A method and device for evaluating the influence of low temperature, continuous rain and cloudy weather on rice

Abstract

The present application relates to the technical field of agricultural meteorological disaster assessment, in particular to a rice low-temperature and continuous-rain compound disaster influence assessment method and device, which can extract low-temperature stress factor sequence and continuous-rain stress factor sequence from daily meteorological sequence data, fit a probability distribution model to obtain a corresponding cumulative probability distribution function after statistical testing, substitute into multiple Copula models to obtain a joint probability sequence, convert into a standard compound disaster index sequence of each event, obtain independent compound disaster events through the sequence, map the cumulative intensity to the rice growth period, and calculate the partial correlation coefficient between the cumulative intensity and the historical yield data of the rice in each growth period by using a multivariate partial correlation analysis model after filtering the covariates. The technical scheme accurately quantifies the joint influence of low temperature and continuous rain, carefully depicts the disaster process, and removes interference factors, so that the assessment result is accurate.

Description

Technical Field

[0001] This invention relates to the field of agricultural meteorological disaster assessment technology, specifically to a method and equipment for assessing the impact of combined disasters of low temperature and continuous rain on rice. Background Technology

[0002] Against the backdrop of global warming, extreme weather events are showing a trend of increasing frequency and becoming more complex. Major rice-producing areas in my country are prone to abnormal weather processes such as prolonged low temperatures and continuous rain due to regional climate conditions. The combined effects of these two factors in time and space, resulting in low-temperature and continuous rain disasters, are one of the most significant meteorological hazards threatening the stability of rice production and food security in my country.

[0003] Currently, assessments of the impact of meteorological disasters on rice cultivation often rely on single meteorological data such as temperature, precipitation, or sunshine duration to construct evaluation indicators. When assessing the impact of disasters on yield, a macro-statistical correlation analysis between annual disaster indicators and the final rice yield is typically employed. However, existing technologies still have the following limitations in assessing combined disasters of low temperatures and continuous overcast / rainy weather affecting rice:

[0004] First, the indicators used are too simplistic, making it difficult to fully quantify the combined effects of multiple weather stresses. Existing technologies rely heavily on single-factor meteorological indicators for assessment, which at the algorithmic level struggles to effectively characterize the nonlinear correlation and combined destructive impacts of low temperatures and prolonged periods of rain on rice. Second, the assessment timescale is too coarse, lacking the ability to finely characterize the disaster process. Existing assessment methods based on annual "disaster-yield" do not fully consider the varying sensitivity of rice at different growth stages to meteorological stresses, making it difficult to accurately distinguish the differentiated destructive consequences caused by multiple disasters occurring at different growth stages. Third, there is a lack of effective mechanisms to control interference from non-target factors. When establishing the statistical mapping relationship between multiple meteorological disasters and yield, existing methods struggle to effectively isolate confounding interference from other non-target meteorological factors such as drought and extreme high temperatures at the model level, leading to biased assessments of the contribution of target disasters.

[0005] In summary, existing technologies for assessing the impact of combined low-temperature and continuous rainy weather disasters on rice still have technical problems, such as difficulty in accurately quantifying the combined disaster-causing effects, insufficient characterization of the growth period response, and susceptibility of assessment results to interference from non-target meteorological factors. Summary of the Invention

[0006] In view of this, the purpose of this invention is to provide a method and equipment for assessing the combined effects of low temperature and continuous rain on rice, so as to solve the problems of existing technologies that cannot quantify the combined disaster-causing effects, have a rough description of the disaster process, and whose assessment results are easily affected by other meteorological factors.

[0007] According to a first aspect of the present invention, a method for assessing the impact of combined disasters of low temperature and continuous rain on rice is provided, comprising: Acquire daily meteorological sequence data and historical rice yield data for the target area, and set monitoring time windows corresponding to multiple growth stages of rice on the time axis based on the rice agricultural phenological calendar; The continuous rain stress events were identified from the daily meteorological sequence data and ordered by time. The low temperature stress characteristics corresponding to each continuous rain stress event were calculated based on the daily meteorological sequence data. The continuous rain stress factor sequence and the low temperature stress factor sequence were determined and statistical tests were performed. The probability distribution models of the low-temperature stress factor sequence and the continuous rain stress factor sequence after passing the test were matched and fitted to obtain the low-temperature cumulative probability distribution function and the continuous rain cumulative probability distribution function. The cumulative probability distribution function of low temperature and the cumulative probability distribution function of continuous rain are substituted into multiple Copula models in the Copula function family, and the optimal Copula model is selected according to the information criterion. Using the optimal Copula model, the joint probability sequence of low temperature-continuous rain superposition in the target area under each continuous rain stress event is calculated. Transform the joint probability sequence into a standard composite disaster index sequence that includes positive and negative intervals; Based on the preset disaster threshold, the standard composite disaster index sequence is truncated to obtain independent composite disaster events; the duration and cumulative intensity of each independent composite disaster event are determined. Based on the duration of independent compound disaster events, the cumulative intensity is mapped and accumulated to the corresponding monitoring time window; For each continuous rain stress event, the covariates of the event and the corresponding rice growth period are calculated based on daily meteorological sequence data. A multivariate partial correlation analysis model was constructed after filtering out the covariate weights. The cumulative intensity accumulated in each monitoring time window and the historical rice yield data were used as inputs to the multivariate partial correlation analysis model to obtain the partial correlation coefficient.

[0008] Preferably, determining the sequence of continuous rain stress factors and the sequence of low temperature stress factors includes: Based on the daily meteorological sequence data of the target area, the daily average heat index for each date in previous years is calculated as the benchmark value of the heat index for the same period of the year for that date; Daily precipitation and sunshine duration are extracted from the daily meteorological sequence data, and a time series is constructed according to the date sequence. Based on the time series, continuous rain stress events are identified, the intensity index of continuous rain stress events is calculated, and the corresponding continuous rain stress factor sequence is formed. For each continuous rain stress event, the actual daily average temperature of each date within the event time window is extracted from the daily meteorological sequence data, and coupled calculations are performed based on the three cardinal temperature parameters for rice growth and development to construct a temperature response function as the actual heat index for that date. For each date within the event time window of each continuous rain stress event, calculate the difference between the actual heat index of the same date and the baseline value of the heat index for the same period in previous years to obtain the heat index anomaly value for each date; Based on the heat index anomaly value of each date within the event time window of each continuous rainy stress event, the event heat index anomaly value corresponding to the continuous rainy stress event is calculated as a low temperature stress feature; the low temperature stress features are sorted according to the time series to obtain the low temperature stress factor sequence.

[0009] Preferably, the intensity index of the continuous rain stress event is calculated, including: Calculate the average daily precipitation within the event time window of each continuous rain stress event; calculate the ratio of the average daily precipitation to the average annual precipitation of the target area during the same period, and combine it with the set precipitation weighting coefficient to obtain the precipitation stress component; Subtract a preset low-sunlight threshold from the daily sunshine hours within each continuous rain stress event to obtain a sunshine difference set; sum the absolute values ​​of all negative differences in the sunshine difference set to obtain the total event deficit; calculate the product of the preset low-sunlight threshold and the total number of days of the continuous rain stress event; calculate the ratio of the total event deficit to the product, and combine it with the set low-sunlight weighting coefficient to obtain the low-sunlight stress component; The total number of days of the continuous rain stress event is extracted, and a nonlinear power function is constructed as the duration cumulative effect factor. The precipitation stress component and the low sunshine stress component are added together and multiplied by the duration cumulative effect factor to obtain the continuous rain intensity index.

[0010] Preferably, the probability distribution models are matched and fitted to the low-temperature stress factor sequence and the continuous rain stress factor sequence after they have passed the test, including: To verify the fitting extreme value distribution model of the low-temperature stress factor sequence after the test, the low-temperature cumulative probability distribution function was obtained; To verify the fitting of the Gamma distribution model to the stress factor sequence after continuous rain, the cumulative probability distribution function of continuous rain was obtained.

[0011] Preferably, the optimal Copula model is selected based on information criteria, including: Calculate the information criterion value for each Copula model, and select the Copula model with the smallest information criterion value as the optimal Copula model.

[0012] Preferably, before constructing the multivariate partial correlation analysis model after filtering out the covariate weights, the method further includes: The historical rice yield data was detrended.

[0013] Preferably, the monitoring time window corresponds to multiple growth stages of rice, including at least the vegetative growth stage and the reproductive growth stage of rice.

[0014] Preferably, the method further includes: The reproductive period corresponding to the monitoring time window where the extreme value of the partial correlation coefficient is located is marked as the disaster prevention sensitive period.

[0015] Preferably, the joint probability sequence is transformed into a standard composite hazard index sequence containing positive and negative intervals, including: Using the inverse function algorithm of the standard normal distribution, the probabilities between 0 and 1 in the joint probability sequence are transformed into a standard composite disaster index sequence containing positive and negative intervals.

[0016] According to a second aspect of the present invention, a device for assessing the impact of combined disasters of low temperature and continuous rain on rice is provided, comprising: The main controller and the memory connected to the main controller; The memory stores program instructions; The main controller is used to execute program instructions stored in the memory and perform any of the methods described above.

[0017] The technical solution provided by this invention may include the following beneficial effects: It is understood that the technical solution presented in this invention can extract the low-temperature stress factor sequence and the continuous rain stress factor sequence from daily meteorological sequence data, obtain the corresponding cumulative probability distribution function by fitting a probability distribution model after statistical testing, substitute it into multiple Copula models to obtain a joint probability sequence, and then convert it into a standard composite disaster index sequence for each continuous rain stress event. Independent composite disaster events are obtained through this sequence, and the cumulative intensity is mapped to the rice growth period. Then, the partial correlation coefficient between the cumulative intensity and the historical rice yield data in each growth period is calculated and output using a multivariate partial correlation analysis model after filtering out covariates.

[0018] This technical solution breaks through the limitations of single-dimensional analysis and accurately quantifies the combined synergistic stress effect. By extracting the sequences of continuous rain and low temperature, and after statistical testing, introducing a Copula function family to fit a joint probability distribution model, it overcomes the limitations of traditional single-factor assessment. At the algorithmic level, it accurately captures and quantifies the nonlinear superimposed disaster-causing effect on rice when low temperature and continuous rain occur simultaneously.

[0019] By coupling agricultural phenology, a refined characterization of disaster processes can be achieved. Innovatively, monitoring time windows are set based on the rice agricultural phenological calendar, and the cumulative intensity of extracted independent compound disaster events is mapped to the corresponding growth stages. This effectively distinguishes the differences in the destructive power of compound disasters at different growth stages of rice, solving the problem of spatiotemporal scale misalignment in traditional methods.

[0020] A filtering mechanism was constructed to effectively eliminate interference from irrelevant meteorological factors. By calculating covariates of the combined disaster event and the corresponding rice growth stage, a multivariate partial correlation analysis model was built after filtering out the weights of the covariates to calculate the partial correlation coefficient. This automatically eliminated the confounding interference from other non-target meteorological factors at the mathematical model level, significantly improving the objectivity and accuracy of the assessment of the disaster's impact on yield.

[0021] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0022] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0023] Figure 1 This is a schematic diagram illustrating the steps of a method for assessing the combined impact of low temperature and continuous rain on rice, according to an exemplary embodiment. Detailed Implementation

[0024] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the invention as detailed in the appended claims.

[0025] In one embodiment, Figure 1 This is a schematic diagram illustrating the steps of a method for assessing the combined impact of low-temperature and continuous rainy weather on rice, according to an exemplary embodiment. See also... Figure 1 According to a first aspect of the present invention, a method for assessing the impact of combined disasters of low temperature and continuous rain on rice is provided, comprising: Step S11: Obtain daily meteorological sequence data and historical rice yield data for the target area, and set monitoring time windows corresponding to multiple growth stages of rice on the time axis according to the rice agricultural phenological calendar.

[0026] In practice, for data acquisition, computer systems can obtain daily meteorological sequence data and historical rice yield data for the target area through data interfaces. Daily meteorological sequence data includes measured physical indicators that directly reflect the characteristics of low temperatures and continuous rainy weather, such as daily average temperature, daily precipitation, and sunshine hours.

[0027] In terms of setting monitoring time windows, the system will divide the time axis into monitoring time windows corresponding to multiple growth stages of rice, based on the agricultural phenological calendar adapted to the rice planting characteristics of the target area. For example, the growth stage of rice covers at least the vegetative growth stage and the reproductive growth stage (in practice, it can be further subdivided into multiple stages such as the vegetative growth stage, the stage of simultaneous vegetative and reproductive growth, the reproductive growth stage, and the grain filling and ripening stage).

[0028] This setting solves the technical defect of misaligned spatiotemporal scales in traditional disaster assessment, allowing subsequent processing to accurately correspond to different growth and development stages of rice.

[0029] Step S12: Identify the continuous rain stress events sorted by time from the daily meteorological sequence data, calculate the low temperature stress characteristics corresponding to each continuous rain stress event based on the daily meteorological sequence data, determine the continuous rain stress factor sequence and the low temperature stress factor sequence respectively, and perform statistical tests.

[0030] In the dual-stress factor sequence extraction stage, the system automatically traverses daily meteorological sequence data to perform targeted extraction. First, it extracts multiple consecutive rainy weather stress events from the daily meteorological sequence data and sorts them by time to form a consecutive rainy weather stress factor sequence. Each consecutive rainy weather stress event consists of daily precipitation and sunshine duration indicators. For each consecutive rainy weather stress event, its event time window is determined. Then, based on this event time window, the corresponding low-temperature stress features are extracted from the daily meteorological sequence data and sorted by time to form a low-temperature stress factor sequence.

[0031] Both the continuous rain stress factor sequence and the low temperature stress factor sequence are event sequences, which are event sequences ordered by time. Each continuous rain stress event corresponds to a low temperature stress feature.

[0032] During the statistical testing phase, the system will call the statistical calculation engine to perform the Kolmogorov-Smirnov (KS) test on the extracted low-temperature stress factor sequence and continuous rain stress factor sequence. The core purpose of this test is to verify whether the sample data distribution characteristics of the two factor sequences meet the prerequisites for subsequent probability distribution model fitting, and at the same time remove outlier data in the sequence. Only the sequence data that passes this statistical test can enter the subsequent probability distribution model matching and fitting stage.

[0033] Step S13: Match and fit the probability distribution models of the low temperature stress factor sequence and the continuous rain stress factor sequence after they have passed the test, and obtain the low temperature cumulative probability distribution function and the continuous rain cumulative probability distribution function.

[0034] Preferably, to obtain the low-temperature cumulative probability distribution function, the extreme value distribution model is fitted to the low-temperature stress factor sequence after verification; and to obtain the continuous rainy season stress factor sequence after verification, the Gamma distribution model is fitted to obtain the continuous rainy season cumulative probability distribution function.

[0035] This step involves verifying the matching and fitting of the extreme value distribution model of the low-temperature stress factor sequence after it has passed the test. The model parameters are solved and fitted using statistical algorithms, and finally the low-temperature cumulative probability distribution function is generated. This function can accurately characterize the cumulative probability of occurrence corresponding to different degrees of low-temperature stress, realizing the transformation of low-temperature stress factors from quantitative values ​​to probability values.

[0036] For the stress factor sequence of continuous rain that passed the test, a Gamma distribution model was matched and fitted. After statistical fitting calculation, the cumulative probability distribution function of continuous rain was obtained. This function can effectively characterize the cumulative probability of occurrence of different degrees of continuous rain stress.

[0037] The two cumulative probability distribution functions obtained in this step both have values ​​between 0 and 1, while retaining the date and timestamp that is completely consistent with the cause subsequence. This step transforms the original quantified stress factor sequence into a probabilistic marginal distribution function, realizing a probabilistic representation of the stress level of a single factor.

[0038] Step S14: Substitute the low temperature cumulative probability distribution function and the continuous rain cumulative probability distribution function into multiple Copula models in the Copula function family, and select the optimal Copula model based on the information criterion; using the optimal Copula model, calculate the joint probability sequence of low temperature-continuous rain superposition in the target area under each continuous rain stress event.

[0039] This step mainly consists of three consecutive operations: model substitution, optimal model selection, and joint probability sequence calculation. The system first matches the cumulative probability distribution function of low temperature and the cumulative probability distribution function of continuous overcast and rainy weather according to daily timestamps, and then substitutes them into multiple classic Copula models (including Clayton, Gumbel, and Frank models) from the Archimedesian Copula family of functions. Each model will construct a joint probability relationship between low temperature and continuous overcast and rainy weather based on the cumulative probabilities of the two single factors, thus characterizing the different correlation features between the two.

[0040] Subsequently, the system calculates the information criterion value for each model based on the AIC / BIC minimization information criterion, and selects the optimal Copula model with the smallest criterion value through numerical comparison. This selection method replaces manual selection with data-driven selection, effectively ensuring the model's fit to the meteorological data of the target area and accurately capturing nonlinear characteristics such as the tail correlation of low temperature and continuous rain.

[0041] Finally, the system uses the selected optimal Copula model to automatically calculate the daily probability sequence of low temperature and continuous rain in the target area. Each value in the sequence corresponds to the probability of low temperature and continuous rain occurring simultaneously on a single day, with values ​​ranging from 0 to 1. This not only preserves the complete time dimension characteristics but also realizes the probability quantification of the combined synergistic stress effect of the two, laying a key mathematical foundation for the subsequent transformation of the joint probability into an intuitive and analyzable composite disaster index.

[0042] Step S15: Transform the joint probability sequence into a standard composite disaster index sequence containing positive and negative intervals.

[0043] This step utilizes the inverse function algorithm of the standard normal distribution to transform the probabilities between 0 and 1 in the joint probability sequence into a standard composite hazard index sequence (LTCRI) containing positive and negative intervals.

[0044] The converted index values ​​have clear physical meanings. Positive numbers indicate that there is a combined stress of low temperature and continuous rain on that day, and the larger the value, the stronger the intensity of the combined disaster. Negative numbers indicate that there is no combined disaster stress on that day, and the meteorological conditions have not created a combined stress of low temperature and continuous rain on rice. This division of positive and negative intervals makes it intuitive and easy to judge whether a combined disaster occurs and its intensity.

[0045] Meanwhile, the converted standard composite disaster index sequence retains the daily timestamps consistent with the original joint probability sequence, and precisely corresponds to the multiple monitoring time windows of rice growth stages set in the early stage. This lays a standardized and homogenized index data foundation for subsequent extraction of the duration and cumulative intensity of independent composite disaster events by setting disaster thresholds and using run theory to truncate the sequence.

[0046] Step S16: Based on the preset disaster threshold, truncate the standard composite disaster index sequence to obtain independent composite disaster events; determine the duration and cumulative intensity of each independent composite disaster event.

[0047] The system first sets up a scientific disaster judgment threshold based on the growth characteristics of rice in the target area and the agricultural meteorological disaster-causing pattern of low temperature and continuous rain. This threshold serves as a quantitative standard to distinguish whether a compound disaster has occurred and is the basis for determining the disaster period and non-disaster period in the index sequence.

[0048] The system then employs the run-length theory algorithm, using the preset threshold as the criterion, to automatically truncate the daily standard composite disaster index sequence. The index segments that continuously exceed the disaster threshold are defined as independent low-temperature continuous rainy composite disaster events, thereby accurately separating all independent and complete composite disaster events from the continuous index sequence.

[0049] For each independent composite disaster event obtained through truncation, the system will further extract two feature values ​​that can fully characterize the degree of disaster based on its corresponding daily-scale standard composite disaster index sequence: one is the duration, that is, the number of consecutive days from the occurrence to the end of the independent disaster event, which quantifies the scope of the disaster's impact in the time dimension; the other is the cumulative intensity, that is, the cumulative value of the daily standard composite disaster index sequence during the duration of the event, which quantifies the degree of disaster's degree of disaster in the intensity dimension.

[0050] These two feature values ​​visualize and quantify the catalytic effect of each independent compound disaster from both time and intensity dimensions. Furthermore, all feature values ​​retain the corresponding date and timestamp, which can accurately match the monitoring time windows set for multiple growth stages of rice in the early stage. This lays the core disaster feature data foundation for subsequently mapping and accumulating the cumulative intensity of the disaster to the corresponding growth stage according to the time node and carrying out a refined disaster impact assessment by growth stage.

[0051] Step S17: Based on the duration of the independent compound disaster events, map and accumulate the cumulative intensity to the corresponding monitoring time window.

[0052] The system first matches the occurrence time of each independent compound disaster event with the monitoring time windows set in the early stages for the vegetative growth and reproductive growth stages of rice based on the complete date interval corresponding to the duration of each event. This clarifies the rice growth stage to which a single disaster event belongs. If a disaster event spans multiple growth stages, the cumulative intensity is broken down according to the actual number of days in different growth stages, and then corresponding monitoring time windows are matched for each stage.

[0053] The system then automatically calculates the cumulative intensity of all independent compound disaster events matched to the same monitoring time window (i.e., the same rice growth period), and obtains the total cumulative intensity of low temperature and continuous rainy compound disasters corresponding to each growth period monitoring time window. This value comprehensively represents the overall disaster intensity after the superposition of all compound disaster events within the corresponding growth period.

[0054] This step achieves the data transformation from "event-scale disaster occurrence" to "accumulated intensity during critical growth periods" through precise matching of timestamps and targeted accumulation of intensity. This fundamentally solves the technical defects of spatiotemporal scale misalignment in traditional disaster assessment. Furthermore, the accumulated intensity of each growth period corresponds one-to-one with the monitoring time window, allowing subsequent disaster damage assessment to accurately target the disaster impact characteristics of different growth stages of rice.

[0055] Step S18: For each continuous rain stress event, calculate the covariate of the event and the corresponding rice growth period based on the daily meteorological sequence data.

[0056] The system first extracts relevant indicators such as precipitation, daily average / maximum temperature, sunshine duration, relative humidity, and wind speed from daily meteorological data. Then, for each continuous rainy season stress event identified in step S12, it integrates and calculates various indicators within the corresponding reproductive period covered by the event, ultimately generating concurrent covariates that match the event and reproductive period. Preferably, the concurrent covariates include the Standardized Precipitation Evapotranspiration Index (SPEI) and the number of extreme high-temperature days (EDD).

[0057] Step S19: Construct a multivariate partial correlation analysis model after filtering out the weights of the covariates. Use the cumulative intensity accumulated in each monitoring time window and the historical rice yield data as inputs to the multivariate partial correlation analysis model to obtain the partial correlation coefficient.

[0058] Preferably, before constructing the multivariate partial correlation analysis model after filtering out the covariate weights, the method further includes: detrending the historical rice yield data.

[0059] Historical rice yield data were detrended to eliminate the influence of non-meteorological factors such as variety improvement and cultivation technology upgrades, ensuring that the yield data used in the analysis is the relative meteorological yield affected only by meteorological factors.

[0060] Meanwhile, the model will align various data according to the monitoring time windows of the rice growth period and clearly define three types of variables: the detrended relative meteorological yield of rice corresponding to each monitoring time window as the dependent variable, the cumulative intensity of the low temperature and continuous rainy weather combined disaster accumulated in each monitoring time window as the core independent variable, and the standardized precipitation evapotranspiration index (SPEI) and extreme high temperature days (EDD) generated in step S18 as covariates as control variables.

[0061] After constructing a multivariate partial correlation analysis model, the system substitutes the above variables into the model and uses statistical algorithms to eliminate the weight interference of non-target meteorological disasters such as drought and extreme high temperature on rice yield while keeping the values ​​of all control variables unchanged. This completely eliminates the confusing effects of such irrelevant meteorological factors, allowing the model to focus only on the correlation analysis between the core independent variable and the dependent variable, and avoids misattributing the yield loss of other meteorological disasters to the combined effects of low temperature and continuous rain.

[0062] After model construction and interference removal are completed, the system will automatically calculate the data according to four monitoring time windows: the vegetative growth period, the period of simultaneous vegetative and reproductive growth, the reproductive growth period, and the grain filling and ripening period. The system will output the partial correlation coefficient between the cumulative intensity of the combined disaster and the historical yield data of rice in each monitoring time window. The value of this coefficient is between -1 and 1. The sign and absolute value of the coefficient can characterize the independent impact of the combined disaster of low temperature and continuous rain on rice yield during the corresponding growth period.

[0063] Finally, the reproductive period corresponding to the monitoring time window where the extreme values ​​of the partial correlation coefficient occur is marked as the disaster prevention sensitive period. Users can take corresponding protective measures in advance based on the partial correlation coefficient or the disaster prevention sensitive period.

[0064] In a preferred embodiment, step S12, extracting the low-temperature stress factor sequence and the continuous rain stress factor sequence from the daily meteorological sequence data, includes: Step S21: Based on the historical daily meteorological sequence data of the target area, calculate the daily average heat index for each date in previous years as the benchmark value of the heat index for the same period of the year for that date; Step S22: Extract daily precipitation and sunshine hours from the daily meteorological sequence data, and construct a time series according to the date sequence; identify continuous rain stress events based on the time series, calculate the intensity index of continuous rain stress events, and form the corresponding continuous rain stress factor sequence. Step S23: For each continuous rain stress event, extract the actual daily average temperature of each date within the event time window from the daily meteorological sequence data, and perform coupled calculations based on the three cardinal temperature parameters (lower limit temperature, suitable temperature, and upper limit temperature) for rice growth and development to construct a temperature response function as the actual heat index for that date; the temperature response function is used to characterize the nonlinear effect of temperature on the growth and development rate of rice. Step S24: For each date within the event time window of each continuous rain stress event, calculate the difference between the actual heat index of the same date and the baseline value of the heat index for the same period in previous years, and obtain the heat index anomaly value for each date. Step S25: Calculate the event heat index anomaly value corresponding to the continuous rain stress event based on the heat index anomaly value of each date within the event time window of each continuous rain stress event, and use it as a low temperature stress feature; sort the low temperature stress features according to the time series to obtain the low temperature stress factor sequence.

[0065] Identifying continuous rain stress events based on the time series includes: The event is triggered when the precipitation at a single station is greater than 0 mm for 3 consecutive days or more (with 1 day of no precipitation allowed during the period), and the sunshine duration on each corresponding day is less than 3 hours. In this case, the continuous rain stress event is determined to have occurred. The event termination condition is as follows: when there are two consecutive days with 0 mm of precipitation, the precipitation process is determined to be over, and the event time window of the continuous rain stress event is locked accordingly, and its duration is calculated.

[0066] This allows us to obtain each continuous rain stress event. By recording each continuous rain stress event that meets the above conditions in chronological order, we can form a continuous rain stress factor sequence.

[0067] Calculate the intensity index of prolonged rainy weather stress events, including: Step S31: Calculate the average daily precipitation within the event time window of each continuous rain stress event, and compare it with the average annual precipitation for the same period in the target area. Combine this ratio with a set precipitation weighting coefficient to obtain the precipitation stress component. .

[0068] The formula is as follows:

[0069] in, The actual precipitation (mm) on day i within the event time window for each continuous rain stress event; The average annual precipitation (mm) for the target area during the same period (same time period); For a preset minimum constant (e.g., 10) -5 ), used to prevent the denominator from being 0; This is the set precipitation weighting coefficient.

[0070] In this step, to prevent the annual precipitation from being extremely low or even zero (which would cause anomalies in the calculation), a very small constant can be added to the annual average precipitation as a smoothing process.

[0071] The purpose of comparing the average value with the average annual precipitation for the same period is to perform dimensionless processing.

[0072] Step S32: Subtract a preset low-sunlight threshold from the daily sunshine hours within each continuous rainy weather stress event to obtain a sunshine difference set; sum the absolute values ​​of all negative differences in the sunshine difference set to obtain the total event deficit; calculate the product of the preset low-sunlight threshold and the total number of days of the continuous rainy weather stress event; calculate the ratio of the total event deficit to the product (for dimensionless processing), and combine it with the set low-sunlight weighting coefficient to obtain the low-sunlight stress component. I S .

[0073] In this step, if the daily sunshine duration is greater than the preset low sunshine threshold, then there is sufficient sunshine and no low sunshine stress; if the daily sunshine duration is less than the threshold, then there is insufficient sunshine and low sunshine stress occurs. Therefore, only the negative difference values ​​of the sunshine difference set are extracted for subsequent calculations. The absolute value of each negative difference value is the daily deficit, and the sum of these values ​​is the total deficit for each consecutive rainy weather stress event.

[0074] The calculation formula for this step is as follows:

[0075] in, Mathematically, it is completely equivalent to the number of hours of sunshine per day. Subtract threshold Filter out negative differences and take their absolute values. When > When, the result is 0 (no coercion); when < At that time, the result was (i.e., the absolute value of the negative difference). The actual sunshine duration (h) on the i-th day within each continuous rainy weather stress event; The preset threshold for low sunshine duration (h) is given. D is the total number of days (d) of this continuous rainy weather stress event. This is the weighting coefficient for the few photos. , The weighting coefficients are mainly obtained through multiple linear regression analysis of historical production reduction data and meteorological factors.

[0076] Step S33: Extract the total number of days of the continuous rain stress event, construct a nonlinear power function as the duration cumulative effect factor; add the precipitation stress component and the low sunshine stress component, and multiply by the duration cumulative effect factor to obtain the continuous rain intensity index.

[0077] The formula is as follows:

[0078] in, is the cumulative effect factor over time (a nonlinear power function). k is the sensitivity index parameter (an empirical constant, usually k>1, used to amplify the cumulative penalty effect of long-term disasters). This is the final intensity index of continuous rain.

[0079] The intensity index of continuous rain obtained in this step reflects the combined effect of precipitation, low sunshine, and duration.

[0080] This technical solution breaks through the limitations of traditional single-factor exponential linear superposition and innovatively introduces the Copula joint probability model to process underlying meteorological data. Through algorithmic logic, it accurately captures the tail correlation generated by the superposition of low temperatures and continuous rain, effectively quantifying the synergistic destructive power of the two. This invention establishes a three-level processing architecture: "daily-scale data modeling → run-length theory event extraction → accumulation of growth period monitoring time windows." This solves the problem of misaligned spatiotemporal scales in macro-interannual data in traditional assessments, and can theoretically pinpoint the sensitive growth period of rice to compound disasters, clarifying the key time window for crop disaster prevention. Unlike traditional direct statistical regression analysis, this invention uses pre-control variable extraction (such as SPEI and EDD) and partial correlation purification logic in the disaster impact assessment module. This gives the system a strong anti-interference capability against non-target meteorological disasters when processing complex actual climate data, effectively ensuring the objectivity and accuracy of disaster damage assessment results.

[0081] According to a second aspect of the present invention, a device for assessing the impact of combined disasters of low temperature and continuous rain on rice is provided, comprising: The main controller and the memory connected to the main controller; The memory stores program instructions; The main controller is used to execute program instructions stored in the memory and perform any of the methods described above.

[0082] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0083] It should be noted that in the description of this invention, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this invention, unless otherwise stated, "a plurality of" means at least two.

[0084] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0085] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0086] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0087] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0088] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0089] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0090] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for assessing the impact of combined disasters of low temperature and continuous rain on rice cultivation, characterized in that, include: Acquire daily meteorological sequence data and historical rice yield data for the target area, and set monitoring time windows corresponding to multiple growth stages of rice on the time axis based on the rice agricultural phenological calendar; The continuous rain stress events were identified from the daily meteorological sequence data and ordered by time. The low temperature stress characteristics corresponding to each continuous rain stress event were calculated based on the daily meteorological sequence data. The continuous rain stress factor sequence and the low temperature stress factor sequence were determined and statistical tests were performed. The probability distribution models of the low-temperature stress factor sequence and the continuous rain stress factor sequence after passing the test were matched and fitted to obtain the low-temperature cumulative probability distribution function and the continuous rain cumulative probability distribution function. The cumulative probability distribution function of low temperature and the cumulative probability distribution function of continuous rain are substituted into multiple Copula models in the Copula function family, and the optimal Copula model is selected according to the information criterion. Using the optimal Copula model, the joint probability sequence of low temperature-continuous rain superposition in the target area under each continuous rain stress event is calculated. Transform the joint probability sequence into a standard composite disaster index sequence that includes positive and negative intervals; Based on the preset disaster threshold, the standard composite disaster index sequence is truncated to obtain independent composite disaster events; the duration and cumulative intensity of each independent composite disaster event are determined. Based on the duration of independent compound disaster events, the cumulative intensity is mapped and accumulated to the corresponding monitoring time window; For each continuous rain stress event, the covariates of the event and the corresponding rice growth period are calculated based on daily meteorological sequence data. A multivariate partial correlation analysis model was constructed after filtering out the covariate weights. The cumulative intensity accumulated in each monitoring time window and the historical rice yield data were used as inputs to the multivariate partial correlation analysis model to obtain the partial correlation coefficient.

2. The method according to claim 1, characterized in that, The sequences of factors causing continuous rain and low temperature stress were determined, including: Based on the daily meteorological sequence data of the target area, the daily average heat index for each date in previous years is calculated as the benchmark value of the heat index for the same period of the year for that date; Daily precipitation and sunshine duration are extracted from the daily meteorological sequence data, and a time series is constructed according to the date sequence. Based on the time series, continuous rain stress events are identified, the intensity index of continuous rain stress events is calculated, and the corresponding continuous rain stress factor sequence is formed. For each continuous rain stress event, the actual daily average temperature of each date within the event time window is extracted from the daily meteorological sequence data, and coupled calculations are performed based on the three cardinal temperature parameters for rice growth and development to construct a temperature response function as the actual heat index for that date. For each date within the event time window of each continuous rain stress event, calculate the difference between the actual heat index of the same date and the baseline value of the heat index for the same period in previous years to obtain the heat index anomaly value for each date; Based on the heat index anomaly value of each date within the event time window of each continuous rainy stress event, the event heat index anomaly value corresponding to the continuous rainy stress event is calculated as a low temperature stress feature; the low temperature stress features are sorted according to the time series to obtain the low temperature stress factor sequence.

3. The method according to claim 2, characterized in that, Calculate the intensity index of prolonged rainy weather stress events, including: Calculate the average daily precipitation within the event time window of each continuous rain stress event; calculate the ratio of the average daily precipitation to the average annual precipitation of the target area during the same period, and combine it with the set precipitation weighting coefficient to obtain the precipitation stress component; Subtract a preset low-sunlight threshold from the daily sunshine hours during the continuous rain stress event to obtain a sunshine difference set; sum the absolute values ​​of all negative differences in the sunshine difference set to obtain the total event deficit; calculate the product of the preset low-sunlight threshold and the total number of days of the continuous rain stress event; calculate the ratio of the total event deficit to the product, and combine it with a set low-sunlight weighting coefficient to obtain the low-sunlight stress component; The total number of days of the continuous rain stress event is extracted, and a nonlinear power function is constructed as the duration cumulative effect factor. The precipitation stress component and the low sunshine stress component are added together and multiplied by the duration cumulative effect factor to obtain the continuous rain intensity index.

4. The method according to claim 1, characterized in that, The probability distribution models were matched and fitted to the low-temperature stress factor sequence and the continuous rain stress factor sequence after they passed the test, including: To verify the fitting extreme value distribution model of the low-temperature stress factor sequence after the test, the low-temperature cumulative probability distribution function was obtained; To verify the fitting of the Gamma distribution model to the stress factor sequence after continuous rain, the cumulative probability distribution function of continuous rain was obtained.

5. The method according to claim 1, characterized in that, The optimal Copula model is selected based on information criteria, including: Calculate the information criterion value for each Copula model, and select the Copula model with the smallest information criterion value as the optimal Copula model.

6. The method according to claim 1, characterized in that, Before constructing the multivariate partial correlation analysis model after filtering out the covariate weights, the following steps are also included: The historical rice yield data was detrended.

7. The method according to claim 1, characterized in that, The monitoring time window corresponds to multiple growth stages of rice, including at least the vegetative growth stage and the reproductive growth stage.

8. The method according to claim 1, characterized in that, Also includes: The reproductive period corresponding to the monitoring time window where the extreme value of the partial correlation coefficient is located is marked as the disaster prevention sensitive period.

9. The method according to claim 1, characterized in that, Transforming the joint probability sequence into a standard composite hazard index sequence that includes positive and negative intervals includes: Using the inverse function algorithm of the standard normal distribution, the probabilities between 0 and 1 in the joint probability sequence are transformed into a standard composite disaster index sequence containing positive and negative intervals.

10. An assessment device for the combined impact of low temperature and continuous rainy weather on rice, characterized in that, include: The main controller and the memory connected to the main controller; The memory stores program instructions; The main controller is used to execute program instructions stored in the memory to perform the method as described in any one of claims 1 to 9.

Images