037] determine the tissue water diffusion coefficient based on the characteristic rate of water diffusion in the tissue;

[0038] determine the tissue water capacity gradient based on the difference in water capacity between different tissue parts.

[0039] As a preferred embodiment, the method for quantifying the time-lag response of extreme precipitation-vegetation regulation and storage includes:

[0040] extract the key feature parameters of response delay, rising rate, peak position, falling rate and recovery degree from the water state change curve;

[0041] establish a relationship model between the precipitation feature parameters and the response curve feature parameters;

[0042] based on the relationship model, predict the regulation and storage response characteristics of vegetation under different intensity and duration of extreme precipitation scenarios;

[0043] determine the critical threshold of vegetation regulation and storage capacity, and predict the possible failure of regulation and storage function beyond the threshold;

[0044] map the multi-scale vegetation regulation and storage efficiency index to geographical space to generate a spatial distribution map containing regulation and storage intensity, regulation and storage duration and regulation and storage efficiency.

[0045] As a preferred embodiment, the method for quantifying the time-lag response of extreme precipitation-vegetation regulation and storage is used for sponge city planning, flood risk management and ecosystem service evaluation.

[0046] The time-lag response quantification system of extreme precipitation-vegetation regulation and storage adopts the method, and is characterized in that it comprises:

[0047] a multi-band spectral data acquisition module for acquiring near-infrared spectral data, mid-infrared spectral data and Raman spectral data of the vegetation sample;

[0048] a water status spectrum constructing module configured to construct a water status spectrum representing water molecule status in the vegetation tissue based on the multi-band spectrum data, the water status spectrum including a free water index, a bound water index, and a hydrogen bond network index;

[0049] a time-lag feature extracting module configured to extract a molecular water dynamics time-lag feature including a water response time, a water balance time, and a water release time from the water status spectrum;

[0050] a microstructure analyzing module configured to obtain vegetation tissue microstructure parameters including epidermis structure parameters, mesophyll structure parameters, and vascular bundle structure parameters;

[0051] a storage and regulation efficiency evaluating module configured to determine multi-scale vegetation storage and regulation efficiency indexes including cell-scale storage and regulation indexes and tissue-scale storage and regulation indexes based on the molecular water dynamics time-lag feature and the vegetation tissue microstructure parameters;

[0052] a prediction applying module configured to generate a regional vegetation storage and regulation response prediction result and a storage and regulation efficiency spatial distribution map based on the multi-scale vegetation storage and regulation efficiency indexes.

[0053] Preferably, the multi-band spectrum data collecting module includes:

[0054] a near-infrared spectrum collecting unit configured to obtain near-infrared spectrum data in a wavelength range of 780-2500 nm;

[0055] a mid-infrared spectrum collecting unit configured to obtain mid-infrared spectrum data in a wavelength range of 2500-25000 nm;

[0056] a Raman spectrum collecting unit configured to obtain Raman spectrum data in a wave number range of 200-4000 cm-1;

[0057] a spectrum preprocessing unit configured to perform calibration, noise elimination, baseline correction, standardization, and differential transformation processing on the multi-band spectrum data.

[0058] Preferably, the prediction applying module includes:

[0059] a response feature parameterizing unit configured to extract key feature parameters from a water status change curve and establish a precipitation-response relationship model;

[0060] a storage and regulation capacity predicting unit configured to predict storage and regulation response features of the vegetation under different extreme precipitation scenarios based on the relationship model;

[0061] The spatial distribution visualization unit is used for mapping the regulation and storage efficiency index to the geographical space to generate a regulation and storage efficiency spatial distribution map.

[0062] The decision support unit is used for providing decision support for sponge city planning, flood risk management and ecosystem service evaluation based on the regulation and storage efficiency evaluation results.

[0063] The beneficial effects of the present application include:

[0064] 1. The accurate characterization of the molecular level water state is realized: through multi-band molecular spectral analysis, the water molecules in different binding states in the vegetation tissue can be distinguished, the molecular mechanism of vegetation water regulation is revealed in depth, and the characterization accuracy is improved by more than 90% compared with traditional methods.

[0065] 2. The limitation of static measurement is broken through, and the continuous dynamic monitoring of the whole process of extreme precipitation is realized: by tracking the changes of water molecules in vegetation tissue before, during and after extreme precipitation, the dynamic process of water absorption, balance and release is fully captured, and the time resolution can reach 15 minutes.

[0066] 3. A complete molecular water dynamics lag time analysis framework is established: the lag time characteristics of vegetation regulation and storage are quantified from the perspective of molecular water dynamics, and the time characteristics of the absorption, balance and release stages are accurately distinguished, which provides a new perspective for understanding the mechanism of vegetation regulation and storage.

[0067] 4. The quantitative relationship between vegetation microstructure and water regulation function is revealed: the correlation between vegetation microstructure and water dynamic parameters is established, and the influence of different structural characteristics on the regulation and storage capacity is revealed, which provides a scientific basis for the optimization selection of vegetation.

[0068] 5. The integrated analysis of regulation and storage characteristics from the molecular scale to the regional scale is realized: by establishing the correlation between molecular spectral characteristics and remote sensing hyperspectral data, the point scale analysis method is extended to the regional scale, which provides a new method for regional hydrological process research.

[0069] 6. A complete application support system is formed: decision support tools for sponge city planning, watershed management and flood control and disaster reduction, ecosystem service evaluation and other aspects are developed, and the research results are transformed into practical application solutions. BRIEF DESCRIPTION OF DRAWINGS

[0070] Figure 1 The flowchart of the extreme precipitation-vegetation regulation and storage lag response quantification method of the present application. DETAILED DESCRIPTION

[0071] The application provides an extreme precipitation-vegetation regulation and storage hysteresis response quantification method and system, which aims to study the extreme precipitation regulation and storage mechanism of vegetation from the perspective of molecular hydrodynamics and quantifies the hysteresis response characteristics.

[0072] With reference to Figure 1 The extreme precipitation-vegetation regulation and storage hysteresis response quantification method of the application comprises the following steps: acquiring multi-band molecular spectrum data of a vegetation sample; constructing a water state spectrum graph representing the state of water molecules in the vegetation tissue; extracting a molecular hydrodynamic hysteresis characteristic; acquiring a vegetation tissue microstructure parameter; determining a multi-scale vegetation regulation and storage efficiency index; and generating a regional vegetation regulation and storage response prediction result and a regulation and storage efficiency spatial distribution graph.

[0073] The application first acquires multi-band molecular spectrum data of a vegetation sample, including near-infrared spectrum data, mid-infrared spectrum data and Raman spectrum data.

[0074] 1. Near-infrared spectrum data acquisition: a spectrometer is used to acquire near-infrared spectrum data with a wavelength range of 780-2500 nm. Preferably, the water molecule absorption peaks near 1450 nm and 1940 nm are focused on, which mainly reflect the combination band and frequency band vibration characteristics of the water molecule -OH bond. During acquisition, the distance between the spectrometer and the plant leaf is kept at 3-5 cm, the sampling angle is 45°, each sample point is measured 10 times and the average value is taken to eliminate the influence of random errors.

[0075] 2. Mid-infrared spectrum data acquisition: a Fourier transform infrared spectrometer is used to acquire mid-infrared spectrum data with a wavelength range of 2500-25000 nm. Preferably, the -OH stretching vibration near 3400 and the H-O-H bending vibration near 1640 are focused on, which can directly reflect the fundamental frequency vibration of water molecules. During acquisition, the resolution is set to 4 , the scanning number is 32, and air is used as the reference for background acquisition.

[0076] 3. Raman spectrum data acquisition: a Raman spectrometer is used to acquire Raman scattering spectrum in the range of 200-4000 . Preferably, the -OH stretching vibration scattering characteristics in the range of 3200-3600 are focused on, which are highly sensitive to the changes in intermolecular hydrogen bond structure. During acquisition, the excitation wavelength uses a 785 nm laser with a power controlled below 10 mW to avoid sample damage, and the integration time is 10 seconds, with each point repeated 5 times.

[0077] In practical application, data collection is divided into three key stages: before extreme precipitation (baseline data is established), during extreme precipitation (data is collected every 15 minutes to capture the characteristics of rapid response), and after extreme precipitation (data is collected every hour for the first 4 hours, then every 4 hours for 24 hours to track the characteristics of the recovery process).

[0078] The collected raw spectral data needs to be pre-processed. In one embodiment of the present application, the pre-processing procedure is as follows:

[0079] 1. Spectral calibration: white reference calibration is performed using a standard reflectance plate (such as Spectralon®) to eliminate the effects of instrument response and ambient light. White reference calibration is performed before each sample measurement to ensure data accuracy.

[0080] 2. Noise elimination: for near-infrared and mid-infrared spectra, the Savitzky-Golay smoothing algorithm is used with a window size of 9 points and a polynomial order of 3 to balance noise elimination and feature preservation. For Raman spectra, the wavelet transform denoising algorithm is used with Daubechies4 wavelet and 5 levels of decomposition to effectively eliminate random noise while preserving the fine features of water molecule vibrations.

[0081] 3. Baseline correction: for near-infrared spectra, multi-point linear interpolation is used for baseline correction; for mid-infrared spectra, the iterative polynomial fitting method is used with a polynomial order of 5 and an iteration number of 10; for Raman spectra, the improved adaptive iterative baseline correction algorithm (ModPoly) is used to accurately eliminate fluorescence background interference.

[0082] 4. Spectral standardization: standard normal variate transformation (SNV) and multi-scattering correction (MSC) methods are used to eliminate the effects of physical factors such as sample thickness, density, and surface scattering. In addition, for spectra collected at different time points, the maximum value normalization method is used for standardization processing to ensure data comparability.

[0083] 5. Spectral differential transformation: first and second derivatives of the spectrum are calculated to enhance the fine features of water molecule vibrations. First derivative is used to enhance absorption edge and peak position changes, and second derivative is used to enhance peak shape and peak width changes. Derivative calculation uses the Savitzky-Golay differential smoothing method with a window size of 11 points and a polynomial order of 3.

[0084] Through the above pre-processing steps, the present application can effectively eliminate various interference factors and extract high-quality spectral information reflecting the state of water molecules in plant tissues.

[0085] Based on the pre-processed multi-band spectral data, the present application constructs a water state spectrum representing the state of water molecules in plant tissues. The specific implementation process is as follows:

[0086] 1. Water molecular vibration feature extraction: The present application firstly extracts the key vibration features from multi-band spectral data, which can characterize the water molecular state in plant tissues.

[0087] Preferably, the water absorption peak features near 1450 nm and 1940 nm are extracted from near-infrared spectral data. The 1450 nm peak mainly corresponds to the second overtone of the first -OH stretching vibration, and the 1940 nm peak corresponds to the combination band of -OH stretching and H-O-H bending. The intensity, position, half-peak width, and peak shape of these two peaks can accurately reflect the binding state and distribution of water molecules in plant tissues. In the extraction process, Gaussian peak fitting method is used to determine the characteristic parameters such as peak position, peak height, peak area, and half-peak width.

[0088] The -OH stretching vibration features near 3400 and the H-O-H bending vibration features near 1640 are extracted from mid-infrared spectral data. These fundamental vibration features are extremely sensitive to the binding state of water molecules and can directly reflect the interaction between water molecules and plant tissue macromolecules. In particular, free water and bound water show obvious differences in these characteristic peaks: the -OH stretching vibration peak of free water is relatively narrow and high in position (about 3600 ), while the peak of bound water is wide and low in position (about 3300 ). In the extraction process, peak decomposition technology is used to separate the contributions of water with different binding states.

[0089] The -OH stretching vibration scattering features in the range of 3200-3600 are extracted from Raman spectral data. Raman spectrum is extremely sensitive to the hydrogen bond network structure of water molecules, and can reflect the strength and order of the hydrogen bond network between water molecules through the profile changes in the -OH stretching vibration region. In the extraction process, a multi-component Gaussian-Lorentz mixed model is used to fit this region to separate the contributions of water molecules with different hydrogen bond strengths.

[0090] 2. Water state parameter calculation: Based on the extracted vibration features, the present application calculates a series of key parameters that characterize the state of water molecules.

[0091] Preferably, the free water index (FWI) is calculated, defined as the ratio of the intensity of high-frequency (3600 near) and low-frequency (3200 near) -OH stretching vibration:

[0092] ,

[0093] where, is 3600 cm -1Nearby Raman scattering intensity, reflecting the contribution of weak hydrogen bond or free OH; FWI = 3200 Nearby Raman scattering intensity, reflecting the contribution of strong hydrogen bond network. The higher the FWI value, the greater the proportion of free water. In practical application, for broad-leaved forest vegetation, the normal range of FWI is 0.4-0.7, which can increase to 0.8-1.2 after extreme precipitation, indicating that the amount of free water increases greatly.

[0094] Calculate the bound water index (BWI), defined as the ratio of bending (1640 ) and stretching (3400 ) vibration intensity:

[0095] BWI = 1640 ,

[0096] Where is the mid-infrared absorption intensity near 1640 , reflecting the H-O-H bending vibration; is the mid-infrared absorption intensity near 3600 , reflecting the -OH stretching vibration. The higher the BWI value, the greater the proportion of water bound to biological macromolecules. In practical application, for broad-leaved forest vegetation, the normal range of BWI is 0.3-0.5, which can decrease to 0.2-0.3 in the early stage after extreme precipitation, and then gradually increase to 0.4-0.6, indicating that water gradually changes from free state to bound state.

[0097] Calculate the hydrogen bond network index (HBNI), defined as the displacement of the -OH stretching vibration peak:

[0098] ,

[0099] Where, is the standard vibration frequency of free -OH (about 3650 ） ; is the center frequency of the measured -OH stretching vibration peak. The higher the HBNI value, the stronger the hydrogen bond network between water molecules. In practical application, for broad-leaved forest vegetation, the normal range of HBNI is 150-250 , which can increase to 250-350 after extreme precipitation, indicating that the hydrogen bond network between water molecules is enhanced.

[0100] Based on spectral decomposition technology, the relative proportions of free water, weakly bound water and strongly bound water in plant tissues are quantitatively calculated. This is done by analyzing the Raman spectrum in the range of 3200-3600 The multi-component fitting is realized in the region, and the typical proportion distribution is: under normal conditions, the strong bound water accounts for 40-50%, the weak bound water accounts for 30-40%, and the free water accounts for 10-20%; after extreme precipitation, the proportion of free water can increase to 40-60% at the initial stage, and then gradually return to the normal proportion.

[0101] 3. Water state spectrum drawing: based on the calculated water state parameters, the water state spectrum is constructed, and the comprehensive state of water molecules in the vegetation tissue is intuitively represented.

[0102] Preferably, a water state dynamic distribution graph is constructed, taking time as the horizontal axis and the proportion of water in different states as the vertical axis, to show the change trend of the proportion of water in various states before and after extreme precipitation. The graph can clearly show that the free water rapidly increases at the initial stage of precipitation, and then gradually decreases, and the corresponding change of the proportion of bound water.

[0103] A hydrogen bond strength dynamic change graph is constructed, taking time as the horizontal axis and the hydrogen bond network index as the vertical axis, to show the dynamic evolution process of the interaction between water molecules. The graph can reveal the formation, reorganization and stabilization process of the hydrogen bond network of water molecules after extreme precipitation.

[0104] A water state conversion graph is constructed to represent the atlas of the mutual conversion relationship of water in different states, and to reflect the dynamic process of the state conversion of water molecules in the vegetation tissue under extreme precipitation conditions. The graph is displayed in the form of a state transition matrix, and the matrix elements represent the probability or rate of conversion from one state to another state.

[0105] The water state spectrum constructed by the above method can comprehensively represent the dynamic change process of water molecules in the vegetation tissue, and lay a foundation for extracting the time delay characteristics.

[0106] Based on the water state spectrum, the molecular water dynamics time delay characteristics are extracted. The specific implementation process is as follows:

[0107] 1. Water dynamic process segmentation: the invention first divides the extreme precipitation response process into clear dynamic stages.

[0108] Preferably, the absorption stage is identified, that is, the stage from the start of precipitation to the maximum water absorption of the vegetation. The characteristics of this stage are that the free water index (FWI) rapidly rises, and the total amount of water rapidly increases. By monitoring the growth rate of FWI, when the growth rate is greater than 0.01 / minute, it is determined that the absorption stage starts, and when the FWI reaches the peak (the change rate is close to 0), it is determined that the absorption stage ends. For broad-leaved forest vegetation, the typical absorption stage lasts for 30-90 minutes, which depends on the precipitation intensity and vegetation characteristics.

[0109] The balance phase, which is the phase of water redistribution within the vegetation, is identified. This phase is characterized by a gradual decrease in the free water index (FWI) and a gradual increase in the bound water index (BWI), with the total water content being relatively stable. The balance phase is determined to start when the FWI begins to decrease (the rate of change becomes negative) and the BWI begins to increase (the rate of change is greater than 0.005 / hour), and to end when the absolute values of the rates of change of the FWI and the BWI are both less than 0.002 / hour. For broadleaf forest vegetation, the typical duration of the balance phase is 2-8 hours.

[0110] The release phase, which is the phase of surplus water release and transpiration, is identified. This phase is characterized by a gradual decrease in all water indices, tending towards the pre-extreme precipitation level. The release phase is determined to start when the FWI, the BWI and the HBNI all begin to decrease, and to end when these indices return to 90% of the pre-precipitation baseline level. For broadleaf forest vegetation, the typical duration of the release phase is 12-48 hours, and is significantly affected by environmental conditions such as temperature, humidity and wind speed.

[0111] 2. Analysis of the migration path of water molecules: The present application analyzes the migration path and conversion rules of water molecules within the vegetation tissue during the extreme precipitation process.

[0112] Preferably, based on the time-series changes in the water state parameters, the conversion path of water molecules from free water to weakly bound water and then to strongly bound water, and the reverse conversion path, are identified. This is achieved by analyzing the trends of changes in the FWI, the BWI and the water state proportions. A typical conversion path includes: at the beginning of the precipitation, external water enters the plant tissue in the form of free water (the FWI increases); subsequently, part of the free water is converted into weakly bound water (the FWI decreases and the BWI increases); finally, part of the weakly bound water is further converted into strongly bound water (the BWI remains high). The release phase exhibits the opposite process.

[0113] The rate of water conversion between different states is calculated to represent the regulation ability of the vegetation tissue to water. The conversion rate is calculated by the time derivative of the water state parameters:

[0114] ,

[0115] wherein, is the rate of conversion from state A to state B, with the unit of % / hour; [A] and [B] are the water proportions of states A and B, respectively. In practical applications, for broadleaf forest vegetation, the conversion rate of free water to bound water is usually 5%-15% / hour.

[0116] The migration of water molecules between different tissues, such as epidermis, mesophyll and vascular bundle, is analyzed by combining the spectral data of different tissue parts. This is achieved by measuring the micro-spectrum of different parts of the plant leaves. Typical migration rules include: water first enters the leaf through the stomata or directly penetrates the cuticle, then diffuses in the intercellular space of mesophyll cells in the form of free water, and then is absorbed by cells to become bound water, and the excess water is transported to other parts of the plant through the vascular system.

[0117] 3. Time lag feature quantification: based on the analysis of molecular water dynamic process, the invention quantifies key time lag feature parameters.

[0118] Preferably, the molecular water response time (T ) is defined as the time interval from the peak of extreme precipitation intensity to the peak of vegetation water status parameter:

[0119] ,

[0120] wherein, ) is the time when the water status parameter (such as FWI) reaches the peak; is the time when the precipitation intensity reaches the peak. Reflects the response speed of vegetation to absorb water. In practical applications, for broad-leaved forest vegetation, usually 15-45 minutes.

[0121] The molecular water balance time (T ) is defined as the time required for the change rate of water status parameter to fall below the preset threshold:

[0122] ,

[0123] wherein, is the start time of the balance stage; is the end time of the balance stage; is the water status parameter (such as FWI or BWI); is the absolute value of the change rate thereof. Reflects the speed of water in the vegetation to reach a balanced state. In practical applications, for broad-leaved forest vegetation, usually 2-8 hours.

[0124] The molecular water release time (T ) is defined as the time required for the water status parameter to drop from the peak to half the peak:

[0125] ,

[0126] wherein, is the time when the water status parameter drops to half the peak. Reflects the ability of vegetation to release excess water. In practical applications, for broad-leaved forest vegetation, Typically 6-24 hours, significantly affected by environmental conditions.

[0127] Determine the state transition time (T ), defined as the characteristic time of water transition from free state to bound state:

[0128] ,

[0129] Where, is the first-order rate constant of free water conversion to bound water, obtained by fitting the exponential decay curve of the proportion of free water. Reflects the ability of vegetation tissue to fix water. In practical applications, for broad-leaved forest vegetation, Typically 2-6 hours.

[0130] Through the above lag time characterization method, the invention can accurately describe the response dynamics of vegetation to extreme precipitation, providing quantitative basis for subsequent storage efficiency evaluation.

[0131] The invention obtains vegetation tissue microstructure parameters and analyzes their relationship with water dynamic characteristics. The specific implementation process is as follows:

[0132] 1. Obtain vegetation tissue microstructure parameters: The invention obtains key tissue microstructure parameters related to vegetation water regulation.

[0133] Preferably, the epidermis structure parameters are obtained, including epidermis thickness (μm), stomatal density (pieces / mm²), cuticle thickness (μm), etc. These parameters are obtained by optical microscope or scanning electron microscope observation. In practical applications, for broad-leaved forest vegetation, the typical epidermis thickness is 15-30 μm, the stomatal density is 100-500 pieces / mm², and the cuticle thickness is 2-8 μm.

[0134] Obtain mesophyll structure parameters, including mesophyll cell arrangement (palisade tissue / spongy tissue ratio), intercellular space ratio (%), cell wall thickness (μm), etc. These parameters are obtained by tissue sectioning and microscopic image analysis. In practical applications, for broad-leaved forest vegetation, the typical palisade / spongy tissue ratio is 0.8-1.5, the intercellular space ratio is 15%-30%, and the cell wall thickness is 0.1-0.5 μm.

[0135] Obtain vascular bundle structure parameters, including vascular bundle density (strips / mm²), vessel diameter (μm), sieve tube characteristics, etc. These parameters are obtained by tissue sectioning and microscopic image analysis. In practical applications, for broad-leaved forest vegetation, the typical vascular bundle density is 3-8 strips / mm², and the vessel diameter is 20-60 μm.

[0136] 2. Microstructure-water relationship analysis: The present invention analyzes the relationship between vegetation microstructure characteristics and water dynamic parameters.

[0137] Preferably, the skin structure parameters and the moisture response time are analyzed ( ) relationship, revealing the influence of microstructural characteristics on the initial absorption of water. The study found that the pore density is related to There was a significant negative correlation (r=-0.75, p<0.01), that is, the greater the stomatal density, the shorter the response time; the thickness of the cuticle was significantly There was a significant positive correlation (r=0.68, p<0.01), indicating that the thicker the cuticle, the longer the response time. This suggests that stomata are the main channel for water to enter vegetation under extreme precipitation conditions, while the cuticle is an obstacle.

[0138] Analysis of mesophyll structural parameters and water balance time ( ), the relationship between water state distribution, revealing the influence of microstructural characteristics on water storage. The study found that the ratio of intercellular space and There was a significant negative correlation (r=-0.62, p<0.01), indicating that a larger interstitial ratio was associated with a shorter equilibrium time. There was also a significant positive correlation (r=0.71, p<0.01) between cell wall thickness and the bound water index (BWI), indicating that thicker cell walls were associated with a higher bound water ratio. This suggests that interstitial spaces primarily influence the rate of water diffusion in tissues, while the cell wall is the primary storage location for bound water.

[0139] The relationship between vascular bundle structural parameters and water migration path and conversion rate was analyzed to reveal the influence of microstructural characteristics on water transport. The study found that vascular bundle density and water release time ( ) showed a significant negative correlation (r=-0.58, p<0.01), indicating that greater vascular bundle density was associated with shorter release times. Vessel diameter also showed a significant negative correlation with the rate of conversion of free water to bound water (r=-0.64, p<0.01), indicating that larger vessel diameters were associated with lower conversion rates. This suggests that the vascular system primarily influences the efficiency of excess water discharge, while vessel characteristics influence long-distance water transport and distribution.

[0140] 3. Analysis of dynamic characteristics of hydraulic conductivity: Based on the changes in spectral characteristics, the present invention analyzes the dynamic regulation characteristics of hydraulic conductivity of vegetation tissue.

[0141] Preferably, the water conductivity of vegetation tissue is evaluated based on the rate of change of water molecule state, and its dynamic changes under extreme precipitation conditions are analyzed. The water conductivity (K) can be estimated by the following formula:

[0142] ,

[0143] in, For water flux, it is estimated by the rate of change of the free water index (FWI); For water potential gradient, it is estimated by the spatial distribution of water status parameters.

[0144] It is found that the water conductivity of vegetation increases significantly (by 200% to 300%) at the beginning of extreme precipitation, and then gradually decreases to the normal level. This dynamic change reflects the ability of vegetation to actively regulate water absorption and transport.

[0145] By comparing the water change rate of different tissue parts, it is found that water is mainly transported along the following path: epidermal stomata / cutin layer → intercellular space of mesophyll cells → cytoplasm / cell wall → vascular system. Under extreme precipitation conditions, vegetation can control water transport efficiency by adjusting stomatal opening, cell membrane permeability and vascular water conductivity.

[0146] By analyzing the change rule of water conductivity resistance of vegetation tissue during extreme precipitation, the mechanism of active regulation of water transport by vegetation is revealed. Water conductivity resistance (R) is the inverse of water conductivity:

[0147] ,

[0148] It is found that the change of water conductivity resistance during extreme precipitation presents a U-shaped curve: it decreases rapidly at the beginning, remains low in the middle, and gradually increases at the end. This change pattern reflects the adaptive strategy of vegetation to reduce resistance to promote water absorption during the absorption stage, and to increase resistance to control water loss during the release stage.

[0149] Through the above microstructure-water coupling analysis, the invention reveals the mechanism of the influence of vegetation microstructure characteristics on water regulation capacity, and provides a scientific basis for understanding the structural basis of vegetation regulation and storage efficiency.

[0150] Based on the molecular hydrodynamic characteristics and the microstructure parameters of vegetation tissue, the invention determines the multi-scale vegetation regulation and storage efficiency index. The specific implementation process is as follows:

[0151] 1. Cell-scale regulation and storage index determination: Based on the molecular hydrodynamic characteristics, the invention constructs a cell-scale regulation and storage efficiency index.

[0152] Preferably, the cell water capacity (CWC) is determined, which is defined as the total amount of water that can be stored in a unit volume of cells, and is calculated by integrating the free water index (FWI) over time:

[0153] ,

[0154] wherein, and are the start and end times of the absorption stage, respectively; CWC is the free water index at time t. CWC reflects the ability of cells to store water, with units of arbitrary units time (e.g. AU h). In practical applications, for broadleaf forest vegetation, typical values of CWC are 0.5-2.0 AU h. The cell water content retention time (CRT) is determined, defined as the sum of the water balance time and the water release time, with a value of 0.5-2.0 AU h: wherein, is the water balance time; is the water release time. CRT reflects the sustained ability of cells to retain water, with units of time (e.g. hours). In practical applications, for broadleaf forest vegetation, typical values of CRT are 8-32 hours.

[0155] The cell water status stability (CSS) is determined, defined as the inverse of the standard deviation of the fluctuations in the water status parameter:

[0156] ,

[0157] wherein, is the standard deviation of the FWI in the equilibrium phase. CSS reflects the ability of cells to maintain a stable water status, with units of dimensionless. In practical applications, for broadleaf forest vegetation, typical values of CSS are 5-20.

[0158] 2. Organ scale storage index determination: The present application is based on the water dynamic characteristics of different tissue sites to construct an organ scale storage efficiency index.

[0159] Preferably, the tissue water diffusion coefficient (TDC) is determined, defined as the characteristic rate of water diffusion in the tissue:

[0160] ,

[0161] wherein, is the characteristic diffusion distance, estimated by the tissue thickness; is the water balance time. TDC reflects the efficiency of water migration within the tissue, with units of mm2 / h. In practical applications, for broadleaf forest vegetation, typical values of TDC are 0.02-0.2 mm 2 / h. The tissue water capacity gradient (TCG) is determined, defined as the difference in water capacity between different tissue sites:

[0162] ,

[0163] wherein, is the difference in bound water index between different tissue sites; is the distance between tissue sites. TCG reflects the unevenness of water spatial distribution in vegetation tissues, with units of mm -1In practical applications, the typical TCG value for broadleaf forest vegetation is 0.05-0.5 mm -1 .

[0164] The tissue water coordination coefficient (TCS) is determined, which is defined as the correlation degree of the water state changes of different tissue parts:

[0165] ,

[0166] wherein, is the FWI covariance of tissue parts i and j; and are the corresponding standard deviations, respectively. The TCS reflects the ability of each part of the vegetation to coordinate water regulation, and the unit is dimensionless, with a value range of [-1, 1]. In practical applications, the typical TCS value for broadleaf forest vegetation is 0.6-0.9.

[0167] 3. Overall vegetation regulation and storage efficiency evaluation: The present application comprehensively evaluates the regulation and storage efficiency of the overall vegetation by combining the cell and tissue scale indicators.

[0168] Preferably, the regulation and storage intensity (RI) is determined, which is defined as the maximum change range of the water state parameters of the vegetation before and after extreme precipitation:

[0169] ,

[0170] wherein, is the water state parameter (such as FWI or BWI). The RI reflects the capacity of the vegetation to respond to extreme precipitation, and the unit is dimensionless. In practical applications, the typical RI value for broadleaf forest vegetation is 0.3-0.8. The regulation and storage duration (RD) is determined, which is defined as the sum of the water response time, the balance time and the release time: ,

[0171] wherein, is the water response time; is the water balance time; is the water release time. The RD reflects the total duration of the regulation and storage process of the vegetation, and the unit is time (such as hours). In practical applications, the typical RD value for broadleaf forest vegetation is 9-36 hours. The regulation and storage efficiency (RE) is determined, which is defined as the ratio of the regulation and storage intensity to the regulation and storage duration:

[0172] ,

[0173] The RE reflects the regulation and storage ability of the vegetation per unit time, and the unit is h -1 In practical applications, the typical RE value for broadleaf forest vegetation is 0.01-0.05 h -1Determine regulatory resilience (RR), defined as the ability of vegetation water status parameters to return to baseline levels, through recovery rate and recovery completeness assessment:

[0174] ,

[0175] in, is the recovery rate; To restore integrity; is the baseline level; Peak level The final level. RR reflects the ability of vegetation to recover from extreme precipitation disturbances. It is dimensionless and ranges from [0, 1]. In practice, for broadleaf forest vegetation, the typical RR value is 0.7-0.95.

[0176] Through the above-mentioned multi-scale vegetation storage efficiency evaluation method, the present invention can comprehensively quantify the vegetation's storage capacity for extreme precipitation, providing a scientific basis for subsequent regional applications.

[0177] Based on multi-scale vegetation storage efficiency indicators, the present invention generates regional vegetation storage response prediction results and storage efficiency spatial distribution maps. The specific implementation process is as follows:

[0178] 1. Response characteristic parameterization: The present invention parameterizes the vegetation water molecule state change curve to establish a precipitation-response model.

[0179] Preferably, key characteristic parameters are extracted from the water state change curve, including response delay (minutes), rise rate (% / minute), peak position (dimensionless), fall rate (% / hour), recovery degree (%), etc. These parameters are obtained through curve fitting methods and can concisely characterize the complex response process.

[0180] Extreme precipitation events are parameterized, including characteristics such as precipitation intensity (mm / h), duration (h), cumulative amount (mm), and temporal distribution (e.g., uniform, front-peak, and back-peak). These parameters can be extracted from meteorological observation data.

[0181] Establish a relationship model between the characteristic parameters of precipitation and the characteristic parameters of the response curve. Preferably, the model is constructed using multiple regression or support vector regression methods:

[0182] ,

[0183] in, is the response characteristic parameter vector; is the precipitation characteristic parameter vector; 、 and is the regression coefficient; is the error term. Model training is based on historical extreme precipitation events and corresponding vegetation response data, and model validation uses cross-validation methods. The typical model prediction accuracy (R²) can reach 0.75-0.85.

[0184] 2. Prediction of vegetation storage capacity: Based on historical response data and parameter relationship models, the invention predicts the storage response of vegetation to future extreme precipitation events.

[0185] Preferably, the storage response characteristics of vegetation are predicted for different intensity and duration of extreme precipitation scenarios. For example, for a precipitation scenario of 30mm / h, lasting 2 hours, the response characteristics of broadleaf forest vegetation may be predicted as: response delay 20 minutes, rising rate 2.5% / minute, peak FWI 0.9, falling rate 15% / hour, and recovery degree 95%.

[0186] Determine the critical threshold of vegetation storage capacity, and predict the failure of storage function beyond the threshold. By analyzing the failure cases of vegetation storage in historical extreme events, the critical threshold is determined. For example, for broadleaf forest vegetation, when the precipitation intensity exceeds 50mm / h and the duration exceeds 3 hours, the storage efficiency may decrease by more than 50%, and partial failure of storage function may occur.

[0187] Predict the time and conditions required for vegetation to recover to normal moisture state after extreme precipitation. Based on the relationship model between storage resilience (RR) index and environmental conditions (such as temperature, humidity, light), the recovery process is predicted. For example, under the conditions of high temperature (30°C) and low humidity (50%) in summer, broadleaf forest vegetation may take 24-36 hours to fully recover from moderate intensity extreme precipitation (30mm / h, 2 hours).

[0188] 3. Regional scale application: The invention extends the point scale analysis method to regional scale, supporting practical application.

[0189] Preferably, the molecular spectral characteristics are associated with remote sensing hyperspectral data to realize rapid evaluation of regional vegetation storage capacity. By establishing a regression model between spectral index and storage efficiency index, the laboratory analysis results are extended to regional scale. For example, by analyzing the 1450nm and 1940nm band characteristics of satellite hyperspectral data, the water response time (τᵣ) and storage intensity (RI) of regional vegetation are estimated.

[0190] Generate a spatial distribution map of regional vegetation storage capacity, including the spatial distribution of parameters such as storage intensity (RI), storage duration (RD), and storage efficiency (RE). The spatial resolution can reach 30m x 30m, which can clearly show the spatial differences in storage capacity under different vegetation types and different terrain conditions.

[0191] Based on the evaluation results of the regulation and storage efficiency, a planning auxiliary system for sponge city is developed. The system can optimize the layout of green space system and vegetation configuration according to the regulation and storage efficiency distribution of urban green space, and improve the response ability of city to extreme precipitation. For example, high-efficiency vegetation coverage can be increased in the area with low regulation and storage efficiency, or the vegetation type combination can be adjusted to improve the overall efficiency.

[0192] The vegetation regulation and storage information is integrated to optimize the flood warning and reservoir scheduling strategy. By considering the regulation and storage capacity of upstream vegetation, the accuracy of flood forecasting and warning time are improved, and the reservoir scheduling decision is optimized. For example, when the upstream vegetation regulation and storage capacity is close to saturation due to previous precipitation, the reservoir can be pre-discharged to increase the flood control storage capacity.

[0193] The ecosystem service value of vegetation regulation and storage is quantified to support the design of ecological compensation mechanism. Based on the regulation and storage efficiency index, the regulation and storage service value of different vegetation types and different regional vegetation is evaluated to provide scientific basis for the development of ecological compensation standard. For example, forest areas with high regulation and storage efficiency (RE) can obtain higher ecological compensation standard.

[0194] Through the above prediction and application methods of regional vegetation regulation and storage response, the research results at molecular scale are transformed into practical application at regional scale, which provides scientific support for flood control and disaster reduction and ecosystem management.

[0195] The present application also provides a time-delay response quantification system for extreme precipitation-vegetation regulation and storage, which comprises the following modules: a multi-band spectral data acquisition module, a water state spectrum construction module, a time-delay feature extraction module, a microstructure analysis module, a regulation and storage efficiency evaluation module, and a prediction and application module.

[0196] The multi-band spectral data acquisition module is used to acquire near-infrared spectrum data, mid-infrared spectrum data and Raman spectrum data of the vegetation sample.

[0197] Preferably, the module includes a near-infrared spectrum acquisition unit for acquiring near-infrared spectrum data with a wavelength range of 780-2500 nm. The unit is equipped with a high-precision near-infrared spectrometer (such as ASD FieldSpec4 Hi-Res) with a spectral resolution of 1 nm and a reflectance accuracy of 0.1%.

[0198] The module also includes a mid-infrared spectrum acquisition unit for acquiring mid-infrared spectrum data with a wavelength range of 2500-25000 nm. The unit is equipped with a Fourier transform infrared spectrometer (such as Bruker VERTEX70) with a spectral resolution of 4 and a high signal-to-noise ratio (>10000:1).

[0199] The module also includes a Raman spectrum acquisition unit for acquiring Raman spectrum data in the wave number range of 200-4000 cm⁻¹. The unit is equipped with a high-sensitivity Raman spectrometer (such as Horiba LabRAM HR Evolution) with a spectral resolution of <1 and a high spatial resolution (<1 μm).

[0200] The module also includes a spectrum preprocessing unit for calibrating, noise removing, baseline correction, standardizing, and differential transforming the multi-band spectrum data. The unit integrates various spectrum preprocessing algorithms and can efficiently process a large amount of spectrum data.

[0201] The water state spectrum construction module is used to construct a water state spectrum representing the state of water molecules in the vegetation tissue based on the preprocessed multi-band spectrum data.

[0202] Preferably, the module realizes the function of extracting water molecule vibration features from near-infrared, mid-infrared, and Raman spectrum data. Through peak fitting and decomposition algorithms, the spectrum features related to the binding state of water molecules are accurately extracted.

[0203] The module also realizes the function of calculating water state parameters such as free water index (FWI), bound water index (BWI), and hydrogen bond network index (HBNI). Through the ratio relationship between spectrum features, the distribution of different state water in the vegetation tissue is quantified.

[0204] The module also realizes the function of drawing dynamic distribution maps of water state, dynamic change maps of hydrogen bond strength, and water state conversion maps. Through visual expression, the dynamic change process of water state is intuitively displayed.

[0205] The lag time feature extraction module is used to extract molecular water dynamics lag time features based on the water state spectrum.

[0206] Preferably, the module realizes the function of dividing the extreme precipitation response process into absorption stage, equilibrium stage, and release stage. Through analyzing the change trend and change rate of water state parameters, the start and end times of each stage are automatically identified.

[0207] The module also realizes the function of analyzing the state conversion path and conversion rate of water molecules in the vegetation tissue. By tracking the dynamic change of the proportion of different state water, the migration law of water molecules in the vegetation tissue is revealed.

[0208] The module also realizes the function of calculating lag time features such as water response time (τᵣ), water balance time (τᵇ), and water release time (τᵈ). Through time series analysis, the time characteristics of vegetation storage are accurately quantified.

[0209] ​​The microstructure analysis module is used to obtain the microstructure parameters of vegetation tissues and analyze their relationship with water dynamic characteristics.

[0210] Preferably, the module realizes the functions of obtaining epidermis structure parameters, mesophyll structure parameters and vascular bundle structure parameters. By importing microscopic image data, key structure parameters are automatically extracted.

[0211] The module also realizes the function of analyzing the relationship between microstructure parameters and water dynamic characteristics. Through correlation analysis and regression model, the influence mechanism of structure characteristics on water regulation is revealed.

[0212] The module also realizes the function of evaluating the water conductivity of vegetation tissues and its dynamic changes. Through water flow and water potential gradient calculation, the water transport efficiency of vegetation is quantified.

[0213] The storage and regulation performance evaluation module is used to determine multi-scale vegetation storage and regulation performance indicators based on molecular hydrodynamic lag characteristics and vegetation tissue microstructure parameters.

[0214] Preferably, the module realizes the functions of calculating cell water capacity (CWC), cell water retention time (CRT) and cell water state stability (CSS) and other cell-scale storage and regulation indicators. Through integral calculation and statistical analysis, the cell water regulation capacity is quantified.

[0215] The module also realizes the functions of calculating tissue water diffusion coefficient (TDC), tissue water capacity gradient (TCG) and tissue water coordination coefficient (TCS) and other tissue-scale storage and regulation indicators. Through spatial analysis and correlation calculation, the tissue water coordination regulation capacity is quantified.

[0216] The module also realizes the functions of calculating storage and regulation intensity (RI), storage and regulation duration (RD), storage and regulation efficiency (RE) and storage and regulation resilience (RR) and other overall vegetation storage and regulation performance indicators. Through comprehensive analysis, the storage and regulation capacity of vegetation is comprehensively evaluated.

[0217] The prediction application module is used to generate regional vegetation storage and regulation response prediction results and storage and regulation performance spatial distribution maps according to multi-scale vegetation storage and regulation performance indicators.

[0218] Preferably, the module includes a response characteristic parameterization unit for extracting key characteristic parameters from water state change curves and establishing a precipitation-response relationship model. The unit integrates curve fitting and multiple regression algorithms and can efficiently process a large amount of response data.

[0219] The module also includes a storage and regulation capacity prediction unit for predicting the storage and regulation response characteristics of vegetation under different extreme precipitation scenarios based on the relationship model. The unit can simulate various precipitation scenarios to provide scientific basis for decision-making.

[0220] The module also includes a spatial distribution visualization unit for mapping the storage efficiency indicators to geographic space, generating storage efficiency spatial distribution maps. The unit supports multiple map projections and spatial interpolation methods, generating high-quality distribution maps.

[0221] The module also includes a decision support unit for providing decision support for sponge city planning, flood risk management, and ecosystem service assessment based on storage efficiency evaluation results. The unit integrates various decision support tools to facilitate practical applications.

[0222] Through the above system architecture, the invention realizes the full-process automation from data collection, processing and analysis to application decision-making, providing systematic technical support for extreme precipitation-vegetation storage lag time response research.

[0223] This embodiment selects typical broad-leaved forest vegetation (such as maple) for detailed storage lag time response analysis.

[0224] 1. Multi-band molecular spectral data acquisition and preprocessing: ASD FieldSpec4 Hi-Res spectrometer is used to collect near-infrared spectral data (780-2500 nm), Bruker VERTEX70 spectrometer is used to collect mid-infrared spectral data (2500-25000 nm), and Horiba LabRAM HREvolution spectrometer is used to collect Raman spectral data (200-4000 ). The collection time points include: before extreme precipitation (baseline), during extreme precipitation (every 15 minutes), and after extreme precipitation (every hour for the first 4 hours, then every 4 hours for 24 hours). The raw spectral data is preprocessed through calibration, noise removal, baseline correction, standardization and differential transformation.

[0225] 2. Water state spectrum construction and analysis: key water vibration features are extracted from the preprocessed spectral data. The intensity and position changes of the 1450 nm and 1940 nm peaks in the near-infrared spectrum reflect the overall water content change; the changes of the 3400 and 1640 peaks in the mid-infrared spectrum reflect the water binding state; the contour changes in the 3200-3600 region of the Raman spectrum reflect the hydrogen bond network structure. Based on these features, the free water index (FWI), bound water index (BWI) and hydrogen bond network index (HBNI) are calculated to construct the water state spectrum. The results show that after extreme precipitation, the FWI increases rapidly (from 0.5 to 1.1) and then decreases slowly; the BWI decreases first (from 0.4 to 0.25) and then rises (to 0.5); the HBNI increases continuously (from 180 to 320 ) and then recovers slowly.

[0226] 3. Molecular hydrodynamics lag-time feature extraction: Based on the water state spectrum, three stages of extreme precipitation response are identified: absorption stage (lasts about 60 minutes), equilibrium stage (lasts about 5 hours), and release stage (lasts about 18 hours). By analyzing the conversion relationship of different state water, it is found that the conversion rate of free water to bound water is about 8% / hour. Key lag-time parameters are calculated: water response time τᵣ=25 minutes, water equilibrium time =5 hours, water release time =12 hours, state conversion time =3.5 hours.

[0227] 4. Vegetation tissue microstructure-water coupling analysis: Obtain the tissue microstructure parameters of maple leaf: epidermis thickness 20 μm, stomatal density 350 / mm 2 , cuticle thickness 5 μm, ratio of palisade / spongy tissue 1.2, intercellular space ratio 22%, cell wall thickness 0.3 μm, vascular bundle density 5 / mm 2 , vessel diameter 35 μm. Analysis results show that stomatal density is significantly negatively correlated with water response time (r=-0.78), intercellular space ratio is significantly negatively correlated with water equilibrium time (r=-0.65), and vascular bundle density is significantly negatively correlated with water release time (r=-0.62). Water conductivity analysis shows that water conductivity increases by about 250% at the beginning of extreme precipitation, and then gradually returns to normal within 12 hours.

[0228] 5. Multi-scale vegetation regulation and storage performance evaluation: Based on lag-time features and microstructure parameters, regulation and storage performance indicators are calculated. Cell-scale indicators: cell water capacity CWC=1.2 AU·h, cell water retention time CRT=17 hours, cell water state stability CSS=12. Tissue-scale indicators: tissue water diffusion coefficient TDC=0.08 mm 2 / h, tissue water capacity gradient TCG=0.2 mm -1 , tissue water synergy coefficient TCS=0.82. Overall regulation and storage performance: regulation intensity RI=0.6, regulation duration RD=24 hours, regulation efficiency RE=0.025 h -1 , regulation resilience RR=0.88.

[0229] 6. Regional vegetation regulation and storage response prediction and application: Based on experimental data, a precipitation-response relationship model is established to predict the response characteristics of maple forest to extreme precipitation of different intensities. For example, for extreme precipitation of 50 mm / h lasting 3 hours, the predicted response delay is 30 minutes, the regulation intensity is 0.7, the regulation duration is 36 hours, and the regulation efficiency is reduced to 0.019 h -1The critical threshold of the regulation and storage capacity is determined: when the rainfall intensity exceeds 60 mm / h and the duration exceeds 4 hours, the regulation and storage efficiency decreases by more than 40%, and partial failure occurs. By correlating laboratory analysis results with remote sensing hyperspectral data, a spatial distribution map of the regulation and storage efficiency of the study area is generated, identifying key regulation and storage areas (28% of the total area) and weak areas (15% of the total area).

[0230] In this example, different ecological types of vegetation (broadleaf forest, coniferous forest, herbaceous plant, shrub) are selected for comparative analysis.

[0231] 1. Multi-band molecular spectral data acquisition and preprocessing: Spectral data acquisition and preprocessing are performed on 4 different types of vegetation (Liquidambar formosana represents broadleaf forest, Pinus massoniana represents coniferous forest, Setaria viridis represents herbaceous plant, Rhododendron represents shrub). Each type selects 3 samples as repeats.

[0232] 2. Water status spectrum construction and comparison: Water status spectra of each vegetation type are constructed, and their characteristic differences are compared. The results show that the change range of FWI of coniferous forest is smaller than that of broadleaf forest (maximum 0.8 vs 1.1), but the change range of BWI is larger than that of broadleaf forest (0.3-0.6 vs 0.25-0.5); the FWI of herbaceous plant changes rapidly (peaks at 15 minutes vs 25 minutes for broadleaf forest), but it also recovers quickly (total duration 12 hours vs 24 hours for broadleaf forest); the change range of HBNI of shrub is between that of broadleaf forest and coniferous forest.

[0233] 3. Difference analysis of lag characteristic parameters between vegetation types: The lag characteristic parameters of each vegetation type are calculated and compared. Water response time τᵣ: herbaceous (15 minutes) < shrub (20 minutes) < broadleaf forest (25 minutes) < coniferous forest (35 minutes); water balance time : herbaceous (2 hours) < shrub (3.5 hours) < broadleaf forest (5 hours) < coniferous forest (7 hours); water release time : herbaceous (6 hours) < shrub (9 hours) < broadleaf forest (12 hours) < coniferous forest (16 hours).

[0234] 4. Analysis of vegetation type-specific microstructure-water relationship: The microstructure characteristics of different vegetation types and their relationship with water dynamics are compared. The epidermis thickness and cuticle thickness of coniferous forest are the largest (30 μm and 8 μm, respectively), corresponding to its longest response time; the proportion of intercellular space of herbaceous plant is the highest (30%), corresponding to its shortest balance time; the vascular bundle density of broadleaf forest is higher (5 / mm²), but the vessel diameter is moderate (35 μm), showing balanced water transport capacity.

[0235] 5. Comparison of vegetation types on the index of regulation and storage efficiency: Calculate and compare the index of regulation and storage efficiency of each vegetation type. Regulation intensity RI: coniferous forest (0.5) < herb (0.55) < shrub (0.58) < broad-leaved forest (0.6); Regulation duration RD: herb (12 hours) < shrub (18 hours) < broad-leaved forest (24 hours) < coniferous forest (30 hours); Regulation efficiency RE: coniferous forest (0.017h -1） < broad-leaved forest (0.025h -1） < shrub (0.032h -1） < herb (0.046h -1） ; Regulation resilience RR: herb (0.75) < shrub (0.82) < broad-leaved forest (0.88) < coniferous forest (0.92).

[0236] 6. Classification of vegetation types on regulation and storage characteristics: Based on the time delay characteristics and the index of regulation and storage efficiency, classify different vegetation types and establish a classification system of vegetation regulation and storage capacity. Vegetation is divided into four categories: Class I (such as coniferous forest) - long duration type, characterized by slow response, long duration, low efficiency, and high resilience; Class II (such as broad-leaved forest) - balanced type, characterized by moderate response, moderate duration, moderate efficiency, and high resilience; Class III (such as shrub) - intermediate type, with indicators between I / II and IV; Class IV (such as herb) - fast response type, characterized by fast response, short duration, high efficiency, and low resilience.

[0237] Through this comparison and analysis of different vegetation types, the invention reveals the differences in regulation and storage mechanisms and efficiency of different vegetation types, providing a scientific basis for optimizing vegetation configuration.

[0238] This embodiment extends the analysis method to the regional scale.

[0239] 1. Selection of typical sample area: Select a typical sample area (about 10 km²) containing multiple vegetation types, including urban parks, suburban forest land, riparian zones, and different ecosystem types.

[0240] 2. Obtain high-spectral remote sensing data: Use an airborne hyperspectral imager (such as AVIRIS) to obtain high-spectral image data of the sample area, with a spatial resolution of 5m, a spectral range of 400-2500nm, and 224 continuous bands.

[0241] 3. Point-to-area scale conversion: Based on laboratory analysis results, establish the correspondence between spectral indices and regulation and storage efficiency indicators. For example, by analyzing the characteristics (such as location, depth, area, asymmetry, etc.) of the 1450nm and 1940nm bands, a regression model (R²=0.82) is established to predict the water response time (τᵣ); by analyzing the first derivative characteristics in the range of 950-1050nm, a regression model (R²=0.79) is established to predict the regulation intensity (RI).

[0242] 4. Regional storage capacity evaluation: Based on the established model, the remote sensing image is processed to generate a regional vegetation storage capacity spatial distribution map. The analysis results show that about 25% of the area in the sample area has high storage efficiency (RE>0.03h -1） , mainly distributed in mixed forest and healthy broad-leaved forest area; about 18% of the area shows low storage efficiency (RE<0.015h -1） , mainly distributed in degraded forest land and artificial green land area. The high storage resilience (RR>0.85) area accounts for about 30%, mainly distributed in mature forest area; the low storage resilience (RR<0.7) area accounts for about 15%, mainly distributed in newly built green land and grassland area.

[0243] 5. Extreme precipitation response prediction: Selecting historical extreme precipitation events, the vegetation response is predicted based on the regional storage capacity distribution map. The prediction results are compared and verified with the actual observation data (obtained through ground monitoring stations and satellite remote sensing), and the average prediction accuracy reaches 83%. Based on the verified model, regional response prediction maps under different intensity extreme precipitation scenarios (20mm / h, 40mm / h, 60mm / h, duration 1-4 hours) are constructed to identify potential high-risk areas.

[0244] The present application introduces molecular spectroscopy analysis technology, and constructs a complete extreme precipitation-vegetation storage lag time response quantification method system, reveals the vegetation storage mechanism from the molecular hydrodynamics, accurately quantifies the lag time response characteristics, and provides a scientific basis for sponge city planning, watershed management and flood control and disaster reduction. Compared with the prior art, the present application realizes accurate characterization of molecular water state, breaks through the limitation of static measurement, establishes a complete molecular hydrodynamics lag time analysis framework, reveals the quantitative relationship between vegetation microstructure and water regulation function, realizes the integration analysis of storage characteristics from molecular scale to regional scale, forms a complete application support system, and has important theoretical and practical value.

[0245] The above only describes the preferred embodiments of the present application and is not used to limit the present application. For those skilled in the art, the present application can have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A quantitative method for the delayed response of extreme precipitation to vegetation storage, characterized by: include: Acquiring multi-band molecular spectral data of a vegetation sample, wherein the multi-band molecular spectral data includes near-infrared spectral data, mid-infrared spectral data, and Raman spectral data; Based on the multi-band molecular spectral data, constructing a water state spectrum that characterizes the state of water molecules in plant tissues, the water state spectrum including a free water index, a bound water index, and a hydrogen bond network index; Extracting molecular water dynamics hysteresis characteristics according to the water state spectrum, wherein the molecular water dynamics hysteresis characteristics include water response time, water equilibrium time and water release time; Acquiring vegetation tissue microstructure parameters, wherein the vegetation tissue microstructure parameters include epidermal structure parameters, mesophyll structure parameters, and vascular bundle structure parameters; Determining a multi-scale vegetation storage efficiency index based on the molecular hydrodynamic hysteresis characteristics and the vegetation tissue microstructure parameters, wherein the multi-scale vegetation storage efficiency index includes a cell-scale storage index and a tissue-scale storage index; Based on the multi-scale vegetation storage efficiency index, regional vegetation storage response prediction results and a storage efficiency spatial distribution map are generated.

2. The method according to claim 1, characterized in that The obtaining of multi-band molecular spectral data of vegetation samples comprises: Acquiring near-infrared spectral data in a wavelength range of 780-2500 nm, wherein the near-infrared spectral data includes vibration characteristics of a combination band and a harmonic band of water molecules; Acquiring mid-infrared spectral data in a wavelength range of 2500-25000 nm, wherein the mid-infrared spectral data includes fundamental frequency vibration characteristics of water molecules; The range of wave number is 200~4000 Raman spectral data comprising scattering characteristics of hydrogen bond structures between water molecules; The multi-band molecular spectral data are preprocessed, and the preprocessing includes spectral calibration, noise elimination, baseline correction, spectral standardization and spectral differential transformation.

3. The method according to claim 1, characterized in that The step of constructing a water state spectrum representing the state of water molecules in plant tissue based on the multi-band molecular spectral data includes: Extracting water absorption peak features near 1450nm and 1940nm from the near-infrared spectral data; Extract 3400 from the mid-infrared spectral data -OH stretching vibration characteristics near 1640 nearby HOH bending vibration characteristics; Extract 3200~3600 from the Raman spectrum data -OH stretching vibration scattering characteristics of the range; Based on the extracted spectral features, the free water index representing the relative content of free water, the bound water index representing the relative content of bound water, and the hydrogen bond network index representing the strength of the hydrogen bond network between water molecules were calculated; Construct a dynamic distribution diagram of moisture status with time as the horizontal axis and the moisture ratio of different states as the vertical axis.

4. The method according to claim 1, wherein The extracting of molecular water dynamics hysteresis characteristics according to the water state spectrum includes: Based on the water state spectrum, the extreme precipitation response process is divided into absorption stage, equilibrium stage and release stage; Analyze the state conversion paths and conversion rates of water molecules in vegetation tissues at each stage; The time interval from the peak of extreme precipitation intensity to the peak of vegetation water state parameters was determined as the water response time; Determine the time required for the rate of change of the moisture state parameter to drop below a preset threshold as the moisture balance time; The time required for the moisture state parameter to drop from the peak value to the half-peak value was determined as the moisture release time.

5. The method according to claim 1, wherein Determining the multi-scale vegetation storage efficiency index based on the molecular hydrodynamic hysteresis characteristics and the vegetation tissue microstructure parameters includes: Analyzing the relationship between the epidermal structural parameters and the moisture response time to determine the effect of the epidermis on the initial absorption of moisture; Analyzing the relationship between the mesophyll structural parameters and the water balance time to determine the effect of the mesophyll on water storage; Analyzing the relationship between the vascular bundle structural parameters and the water conversion rate to determine the impact of the vascular bundle on water transport; Determine the water capacity of a cell based on the total amount of water that can be stored per unit volume of the cell; The cell water retention time is determined based on the sum of the water balance time and the water release time; Based on the characteristic rate of water diffusion in tissue, the tissue water diffusion coefficient is determined; Based on the differences in water content between different tissue parts, the tissue water content gradient is determined.

6. The method according to claim 1, characterized in that Generating a regional vegetation storage response prediction result and a storage efficiency spatial distribution map based on the multi-scale vegetation storage efficiency index includes: Extract key characteristic parameters such as response delay, rising rate, peak position, falling rate and recovery degree from the moisture state change curve; Establish a relationship model between precipitation characteristic parameters and response curve characteristic parameters; Based on the relationship model, the regulation and storage response characteristics of vegetation under extreme precipitation scenarios of different intensities and durations are predicted; Determine the critical threshold of vegetation storage capacity and predict the potential failure of storage function if the threshold is exceeded; The multi-scale vegetation storage efficiency index is mapped to geographic space to generate a spatial distribution map including storage intensity, storage duration and storage efficiency.

7. The method according to claim 1, characterized in that The proposed method for quantifying the delayed response of extreme precipitation to vegetation storage is used for sponge city planning, flood risk management, and ecosystem service assessment.

8. A time-delayed response quantitative system for extreme precipitation-vegetation regulation, using the method according to any one of claims 1 to 7, characterized in that: include: Multi-band spectral data acquisition module, used to obtain near-infrared spectral data, mid-infrared spectral data and Raman spectral data of vegetation samples; A water state spectrum construction module is used to construct a water state spectrum characterizing the state of water molecules in plant tissue based on the multi-band spectral data, wherein the water state spectrum includes a free water index, a bound water index, and a hydrogen bond network index; A hysteresis feature extraction module, configured to extract molecular hydrodynamic hysteresis features according to the moisture state spectrum, wherein the molecular hydrodynamic hysteresis features include moisture response time, moisture equilibrium time, and moisture release time; A microstructure analysis module is used to obtain vegetation tissue microstructure parameters, wherein the vegetation tissue microstructure parameters include epidermal structure parameters, mesophyll structure parameters and vascular bundle structure parameters; A regulation and storage efficiency evaluation module is used to determine a multi-scale vegetation regulation and storage efficiency index based on the molecular hydrodynamic hysteresis characteristics and the vegetation tissue microstructure parameters, wherein the multi-scale vegetation regulation and storage efficiency index includes a cell-scale regulation and storage index and a tissue-scale regulation and storage index; The prediction application module is used to generate regional vegetation storage response prediction results and a storage efficiency spatial distribution map based on the multi-scale vegetation storage efficiency index.

9. The system according to claim 8, characterized in that The multi-band spectral data acquisition module includes: A near-infrared spectrum acquisition unit, used to acquire near-infrared spectrum data in the wavelength range of 780-2500nm; A mid-infrared spectrum acquisition unit, used to acquire mid-infrared spectrum data in the wavelength range of 2500-25000nm; Raman spectrum acquisition unit, used to obtain wavenumber range of 200 to 4000 Raman spectral data of The spectrum preprocessing unit is used to perform calibration, noise elimination, baseline correction, standardization and differential transformation on the multi-band spectrum data.

10. The system according to claim 8, wherein: The prediction application module includes: Response characteristic parameterization unit, used to extract key characteristic parameters from the water state change curve and establish a precipitation-response relationship model; A storage capacity prediction unit, configured to predict the storage response characteristics of vegetation under different extreme precipitation scenarios based on the relationship model; Spatial distribution visualization unit, used to map the storage efficiency index to the geographic space and generate a spatial distribution map of storage efficiency; The decision support unit is used to provide decision support for sponge city planning, flood risk management and ecosystem service assessment based on the results of storage efficiency assessment.

Images