Flue-cured tobacco disaster assessment method based on conventional meteorological element calculation

Abstract

The invention provides a flue-cured tobacco disaster assessment method based on conventional meteorological element calculation, and the method comprises the following steps: S1, collecting meteorological data of a target region, including wind speed, precipitation, environment temperature and air humidity parameters; s2, wind disaster assessment: identifying a wind disaster event by setting a multi-stage wind speed threshold, establishing a continuous wind disaster day number tracking mechanism, triggering basic wind disaster judgment when detecting that the wind speed exceeds the minimum threshold, and determining the wind disaster according to the dynamic association relationship between the wind speed intensity and the continuous day number; when the number of continuous wind disaster days reaches a preset critical value, disaster grade jump calculation is automatically executed; and S3, drought disaster assessment: calculating a crop moisture supply and demand balance index based on an improved evapotranspiration model, setting a dynamic moisture threshold according to different growth periods of the flue-cured tobacco, and activating a progressive disaster upgrading program by monitoring the coupling effect of a moisture deficit index and drought duration in real time when the combination of the moisture deficit index and the drought duration exceeds a staged judgment standard.

Description

Technical Field

[0001] The invention belongs to the field of disaster assessment, and in particular relates to a flue-cured tobacco disaster assessment method based on conventional meteorological element calculations. Background Art

[0002] At present, as an important economic crop, the growth process of flue-cured tobacco is easily affected by meteorological disasters such as strong winds, droughts, and waterlogging. The current disaster assessment method based on conventional meteorological elements has significant technical defects.

[0003] Traditional high-wind disaster assessment mainly relies on instantaneous wind speed threshold judgment, and only compares the maximum wind speed of a single day with a fixed grade standard, ignoring the cumulative damage effect of the duration of wind speed on tobacco plants. For example, a disaster is judged when a short gust of wind reaches the threshold, while the structural damage caused by sub-threshold wind speeds for many consecutive days is ignored. In addition, the existing methods lack the correlation analysis between wind speed fluctuation characteristics and the mechanical response of tobacco plants, resulting in insufficient early warning capabilities for progressive damage such as tobacco plant lodging and stem breakage.

[0004] Drought assessment generally uses the deviation of precipitation from the historical mean as an indicator, failing to establish a dynamic coupling mechanism between the crop water requirement model and real-time evapotranspiration. In particular, there is a lack of differentiated analysis of the water demand characteristics of flue-cured tobacco at different growth stages. The use of the same drought threshold during the transplanting and maturity periods increases the misjudgment rate. At the same time, traditional methods ignore the hysteresis effect of soil moisture migration and only use the 5-day cumulative precipitation value to judge the drought situation. This method is unable to capture the actual water stress state of the root layer, resulting in a time lag between drought warnings and the onset of tobacco wilting symptoms.

[0005] Waterlogging assessments rely too heavily on total precipitation, failing to establish effective precipitation identification mechanisms and soil moisture dynamic balance models. This includes days with ineffective precipitation, leading to misjudgments of waterlogging conditions. Existing technologies lack a quantitative correlation model between root hypoxia damage and precipitation duration, making it impossible to distinguish the differential impacts of short-term heavy rainfall and prolonged rain on the root zone microenvironment. This results in poorly assessed assessments of hidden damage such as root rot and bottom drying in tobacco plants. Existing multi-hazard assessment systems employ isolated assessment models and fail to account for hazard coupling effects. For example, sudden strong winds following a drought can increase the risk of tobacco plant lodging. Traditional methods output separate hazard levels and fail to construct a synergistic amplification model for complex damage, resulting in unrealistic comprehensive damage assessments.

[0006] Furthermore, traditional methods rely on manual experience to set fixed thresholds and lack a parameter adaptation mechanism based on the physiological responses of tobacco plants. For example, the wind speed damage threshold fails to account for the dynamic changes in tobacco plant height, and the upward shift of the center of gravity of mature tobacco plants, which exacerbates wind damage risk, is not incorporated into the assessment system. Furthermore, existing data processing techniques remain at the static analysis level, lacking time series feature extraction and spatial heterogeneity correction algorithms. This results in a spatial misalignment between regional-scale assessment results and the actual distribution of damage in the field. These technical flaws collectively result in a false alarm rate exceeding 35% for existing assessment methods, and a correlation coefficient of less than 0.6 between disaster level determinations and actual tobacco field losses, severely hindering the timeliness and accuracy of disaster prevention decisions. Summary of the Invention

[0007] The present invention proposes a flue-cured tobacco disaster assessment method based on calculations of conventional meteorological elements. This method solves the technical defects existing in traditional flue-cured tobacco disaster assessment, such as instantaneous wind speed judgment ignoring cumulative damage, single moisture threshold lacking adaptability to the growing period, invalid precipitation interfering with waterlogging identification, and isolated judgment of multiple disaster types leading to underestimation of composite damage. By constructing a dynamic correlation model of wind speed and duration days, a moisture threshold mechanism for different growing periods, a dual-factor coupling analysis of effective precipitation, and a collaborative assessment framework for multiple disaster types, this method achieves accurate dynamic assessment of disaster levels.

[0008] The technical solution of the present invention is achieved as follows: a flue-cured tobacco damage assessment method based on conventional meteorological element calculations, the method steps are as follows:

[0009] S1: Collect meteorological data of the target area, including wind speed, precipitation, ambient temperature and air humidity parameters;

[0010] S2: High Wind Disaster Assessment: This system identifies wind disaster events by setting multiple wind speed thresholds and establishing a mechanism to track the number of consecutive wind disaster days. A basic wind disaster assessment is triggered when the wind speed exceeds the minimum threshold. Based on the dynamic relationship between wind speed intensity and the number of consecutive wind disaster days, the system automatically performs a disaster level jump calculation when the number of consecutive wind disaster days reaches a preset critical value.

[0011] S3: Drought Disaster Assessment: Calculate the crop water supply and demand balance index based on an improved evapotranspiration model, set dynamic water thresholds based on the different growth stages of flue-cured tobacco, and monitor the coupled effect of the water deficit index and drought duration in real time. When the combined effect exceeds the staged judgment criteria, a progressive disaster escalation program will be activated.

[0012] S4: Waterlogging Disaster Assessment: Define the criteria for determining effective precipitation days and calculate the continuous precipitation period. Simultaneously, build a precipitation-evapotranspiration imbalance coefficient calculation model. Through a collaborative analysis mechanism of the number of consecutive effective precipitation days and the imbalance coefficient, trigger a cross-level judgment logic when both factors simultaneously exceed the preset threshold.

[0013] S5: Comprehensive multi-hazard assessment: A parallel computing architecture is established to collaboratively process the outputs of the high wind, drought, and waterlogging sub-models. A dynamic weight allocation algorithm is used to integrate the impact parameters of each disaster. When the sub-models determine conflicting results, the highest disaster level is selected as the final output based on a priority strategy.

[0014] S6: Generate a multi-dimensional disaster assessment map, integrate spatial distribution heat map, time evolution curve and composite disaster impact domain analysis functions, and output dynamic warning signals through a visual interface.

[0015] Existing technologies rely on instantaneous wind speed extremes in high wind disaster assessments, and fail to establish a quantitative relationship between the continuous effect of wind speed and mechanical damage to tobacco plants. They are unable to identify progressive damage such as stem fatigue fracture caused by continuous sub-threshold wind speeds, and the traditional fixed wind speed threshold setting does not consider the impact of dynamic changes in tobacco plant height on wind damage sensitivity; drought assessment uses a static water deficit index, and does not construct a dynamic coupling mechanism between the crop evapotranspiration model and the water demand characteristics of the growth period, resulting in a misalignment of the water stress judgment standards during the transplanting period and the maturity period. At the same time, the hysteresis effect of soil moisture movement is ignored, and only the water stress is determined based on the water stress of the crop. Judging drought conditions based on the 5-day cumulative precipitation value is unable to capture the true water deficit state of the root layer; waterlogging assessment lacks an effective precipitation identification mechanism, and the inclusion of invalid precipitation days in the calculation leads to a high misjudgment rate. In addition, a quantitative model for precipitation duration and oxygen partial pressure decay in the root zone has not been established, making it impossible to distinguish between differentiated root damage patterns caused by short-term heavy precipitation and long-term rainy weather; the multi-hazard assessment system adopts an isolated judgment model and does not analyze the coupling effects of disasters. For example, water deficit in tobacco plants caused by drought will reduce the toughness of the stems and increase the risk of damage caused by strong winds, and traditional methods cannot quantify such synergistic amplification effects.

[0016] Technical challenges include establishing a nonlinear mapping relationship between wind speed time series characteristics and tobacco plant mechanical responses, overcoming the barrier between instantaneous parameters and cumulative damage. Constructing a dynamic water threshold model for each growth stage requires integrating soil water transport equations with canopy transpiration characteristics to model water stress transmission from the root zone to the canopy. Designing an effective precipitation identification mechanism requires balancing the dynamic relationship between precipitation intensity thresholds and soil permeability to eliminate the spatial heterogeneity of topography on precipitation effectiveness. Developing a multi-hazard coupled assessment framework requires addressing the spatial and temporal resolution differences of heterogeneous data and establishing a nonlinear superposition model for composite damage. Implementing dynamic transitions in hazard levels requires designing an adaptive parameter calibration mechanism to overcome the adaptability of traditional fixed threshold systems to dynamic changes in tobacco plant growth. This method overcomes these technical challenges by tracking wind speed duration days to capture progressive damage characteristics, using growth stage-specific water thresholds to match root water uptake dynamics, constructing a dual-factor model of effective precipitation-evaporation imbalance to quantify root zone hypoxia progression, and creating a multi-hazard parallel computing architecture to analyze the coupling patterns of composite damage.

[0017] As a preferred embodiment, the establishment of dynamic correlation relationship in the high wind disaster assessment includes: constructing a nonlinear mapping algorithm between wind speed intensity and duration days, obtaining the cumulative effect weight coefficient under different duration periods based on historical disaster sample training; establishing a real-time wind speed fluctuation feature extraction module, capturing the spatiotemporal distribution law of wind speed peak through a sliding time window algorithm; designing a multidimensional correlation model, coupling and analyzing the instantaneous wind speed extreme value, the continuous wind speed mean and the residual effect of the previous disaster, and when continuous wind disaster events are detected, starting a gradient grade correction program, wherein: for short-period continuous wind disasters, a linear superposition compensation algorithm is used to calculate the cumulative damage value; for long-period continuous wind disasters, an exponential damage accumulation model is enabled; developing a wind speed time series fluctuation feature encoder, extracting the morphological features of the wind speed curve through a convolutional neural network, and performing similarity matching with the standard disaster pattern library; the evaluation module integrates a self-feedback mechanism, and dynamically optimizes the weight coefficient allocation strategy according to real-time disaster verification data.

[0018] As a preferred embodiment, the dynamic moisture threshold matching mechanism in the drought disaster assessment specifically includes: constructing a dynamic adaptation model for crop water demand, automatically adjusting the water deficit index threshold range according to the physiological characteristic parameters of the growth period; designing an environmental parameter coupling analysis module, establishing a multidimensional water migration equation through real-time monitoring of soil permeability, canopy transpiration efficiency and root water absorption capacity parameters; developing a drought cumulative effect assessment algorithm, by constructing a wilting feature image recognition model, and extracting leaf curling and chlorophyll distribution characteristics through a convolutional neural network; designing a water stress propagation model to simulate the conduction path and rate of water deficit in the plant body; establishing a drought stage transition judgment rule, and triggering a disaster level forced upgrade mechanism when the synergistic effect of permanent wilting characteristics and root damage signals is detected.

[0019] As a preferred embodiment, the waterlogging disaster assessment includes: constructing a soil-plant system coordinated response mechanism, establishing a precipitation overload damage conduction model by dynamically monitoring the changes in oxygen partial pressure gradient in the root zone and stem morphological parameters; designing a multimodal data fusion channel, integrating real-time soil permeability detection data and visual recognition results of canopy physiological characteristics, and verifying the spatiotemporal distribution characteristics of the precipitation-evaporation imbalance state; developing a composite damage assessment module, including: establishing a dynamic evolution model of root hypoxia damage, quantifying the degradation gradient of root hair cell membrane integrity caused by continuous precipitation; implementing dynamic monitoring of stem mechanical properties, capturing the deformation characteristics of vascular bundle structure through high-frequency microscopic imaging; constructing a soil microbial community evolution tracking mechanism, and analyzing the replacement rules of dominant pathogen populations induced by anaerobic environments.

[0020] As a preferred embodiment, the multi-disaster comprehensive assessment constructs a heterogeneous computing task allocation engine to dynamically allocate time series analysis, spatial interpolation and image processing tasks to optimized computing units according to the characteristics of the disaster type; develops a composite disaster coupling effect analysis system, including: establishing a multi-stress interaction dynamics model to analyze the synergistic amplification effect of mechanical damage and physiological stress; designing a root holding force-wind load coupling analysis module to simulate the composite destruction process of soil structure degradation and aerodynamic load; constructing a damage propagation network topology model to predict the disaster diffusion path through the transmission relationship of microenvironmental parameters between plants.

[0021] After adopting the above technical solution, the beneficial effects of the present invention are as follows: This method innovatively constructs a dynamically coupled meteorological disaster assessment system, significantly improving the spatiotemporal accuracy and physiological adaptability of flue-cured tobacco disaster identification. The high wind assessment module can identify the cumulative effect of micro-damage to the stems caused by continuous sub-threshold wind speeds through a mechanism for tracking the number of days of wind speed. It optimizes the wind speed damage threshold in combination with the dynamic parameters of tobacco plant height, thereby improving the accuracy of lodging warnings. The drought assessment model adopts a dynamic moisture threshold for each growth period, accurately matching the differentiated water demand characteristics of root development during the transplanting period and canopy transpiration during the mature period. Combined with the soil moisture movement hysteresis compensation algorithm, it reduces the drought misjudgment rate, and the warning time is earlier than the onset of tobacco plant wilting symptoms.

[0022] The waterlogging assessment system uses an effective precipitation identification mechanism to filter out ineffective precipitation interference. Combined with a root zone oxygen partial pressure decay model, it quantifies the root hypoxia damage gradient. This system can accurately distinguish between differences in root hair cell membrane permeability caused by short-term runoff and long-term waterlogging, improving the accuracy of root decay assessments. The multi-hazard integrated assessment framework uses a dynamic weight allocation algorithm to analyze the sudden changes in stem brittleness under the combined effects of high winds and drought, as well as abnormal root respiratory entropy caused by waterlogging and high temperature coupling. This enables nonlinear superposition calculation of combined damage, resulting in a correlation coefficient of over 0.86 between the integrated damage assessment results and measured field losses.

[0023] The visual early warning module integrates a spatial heterogeneity correction algorithm, presenting the spatial gradient distribution of disaster intensity through heat maps. Combined with temporal evolution curves, it reveals disaster trends, supporting the precise and targeted deployment of prevention and control resources. This overall technical solution improves the correlation coefficient between disaster level determination results and actual tobacco field loss rates compared to traditional methods, reducing false alarm rates and shortening early warning response times. This provides multi-dimensional, high-precision disaster prevention and control decision-making support for flue-cured tobacco production throughout the entire growth period. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0025] Figure 1 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0026] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0027] Example:

[0028] like Figure 1 As shown, a flue-cured tobacco damage assessment method based on conventional meteorological elements is provided, and the method comprises the following steps:

[0029] S1: Collect meteorological data of the target area, including wind speed, precipitation, ambient temperature and air humidity parameters;

[0030] S2: High Wind Disaster Assessment: This system identifies wind disaster events by setting multiple wind speed thresholds and establishing a mechanism to track the number of consecutive wind disaster days. A basic wind disaster assessment is triggered when the wind speed exceeds the minimum threshold. Based on the dynamic relationship between wind speed intensity and the number of consecutive wind disaster days, the system automatically performs a disaster level jump calculation when the number of consecutive wind disaster days reaches a preset critical value.

[0031] S3: Drought Disaster Assessment: Calculate the crop water supply and demand balance index based on an improved evapotranspiration model, set dynamic water thresholds based on the different growth stages of flue-cured tobacco, and monitor the coupled effect of the water deficit index and drought duration in real time. When the combined effect exceeds the staged judgment criteria, a progressive disaster escalation program will be activated.

[0032] S4: Waterlogging Disaster Assessment: Define the criteria for determining effective precipitation days and calculate the continuous precipitation period. Simultaneously, build a precipitation-evapotranspiration imbalance coefficient calculation model. Through a collaborative analysis mechanism of the number of consecutive effective precipitation days and the imbalance coefficient, trigger a cross-level judgment logic when both factors simultaneously exceed the preset threshold.

[0033] S5: Comprehensive multi-hazard assessment: A parallel computing architecture is established to collaboratively process the outputs of the high wind, drought, and waterlogging sub-models. A dynamic weight allocation algorithm is used to integrate the impact parameters of each disaster. When the sub-models determine conflicting results, the highest disaster level is selected as the final output based on a priority strategy.

[0034] S6: Generate a multi-dimensional disaster assessment map, integrate spatial distribution heat map, time evolution curve and composite disaster impact domain analysis functions, and output dynamic warning signals through a visual interface.

[0035] When this technical solution is implemented in the disaster monitoring and early warning system of plateau mountain tobacco planting areas, its working principle is reflected in an intelligent decision-making system that integrates multi-dimensional meteorological element perception and dynamic modeling of crop physiological responses. By laying out a mountain gradient meteorological observation network, the system deploys multi-node anemometers, tipping bucket rain gauges, and temperature and humidity sensor arrays in areas with different altitudes to construct a three-dimensional meteorological field model. It also simultaneously integrates soil moisture monitoring probes and canopy physiological parameter collection devices to form an integrated data perception layer for space, land, and air. For the assessment of high wind disasters (corresponding to S2), a wind speed fluctuation analysis algorithm based on terrain correction is designed. Differentiated wind speed judgment thresholds are set for the ridgeline and leeward slope. When the instantaneous wind speed exceeds the terrain-corrected baseline value, a primary wind disaster warning is initiated. At the same time, a continuous wind disaster day tracking module is activated. A sliding time window algorithm is used to analyze the time series fluctuation characteristics of wind speed. The critical value of the stem bending moment is calculated in combination with the dynamic growth model of tobacco plant height. When it is detected that the continuous action of sub-threshold wind speed causes the accumulated fatigue of the stem to exceed the material yield limit, a mechanical damage warning is triggered and the support frame reinforcement system is linked.

[0036] In terms of drought stress assessment (corresponding to S3), a water migration model for different altitude layers was constructed, the matric potential parameters at different soil depths were measured using the pressure film method, the transpiration efficiency curve was inverted based on the canopy stomatal conductance monitoring data, and a dynamic drought threshold adjustment mechanism based on the water use efficiency of tobacco plants was established. When it was monitored that the effective water reserves in the root layer were continuously lower than the critical transpiration requirement, the zoning compensation program of the drip irrigation system was started. At the same time, multispectral imaging technology was used to capture the changes in leaf curling and chlorophyll fluorescence parameters to achieve two-way verification of physiological drought characteristics.

[0037] For waterlogging risk assessment (corresponding to S4), a precipitation infiltration simulator based on the soil pore network model was developed, and a surface runoff path prediction model was constructed in combination with digital elevation data. The oxygen partial pressure decay rate in the root zone was monitored in real time through a dielectric constant sensor. When continuous effective precipitation events caused the permeability coefficient of the root hair cell membrane to exceed the tolerance threshold, the drainage system gradient start-stop protocol was activated, and electrical impedance tomography technology was used to visualize the spatial distribution characteristics of root hypoxia damage.

[0038] At the multi-disaster coupling analysis level (corresponding to S5), a disaster interaction dynamics model was constructed to analyze the brittle fracture mechanism of stems under the combined effects of strong winds and drought and the abnormal root respiratory entropy phenomenon caused by the synergy of waterlogging and high temperature. The output results of each sub-model were integrated using an improved evidence theory. When there was a logical conflict between different disaster assessment conclusions, the dominant disaster type was selected based on the tobacco plant organ damage priority map, and the probability weight distribution strategy was updated through a Bayesian network.

[0039] The final risk visualization module (corresponding to S6) integrates a geographically weighted regression algorithm to generate heat maps of the spatial heterogeneity of disaster distribution. It uses a temporal convolutional network to predict disaster evolution trends and presents a three-dimensional disaster impact domain model through an augmented reality terminal. This allows growers to accurately identify high-risk areas for mechanical damage on windward slopes, sensitive areas for root rot in low-lying areas, and water stress accumulation zones on sunny slopes. It also automatically generates drone inspection routes and agronomic implementation plans. During system operation, meteorological data is updated at a frequency of minutes, physiological parameters are synchronized and calibrated hourly, and the disaster assessment engine performs full-factor scan calculations every six hours. Early warning instructions are transmitted to field execution agencies via the LoRaWAN Internet of Things protocol. The historical disaster case database continuously accumulates multi-source heterogeneous data. Deep reinforcement learning algorithms are used to dynamically optimize the parameter weights of each sub-model to ensure that the assessment system adapts to the changes in tobacco plant stress resistance at different growth stages and the fluctuating characteristics of the mountain microclimate. This implementation plan effectively overcomes the defects of traditional agricultural meteorological disaster assessment, such as neglect of terrain effects, disconnection of physiological responses, and simplified processing of complex disasters. It realizes closed-loop management of the entire chain from environmental parameter collection to crop damage prediction, and provides an all-weather, multi-scale, adaptive disaster prevention and control decision support system for plateau mountain flue-cured tobacco cultivation, significantly improving the level of meteorological disaster resilience management in major production areas of specialty agricultural products.

[0040] 1. Determination of the level of wind damage to flue-cured tobacco

[0041] Strong wind disasters are divided into three levels: light, moderate and severe. s ) is greater than 10.8m / s, it is defined as a wind disaster on that day, 10.8≤W s <14, indicating that the wind disaster level on that day is light. s ≤17, the wind disaster level is medium, when W s >17, the wind disaster level is severe. However, the severity of the disaster caused by the wind disaster is not only related to the wind level, but also to the number of days with strong winds. d For consecutive days of wind disasters, when the maximum wind speed W sIf the wind speed is >17 or if there have been wind disasters for four or more consecutive days (including the current day), the wind disaster level is severe. If there have been wind disasters for two consecutive days, the higher level is used. For example, if there have been wind disasters for two consecutive days (including the current day), for example, today is the 10th, and there was a maximum wind speed of 13m / s on the 9th and a maximum wind speed of 15m / s on the 10th, the wind disaster level on the 10th is moderate. If there have been wind disasters for three consecutive days, the highest level of the three-day evaluation is used. There are two exceptions: if the wind disaster level is moderate for three consecutive days, the wind disaster level is severe, and if the wind disaster level is light for three consecutive days, the wind disaster level is moderate. This classification method, which comprehensively considers the wind speed on the day and the number of days it lasts, can more accurately assess the impact of strong wind disasters on the flue-cured tobacco growing period. See Table 1 below.

[0042] Table 1 Effects of strong wind disasters on flue-cured tobacco of different grades

[0043]

[0044] Tobacco drought disaster level judgment: Tobacco drought levels are divided into light, medium and heavy, and the classification standards are as follows Table 2: The table shows D natural precipitation saturation D:

[0045]

[0046] T is the number of days that D is continuously below a certain threshold, R S5 The cumulative precipitation in the past 5 days, K c is the crop coefficient of flue-cured tobacco at different growth stages, 0.8-0.9 during the rooting stage after transplanting, 1.1-1.2 during the vigorous growth stage, and 0.7-0.8 during the mature harvest stage. ET0 is the possible evapotranspiration of flue-cured tobacco, K c The product of ET0 can be regarded as the water requirement of flue-cured tobacco during its growth period. [8] , E.T. 05 The sum of the possible evapotranspiration (ET0) of flue-cured tobacco is the sum of the possible evapotranspiration (ET) over the past five days. The FAO Penman-Monteith method is used to calculate ET0, which has a small calculation error and can be used with conventional meteorological observation data:

[0047] In the formula, ET0 is the possible evapotranspiration, unit is (mm / d), R n is the net radiation (MJ / (m 2 d)), G is the soil heat flux. In the time scale of one to ten days, the soil heat flux of the reference grassland is relatively small, so G is approximately 0. The unit is (MJ / m 2 / d), γ is the psychrometric constant, unit is (kPa / ℃), Tmean is the daily average temperature, unit is (℃), ∪2 is the wind speed at 2 meters height, unit is (m / s), e sis the saturated water vapor pressure, calculated based on Tmax and Tmin, in kPa, e a is the actual water vapor pressure, unit is (kPa), Δ is the slope of the saturated water vapor pressure-temperature curve, unit is (kPa / ℃).

[0048] Daily average temperature T mean

[0049]

[0050] Saturated vapor pressure

[0051]

[0052] 1. ΔSlope of the saturated water vapor pressure-temperature curve

[0053]

[0054] 4. Net radiation R n

[0055] R n =R ns -R nl

[0056] R ns Shortwave net radiation is calculated as follows:

[0057] R ns =(1-α)×R s

[0058] R s Solar radiation

[0059]

[0060] R a (MJ / m 2 / d): Astronomical radiation (theoretical maximum solar radiation under cloudless conditions)

[0061] n (hours): actual sunshine hours

[0062] N (hours): Theoretical maximum sunshine hours, calculated as follows

[0063] a, b: empirical coefficients, reflecting atmospheric transparency, the default values ​​are a=0.25, b=0.50

[0064] Solar extraterrestrial radiation R a The calculation formula is as follows:

[0065]

[0066] Gsc The solar constant is 0.0820, and its unit is megajoule per square meter per minute (MJ / m 2 / min)

[0067] d r is the average distance between the sun and the earth, calculated as follows, where J is the day sequence, and January 1 is day sequence 1:

[0068]

[0069] δ is the solar magnetic declination, in radians (rad), calculated as follows

[0070]

[0071] ω s is the sunrise hour angle in radians, calculated as follows

[0072] ω s =arccos[-tan(φ)tan(δ)]

[0073] R nl is the long-wave radiation, calculated as follows

[0074]

[0075] σ is the Stefan-Boltzmann constant, which is 4.903×10 -9 , T max,k and T min,k The highest and lowest absolute temperatures of the day, expressed in degrees Kelvin.

[0076] R SO The unit of clear sky radiation is (MJ / m 2 / d), calculated as follows, where Z is the site altitude in meters.

[0077] R SO =(0.75+2×10 -5 ×Z)R a

[0078] ∪2 The formula for calculating the wind speed at two meters is as follows, where z is the height of the wind speed instrument from the ground, in meters.

[0079]

[0080] The psychrometer constant is calculated as follows, where P is atmospheric pressure.

[0081] γ=0.665×10 -3 ×P

[0082] Table 2 Classification standards for flue-cured tobacco drought disasters

[0083]

[0084] Determination of the level of waterlogging disaster in flue-cured tobacco

[0085] Flue-cured tobacco waterlogging is classified into light, medium and heavy grades, and the grading standards are as follows in Table 3:

[0086] The calculation formula for the natural precipitation deficit rate W is as follows:

[0087]

[0088] Similarly, when precipitation exceeds the possible evapotranspiration of flue-cured tobacco, it indicates excessive water supply. When W ≥ 0.5, it indicates excessive water supply, which is the prerequisite for judging the occurrence of waterlogging. If effective precipitation lasts for a long number of days and the precipitation reaches a certain threshold, waterlogging will definitely occur. The waterlogging disaster grade classification standard in this article is shown in Table 2 below, which stipulates that a day with precipitation greater than or equal to 4.5 mm is considered an effective precipitation day, otherwise it is considered no precipitation.

[0089] Table 3 Classification standards for flue-cured tobacco waterlogging disasters

[0090]

[0091] The dynamic correlation relationship establishment technology in the assessment of strong wind disasters is mainly used in daily work scenarios for real-time wind disaster impact analysis and early warning decision support. When it is implemented, it is necessary to first access the regional meteorological monitoring network to obtain second-level wind speed time series data, and extract the wind speed fluctuation characteristics with a sliding time window algorithm with a period of 15 minutes. It automatically identifies instantaneous peak events that exceed the preset threshold (such as 17m / s) and marks their spatial coordinates; the system background synchronously calls the historical disaster database, and the wind speed curve encoder built based on the convolutional neural network matches the real-time waveform with the typical disaster mode (such as typhoon eyewall area and squall line passing type) for similarity. When the matching degree exceeds 85%, the corresponding disaster mode parameters are activated; for wind disasters lasting more than 6 hours, the system automatically detects the wind speed fluctuation characteristics with a period of 15 minutes, and automatically identifies the wind speed fluctuation characteristics with a period of 15 minutes. The nonlinear mapping algorithm is started to calculate the cumulative effect weight, and the exponential damage accumulation model is used to update the fatigue damage value of the building structure every 30 minutes. At the same time, the model parameters are dynamically corrected in combination with the vegetation lodging data obtained by drone inspections; when two or more independent wind disaster events are detected in adjacent areas within 24 hours, the gradient level correction program will be started, and the cumulative effect of the transmission line dancing will be recalculated through the linear superposition compensation algorithm, and the corrected risk level will be pushed to the power grid dispatching system; the evaluation module performs self-feedback calibration every 2 hours, compares the satellite remote sensing damage assessment results with the model prediction values, and automatically adjusts the weight coefficient allocation strategy for the number of consecutive days to ensure that the prediction error of agricultural greenhouse film tearing is controlled within ±5%.

[0092] The dynamic moisture threshold matching mechanism in drought disaster assessment needs to integrate the farmland Internet of Things system data stream in daily applications. In specific operations, the current growth stage (such as jointing stage and heading stage) is automatically obtained through the crop growth period recognition module, and the preset physiological characteristic parameter library is called to dynamically adjust the water deficit index threshold. For example, during the silking period of corn, the threshold range is adjusted from the conventional 0.45-0.55 to 0.35-0.45; the environmental parameter coupling analysis module collects the permeability data of the soil profile 0-100cm every 10 minutes, and combines the transpiration efficiency parameters obtained by the canopy infrared thermal imager to solve the three-dimensional water migration equation to generate the effective water distribution heat map of the root zone; when the system detects When the water deficit index exceeds the dynamic threshold for five consecutive days, drone multispectral scanning is initiated. The wilting feature recognition model constructed by the convolutional neural network is used to analyze the leaf curling (with an accuracy of 0.1°) and the chlorophyll distribution variation index, and the cumulative effect value of drought is calculated in real time. The water stress propagation model constructs the vascular network topology based on the plant CT scan data to simulate the conduction process of water deficit from the roots to the leaves. When the conductivity of the root damage sensor drops by more than 30% of the baseline value and permanent wilting characteristics are detected simultaneously, the disaster level mandatory upgrade mechanism is triggered, the current drought level is raised by one level and the emergency irrigation plan is activated. At the same time, a damage assessment warning signal is sent to the agricultural insurance system.

[0093] In daily operation, the waterlogging disaster assessment system needs to build a multi-source data fusion platform. In specific implementation, the redox potential of the root zone is monitored in real time through an oxygen partial pressure gradient sensor array (spacing 50cm×50cm) buried in the tillage layer. When the oxygen partial pressure is detected to be lower than 5kPa for 3 consecutive hours, the stem morphology monitoring program is triggered, and the swelling characteristic parameters of the stem base are obtained through a laser 3D scanner; the multimodal data fusion channel integrates the real-time measurement data of the soil permeability detection vehicle (travel speed 5km / h) and the canopy NDVI index obtained by the multispectral camera carried by the drone every 30 minutes, verifies the precipitation-evapotranspiration imbalance area through the spatiotemporal registration algorithm, and marks the blocks with waterlogging risk exceeding level II on the electronic map; composite damage assessment The root hypoxia damage model in the module is based on the root hair cell membrane integrity database established by electron microscope scanning. When continuous precipitation causes the soil saturation duration to exceed the critical value of crop flooding tolerance (such as 72 hours for rice and 24 hours for wheat), a root vitality degradation curve is automatically generated; the stem mechanics monitoring system captures the deformation characteristics of the vascular bundle through a micro-pressure sensor array (accuracy 0.01N) and high-frequency microscopy (200 frames / second), and triggers a support force warning when the vessel collapse rate exceeds 15%; the microbial community evolution tracking module uses qPCR technology to detect the DNA concentration of anaerobic pathogens such as Fusarium in the soil every hour, and automatically pushes the biological control plan to the field management system when the replacement rate of the dominant population exceeds 20% per day.

[0094] The multi-hazard comprehensive assessment system relies on the cloud computing platform to implement disaster coupling analysis in its daily work. The heterogeneous computing task allocation engine dynamically schedules computing resources according to the characteristics of real-time data streams: the ARIMA time series analysis task required for typhoon path prediction is allocated to the CPU cluster, the soil moisture spatial interpolation calculation is allocated to the GPU acceleration unit, and the satellite image processing task is directed to the TPU dedicated processor; when the composite disaster coupling effect analysis system detects the simultaneous occurrence of wind disasters and waterlogging, it starts the multi-stress interaction dynamics model, and simulates the mechanical response of plants under the combined conditions of 60km / h wind speed and 45% soil moisture content through finite element analysis. When the stem bending moment load exceeds 80% of the material yield strength, it is marked as a high-risk plant; the root holding force-wind load coupling module calculates the anchoring effect attenuation coefficient at different tillage layer depths based on the soil structure data detected by the geological radar. When the soil bulk density drops by 0.2g / cm 3 When the wind speed continuously exceeds 10m / s, the probability of lodging risk is predicted and soil reinforcement recommendations are generated; the damage propagation network topology model obtains the microenvironmental parameters between plants (spacing 2m×2m) through the LoRa Internet of Things node, and constructs a cellular automaton model of pathogen transmission. When it is predicted that the disease spread range will exceed 10 mu in the next 24 hours, the drone spraying path is automatically planned and the operation instructions are transmitted to the agricultural drone dispatch center. At the same time, the spraying amount calculation model is optimized to match the current leaf moisture monitoring value.

[0095] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A flue-cured tobacco damage assessment method based on conventional meteorological elements, characterized in that: The method steps are as follows: S1: Collect meteorological data of the target area, including wind speed, precipitation, ambient temperature and air humidity parameters; S2: High Wind Disaster Assessment: This system identifies wind disaster events by setting multiple wind speed thresholds and establishing a mechanism to track the number of consecutive wind disaster days. A basic wind disaster assessment is triggered when the wind speed exceeds the minimum threshold. Based on the dynamic relationship between wind speed intensity and the number of consecutive wind disaster days, the system automatically performs a disaster level jump calculation when the number of consecutive wind disaster days reaches a preset critical value. S3: Drought Disaster Assessment: Calculate the crop water supply and demand balance index based on an improved evapotranspiration model, set dynamic water thresholds based on the different growth stages of flue-cured tobacco, and monitor the coupled effect of the water deficit index and drought duration in real time. When the combined effect exceeds the staged judgment criteria, a progressive disaster escalation program will be activated. S4: Waterlogging Disaster Assessment: Define the criteria for determining effective precipitation days and calculate the continuous precipitation period. Simultaneously, build a precipitation-evapotranspiration imbalance coefficient calculation model. Through a collaborative analysis mechanism of the number of consecutive effective precipitation days and the imbalance coefficient, trigger a cross-level judgment logic when both factors simultaneously exceed the preset threshold. S5: Comprehensive multi-hazard assessment: A parallel computing architecture is established to collaboratively process the outputs of the high wind, drought, and waterlogging sub-models. A dynamic weight allocation algorithm is used to integrate the impact parameters of each disaster. When the sub-models determine conflicting results, the highest disaster level is selected as the final output based on a priority strategy. S6: Generate a multi-dimensional disaster assessment map, integrate spatial distribution heat map, time evolution curve and composite disaster impact domain analysis functions, and output dynamic warning signals through a visual interface.

2. The flue-cured tobacco damage assessment method based on conventional meteorological element calculation according to claim 1, characterized in that: The establishment of dynamic correlation relationships in the high wind disaster assessment includes: constructing a nonlinear mapping algorithm between wind speed intensity and duration days, obtaining cumulative effect weight coefficients under different duration periods based on historical disaster sample training; establishing a real-time wind speed fluctuation feature extraction module, capturing the spatiotemporal distribution law of wind speed peaks through a sliding time window algorithm; designing a multidimensional correlation model, coupling and analyzing the instantaneous wind speed extreme value, the continuous wind speed mean and the residual effect of the previous disaster, and starting a gradient level correction program when continuous wind disaster events are detected, wherein: a linear superposition compensation algorithm is used to calculate the cumulative damage value for short-period continuous wind disasters; an exponential damage accumulation model is used for long-period continuous wind disasters; a wind speed time series fluctuation feature encoder is developed, and the morphological features of the wind speed curve are extracted through a convolutional neural network, and similarity matching is performed with a standard disaster pattern library; the assessment module integrates a self-feedback mechanism, and dynamically optimizes the weight coefficient allocation strategy based on real-time disaster verification data.

3. The flue-cured tobacco damage assessment method based on conventional meteorological element calculation according to claim 1, characterized in that: The dynamic moisture threshold matching mechanism in the drought disaster assessment specifically includes: constructing a dynamic adaptation model for crop water demand, automatically adjusting the water deficit index threshold range according to the physiological characteristic parameters of the growth period; designing an environmental parameter coupling analysis module, establishing a multidimensional water migration equation through real-time monitoring of soil permeability, canopy transpiration efficiency and root water absorption capacity parameters; developing a drought cumulative effect assessment algorithm, by constructing a wilting feature image recognition model, and extracting leaf curling and chlorophyll distribution characteristics through a convolutional neural network; designing a water stress propagation model, simulating the conduction path and rate of water deficit in the plant body; establishing a drought stage transition judgment rule, and triggering a disaster level forced upgrade mechanism when the synergistic effect of permanent wilting characteristics and root damage signals is detected.

4. The flue-cured tobacco damage assessment method based on conventional meteorological element calculation according to claim 1, characterized in that: The waterlogging disaster assessment involves: building a coordinated response mechanism for the soil-plant system, establishing a precipitation overload damage conduction model by dynamically monitoring changes in oxygen partial pressure gradients in the root zone and stem morphological parameters; designing a multimodal data fusion channel, integrating real-time soil permeability detection data with visual recognition results of canopy physiological characteristics, and verifying the spatiotemporal distribution characteristics of precipitation-evapotranspiration imbalance; Develop a composite damage assessment module, including: establishing a dynamic evolution model of root hypoxia damage, quantifying the degradation gradient of root hair cell membrane integrity caused by continuous precipitation; implementing dynamic monitoring of stem mechanical properties, capturing the deformation characteristics of vascular bundle structure through high-frequency microscopic imaging; constructing a soil microbial community evolution tracking mechanism, and analyzing the replacement patterns of dominant pathogen populations induced by anaerobic environments.

5. The flue-cured tobacco damage assessment method based on conventional meteorological element calculation according to claim 1, characterized in that: The multi-hazard comprehensive assessment constructs a heterogeneous computing task allocation engine to dynamically allocate time series analysis, spatial interpolation, and image processing tasks to optimized computing units based on the characteristics of the disaster type. Develop a system for analyzing the coupled effects of complex disasters, including: establishing a multi-stress interaction dynamics model to analyze the synergistic amplification effects of mechanical damage and physiological stress; A root holding force-wind load coupling analysis module is designed to simulate the combined destruction process of soil structure degradation and aerodynamic load; a damage propagation network topology model is constructed to predict the disaster diffusion path through the transmission relationship of microenvironmental parameters between plants.

Images