Method for predicting corn yield in target area based on physical mechanism of combined high temperature and drought influence

Abstract

This invention discloses a method for predicting maize yield in a target area based on the physical mechanisms of combined high temperature and drought effects, belonging to the field of agricultural yield prediction technology. It includes acquiring maize yield per unit area (MAIZE) data for the target area in year Y-1. Y‑1 ; Obtain the forecast factor index for the early stage of forecast year Y; Input the obtained forecast factor index for the early stage of forecast year Y into the pre-constructed forecast model of forecast quantity DMAIZE to predict the DMAIZE forecast value for year Y; ; Maize the data Y‑1 Adding the predicted DMAIZE value for year Y, we obtain the predicted maize yield MAIZE for year Y. Y This invention uses the interannual increment method to predict corn yield in the target area, significantly improving the accuracy of prediction.

Description

Technical Field

[0001] This invention relates to a method for predicting maize yield in a target area based on the physical mechanisms of combined high temperature and drought effects, and belongs to the field of agricultural yield prediction technology. Background Technology

[0002] Maize is a major food crop, cash crop, feed source, and industrial raw material. However, it is vulnerable to climate change and faces serious threats from extreme events such as heat waves and droughts. Maize growth depends on suitable temperatures; heat during the growing season is crucial for avoiding chilling injury, increasing photosynthetic area and duration, and improving yield. However, high-temperature stress caused by global warming has already posed a serious threat to maize production. It is estimated that for every 1 degree Celsius increase in temperature, maize yields decrease by about 10%. In particular, maize during the flowering and grain-filling stages is most sensitive to high temperatures, and these two stages coincide with summer, a time when extreme combined heat and drought events are frequent. Drought also severely inhibits maize growth, potentially leading to yield reductions of 25-30%. Against the backdrop of global warming, the frequency and intensity of heat and drought events are expected to increase, exacerbating the threats to maize growth and regional food security. Therefore, it is crucial to construct a target region maize yield forecasting model based on the physical mechanisms of the effects of combined heat and drought.

[0003] Currently, the approved maize yield prediction models are mainly based on remote sensing of field growth conditions. For example, Chinese patent application CN116665073A discloses a remote sensing estimation method for maize yield based on multi-source data; and Chinese patent application CN117173524A discloses a method for predicting maize yield based on multimodal unmanned aerial vehicles (UAVs). Some models are based on meteorological factors during the maize growing season, such as Chinese patent application CN116777081A, which discloses a spring maize yield prediction model based on meteorological conditions at different growth stages. In addition, some models are based on artificial intelligence technology to predict maize yield. For example, Chinese patent application CN118521008A discloses a maize yield prediction method based on a deep learning attention mechanism; and Chinese patent application CN118397297A discloses a maize yield prediction method and system based on machine learning. However, current maize yield prediction models still have significant limitations in forecast timeliness; their prediction results are often limited to the maize growing season and cannot be significantly advanced. This shortcoming in forecast timeliness weakens the practical value of the models, making it difficult for agricultural departments to formulate comprehensive planting plans and preventative measures in a timely manner. Of particular note is that some forecasting models based on artificial intelligence, while showing potential, have failed to clearly reveal the physical mechanisms underlying their maize yield prediction models; therefore, their reliability and stability require further verification. Research indicates that high-temperature drought events have become a key factor restricting maize growth, and that early meteorological factors such as sea ice and sea temperature can influence the circulation patterns in China through physical processes. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for predicting maize yield in a target area based on the physical mechanism of the combined effects of high temperature and drought. The method constructs a prediction model based on the physical mechanism of the combined effects of high temperature and drought on the target area by previous meteorological factors, which can achieve the goal of predicting maize yield several months in advance, thereby effectively addressing the shortcomings of the prior art.

[0005] To achieve the above objectives / to solve the above technical problems, the present invention is implemented using the following technical solution: A method for predicting maize yield in a target area based on the physical mechanisms of combined high temperature and drought effects, the method comprising: Obtain maize yield per unit area data for the target region in year Y-1 (MAIZE) Y-1 ; Obtain the forecast factor index in the early stage of the forecast year Y; The forecast factor index of the early stage of the forecast year Y will be obtained and input into the pre-constructed forecast model of the forecast amount DMAIZE to predict the DMAIZE forecast value of the forecast year Y. MAIZE the data Y-1 Adding the predicted DMAIZE value for forecast year Y, we obtain the predicted maize yield MAIZE for forecast year Y. Y .

[0006] Optionally, the step of obtaining maize yield data per unit area for the target region in year Y-1 (MAIZE) Y-1 ,include: Collect annual maize yield data per unit area for the target region over N years; The annual yield data per unit area of ​​corn are averaged to form a time series of corn yield for the overall region. After standardizing the time series of maize yield for the entire predicted region, we obtain the maize yield per unit area (MAIZE) for the target region from year YN to year Y-1, where the maize yield per unit area for the target region in year Y-1 is the data MAIZE. Y-1 Y represents the forecast year, and N represents the data collection period in years.

[0007] Optionally, the method for constructing the forecast model for the forecast quantity DMAIZE includes: Collect annual maize yield data per unit area for the target region over N years; The annual yield data per unit area of ​​corn are averaged to form a time series of corn yield for the overall region. Global monthly gridded sea ice concentration data from YN to Y-1 were collected through the Hadley Centre, and global monthly gridded sea surface temperature data, 0-10cm soil moisture data and sea level pressure data from YN to Y-1 were collected through ERA5. After standardizing the time series of maize yield for the overall predicted region, we obtain the maize yield per unit area (MAIZE) data for the standardized target region from YN to Y-1. The data MAIZE was processed using the interannual increment method to calculate the interannual increment data DMAIZE of maize yield per unit area from year YN-1 to year Y-1 in the target area. Obtain forecast factor indices for the early stage of forecast year Y, including sea ice concentration index, sea surface temperature index, soil moisture index, and North Atlantic Oscillation index; The sea ice concentration index, sea surface temperature index, soil moisture index, and North Atlantic Oscillation index were used as forecasting factors, and the data DMAIZE was used as the forecast quantity. A forecasting model for the forecast quantity DMAIZE was constructed based on the stepwise linear regression method.

[0008] Optionally, the calculation formula for averaging the annual corn yield per unit area is as follows: ; In the formula: y is the year, M is the corn yield per unit area data of the target region from year YN to year Y-1, region1 is the corn yield per unit area data of region one from year YN to year Y-1, region2 is the corn yield per unit area data of region two from year YN to year Y-1, and regionX is the corn yield per unit area data of region X from year YN to year Y-1.

[0009] Optionally, the formula for calculating the maize yield per unit area data (MAIZE) of the standardized target area from year YN to year Y-1 after standardizing the time series of maize yield for the entire predicted region is as follows: ; In the formula: y is the year, MAIZE is the standardized maize yield per unit area data of the target area from year YN to year Y-1, Z is the standardization process, M is the maize yield per unit area data of the target area from year YN to year Y-1, μ(M) is the average value of M data for a total of N years from year YN to year Y-1, σ(M) is the standard deviation of M data for a total of N years from year YN to year Y-1, and N is the time length of data collection in years.

[0010] Optionally, the step of processing the data MAIZE using the interannual increment method to calculate the standardized maize yield per unit area (DMAIZE) data from year YN-1 to year Y-1 in the target area includes: ; In the formula: DMAIZE represents the annual increment of standardized maize yield per unit area in the target region from year YN-1 to year Y-1. The standardized maize yield per unit area represents the interannual increment data for the target region from year YN-1 to year Y-1, while MAIZE represents the standardized maize yield per unit area for the target region from year YN to year Y-1. This provides standardized maize yield per unit area data for the target region in year y from year YN to year Y-1. This refers to the standardized maize yield per unit area in the target region from year YN to year Y-1, where Y is the forecast year, y is the year, and the range is any year from year YN to year Y-1. N is the length of time for data collection, in years.

[0011] Optionally, obtaining the forecast factor index for the early stage of the forecast year Y includes: The collected sea ice concentration data, sea surface temperature data, soil moisture data, and sea level pressure data were preprocessed using the interannual increment method, and converted into interannual increment values ​​for sea ice concentration (DSIC), sea surface temperature (DSST), soil moisture (DSW), and sea level pressure (DSLP), respectively. The calculation formulas are as follows: ; In the formula: y is the year, m is the month, m=1,2,3,...,12, t is the t-th latitude, t=1,2,3,...,181, p is the p-th longitude, p=1,2,3,...,360; DXXX is the interannual increment data of meteorological factors from year YN-1 to year Y-1, DXXX=DSIC, DSST, DW, DSLP; XXX is the meteorological factor data from year YN to year Y-1, XXX=SIC, SST, SW,SLP; The sea ice concentration index includes: extracted from the monthly scale of the interannual sea ice concentration increment (DSIC) value. Monthly sea ice concentration data, denoted as SICM, is calculated using the following formula: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for sea ice concentration are from year YN-1 to year Y-1. Monthly sea ice concentration interannual increment data, DSIC from YN-1 to Y-1 Monthly data on the annual increase in sea ice concentration; exist Based on monthly sea ice concentration (SICM) data, the SICM data of latitude and longitude grid points simultaneously located within the critical sea ice concentration region are extracted and averaged to obtain the critical sea ice concentration region from year YN-1 to year Y-1. The formula for calculating the interannual increase in sea ice concentration (SICI) for each month is as follows: ; In the formula: y represents the year, n1 represents the number of grid points whose latitude and longitude fall within the critical sea ice concentration zone, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical sea ice concentration zone, i=1,2,3,...,n1, and SICI represents the period from YN-1 to Y-1. Monthly sea ice concentration key area interannual increment data, SICM from year YN-1 to year Y-1 Monthly data on the annual increase in sea ice concentration; From year YN-1 to year Y-1 The interannual increment data of sea ice concentration in the key monthly sea ice concentration region were standardized using SICI to obtain the data from year YN-1 to year Y-1. The Standardized Interannual Increment Index of Sea Ice Concentration (ZSICI) for key areas of monthly sea ice concentration is defined and calculated using the following formula: ; In the formula: y represents the year, and ZSICI represents the years from YN-1 to Y-1. Standardized interannual increments of sea ice concentration in key areas of monthly sea ice concentration, where Z represents the standardization process and SICI represents the transition from year YN-1 to year Y-1. The annual increment data of sea ice concentration in the key area of ​​monthly sea ice concentration, μ(SICI) is the average value of SICI data for N-1 years from YN-1 to Y-1, and σ(SICI) is the standard deviation of SICI data for N-1 years from YN-1 to Y-1.

[0012] Optionally, the sea surface temperature index includes: Extracting from the monthly scale of the annual sea surface temperature increment (DSST) Monthly sea surface temperature data is denoted as SSTM, and its calculation formula is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for sea surface temperature (SSTM) are from year YN-1 to year Y-1. Monthly sea surface temperature annual increment data, DSST from year YN-1 to year Y-1 Monthly sea surface temperature annual increase data; exist Based on monthly sea surface temperature (SST) data from SSTM, SSTM data where both latitude and longitude grid points are located within the critical SST region are extracted and averaged to obtain the critical SST region from year YN-1 to year Y-1. The formula for calculating the monthly interannual sea surface temperature increment (SSTI) is as follows: ; In the formula: y represents the year, n2 represents the number of grid points whose latitude and longitude fall within the critical sea surface temperature zone, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical sea surface temperature zone, i=1,2,3,...,n2, and SSTI represents the time from YN-1 to Y-1. Monthly sea surface temperature (SST) key area interannual SST increment data, SSTM from year YN-1 to year Y-1 Monthly sea surface temperature annual increase data; From year YN-1 to year Y-1 The interannual sea surface temperature increment data of key monthly sea surface temperature areas were standardized using SSTI to obtain the data from year YN-1 to year Y-1. The standardized interannual sea surface temperature increment index (ZSSTI) for key monthly sea surface temperature zones is calculated using the following formula: ; In the formula: y represents the year, and ZSSTI represents the years from YN-1 to Y-1. Standardized interannual sea surface temperature increment index (SSTI) for key monthly sea surface temperature zones, where Z represents the standardization process and SSTI represents the change from year YN-1 to year Y-1. The annual sea surface temperature increment data for key monthly sea surface temperature areas, μ(SSTI) is the average value of the SSTI data for N-1 years from YN-1 to Y-1, and σ(SSTI) is the standard deviation of the SSTI data for N-1 years from YN-1 to Y-1.

[0013] Optionally, the soil moisture index includes: Extracting from the monthly scale of the interannual increase in soil moisture (DSW) The formula for calculating monthly soil moisture data (SWM) is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for soil moisture are from year YN-1 to year Y-1. Monthly data on interannual soil moisture increments, DSW from year YN-1 to year Y-1. Monthly data on the annual increase in soil moisture; exist Based on the monthly soil moisture data (SWM data), SWM data where both latitude and longitude grid points are located within the critical soil moisture zone are extracted and averaged within the same time frame to obtain the data from year YN-1 to year Y-1. The annual increment of soil moisture (SWI) in key monthly soil moisture zones is calculated using the following formula: ; In the formula: y represents the year, n3 represents the number of grid points whose latitude and longitude fall within the critical soil moisture zone, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical soil moisture zone, i=1,2,3,...,n3, and SWI represents the period from YN-1 to Y-1. Monthly soil moisture key zone interannual increment data, SWM from year YN-1 to year Y-1 Monthly data on annual increases in soil moisture; From year YN-1 to year Y-1 The interannual increment data of soil moisture in key areas of monthly soil moisture were standardized using SWI to obtain the data from year YN-1 to year Y-1. The standardized interannual soil moisture increment index (ZSWI) for key areas of soil moisture in a given month is calculated using the following formula: ; In the formula: y is the year, and ZSWI is from year YN-1 to year Y-1. Standardized interannual soil moisture increment index for key soil moisture zones in a given month, where Z represents the standardization process and SWI represents the change from year YN-1 to year Y-1. The interannual increment data of soil moisture in key areas of soil moisture in the month, μ(SWI) is the average value of the SWI data for N-1 years from YN-1 to Y-1, and σ(SWI) is the standard deviation of the SWI data for N-1 years from YN-1 to Y-1.

[0014] Optionally, the North Atlantic Oscillation Index includes: Extracting from the monthly scale of the annual sea level pressure increment (DSLP) Monthly sea level pressure data is denoted as SLPM, and its calculation formula is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for the North Atlantic Oscillation (SLPM) are from year N-1 to year Y-1. Monthly sea level pressure annual increment data, DSLP from year YN-1 to year Y-1 Monthly data on the annual increase in sea level pressure; exist Based on monthly sea level pressure (SLPM) data, SLPM data where both latitude and longitude grid points fall within the key North Atlantic sea level pressure regions S1 and S2 are extracted and averaged within the same time stratum to obtain the North Atlantic region data for years YN-1 to Y-1. The annual sea level pressure increments for each month, SLPS1 and SLPS2, are calculated using the following formulas: ; In the formula: y represents the year, n4 represents the number of grid points whose latitude and longitude fall within the critical region S1 of the North Atlantic sea level pressure, (ti,pi) represents the i-th grid point whose latitude and longitude fall within the critical region S1 of the North Atlantic sea level pressure, i=1,2,3,...,n4, and SLPS1 represents the period from YN-1 to Y-1. Monthly interannual sea level pressure increment data for the North Atlantic S1 region, SLPM from year N-1 to year Y-1 Monthly sea level pressure annual increment data; ; In the formula: y represents the year, n5 represents the number of grid points whose latitude and longitude fall within the critical region S2 of the North Atlantic Sea Level Pressure, (ti,pi) represents the i-th grid point whose latitude and longitude fall within the critical region S2 of the North Atlantic Sea Level Pressure, i=1,2,3,...,n5, and SLPS2 represents the period from YN-1 to Y-1. Monthly interannual sea level pressure increment data for the North Atlantic S2 region, SLPM from year N-1 to year Y-1. Monthly sea level pressure annual increment data; Subtracting SLPS2 from SLPS1 in year YN-1 to year Y-1 reveals the sea-level pressure characteristics of the North Atlantic Oscillation's north-south "seesaw" pattern, as shown in the following equation: ; In the formula: y represents the year, and NAOI represents the years from YN-1 to Y-1. Monthly North Atlantic Oscillation data for the North Atlantic region, SLPS1 for year YN-1 to year Y-1 Monthly interannual sea level pressure increment data for the North Atlantic S1 region, SLPS2 for year N-1 to year Y-1. Monthly data on annual sea level pressure increments in the North Atlantic S2 region; From year YN-1 to year Y-1 The NAOI data of the North Atlantic Oscillation in the North Atlantic region for each month were standardized to obtain the data from year YN-1 to year Y-1. The standardized interannual increment index of the North Atlantic Oscillation (ZNAOI) for the North Atlantic region in a given month is defined by the following equation: ; In the formula: y represents the year, and ZNAOI represents the years from YN-1 to Y-1. The standardized interannual increment index of the North Atlantic Oscillation (NAOI) for the North Atlantic region in months, where Z represents the standardized process and NAOI represents the period from year YN-1 to year Y-1. The monthly North Atlantic Oscillation (NAOI) data for the North Atlantic region, where μ(NAOI) is the average NAOI data for the N-1 years from YN-1 to Y-1, and σ(NAOI) is the standard deviation of the NAOI data for the N-1 years from YN-1 to Y-1.

[0015] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: The model constructed in this invention can predict corn yield at least several months in advance, significantly enhancing the timeliness and practicality of the forecast model; This invention uses the interannual increment method to predict maize yield in a target area, significantly improving prediction accuracy. This invention establishes a maize yield forecasting model by studying the physical mechanism of the impact of previous meteorological factors on the combined high temperature and drought event in the target area. Compared with the models constructed by previous researchers, the applicability of this invention model has been significantly improved in the present and future. The model of this invention has more important practical significance for the prediction of maize yield in practical applications and can provide more accurate guidance and support for agricultural production. Attached Figure Description

[0016] Figure 1 This is a diagram illustrating the key area range for defining the Barents Sea ice index in this embodiment of the invention. Figure 2 This is a diagram illustrating the key regional range of the La Niña-like sea surface temperature index defined in this embodiment of the invention; Figure 3 This is a diagram illustrating the key area range for soil moisture index definition in northwestern Siberia in this embodiment of the invention. Figure 4This is a diagram illustrating the key regional range of the North Atlantic Oscillation Index defined in this embodiment of the invention; Figure 5 This is a graph showing the correlation coefficients between (a) DMAIZE and (b) MAIZE predicted values ​​(dashed line) and actual values ​​(solid line) of the 2021 maize yield forecast model for Northeast China constructed using data from 1979 to 2020 in an embodiment of the present invention. Figure 6 This is a schematic diagram of the process for predicting corn yield in Northeast China in an embodiment of the present invention; Figure 7 This is a schematic diagram of the method flow of the present invention. Detailed Implementation

[0017] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0018] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, are used only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined with "first," "second," etc., may explicitly or implicitly include one or more of that feature. In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0019] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art will understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0020] This embodiment provides a method for predicting maize yield in a target region based on the physical mechanisms of combined high temperature and drought effects. Using actual maize yield data and meteorological data for Northeast China from 1979 to 2020, along with global sea level pressure data for February 2021, global sea surface temperature data for March 2021, soil moisture data for Siberia for April 2021, and global sea ice concentration data for May 2021, the specific steps for predicting the maize yield to be harvested in Northeast China at the end of September and beginning of October 2021 are as follows: Step 1: Obtain maize yield per unit area data for the target region in year Y-1 (MAIZE) Y-1 ; Step 2: Obtain the forecast factor index for the early stage of the forecast year Y; Step 3: Obtain the forecast factor index of the early stage of the forecast year Y, input it into the pre-constructed forecast model of the forecast quantity DMAIZE, and predict the DMAIZE forecast value of the forecast year Y. Step 4, MAIZE the data Y-1 Adding the predicted DMAIZE value for forecast year Y, we obtain the predicted maize yield MAIZE for forecast year Y. Y .

[0021] Regarding step 1, in the specific implementation process of this embodiment, the step of obtaining the maize yield per unit area data (MAIZE) for the target region in year Y-1 is... Y-1 ,include: Collect annual maize yield data per unit area for the target region over N years; The annual yield data per unit area of ​​corn are averaged to form a time series of corn yield for the overall region. After standardizing the time series data of maize yield for the overall predicted region, we obtain the maize yield per unit area (MAIZE) for the target region from year YN to year Y-1, where the maize yield per unit area for the target region in year Y-1 is the data MAIZE. Y-1 Y represents the forecast year, and N represents the data collection period in years.

[0022] In the specific implementation of this embodiment, the method for constructing the forecast model of the forecast quantity DMAIZE includes: Collect annual maize yield data per unit area for the target region over N years; The annual yield data per unit area of ​​corn are averaged to form a time series of corn yield for the overall region. Global monthly gridded sea ice concentration data from YN to Y-1 were collected through the Hadley Centre, and global monthly gridded sea surface temperature data, 0-10cm soil moisture data and sea level pressure data from YN to Y-1 were collected through ERA5. After standardizing the time series of maize yield for the overall predicted region, we obtain the maize yield per unit area (MAIZE) data for the standardized target region from YN to Y-1. The data MAIZE was processed using the interannual increment method to calculate the interannual increment data DMAIZE of maize yield per unit area from year YN-1 to year Y-1 in the target area. Obtain forecast factor indices for the early stage of forecast year Y, including sea ice concentration index, sea surface temperature index, soil moisture index, and North Atlantic Oscillation index; The sea ice concentration index, sea surface temperature index, soil moisture index, and North Atlantic Oscillation index were used as forecasting factors, and the data DMAIZE was used as the forecast quantity. A forecasting model for the forecast quantity DMAIZE was constructed based on the stepwise linear regression method.

[0023] In this embodiment, the calculation formula for averaging the annual corn yield per unit area is as follows: ; In the formula: y is the year, Y is the forecast year, N is the data collection time in years, M is the corn yield per unit area in the target region from year YN to year Y-1, region1 is the corn yield per unit area in region one from year YN to year Y-1, region2 is the corn yield per unit area in region two from year YN to year Y-1, and regionX is the corn yield per unit area in region X from year YN to year Y-1.

[0024] In this embodiment, after standardizing the time series of maize yield for the entire predicted region, the maize yield per unit area data of the standardized target area from year YN to year Y-1 is denoted as data MAIZE. The calculation formula is as follows: ; In the formula: y is the year, MAIZE is the standardized maize yield per unit area data of the target area from year YN to year Y-1, Z is the standardization process, M is the maize yield per unit area data of the target area from year YN to year Y-1, μ(M) is the average value of M data for a total of N years from year YN to year Y-1, and σ(M) is the standard deviation of M data for a total of N years from year YN to year Y-1.

[0025] In step 1, the data MAIZE is processed using the interannual increment method to calculate the standardized interannual increment data DMAIZE of maize yield per unit area from year YN-1 to year Y-1 in the target area, including: ; In the formula: DMAIZE represents the annual increment of standardized maize yield per unit area in the target region from year YN-1 to year Y-1. The standardized maize yield per unit area represents the interannual increment data for the target region from year YN-1 to year Y-1, while MAIZE represents the standardized maize yield per unit area for the target region from year YN to year Y-1. This provides standardized maize yield per unit area data for the target region in year y from year YN to year Y-1. This refers to the standardized maize yield per unit area in the target region from year YN to year Y-1, where Y is the forecast year, y is the year, and the range is any year from year YN to year Y-1. N is the length of time for data collection, in years.

[0026] In this embodiment, step 2 is specifically explained, which involves obtaining the forecast factor index in the early stage of the forecast year Y, including: The collected sea ice concentration data, sea surface temperature data, soil moisture data, and sea level pressure data were preprocessed using the interannual increment method, and converted into interannual increment values ​​for sea ice concentration (DSIC), sea surface temperature (DSST), soil moisture (DSW), and sea level pressure (DSLP), respectively. The calculation formulas are as follows: ; In the formula: y is the year, m is the month, m=1,2,3,...,12, t is the t-th latitude, t=1,2,3,...,181, p is the p-th longitude, p=1,2,3,...,360; DXXX is the interannual increment data of meteorological factors from year YN-1 to year Y-1, DXXX=DSIC, DSST, DW, DSLP; XXX is the meteorological factor data from year YN to year Y-1, XXX=SIC, SST, SW,SLP; The sea ice concentration index includes: extracted from the monthly scale of the interannual sea ice concentration increment (DSIC) value. Monthly sea ice concentration data, denoted as SICM, is calculated using the following formula: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for sea ice concentration are from year YN-1 to year Y-1. Monthly sea ice concentration interannual increment data, DSIC from YN-1 to Y-1 Monthly data on the annual increase in sea ice concentration; exist Based on monthly sea ice concentration (SICM) data, the SICM data of latitude and longitude grid points simultaneously located within the critical sea ice concentration region are extracted and averaged to obtain the critical sea ice concentration region from year YN-1 to year Y-1. The formula for calculating the interannual increase in sea ice concentration (SICI) for each month is as follows: ; In the formula: y represents the year, n1 represents the number of grid points whose latitude and longitude fall within the critical sea ice concentration zone, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical sea ice concentration zone, i=1,2,3,...,n1, and SICI represents the period from YN-1 to Y-1. Monthly sea ice concentration key area interannual increment data, SICM from year YN-1 to year Y-1 Monthly data on the annual increase in sea ice concentration; From year YN-1 to year Y-1 The interannual increment data of sea ice concentration in the key monthly sea ice concentration region were standardized using SICI to obtain the data from year YN-1 to year Y-1. The Standardized Interannual Increment Index of Sea Ice Concentration (ZSICI) for key areas of monthly sea ice concentration is defined and calculated using the following formula: ; In the formula: y represents the year, and ZSICI represents the years from YN-1 to Y-1. Standardized interannual increments of sea ice concentration in key areas of monthly sea ice concentration, where Z represents the standardization process and SICI represents the transition from year YN-1 to year Y-1. The annual increment data of sea ice concentration in the key area of ​​monthly sea ice concentration, μ(SICI) is the average value of SICI data for N-1 years from YN-1 to Y-1, and σ(SICI) is the standard deviation of SICI data for N-1 years from YN-1 to Y-1.

[0027] In this embodiment, the sea surface temperature index is of the El Niño-like type, and the calculation method includes: Extracting from the monthly scale of the annual sea surface temperature increment (DSST) Monthly sea surface temperature data is denoted as SSTM, and its calculation formula is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for sea surface temperature (SSTM) are from year YN-1 to year Y-1. Monthly sea surface temperature annual increment data, DSST from year YN-1 to year Y-1 Monthly sea surface temperature annual increase data; exist Based on monthly sea surface temperature (SST) data from SSTM, SSTM data where both latitude and longitude grid points are located within the critical SST region are extracted and averaged to obtain the critical SST region from year YN-1 to year Y-1. The formula for calculating the monthly interannual sea surface temperature increment (SSTI) is as follows: ; In the formula: y represents the year, n2 represents the number of grid points whose latitude and longitude fall within the critical sea surface temperature zone, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical sea surface temperature zone, i=1,2,3,...,n2, and SSTI represents the time from YN-1 to Y-1. Monthly sea surface temperature (SST) key area interannual SST increment data, SSTM from year YN-1 to year Y-1 Monthly sea surface temperature annual increase data; From year YN-1 to year Y-1 The interannual sea surface temperature increment data of key monthly sea surface temperature areas were standardized using SSTI to obtain the data from year YN-1 to year Y-1. The standardized interannual sea surface temperature increment index (ZSSTI) for key monthly sea surface temperature zones is calculated using the following formula: ; In the formula: y represents the year, and ZSSTI represents the years from YN-1 to Y-1. Standardized interannual sea surface temperature increment index (SSTI) for key monthly sea surface temperature zones, where Z represents the standardization process and SSTI represents the change from year YN-1 to year Y-1. The annual sea surface temperature increment data for key monthly sea surface temperature areas, μ(SSTI) is the average value of the SSTI data for N-1 years from YN-1 to Y-1, and σ(SSTI) is the standard deviation of the SSTI data for N-1 years from YN-1 to Y-1.

[0028] In this embodiment, the soil moisture index includes: Extracting from the monthly scale of the interannual increase in soil moisture (DSW) The formula for calculating monthly soil moisture data (SWM) is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for soil moisture are from year YN-1 to year Y-1. Monthly data on interannual soil moisture increments, DSW from year YN-1 to year Y-1. Monthly data on the annual increase in soil moisture; exist Based on the monthly soil moisture data (SWM data), SWM data where both latitude and longitude grid points are located within the critical soil moisture zone are extracted and averaged within the same time frame to obtain the data from year YN-1 to year Y-1. The annual increment of soil moisture (SWI) in key monthly soil moisture zones is calculated using the following formula: ; In the formula: y represents the year, n3 represents the number of grid points whose latitude and longitude fall within the critical soil moisture zone, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical soil moisture zone, i=1,2,3,...,n3, and SWI represents the period from YN-1 to Y-1. Monthly soil moisture key zone interannual increment data, SWM from year YN-1 to year Y-1 Monthly data on annual increases in soil moisture; From year YN-1 to year Y-1 The interannual increment data of soil moisture in key areas of monthly soil moisture were standardized using SWI to obtain the data from year YN-1 to year Y-1. The standardized interannual soil moisture increment index (ZSWI) for key areas of soil moisture in a given month is calculated using the following formula: ; In the formula: y is the year, and ZSWI is from year YN-1 to year Y-1. Standardized interannual soil moisture increment index for key soil moisture zones in a given month, where Z represents the standardization process and SWI represents the change from year YN-1 to year Y-1. The interannual increment data of soil moisture in key areas of soil moisture in the month, μ(SWI) is the average value of the SWI data for N-1 years from YN-1 to Y-1, and σ(SWI) is the standard deviation of the SWI data for N-1 years from YN-1 to Y-1.

[0029] In this embodiment, the North Atlantic Oscillation Index includes: Extracting from the monthly scale of the annual sea level pressure increment (DSLP) Monthly sea level pressure data is denoted as SLPM, and its calculation formula is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for the North Atlantic Oscillation (SLPM) are from year N-1 to year Y-1. Monthly sea level pressure annual increment data, DSLP from year YN-1 to year Y-1 Monthly data on the annual increase in sea level pressure; exist Based on monthly sea level pressure (SLPM) data, SLPM data where both latitude and longitude grid points fall within the key North Atlantic sea level pressure regions S1 and S2 are extracted and averaged within the same time stratum to obtain the North Atlantic region data for years YN-1 to Y-1. The annual sea level pressure increments for each month, SLPS1 and SLPS2, are calculated using the following formulas: ; In the formula: y represents the year, n4 represents the number of grid points whose latitude and longitude fall within the critical region S1 of the North Atlantic sea level pressure, (ti,pi) represents the i-th grid point whose latitude and longitude fall within the critical region S1 of the North Atlantic sea level pressure, i=1,2,3,...,n4, and SLPS1 represents the period from YN-1 to Y-1. Monthly interannual sea level pressure increment data for the North Atlantic S1 region, SLPM from year N-1 to year Y-1 Monthly sea level pressure annual increment data; ; In the formula: y represents the year, n5 represents the number of grid points whose latitude and longitude fall within the critical region S2 of the North Atlantic Sea Level Pressure, (ti,pi) represents the i-th grid point whose latitude and longitude fall within the critical region S2 of the North Atlantic Sea Level Pressure, i=1,2,3,...,n5, and SLPS2 represents the period from YN-1 to Y-1. Monthly interannual sea level pressure increment data for the North Atlantic S2 region, SLPM from year N-1 to year Y-1. Monthly sea level pressure annual increment data; Subtracting SLPS2 from SLPS1 in year YN-1 to year Y-1 reveals the sea-level pressure characteristics of the North Atlantic Oscillation's north-south "seesaw" pattern, as shown in the following equation: ; In the formula: y represents the year, and NAOI represents the years from YN-1 to Y-1. Monthly North Atlantic Oscillation data for the North Atlantic region, SLPS1 for year YN-1 to year Y-1 Monthly interannual sea level pressure increment data for the North Atlantic S1 region, SLPS2 for year N-1 to year Y-1. Monthly data on annual sea level pressure increments in the North Atlantic S2 region; From year YN-1 to year Y-1 The NAOI data of the North Atlantic Oscillation in the North Atlantic region for each month were standardized to obtain the data from year YN-1 to year Y-1. The standardized interannual increment index of the North Atlantic Oscillation (ZNAOI) for the North Atlantic region in a given month is defined by the following equation: ; In the formula: y represents the year, and ZNAOI represents the years from YN-1 to Y-1. The standardized interannual increment index of the North Atlantic Oscillation (NAOI) for the North Atlantic region in months, where Z represents the standardized process and NAOI represents the period from year YN-1 to year Y-1. The monthly North Atlantic Oscillation (NAOI) data for the North Atlantic region, where μ(NAOI) is the average NAOI data for the N-1 years from YN-1 to Y-1, and σ(NAOI) is the standard deviation of the NAOI data for the N-1 years from YN-1 to Y-1.

[0030] This example illustrates how to predict the 2021 corn yield in Northeast China using annual corn yield data from Heilongjiang, Jilin, Liaoning, and Inner Mongolia provinces over a 42-year period from 1979 to 2020. Step 1: Collect data; Data on annual maize yield per unit area in Heilongjiang, Jilin, Liaoning, and Inner Mongolia provinces were collected from the National Bureau of Statistics over a 42-year period, covering 1979 to 2020. Based on this data, the annual maize yield per unit area data for these four regions were averaged to define a time series of maize production in Northeast China from 1979 to 2020. The calculation method is as follows: ; In the formula: y is the year, M is the corn yield per unit area in Northeast China from 1979 to 2020, heilongjiang is the corn yield per unit area in Heilongjiang Province from 1979 to 2020, jilin is the corn yield per unit area in Jilin Province from 1979 to 2020, liaoning is the corn yield per unit area in Liaoning Province from 1979 to 2020, and neimenggu is the corn yield per unit area in Inner Mongolia from 1979 to 2020. Global monthly gridded sea ice concentration data from 1979 to 2020 were collected through the Hadley Centre; global monthly gridded sea surface temperature data, 0-10 cm soil moisture data, and sea level pressure data from 1979 to 2020 were collected through ERA5.

[0031] Step 2: Preprocess the data; Standardize the maize yield per unit area (M) data for Northeast China from 1979 to 2020, and denote the standardized data as MAIZE. The standardization calculation formula is as follows: ; In the formula: y is the year, MAIZE is the standardized maize yield per unit area data in Northeast China from 1979 to 2020, Z is the standardization process, M is the maize yield per unit area data in Northeast China from 1979 to 2020, μ(M) is the average of M data for a total of 42 years from 1979 to 2020, and σ(M) is the standard deviation of M data for a total of 42 years from 1979 to 2020; The standardized maize yield data for Northeast China obtained above was processed using the interannual increment (DY) method. The standardized maize yield per unit area (MAIZE) data for Northeast China from 1979 to 2020 was transformed into the interannual increment data (DMAIZE) for standardized maize yield per unit area for Northeast China from 1980 to 2020 after interannual increment processing. The calculation formula is as follows: ; In the formula: y represents the year, and DMAIZE represents the interannual increment of standardized maize yield per unit area in Northeast China from 1980 to 2020. The first part represents the interannual increment of standardized maize yield per unit area in Northeast China from 1980 to 2020, while the second part represents the standardized maize yield per unit area in Northeast China from 1979 to 2020. Standardized maize yield per unit area in Northeast China from 1979 to 2020 (MAIZE) y-1 Standardized maize yield per unit area in Northeast China from 1979 to 2020; The meteorological data collected in step one were processed using the interannual increment method: the sea ice concentration (SIC), sea surface temperature (SST), soil moisture (SW), and sea level pressure (SLP) data from step one were transformed into interannual increment data for sea ice concentration (DSIC), sea surface temperature (DSST), soil moisture (DSW), and sea level pressure (DSLP), respectively. The calculation formulas are as follows: ; In the formula: y is the year, m is the month, m=1,2,3,...,12, t is the t-th latitude, t=1,2,3,...,181, p is the p-th longitude, p=1,2,3,...,360; DXXX is the interannual increment data of previous meteorological factors from 1980 to 2020, DXXX=DSIC, DSST, DW, DSLP; XXX is the previous meteorological factor data from 1979 to 2020, XXX=SIC, SST, SW,SLP.

[0032] Step 3: Establish forecast factor indices; Establish forecast factor one: sea ice concentration index; In step two, the global sea ice concentration interannual increment data (DSIC) from 1980 to 2020, obtained through processing, is used to extract the sea ice concentration data for May from the monthly scale of the DSIC, which is then calculated as SICM. The formula is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,…,181), p represents the p-th longitude (p=1,2,3,…,360), and SICM represents the interannual increase in sea ice concentration from 1980 to May 2020. Data on the interannual increase in sea ice concentration from 1980 to May 2020; The key area for sea ice concentration in the Barents Sea is defined as: 20°E–47°E, 72°N–78°N. Figure 1 The area within the black rectangle (indicated by the above calculation) forms the basis for the global sea ice concentration data (SICM) for May. SICM data where both latitude and longitude grid points are located within the Barents Sea sea ice concentration key area are extracted and averaged to obtain the interannual increment data (SICI) of sea ice concentration in the Barents Sea region from 1980 to May 2020. The calculation formula is as follows: ; In the formula: y represents the year, and n1 represents the number of grid points whose latitude and longitude fall within the key area of ​​sea ice concentration in the Barents Sea. Figure 1 The number of grid points in the black box, (ti, pi) is the i-th grid point whose latitude and longitude fall within the key area of ​​sea ice concentration in the Barents Sea, i=1,2,3,…,n1, SICI is the interannual increment data of sea ice concentration in the Barents Sea region from 1980 to May 2020, and SICM is the interannual increment data of sea ice concentration from 1980 to May 2020; Further standardization of the interannual sea ice concentration increment data (SICI) in the Barents Sea region from 1980 to May 2020 yields the standardized interannual sea ice concentration increment index (ZSICI) for the Barents Sea region from 1980 to May 2020. The calculation formula is as follows: ; In the formula: y is the year, ZSICI is the standardized interannual increase index of sea ice concentration in the Barents Sea region from 1980 to May 2020, Z is the standardization process, SICI is the interannual increase data of sea ice concentration in the Barents Sea region from 1980 to May 2020, μ(SICI) is the average value of SICI data for a total of 41 years from 1980 to 2020, and σ(SICI) is the standard deviation of SICI data for a total of 41 years from 1980 to 2020; Establish forecast factor two: El Niño-like sea surface temperature index; In step two, the global sea surface temperature (DSST) interannual increment data from 1980 to 2020 obtained through processing are used to extract the sea surface temperature data for March from the monthly scale of the DSST, denoted as SSTM. The calculation formula is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), p represents the p-th longitude (p=1,2,3,...,360), and SSTM represents the interannual sea surface temperature increment data from 1980 to March 2020. This data represents the interannual increase in sea surface temperature from 1980 to March 2020. The key sea surface temperature zone in the eastern equatorial Pacific is defined as: 130°W–80°W, 10°S–15°N ( Figure 2 The area within the black rectangle (in the middle) serves as the basis for preparing the global El Niño-like sea surface temperature (SSTM) data for March, calculated above. SSTM data where both latitude and longitude grid points are located within the key sea surface temperature region of the eastern equatorial Pacific are extracted and averaged to obtain the interannual increment data of El Niño-like sea surface temperature (SSTI) for the eastern equatorial Pacific region in March from 1980 to 2020. The calculation formula is as follows: ; In the formula: y is the year, and n2 is the number of grid points whose latitude and longitude fall within the key sea surface temperature region of the eastern equatorial Pacific Ocean. Figure 2The number of grid points in the black box, (ti, pi) is the i-th grid point whose latitude and longitude fall within the key sea surface temperature region of the eastern equatorial Pacific, i=1,2,3,...,n2, SSTI is the interannual increment data of El Niño-like sea surface temperature in the eastern equatorial Pacific from 1980 to March 2020, and SSTM is the interannual increment data of El Niño-like sea surface temperature from 1980 to March 2020; Further standardization of the El Niño-like interannual sea surface temperature (SSTI) increment data for the eastern equatorial Pacific from 1980 to March 2020 yields the Standardized El Niño-like Interannual Sea Surface Temperature (SSTI) Indicator for the eastern equatorial Pacific from 1980 to March 2020, calculated as follows: ; In the formula: y is the year, ZSSTI is the standardized El Niño-like interannual sea surface temperature increment index for the eastern equatorial Pacific from 1980 to March 2020, Z is the standardization process, SSTI is the data on the El Niño-like interannual sea surface temperature increment for the eastern equatorial Pacific from 1980 to March 2020, μ(SSTI) is the average of the SSTI data for the 41 years from 1980 to 2020, and σ(SSTI) is the standard deviation of the SSTI data for the 41 years from 1980 to 2020.

[0033] Establish forecast factor three: Soil moisture index; In step two, we prepared the global interannual soil moisture increment data (DSW) from 1980 to 2020. From the monthly scale of the DSW, we extracted the soil moisture data (SWM) for April. The calculation formula is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,…,181), p represents the p-th longitude (p=1,2,3,…,360), and SWM represents the interannual increase in soil moisture from 1980 to April 2020. Data on interannual increases in soil moisture from 1980 to April 2020; The critical zone for soil moisture in northwestern Siberia is defined as: 30°E–110°E, 60°N–75°N. Figure 3 Based on global April soil moisture data (within the black rectangle), SWM data from regions where both latitude and longitude grid points are located within the key soil moisture zone of northwestern Siberia were extracted and averaged to obtain the interannual increment of soil moisture (SWI) in northwestern Siberia from 1980 to April 2020. The calculation formula is as follows: ; In the formula: y is the year, and n3 is the number of grid points whose latitude and longitude fall within the key soil moisture zone in northwestern Siberia. Figure 3 The number of grid points in the black box, (ti, pi) is the i-th grid point whose latitude and longitude fall within the key soil moisture zone of northwestern Siberia, i=1,2,3,…,n3, SWI is the interannual increment data of soil moisture in northwestern Siberia from 1980 to April 2020, and SWM is the interannual increment data of soil moisture from 1980 to April 2020; Further standardization of the interannual soil moisture increment data (SWI) in northwestern Siberia from 1980 to April 2020 yielded the standardized interannual soil moisture increment index (ZSWI) for northwestern Siberia from 1980 to April 2020. The calculation formula is as follows: ; In the formula: y is the year, ZSWI is the standardized interannual increase index of soil moisture in northwestern Siberia from 1980 to April 2020, Z is the standardization process, SWI is the interannual increase data of soil moisture in northwestern Siberia from 1980 to April 2020, μ(SWI) is the average of SWI data for a total of 41 years from 1980 to 2020, and σ(SWI) is the standard deviation of SWI data for a total of 41 years from 1980 to 2020.

[0034] Establish forecast factor four: North Atlantic Oscillation Index; In step two, the global sea level pressure interannual increment data (DSLP) from 1980 to 2020 is processed. The sea level pressure data for February, denoted as SLPM, is extracted from the monthly scale of the DSLP. The calculation formula is as follows: ; In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), p represents the p-th longitude (p=1,2,3,...,360), and SLPM represents the interannual increase in sea level pressure from 1980 to February 2020. This data represents the interannual increase in sea level pressure from 1980 to February 2020. The key area of ​​sea-level pressure in the North Atlantic is defined as: 60°W–0°W, 70°N–75°N (Region S1) Figure 4 (Black solid line rectangle) and 80°W–0°W, 15°N–36°N (Region S2) Figure 4Based on the global February sea level pressure (SLPM) data (within the black dashed rectangle), averaged the SLPM data where both latitude and longitude grid points fall within the key North Atlantic sea level pressure regions S1 and S2, to obtain two interannual sea level pressure increment data sets for the North Atlantic region from 1980 to February 2020: SLPS1 and SLPS2. The calculation formula is as follows: ; In the formula: y is the year, and n4 is the number of grid points whose latitude and longitude fall within the critical region S1 of the North Atlantic sea level pressure. Figure 4 The number of grid points in the black solid box, (ti, pi) is the i-th grid point whose latitude and longitude fall in the key sea level pressure region S1 of the North Atlantic, i=1,2,3,...,n4, SLPS1 is the interannual increment data of sea level pressure in the North Atlantic S1 region from 1980 to February 2020, and SLPM is the interannual increment data of sea level pressure from 1980 to February 2020; ; In the formula: y is the year, and n5 is the number of grid points whose latitude and longitude fall within the critical region S2 of the North Atlantic sea level pressure. Figure 4 The number of grid points in the black dashed box is (ti, pi), which is the i-th grid point whose latitude and longitude fall within the critical sea level pressure region S2 in the North Atlantic, i=1,2,3,...,n5. SLPS2 is the interannual increment data of sea level pressure in the North Atlantic S2 region from 1980 to February 2020, and SLPM is the interannual increment data of sea level pressure from 1980 to February 2020.

[0035] To characterize the strength of the North Atlantic Oscillation (NAO), the NAO data is defined using the interannual sea level pressure increments (SLPS1 and SLPS2) of the key North Atlantic sea level pressure regions S1 and S2 as defined above. Specifically, subtracting the SLPS2 value from the SLPS1 value from 1980 to 2020 reveals the north-south "seesaw" sea level pressure characteristics of the NAO. The calculation formula is as follows: ; In the formula: y represents the year, NAOI represents the North Atlantic Oscillation data for the North Atlantic region from 1980 to February 2020, and SLPS1 represents the interannual increase in sea level pressure in the S1 region of the North Atlantic from 1980 to February 2020. Data on the interannual increase in sea level pressure in the North Atlantic S2 region from 1980 to February 2020; Further standardization of the North Atlantic Oscillation (NAOI) interannual increment data for the North Atlantic region from 1980 to February 2020 yields the standardized North Atlantic Oscillation Interannual Increment Index (ZNAOI) for the North Atlantic region from 1980 to February 2020, calculated using the following formula: ; In the formula: y is the year, ZNAOI is the standardized interannual increment index of the North Atlantic Oscillation in the North Atlantic region from 1980 to February 2020, Z is the standardization process, NAOI is the interannual increment data of the North Atlantic Oscillation in the North Atlantic region from 1980 to February 2020, μ(NAOI) is the average value of the NAOI data for a total of 41 years from 1980 to 2020, and σ(NAOI) is the standard deviation of the NAOI data for a total of 41 years from 1980 to 2020.

[0036] Step 4: Establish a forecast model; Using the four standardized early forecast factor indices ZSICI, ZSSTI, ZSWI, and ZNAOI constructed in step three above as forecast factors, and the standardized interannual increment data of maize yield per unit area in Northeast China from 1980 to 2020 constructed in step two above as the forecast quantity, a forecast model for the forecast quantity DMAIZE is constructed based on the stepwise linear regression method, wherein the forecast equation is: ; Where: DMAIZE is the predicted interannual increase in standardized maize yield per unit area in Northeast China from 1980 to 2020; ZSSTI is the standardized interannual increase index of El Niño-like sea surface temperature in the eastern equatorial Pacific from 1980 to March 2020; and ZSWI is the standardized interannual increase index of soil moisture in northwestern Siberia from 1980 to April 2020. The correlation coefficient between DMAIZE predicted values ​​and observed values ​​from 1980 to 2020 was 0.73 (passing the 99% significance t-test), and the correlation coefficient between MAIZE predicted values ​​and observed values ​​from 1980 to 2020 was 0.91 (passing the 99% significance test). In cross-validation, the correlation coefficient between DMAIZE predicted values ​​and observed values ​​from 1980 to 2020 was 0.68, and the correlation coefficient between MAIZE predicted values ​​and observed values ​​was 0.90. In independent posterior, the correlation coefficient between DMAIZE predicted values ​​and observed values ​​from 1998 to 2020 was 0.75, and the correlation coefficient between MAIZE predicted values ​​and observed values ​​was 0.79. All of the above correlation coefficients passed the 99% significance t-test. The forecast year is 2021. By the end of March 2021, the sea surface temperature (SST) data for 2021 can be obtained and added to the SST data from 1979 to 2020 in step one. Then, we re-enter steps two and three, using the data from 1979 to 2021 to calculate the SSTM data from 1979 to 2021. Then, according to the process of defining SSTI in step three, the SSTM data of latitude and longitude grid points that are simultaneously located within the key SST region of the eastern equatorial Pacific are extracted, averaged, and standardized to obtain the standardized El Niño-like interannual SST increment index ZSSTI for the eastern equatorial Pacific region from 1979 to March 2021. The results show that ZSSTI (2021) is -0.39134869. The forecast year is 2021. At the end of April 2021, the soil moisture data for 2021 can be added to the soil moisture data from 1979 to 2020 in step one, and then steps two and three can be restarted. The SWM data from 1979 to 2021 can be calculated using the data from 1979 to 2021. Then, according to the process of defining SWI in step four, the SWM data of latitude grid points and longitude grid points located within the key soil moisture area of ​​northwestern Siberia are extracted, averaged and standardized to obtain the standardized interannual soil moisture increment index ZSWI for northwestern Siberia from 1979 to April 2021. The result shows that ZSWI (2021) is -0.01044977. By the end of April 2021, all the early forecast factor indices for 2021 had been calculated. These early forecast factor indices were then incorporated into the stepwise linear regression model constructed in step four based on data from 1979 to 2020 to derive the DMAIZE forecast value for 2021. The calculation equation is as follows: ; Where: DMAIZE (2021) is the predicted interannual increment of standardized maize yield per unit area in Northeast China in 2021, ZSSTI (2021) is the interannual increment index of standardized sea ice concentration in the Barents Sea region in May 2021, and ZSWI is the interannual increment index of standardized soil moisture in northwestern Siberia in April 2021. The calculation results show that DMAIZE (2021) is 0.002203; In step two of the preparation process, the calculated MAIZE (2020) is 1.43013. Adding the predicted interannual increase in standardized maize yield per unit area in Northeast China for 2021 (DMAIZE (2021)) to the predicted value of MAIZE (2021) for the forecast year, we obtain the predicted value of MAIZE (2021) for Northeast China in 2021. The calculation formula is as follows: ; In the formula: 2021 is the forecast year, MAIZE(2021) is the forecast value of standardized maize yield per unit area in Northeast China in 2021, DMAIZE(2021) is the forecast value of the interannual increase of standardized maize yield per unit area in Northeast China in 2021, and MAIZE(2020) is the actual value of standardized maize yield per unit area in Northeast China in 2020. The calculation results show that MAIZE (2021) is 1.432333.

[0037] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for predicting maize yield in a target area based on the physical mechanisms of the combined effects of high temperature and drought, characterized in that, The method includes: Obtain maize yield per unit area data for the target region in year Y-1 (MAIZE) Y-1 ; Obtain the forecast factor index for the early stage of forecast year Y; The forecast factor index of the early stage of the forecast year Y will be obtained and input into the pre-built forecast model of the forecast amount DMAIZE to predict the forecast value of DMAIZE for year Y. MAIZE the data Y-1 Adding the predicted DMAIZE value for year Y, we obtain the predicted maize yield MAIZE for year Y. Y Where Y is the predicted year; The method for constructing the forecast model for the forecast quantity DMAIZE includes: Collect annual maize yield data per unit area for the target region over N years; The annual yield data per unit area of ​​corn are averaged to form a time series of corn yield for the overall region. Global monthly gridded sea ice concentration data from YN to Y-1 were collected through the Hadley Centre, and global monthly gridded sea surface temperature data, 0-10cm soil moisture data and sea level pressure data from YN to Y-1 were collected through ERA5. After standardizing the time series of maize yield for the overall predicted region, we obtain the maize yield per unit area (MAIZE) data for the standardized target region from YN to Y-1. The data MAIZE was processed using the interannual increment method to calculate the interannual increment data DMAIZE of maize yield per unit area from year YN-1 to year Y-1 in the target area. Obtain the forecast factor indices for the early stage of forecast year Y, including the sea ice concentration index, sea surface temperature index, soil moisture index, and North Atlantic Oscillation index. The sea ice concentration index, sea surface temperature index, soil moisture index, and North Atlantic Oscillation index were used as forecasting factors, and the data DMAIZE was used as the forecast quantity. A forecasting model for the forecast quantity DMAIZE was constructed based on the stepwise linear regression method. The calculation formula for averaging the annual corn yield per unit area is as follows: ； In the formula: y is the year, M is the corn yield per unit area data of the target region from year YN to year Y-1, region1 is the corn yield per unit area data of region one from year YN to year Y-1, region2 is the corn yield per unit area data of region two from year YN to year Y-1, and regionX is the corn yield per unit area data of region X from year YN to year Y-1.

2. The method for predicting maize yield in a target area based on the physical mechanism of combined high temperature and drought effects according to claim 1, characterized in that, The acquisition of maize yield per unit area data for the target region in year Y-1 (MAIZE) Y-1 ,include: Collect annual maize yield data per unit area for the target region over N years; The annual yield data per unit area of ​​corn are averaged to form a time series of corn yield for the overall region. After standardizing the time series of maize yield for the entire predicted region, we obtain the maize yield per unit area (MAIZE) for the target region from year YN to year Y-1, where the maize yield per unit area for the target region in year Y-1 is the data MAIZE. Y-1 Y represents the forecast year, and N represents the data collection period in years.

3. The method for predicting maize yield in a target area based on the physical mechanism of combined high temperature and drought effects according to claim 1, characterized in that, The formula for calculating the maize yield per unit area (MAIZE) of the standardized target area from year YN to year Y-1 after standardizing the time series of maize yield for the entire predicted region is as follows: ； In the formula: y is the year, MAIZE is the standardized maize yield per unit area data of the target area from year YN to year Y-1, Z is the standardization process, M is the maize yield per unit area data of the target area from year YN to year Y-1, μ(M) is the average value of M data for a total of N years from year YN to year Y-1, σ(M) is the standard deviation of M data for a total of N years from year YN to year Y-1, and N is the time length of data collection in years.

4. The method for predicting maize yield in a target area based on the physical mechanism of combined high temperature and drought effects according to claim 1, characterized in that, The method of processing data MAIZE using the interannual increment method to calculate the standardized maize yield per unit area (DMAIZE) data from year YN-1 to year Y-1 for the target region includes: ； In the formula: DMAIZE represents the annual increment of standardized maize yield per unit area in the target region from year YN-1 to year Y-1. The standardized maize yield per unit area represents the interannual increment data for the target region from year YN-1 to year Y-1, while MAIZE represents the standardized maize yield per unit area for the target region from year YN to year Y-1. This provides standardized maize yield per unit area data for the target region in year y from year YN to year Y-1. This refers to the standardized maize yield per unit area data for the target region in year y-1 from year YN to year Y-1, where Y is the forecast year, y is the year, and the range is any year from year YN to year Y-1. N is the length of time for data collection, in years.

5. The method for predicting maize yield in a target area based on the physical mechanism of combined high temperature and drought effects according to claim 1, characterized in that, The acquisition of the forecast factor index in the early stage of the forecast year Y includes: The collected sea ice concentration data, sea surface temperature data, soil moisture data, and sea level pressure data were preprocessed using the interannual increment method, and converted into interannual increment values ​​for sea ice concentration (DSIC), sea surface temperature (DSST), soil moisture (DSW), and sea level pressure (DSLP), respectively. The calculation formulas are as follows: ； In the formula: y is the year, m is the month, m=1,2,3,...,12, t is the t-th latitude, t=1,2,3,...,181, p is the p-th longitude, p=1,2,3,...,360; DXXX is the interannual increment data of meteorological factors from year YN-1 to year Y-1, DXXX=DSIC, DSST, DW, DSLP; XXX is the meteorological factor data from year YN to year Y-1, XXX=SIC, SST, SW,SLP; The sea ice concentration index includes: extracted from the monthly scale of the interannual sea ice concentration increment (DSIC) value. Monthly sea ice concentration data, denoted as SICM, is calculated using the following formula: ； In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for sea ice concentration are from year YN-1 to year Y-1. Monthly sea ice concentration interannual increment data, DSIC from YN-1 to Y-1 Monthly data on the annual increase in sea ice concentration; exist Based on monthly sea ice concentration (SICM) data, the SICM data of latitude and longitude grid points simultaneously located within the critical sea ice concentration region are extracted and averaged to obtain the critical sea ice concentration region from year YN-1 to year Y-1. The formula for calculating the interannual increase in sea ice concentration (SICI) for each month is as follows: ； In the formula: y represents the year, n1 represents the number of grid points whose latitude and longitude fall within the critical sea ice concentration zone, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical sea ice concentration zone, i=1,2,3,...,n1, and SICI represents the period from YN-1 to Y-1. Monthly sea ice concentration key area interannual increment data, SICM from year YN-1 to year Y-1 Monthly data on the annual increase in sea ice concentration; From year YN-1 to year Y-1 The interannual increment data of sea ice concentration in the key monthly sea ice concentration region were standardized using SICI to obtain the data from year YN-1 to year Y-1. The Standardized Interannual Increment Index of Sea Ice Concentration (ZSICI) for key areas of monthly sea ice concentration is defined and calculated using the following formula: ； In the formula: y represents the year, and ZSICI represents the years from YN-1 to Y-1. Standardized interannual increments of sea ice concentration in key areas of monthly sea ice concentration, where Z represents the standardization process and SICI represents the transition from year YN-1 to year Y-1. The annual increment data of sea ice concentration in the key area of ​​monthly sea ice concentration, μ(SICI) is the average value of SICI data for N-1 years from YN-1 to Y-1, and σ(SICI) is the standard deviation of SICI data for N-1 years from YN-1 to Y-1.

6. The method for predicting maize yield in a target area based on the physical mechanism of combined high temperature and drought effects according to claim 5, characterized in that, The sea surface temperature index includes: Extracting from the monthly scale of the annual sea surface temperature increment (DSST) Monthly sea surface temperature data is denoted as SSTM, and its calculation formula is as follows: ； In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for sea surface temperature (SSTM) are from year YN-1 to year Y-1. Monthly sea surface temperature annual increment data, DSST from year YN-1 to year Y-1 Monthly sea surface temperature annual increase data; exist Based on monthly sea surface temperature (SST) data from SSTM, SSTM data where both latitude and longitude grid points are located within the critical SST region are extracted and averaged to obtain the critical SST region from year YN-1 to year Y-1. The formula for calculating the monthly interannual sea surface temperature increment (SSTI) is as follows: ； In the formula: y represents the year, n2 represents the number of grid points whose latitude and longitude fall within the critical sea surface temperature zone, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical sea surface temperature zone, i=1,2,3,...,n2, and SSTI represents the time from YN-1 to Y-1. Monthly sea surface temperature (SST) key area interannual SST increment data, SSTM from year YN-1 to year Y-1. Monthly sea surface temperature annual increment data; From year YN-1 to year Y-1 The interannual sea surface temperature increment data of key monthly sea surface temperature areas were standardized using SSTI to obtain the data from year YN-1 to year Y-1. The standardized interannual sea surface temperature increment index (ZSSTI) for key monthly sea surface temperature zones is calculated using the following formula: ； In the formula: y represents the year, and ZSSTI represents the years from YN-1 to Y-1. Standardized interannual sea surface temperature (SSTI) index for key monthly sea surface temperature zones, where Z represents the standardization process and SSTI represents the change from year YN-1 to year Y-1. The annual sea surface temperature increment data for key monthly sea surface temperature areas, μ(SSTI) is the average value of the SSTI data for N-1 years from YN-1 to Y-1, and σ(SSTI) is the standard deviation of the SSTI data for N-1 years from YN-1 to Y-1.

7. The method for predicting maize yield in a target area based on the physical mechanism of combined high temperature and drought effects according to claim 5, characterized in that, The soil moisture index includes: Extracting from the monthly scale of the interannual increase in soil moisture (DSW) The formula for calculating monthly soil moisture data (SWM) is as follows: ； In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for soil moisture are from year YN-1 to year Y-1. Monthly data on interannual soil moisture increments, DSW from year YN-1 to year Y-1. Monthly data on the annual increase in soil moisture; exist Based on the monthly soil moisture data (SWM data), SWM data where both latitude and longitude grid points are located within the critical soil moisture zone are extracted and averaged within the same time frame to obtain the data from year YN-1 to year Y-1. The annual increment of soil moisture (SWI) in key monthly soil moisture zones is calculated using the following formula: ； In the formula: y represents the year, n3 represents the number of grid points whose latitude and longitude fall within the critical soil moisture zone, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical soil moisture zone, i=1,2,3,...,n3, and SWI represents the period from YN-1 to Y-1. Monthly soil moisture key zone interannual increment data, SWM from year YN-1 to year Y-1 Monthly data on annual increases in soil moisture; From year YN-1 to year Y-1 The interannual increment data of soil moisture in key areas of monthly soil moisture were standardized using SWI to obtain the data from year YN-1 to year Y-1. The standardized interannual soil moisture increment index (ZSWI) for key areas of soil moisture in a given month is calculated using the following formula: ； In the formula: y is the year, and ZSWI is from year YN-1 to year Y-1. Standardized interannual soil moisture increment index for key soil moisture zones in a given month, where Z represents the standardization process and SWI represents the change from year YN-1 to year Y-1. The interannual increment data of soil moisture in key areas of soil moisture in the month, μ(SWI) is the average value of the SWI data for N-1 years from YN-1 to Y-1, and σ(SWI) is the standard deviation of the SWI data for N-1 years from YN-1 to Y-1.

8. The method for predicting maize yield in a target area based on the physical mechanism of combined high temperature and drought effects according to claim 5, characterized in that, The North Atlantic Oscillation Index includes: Extracting from the monthly scale of the annual sea level pressure increment DSLP Monthly sea level pressure data is denoted as SLPM, and its calculation formula is as follows: ； In the formula: y represents the year, t represents the t-th latitude (t=1,2,3,...,181), and p represents the p-th longitude (p=1,2,3,...,360). The key months for the North Atlantic Oscillation (SLPM) are from year N-1 to year Y-1. Monthly sea level pressure annual increment data, DSLP from year YN-1 to year Y-1 Monthly data on the annual increase in sea level pressure; exist Based on monthly sea level pressure (SLPM) data, SLPM data where both latitude and longitude grid points fall within the key North Atlantic sea level pressure regions S1 and S2 are extracted and averaged within the same time stratum to obtain the North Atlantic region data for years YN-1 to Y-1. The annual sea level pressure increments for each month, SLPS1 and SLPS2, are calculated using the following formulas: ； In the formula: y represents the year, n4 represents the number of grid points whose latitude and longitude fall within the critical region S1 of the North Atlantic sea level pressure, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical region S1 of the North Atlantic sea level pressure, i=1,2,3,...,n4, and SLPS1 represents the period from YN-1 to Y-1. Monthly interannual sea level pressure increment data for the North Atlantic S1 region, SLPM from year N-1 to year Y-1. Monthly sea level pressure annual increment data; ； In the formula: y represents the year, n5 represents the number of grid points whose latitude and longitude fall within the critical region S2 of the North Atlantic sea level pressure, (ti, pi) represents the i-th grid point whose latitude and longitude fall within the critical region S2 of the North Atlantic sea level pressure, i=1,2,3,...,n5, and SLPS2 represents the period from YN-1 to Y-1. Monthly interannual sea level pressure increment data for the North Atlantic S2 region, SLPM from year N-1 to year Y-1. Monthly sea level pressure annual increment data; Subtracting SLPS2 from SLPS1 in year YN-1 to year Y-1 reveals the north-south "seesaw" sea-level pressure characteristics of the North Atlantic Oscillation, as shown in the following equation: ； In the formula: y represents the year, and NAOI represents the years from YN-1 to Y-1. Monthly North Atlantic Oscillation data for the North Atlantic region, SLPS1 for year YN-1 to year Y-1 Monthly interannual sea level pressure increment data for the North Atlantic S1 region, SLPS2 for year N-1 to year Y-1. Monthly data on annual sea level pressure increments in the North Atlantic S2 region; From year YN-1 to year Y-1 The NAOI data of the North Atlantic Oscillation in the North Atlantic region for each month were standardized to obtain the data from year YN-1 to year Y-1. The standardized interannual increment index of the North Atlantic Oscillation (ZNAOI) for the North Atlantic region in a given month is defined by the following equation: ； In the formula: y represents the year, and ZNAOI represents the years from YN-1 to Y-1. The standardized interannual increment index of the North Atlantic Oscillation (NAOI) for the North Atlantic region in months, where Z represents the standardized process and NAOI represents the period from year YN-1 to year Y-1. The monthly North Atlantic Oscillation (NAOI) data for the North Atlantic region, where μ(NAOI) is the average NAOI data for the N-1 years from YN-1 to Y-1, and σ(NAOI) is the standard deviation of the NAOI data for the N-1 years from YN-1 to Y-1.

Images