Dynamic prediction method of irrigation water requirement based on water balance equation and crop water production function

Abstract

This invention discloses a method for dynamic prediction of irrigation water demand based on the water balance equation and crop water production function, comprising the following steps: Step 1, multi-source data processing; Step 2, calculation of net irrigation water volume based on the water balance equation; Step 3, determination of evapotranspiration demand based on the crop water production function; Step 4, determination of precipitation thresholds for different drought levels; Step 5, prediction of regional irrigation water demand under different drought scenarios. The method of this invention estimates the spatiotemporal dynamic changes of farmland irrigation water volume at a regional scale, combines this with the crop water production function to determine crop evapotranspiration demand at different yield levels, and then predicts the spatiotemporal changes of irrigation water demand under different drought levels based on the water balance principle. Compared with traditional irrigation systems, this method considers the potential impact of insufficient water on crop yield and distinguishes the spatiotemporal differences in irrigation water demand, providing effective technical support for dynamic prediction of agricultural irrigation water use and meeting the requirements of agricultural drought prevention.

Description

Technical Field

[0001] This invention belongs to the field of agricultural drought prevention and water resource allocation technology, and in particular relates to a method for dynamic prediction of irrigation water demand based on water balance equation and crop water production function. Background Technology

[0002] Due to the uneven spatial and temporal distribution of rainfall in my country, agricultural production in most areas relies on irrigation to ensure the water needed for crop growth. Approximately 50% of the country's arable land depends on irrigation systems, and irrigated areas contribute about 75% of the nation's grain output. At the same time, my country's agricultural water supply still faces enormous pressure. In recent years, under the influence of global warming, extreme weather events have become more frequent, stronger, and more widespread, with droughts becoming increasingly extreme and anomalous, posing an increasingly severe challenge to food security and agricultural water supply security. Accurately quantifying the spatial and temporal changes in crop water requirements to estimate irrigation water needs has become an important means of resolving the contradiction between agricultural water supply and demand and ensuring food security in my country.

[0003] Crop water requirement refers to the total amount of water consumed by crops throughout their growth cycle to maintain normal physiological activities (such as photosynthesis and transpiration) and growth and development. Its core is evapotranspiration (ET), which is the sum of crop transpiration and soil evaporation. Crop irrigation water requirement refers to the water transported to the field through the irrigation system, its core purpose being to supplement insufficient natural precipitation and ensure that crop water requirements are met. Crop irrigation water requirement is not directly equal to crop water requirement. Traditionally, reference crop evapotranspiration (ET0) is calculated using meteorological data from weather stations and the Penman-Monteith equation, and then potential crop evapotranspiration (ET) is calculated using the crop coefficient method. Irrigation regimes are formulated based on the calculated potential crop evapotranspiration and set precipitation scenarios to guide agricultural irrigation. However, the irrigation regimes obtained in this way are difficult to reflect the spatiotemporal differences in irrigation water volume. Furthermore, it is impossible to dynamically assess changes in regional irrigation water requirement under different drought scenarios, resulting in insufficient precision in water resource management and weak scientific drought resistance and disaster reduction capabilities during agricultural drought prevention.

[0004] Drought development, especially severe drought, exhibits creep characteristics, developing gradually over time. During the progression from mild to severe drought, water demand gradually increases while available water gradually decreases, displaying significant spatiotemporal variability. Agriculture is particularly sensitive to drought and water scarcity; drought leads to insufficient soil moisture, limiting crop water absorption. At different stages of crop growth, including germination, growth, development, and maturity, drought and water scarcity can cause stunted growth, yellowing leaves, and in severe cases, complete crop failure. Therefore, irrigation becomes the primary source of agricultural water during drought development. However, the dynamic changes in agricultural water scarcity caused by drought complicate the supply and demand imbalance of agricultural water resources. Therefore, how to achieve dynamic prediction of regional irrigation water demand based on conventional irrigation systems is a pressing technical problem that needs to be solved in this field. Summary of the Invention

[0005] To address the problems existing in the prior art and to overcome the difficulty in obtaining or lacking data on regional irrigation water demand under varying drought scenarios, this invention proposes a dynamic prediction method for irrigation water demand based on the water balance equation and crop water production function, in order to solve the aforementioned technical problems.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] This invention discloses a method for dynamically predicting irrigation water demand based on the water balance equation and crop water production function. The method includes the following steps:

[0008] Step 1: Multi-source data processing: A plain area was selected as the study area, and multi-source data for the study area were acquired, including remote sensing data, reanalysis data, and statistical data. The remote sensing data included raster-scale land use data, crop distribution data, actual evapotranspiration data, and crop yield data; the reanalysis data included raster-scale precipitation data and soil water storage data; the statistical data included historical yield data, agricultural water consumption, farmland irrigation water consumption data, and irrigation water consumption data for various cities and prefectures; the spatiotemporal resolution of the actual evapotranspiration data, precipitation data, and soil water storage data was unified, and the agricultural irrigation water consumption coefficient and irrigation water consumption coefficient for various cities and prefectures were determined based on the statistical data.

[0009] Step 2: Calculate net irrigation water volume based on the water balance equation: Using spatiotemporally continuous actual evapotranspiration data, precipitation data, and soil water storage data, calculate the net irrigation water volume of farmland in the study area according to the water balance equation at a set step size, thereby obtaining the net irrigation water volume at the annual and monthly scales; Combine the irrigation water consumption coefficient and irrigation water consumption coefficient to convert the agricultural water consumption of various cities in the statistical data into farmland irrigation water consumption, i.e., net irrigation water volume, and analyze the spatiotemporal dynamic changes of net irrigation water volume at the monthly and kilometer scales after comparing and verifying the calculation results.

[0010] The water balance equation is as follows:

[0011] (1)

[0012] In the formula, S i P represents the root zone soil water storage in mm during the i-th time period of the calculation process, where the time period corresponds to the set step size. i ET a_i R i and Irri i These represent precipitation, actual evapotranspiration, runoff, and net irrigation water volume for the same period, all in mm. In plain areas where farmland is flat, the impact of runoff on irrigation is ignored; therefore, the net irrigation water volume is calculated as follows:

[0013] (2)

[0014] Step 3: Determine evapotranspiration demand based on the crop water production function: Based on the crop water production function recommended in the FAO manual, quantify the relationship between yield and evapotranspiration during the crop growing season, and then calculate the changes in evapotranspiration demand during the crop growing season under different yield levels. Specific steps include:

[0015] S31, based on historical crop yield Y and evapotranspiration ET sequences over many years ( (year), calculate the multi-year average (year) , ) and extract the maximum value ( , ):

[0016] (5)

[0017] (6)

[0018] (7)

[0019] (8)

[0020] S32. The specific formula for calculating the crop water production function is as follows:

[0021] (9)

[0022] In the formula, Ky is the crop yield response coefficient, which is obtained by fitting the crop water production function or by looking it up in the manual; This represents actual crop yield, in kg / ha; The evapotranspiration water requirement during the crop growth period corresponding to the crop yield, in mm;

[0023] S33, Average annual yield As a benchmark, different production targets (100%) are set. 90% 80% 70% );

[0024] S34. Based on the constructed crop water production function, substitute different yield targets into the crop water production function to calculate crop evapotranspiration under different yield targets. ET 90% ET 80% ET 70% ), of which the multi-year average yield (100%) The default crop evapotranspiration water requirement under the target is: ET 90% ET 80% and ET 70% 90% respectively 80% and 70% Substitute the evapotranspiration obtained from the crop water production function;

[0025] Step 4: Determining precipitation thresholds for different drought levels: Based on the multi-year precipitation series compiled in Step 1, calculate the historical multi-year average results and standard deviation for each month, and use the standardized precipitation index (SPI) to calculate the precipitation thresholds for different drought levels at the raster scale.

[0026] The specific formula for calculating the Standardized Precipitation Index (SPI) is as follows:

[0027] (10)

[0028] In the formula, P rcp This refers to monthly precipitation. σ represents the multi-year average monthly precipitation, and σ is the standard deviation of the multi-year monthly precipitation series.

[0029] Step 5: Regional irrigation water demand prediction under different drought scenarios: Based on the evapotranspiration water demand calculated in Step 3, calculate the change in evapotranspiration relative to the multi-year average at different yield levels. Similarly, based on the precipitation thresholds for different drought levels in step 4, calculate the change in precipitation relative to the historical multi-year average. Assuming that soil water storage remains constant under different scenarios, the water balance theory is used to dynamically predict changes in crop irrigation water demand under different drought scenarios. :

[0030] (11)

[0031] In the formula, and Evapotranspiration at different production levels ( ET 90% ET 80% ET 70% ) and precipitation at different drought levels (P D1 P D2 P D3 P D4 ) relative to the historical multi-year average ( and The difference between 16 possible scenarios; n represents a total of 16 possible combinations.

[0032] Then, based on the changes in irrigation water demand during the crop growth period... and the average annual irrigation water volume to obtain the finally predicted regional crop irrigation water requirement :

[0033] (12)

[0034] In the formula, the average annual irrigation water volume is the historical average annual net irrigation water volume.

[0035] Furthermore, the irrigation water consumption coefficient mentioned in step 1 is the ratio of the farmland irrigation water consumption to the agricultural water consumption; the irrigation water loss coefficient is the ratio of the farmland irrigation water loss to the irrigation water consumption.

[0036] Furthermore, for the calculation of the net irrigation water volume in step 2, two sets of limiting conditions are defined:

[0037] 1) When the precipitation in a certain period meets the crop evapotranspiration demand in that period, Irri is set to 0, that is:

[0038] (3)

[0039] 2) Considering the influence of the continuous effect of rainstorm infiltration on the irrigation in the next period, if the cumulative precipitation in two consecutive periods can meet the crop evapotranspiration demand in the subsequent two periods, irrigation is not required, that is:

[0040] (4)

[0041] In the formula, represents the precipitation in the (i - 1)-th period, mm; represents the actual evapotranspiration in the (i + 1)-th period, mm.

[0042] Furthermore, the different drought degrees in step 4 include: mild drought D1, moderate drought D2, severe drought D3, and extreme drought D4;

[0043] The drought level division based on SPI and the precipitation threshold are respectively:

[0044] 1) Mild drought D1, -1.3 < SPI ≤ -0.8, and the precipitation threshold is P D1 ;

[0045] 2) Moderate drought D2, -1.6 < SPI ≤ -1.3, and the precipitation threshold is P D2 ;

[0046] 3) Severe drought D3, -2.0 < SPI ≤ -1.6, and the precipitation threshold is P D3 ;

[0047] 4) Severe drought D4, SPI≤-2.0, precipitation threshold is P D4 .

[0048] The beneficial effects of this invention are as follows: The method described in this invention estimates the spatiotemporal dynamic changes of farmland irrigation water at a regional scale, determines crop evapotranspiration demand under different yield levels by combining it with crop water production functions, and then predicts the spatiotemporal changes of irrigation water demand under different drought conditions based on the principle of water balance. Compared with traditional irrigation systems, this method considers the potential impact of insufficient water on crop yield, while distinguishing the spatiotemporal differences in irrigation water demand, providing effective technical support for dynamic prediction of agricultural irrigation water use and meeting the requirements of agricultural drought prevention. Specifically, it includes the following two points:

[0049] 1. In the past, farmland irrigation water use was mainly recorded at a specific point through canal systems or field measurement facilities, while statistical data mainly focused on irrigation water use within an annual administrative region. Agricultural water resource management still lacks high-resolution agricultural irrigation water volume data at a regional scale. This invention uses multi-source remote sensing data and reanalysis data to estimate the spatiotemporally high-resolution net irrigation water volume of farmland in plain areas using a water balance equation, providing a data foundation for accurately quantifying the spatiotemporal variations of agricultural irrigation water and promoting efficient water resource utilization.

[0050] 2. The method described in this invention further incorporates a crop water production function, using the quantitative relationship between crop water consumption and yield during the crop growth period as a basis to quantify changes in crop evapotranspiration water demand under different yield levels, and calculates the changes in precipitation deficit caused by different degrees of meteorological drought based on a standardized precipitation index. Compared to previous methods that only characterize static crop irrigation water demand based on the difference between potential crop evapotranspiration water demand and precipitation, the method described in this invention considers the potential impact of different water conditions on yield, realizing dynamic prediction of crop irrigation water demand at the regional scale. This provides scientific and technological support for major national needs such as emergency scheduling of agricultural water resources and drought relief, and has significant value for widespread application.

[0051] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0052] Figure 1 This is a schematic diagram of the method flow described in this invention;

[0053] Figure 2 A comparison of historical irrigation water consumption statistics and net irrigation water volume estimates for various cities and prefectures in the central and southern Hebei region;

[0054] Figure 3 The relationship between evapotranspiration and yield for winter wheat and summer maize fitted according to the crop water production function;

[0055] Figure 4The monthly dynamic forecast results of irrigation water demand under different yield targets (100%, 90%, 80%, and 70% of average yield) in mild to severe drought scenarios (winter wheat for February-May and summer maize for June-September). Detailed Implementation

[0056] This invention discloses a method for dynamically predicting irrigation water demand based on the water balance equation and crop water production function, such as... Figure 1 As shown, it includes the following steps:

[0057] Step 1, Multi-source data processing:

[0058] The study area was selected as a plain region, and multi-source data were acquired, including remote sensing data, reanalysis data, and statistical data. Remote sensing data included raster-scale land use data, crop distribution data, actual evapotranspiration data, and crop yield data; reanalysis data included raster-scale precipitation data and soil water storage data; and statistical data included historical yield data, agricultural water consumption, farmland irrigation water consumption data, and irrigation water consumption data for each city. The spatiotemporal resolution of the actual evapotranspiration data, precipitation data, and soil water storage data was standardized, and the agricultural irrigation water consumption coefficient and irrigation water consumption coefficient for each city were determined based on the statistical data. The irrigation water consumption coefficient is the ratio of statistically recorded farmland irrigation water consumption to agricultural water consumption, and irrigation water consumption is the difference between gross water consumption and the amount of water returning to surface water and groundwater; therefore, the irrigation water consumption coefficient is the ratio of statistically recorded farmland irrigation water consumption to irrigation water consumption.

[0059] Step 2, calculate the net irrigation water volume based on the water balance equation:

[0060] Using spatiotemporally continuous data on precipitation, soil water storage, and actual evapotranspiration, the net irrigation water volume of farmland in the study area was calculated every 8 days according to the water balance equation, thus obtaining the net irrigation water volume on an annual and monthly scale. The historical multi-year average net irrigation water volume can be used as the irrigation water demand under normal conditions and as a benchmark input for assessing irrigation water demand under different drought scenarios. Combining the irrigation water consumption coefficient and the irrigation water consumption coefficient, the statistical agricultural water consumption of various cities was converted into farmland irrigation water consumption (i.e., net irrigation water volume), and the spatiotemporal dynamic changes of net irrigation water volume on a monthly and kilometer scale were analyzed after comparison and verification with the calculation results.

[0061] The farmland water balance equation is as follows:

[0062] (1)

[0063] In the formula, S i P represents the root zone soil water storage (mm) at the i-th time period in the calculation process, where the time period refers to each time period or step calculated with an 8-day time step; iET a_i R i and Irri i This represents the precipitation (mm), actual evapotranspiration (mm), runoff (mm), and net irrigation water volume (mm) within the same time period. In plain areas, farmland is flat, and the impact of runoff on irrigation is ignored; the net irrigation water volume can be calculated as follows:

[0064] (2)

[0065] Taking into account actual irrigation water usage, two sets of constraints were defined to reduce errors caused by data uncertainty.

[0066] 1) When the precipitation in a certain period meets the evapotranspiration demand of crops in that period, Irri is set to 0.

[0067] (3)

[0068] 2) At the same time, the continuous effect of rainstorm infiltration will affect the irrigation of the next period. If the cumulative rainfall of the two consecutive periods can meet the evapotranspiration demand of the crops in the following two periods, then irrigation is not required.

[0069] (4)

[0070] In the formula, This represents the precipitation in the (i-1)th time period, in mm; This represents the actual evapotranspiration during the (i+1)th time period, in mm.

[0071] Step 3: Determine evapotranspiration demand based on crop water production function:

[0072] Based on the crop water production function recommended in FAO IRRIGATION AND DRAINAGE PAPER 33 (Cropyield response to water), the relationship between yield and evapotranspiration during the crop growing season is quantified, and then the changes in evapotranspiration demand during the crop growing season under different yield levels are calculated. Specific steps include:

[0073] S31, based on historical crop yield Y and evapotranspiration ET sequences over many years ( (year), calculate the multi-year average (year) , ) and extract the maximum value ( , ):

[0074] (5)

[0075] (6)

[0076] (7)

[0077] (8)

[0078] S32. The specific formula for calculating the crop water production function is as follows:

[0079] (9)

[0080] In the formula, Ky is the crop yield response coefficient, which is obtained by fitting the crop water production function or by looking up the formula in the FAO-33 manual. Actual crop yield (kg / ha); The evapotranspiration water requirement (mm) during the crop growth period corresponds to the crop yield. and For potential maximum yield and evapotranspiration, this invention uses the highest crop yield and evapotranspiration in historical periods.

[0081] S33, Average annual yield As a benchmark, different production targets (100%) are set. 90% 80% 70% );

[0082] S34. Based on the constructed crop water production function, substitute different yield targets into the crop water production function to calculate crop evapotranspiration under different yield targets. ET 90% ET 80% ET 70% ), of which the multi-year average yield (100%) The default crop evapotranspiration water requirement under the target is: ET 90% ET 80% and ET 70% 90% respectively 80% and 70% Substitute the evapotranspiration obtained from the crop water production function.

[0083] Step 4: Determine the precipitation thresholds for different drought levels:

[0084] Based on the multi-year precipitation series compiled in Step 1, the historical multi-year average and standard deviation for each month were calculated. The standardized precipitation index (SPI) was used to calculate the precipitation threshold (P) at different drought levels (mild drought D1, moderate drought D2, severe drought D3, and extreme drought D4) at the raster scale.D1 P D2 P D3 P D4 ).

[0085] The specific formula for calculating the Standardized Precipitation Index (SPI) is as follows:

[0086] (10)

[0087] In the formula, P rcp This refers to monthly precipitation. σ represents the multi-year average monthly precipitation, and σ is the standard deviation of the multi-year monthly precipitation series.

[0088] Based on the SPI calculation formula and drought level classification, calculate the corresponding precipitation thresholds (P) for different drought levels. D1 P D2 P D3 P D4 (as shown in Table 1).

[0089] Table 1. Drought Level Classification and Precipitation Thresholds Based on SPI

[0090]

[0091] Step 5, Predicting regional irrigation water demand under different drought scenarios:

[0092] Based on the evapotranspiration water demand calculated in step 3, calculate the change in evapotranspiration relative to the multi-year average at different production levels. Similarly, based on the precipitation thresholds for different drought levels in step 4, the change in precipitation relative to the historical multi-year average is calculated. Assuming that soil water storage remains constant across different scenarios, water balance theory is used to dynamically predict changes in crop irrigation water demand under different drought conditions. :

[0093] (11)

[0094] In the formula, and Evapotranspiration at different production levels ( ET 90% ET 80% ET 70% ) and precipitation at different drought levels (P D1 P D2 P D3 P D4 ) relative to the historical multi-year average ( and The difference between 16 and 16; n represents a total of 16 possible combinations.

[0095] Then, based on the changes in irrigation water demand during the crop growth period... and average annual irrigation water volume The final predicted regional crop irrigation water demand was obtained. :

[0096] (12)

[0097] In the formula, the average annual irrigation water volume is... This refers to the historical average net irrigation water volume over many years.

[0098] Example 1

[0099] This embodiment is a specific application example of the above method. Taking the main crop-growing area of ​​Hebei Province, namely the South Central Plain of Hebei, as the study area, this embodiment quantifies the spatiotemporal high-resolution changes of actual irrigation water volume in farmland in the South Central Plain of Hebei during historical periods and dynamically predicts the changes in irrigation water demand under different drought levels.

[0100] This embodiment discloses a method for dynamic prediction of irrigation water demand based on the water balance equation and crop water production function, including the following steps:

[0101] Step 1, Multi-source data processing:

[0102] Data on land use and spatial distribution of staple crops obtained through remote sensing inversion in the central and southern Hebei region were compiled. Farmland and winter wheat and summer maize planting areas were extracted. Further reanalysis data on regional precipitation and soil water storage, as well as remote sensing evapotranspiration data, were acquired for farmland in the central and southern Hebei region. Historical statistical data on yield, agricultural irrigation water use, and irrigation water consumption were collected from various cities in the central and southern Hebei region. Raster data on precipitation, soil water storage, and actual evapotranspiration at different resolutions were resampled to a 5km resolution, with the time step standardized to 8 days based on the maximum time interval of the actual evapotranspiration data. Specific information regarding different data sources and resolutions is shown in Table 2.

[0103] Table 2. Different Data Sources and Resolutions

[0104]

[0105] Step 2, calculate the net irrigation water volume based on the water balance equation:

[0106] Calculate the net irrigation water volume for farmland in the central and southern Hebei region using the water balance equation:

[0107] (1)

[0108] Because the terrain in the southern Hebei region is relatively flat, the impact of runoff on irrigation can be ignored. Therefore, the calculation of net irrigation water volume can be expressed as follows:

[0109] (2)

[0110] Considering actual irrigation water conditions, two sets of constraints were defined to reduce errors caused by data uncertainty. Irri is set to 0 when the rainfall in a given period meets the crop's evapotranspiration requirement for that period.

[0111] (3)

[0112] At the same time, considering the continuous effect of rainstorm infiltration on irrigation in the next stage, if the cumulative rainfall in two consecutive periods can meet the evapotranspiration demand of crops in the following two periods, then irrigation is not required.

[0113] (4)

[0114] Based on the aforementioned water balance equation, the high-resolution spatiotemporal net irrigation water volume per 5 km every 8 days in the southern Hebei region from 2001 to 2022 was estimated. To verify the accuracy of the estimation results, the calculated net irrigation water volume was summed year by year, and the total net irrigation water volume (in billions of cubic meters) for the region was calculated based on the farmland area over the years. This was then compared with the irrigation water consumption statistics in the water resources bulletin. Figure 2 As shown. According to the definition in the Water Resources Bulletin, irrigation water consumption refers to the difference between gross water consumption and the amount of water returning to surface water and groundwater, which is equivalent to the net irrigation water consumption calculated in this study. The Hebei Provincial Water Resources Bulletin only published agricultural water consumption data for various cities in 2002, 2003, 2005, 2020, 2021, and 2022. The linear regression equation fitted by this method and the statistical data is... , This indicates that the estimation results are highly accurate. Among them, the irrigation water consumption coefficient can be obtained from the ratio of farmland irrigation water consumption to irrigation water consumption (i.e., gross irrigation water volume) in the bulletin, and is used for the conversion between net irrigation water volume and gross irrigation water volume.

[0115] The estimated spatiotemporal high-resolution net irrigation water volume was analyzed according to the main crop growing season (June-September for summer maize and October-May of the following year for winter wheat). The peak irrigation volume during the summer maize growing season occurred in June, with an average of 36.97 mm, accounting for approximately 49.92% of the total. Irrigation water volume decreased significantly thereafter as precipitation increased; August saw the lowest irrigation water volume, but in drought years (such as 2002 and 2015), some areas still required irrigation. The peak irrigation periods during the winter wheat growing season were mainly in April and May, with averages of 21.20 mm and 36.40 mm, respectively, accounting for 23.25% and 39.93% of the total net irrigation water volume. Irrigation water volume in winter (December and January) was close to zero. This is partly due to the possibility of some irrigation in certain years or regions, and partly due to uncertainties in precipitation and soil moisture data caused by winter snowfall or freezing. Spatially, farmland irrigation water volume was mainly concentrated in the winter wheat-summer maize rotation planting area. During the summer maize season, irrigation water volume was relatively high in the central plains, with net irrigation water exceeding 75 mm in Hengshui, Xingtai, and Handan, cities with lower rainfall. During the winter wheat season, the spatial variation in net irrigation water volume was more significant, with higher irrigation water volume concentrated in the central and southeastern regions, exceeding 125 mm.

[0116] Step 3: Determine evapotranspiration demand based on crop water production function:

[0117] Based on the crop water production function recommended in the FAO-33 manual, the relationship between yield and evapotranspiration during the crop growing season is quantified, and then the changes in evapotranspiration demand during the crop growing season under different yield levels are estimated. The specific calculation steps are as follows:

[0118] S31, based on historical multi-year crop yield Y and evapotranspiration ET sequences during the growing season ( (year), calculate the multi-year average (year) , ) and extract the maximum value ( , ):

[0119] (5)

[0120] (6)

[0121] (7)

[0122] (8)

[0123] S32, the specific formula for calculating the crop water production function is as follows:

[0124] (9)

[0125] In the formula, Ky is the crop yield response coefficient; Actual crop yield (kg / ha); This represents the cumulative evapotranspiration (mm) during the crop's growing season.

[0126] Based on historical gridded yield data of winter wheat and summer maize in the central and southern Hebei region and actual evapotranspiration sequences during crop growth periods, crop water production functions were fitted, yielding yield response coefficients Ky of 1.25 and 0.49 for maize and winter wheat, respectively. Figure 3 As shown, the fitted Ky value for maize is consistent with the recommended value in the FAO-33 manual; in this embodiment, the evapotranspiration of winter wheat during its growth period is accumulated from the greening stage in February to the harvest period, and the fitted Ky value is very close to the recommended Ky value (0.50) for the mid-growth stage of winter wheat. Therefore, the fitted Ky values ​​for the two main crops in southern Hebei can be used to estimate crop evapotranspiration water requirements.

[0127] S33, average annual production As a benchmark, different production targets (100%) are set. 90% 80% 70% );

[0128] S34, Based on the constructed crop water production function, calculate the crop evapotranspiration water requirement under different yield targets. ET 90% ET 80% ET 70% ), of which the multi-year average yield (100%) The default crop evapotranspiration water requirement under the target is: ET 90% ET 80% and ET 70% 90% respectively 80% and 70% Substitute the evapotranspiration obtained from the crop water production function.

[0129] Step 4: Determine the precipitation thresholds for different drought levels:

[0130] Based on the CHIRPS Daily precipitation sequence data compiled in step 1, the historical multi-year average results and standard deviations for each month were calculated. The standardized precipitation index SPI was used to back-infer the precipitation thresholds at different drought levels (mild drought D1, moderate drought D2, severe drought D3, and extreme drought D4) at the raster scale.

[0131] (10)

[0132] In the formula, P rcp This refers to monthly precipitation. σ represents the multi-year average monthly precipitation, and σ is the standard deviation of the multi-year monthly precipitation series.

[0133] Based on the SPI calculation formula and drought level classification, calculate the corresponding precipitation thresholds (P) for different drought levels. D1 P D2 P D3 P D4 ).

[0134] Step 5, Predicting regional irrigation water demand under different drought scenarios:

[0135] Based on the evapotranspiration water demand calculated in step 3, calculate the change in evapotranspiration relative to the multi-year average at different production levels. Similarly, based on the precipitation thresholds for different drought levels in step 4, the change in precipitation relative to the historical multi-year average is calculated. Assuming that soil water storage remains constant across different scenarios, water balance theory is used to dynamically predict changes in crop irrigation water demand under different drought conditions. :

[0136] (11)

[0137] In the formula, and Evapotranspiration at different production levels ( ET 90% ET 80% ET 70% ) and precipitation at different drought levels (P D1 P D2 P D3 P D4 ) relative to the historical multi-year average ( and The difference between 16 and 16; n represents a total of 16 possible combinations.

[0138] Then, based on the changes in irrigation water demand during the crop growth period... and average annual irrigation water volume The final predicted regional crop irrigation water demand was obtained. :

[0139] (12)

[0140] In the formula, the average annual irrigation water volume is... This refers to the historical average net irrigation water volume over many years.

[0141] When a drought occurs in a given month, the irrigation water requirement for crops in that month is dynamically predicted based on the severity of the drought (from mild to severe) and the crop yield protection target. Figure 4 As shown, the irrigation water requirements for winter wheat are predicted from February to May, and for maize from June to September. Under the same yield target, the irrigation water requirements for both winter wheat and summer maize in the southern Hebei region increase with the severity of drought. If the average yield target for winter wheat is set, the irrigation water volume in May under mild, moderate, severe, and extreme drought scenarios are 70.8, 77.1, 80.9, and 85.9 mm, respectively. Summer maize receives more rainfall during its growing season. If drought occurs, the significantly reduced rainfall leads to higher irrigation water volume, with the highest volume in July. Furthermore, under the same drought scenario, the irrigation water volume gradually decreases as the yield guarantee target decreases from 100% to 70% of the average yield. For example, when winter wheat experiences mild drought in May, the irrigation water volume gradually decreases from 70.8 mm to 20.0 mm.

[0142] Considering that actual irrigation in the central and southern Hebei region mainly occurs during the vigorous growth stage of winter wheat and before or during the seedling stage of summer maize sowing, the spatial distribution of irrigation water under different yield guarantee levels was further analyzed when a severe drought occurred during this period (March-June). In March, as the yield target was reduced from 100% to 70% of the average yield, the irrigation water for winter wheat gradually decreased from about 20 mm to below 10 mm, and the irrigated area decreased significantly. Meanwhile, from March to May, as winter wheat grew to maturity, both the irrigated area and the irrigation water gradually increased, with the Shijiazhuang area in the central region having significantly higher irrigation water than other areas. The spatial distribution of irrigation water in June was similar to that in May, but the area was larger, with an overall rainfall exceeding 60 mm. The Shijiazhuang area had an overall irrigation water rainfall exceeding 100 mm, followed by the Baoding area in the north, which also had relatively high irrigation water.

[0143] Finally, it should be noted that the above description is only used to illustrate the technical solutions of the present invention and is not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention.

Claims

1. A method for dynamic prediction of irrigation water demand based on water balance equations and crop water production functions, characterized in that, The method includes the following steps: Step 1: Multi-source data processing: A plain area was selected as the study area, and multi-source data for the study area were acquired, including remote sensing data, reanalysis data, and statistical data. The remote sensing data included raster-scale land use data, crop distribution data, actual evapotranspiration data, and crop yield data; the reanalysis data included raster-scale precipitation data and soil water storage data; the statistical data included historical yield data, agricultural water consumption, farmland irrigation water consumption data, and irrigation water consumption data for various cities and prefectures; the spatiotemporal resolution of the actual evapotranspiration data, precipitation data, and soil water storage data was unified, and the agricultural irrigation water consumption coefficient and irrigation water consumption coefficient for various cities and prefectures were determined based on the statistical data. Step 2: Calculate net irrigation water volume based on the water balance equation: Using spatiotemporally continuous actual evapotranspiration data, precipitation data, and soil water storage data, calculate the net irrigation water volume of farmland in the study area according to the water balance equation at a set step size, thereby obtaining the net irrigation water volume at the annual and monthly scales; Combine the irrigation water consumption coefficient and the irrigation water consumption coefficient to convert the statistical agricultural water consumption of various cities into farmland irrigation water consumption, i.e., net irrigation water volume, and analyze the spatiotemporal dynamic changes of net irrigation water volume at the monthly and kilometer scales after comparing and verifying the calculation results. Step 3: Determine evapotranspiration demand based on the crop water production function: Based on the crop water production function recommended in the FAO manual, quantify the relationship between yield and evapotranspiration during the crop growing season, and then calculate the changes in evapotranspiration demand during the crop growing season under different yield levels; specific steps include: S31, based on historical crop yield Y and evapotranspiration ET sequences over many years Calculate the multi-year average. and Extract the maximum value and : （5） （6） （7） （8） S32. The specific formula for calculating the crop water production function is as follows: （9） In the formula, Ky is the crop yield response coefficient, which is obtained by fitting the crop water production function or by looking it up in the manual; This represents actual crop yield, in kg / ha; The evapotranspiration water requirement during the crop growth period corresponding to the crop yield, in mm; S33, Average annual yield As a benchmark, different production targets are set, including 100%. 90% 80% 70% ; S34. Based on the constructed crop water production function, substitute different yield targets into the crop water production function to calculate the crop evapotranspiration under different yield targets. ET 90% ET 80% ET 70% The average annual yield is 100%. The default corresponding crop evapotranspiration water requirement under the target is ET 90% ET 80% and ET 70% 90% respectively 80% and 70% Substitute the evapotranspiration obtained from the crop water production function; Step 4: Determining precipitation thresholds for different drought levels: Based on the multi-year precipitation series compiled in Step 1, calculate the historical multi-year average results and standard deviation for each month, and use the standardized precipitation index (SPI) to calculate the precipitation thresholds for different drought levels at the raster scale. The specific formula for calculating the Standardized Precipitation Index (SPI) is as follows: （10） In the formula, P rcp This refers to monthly precipitation. σ represents the multi-year average monthly precipitation, and σ is the standard deviation of the multi-year monthly precipitation series. Step 5: Regional irrigation water demand prediction under different drought scenarios: Based on the evapotranspiration water demand calculated in Step 3, calculate the change in evapotranspiration relative to the multi-year average at different yield levels. Similarly, based on the precipitation thresholds for different drought levels in step 4, calculate the change in precipitation relative to the historical multi-year average. Assuming that soil water storage remains constant under different scenarios, the water balance theory is used to dynamically predict changes in crop irrigation water demand under different drought scenarios. : （11） In the formula, and Evapotranspiration at different production levels ET 90% ET 80% ET 70% Precipitation P at different levels of drought D1 P D2 P D3 P D4 relative to historical multi-year average and The difference; n represents a total of 16 possible combinations; Then, based on the changes in irrigation water demand during the crop growth period... and average annual irrigation water volume The final predicted regional crop irrigation water demand was obtained. : （12） In the formula, the average annual irrigation water volume is... This refers to the historical average net irrigation water volume over many years.

2. The method for dynamic prediction of irrigation water demand based on water balance equation and crop water production function according to claim 1, characterized in that, The irrigation water consumption coefficient mentioned in step 1 is the ratio of farmland irrigation water consumption to agricultural water consumption; the irrigation water consumption coefficient is the ratio of farmland irrigation water consumption to irrigation water consumption.

3. The method for dynamic prediction of irrigation water demand based on water balance equation and crop water production function according to claim 1, characterized in that, The water balance equation in step 2 is: （1） In the formula, S i P represents the root zone soil water storage in mm during the i-th time period of the calculation process, where the time period corresponds to the set step size. i ET a_i R i and Irri i These represent precipitation, actual evapotranspiration, runoff, and net irrigation water volume for the same period, all in mm. In plain areas where farmland is flat, the impact of runoff on irrigation is ignored; therefore, the net irrigation water volume is calculated as follows: （2） For the calculation of net irrigation water volume, two sets of constraints are specified: 1) When the precipitation in a certain period meets the evapotranspiration demand of crops in that period, Irri is set to 0, that is: （3） 2) Considering the sustained effect of rainstorm infiltration on irrigation in the next period, if the cumulative rainfall over two consecutive periods is sufficient to meet the crop evapotranspiration requirements for the following two periods, then irrigation is unnecessary. （4） In the formula, This represents the precipitation in the (i-1)th time period, in mm; This represents the actual evapotranspiration during the (i+1)th time period, in mm.

4. The method for dynamic prediction of irrigation water demand based on water balance equation and crop water production function according to claim 1, characterized in that, The different drought levels in step 4 include: mild drought D1, moderate drought D2, severe drought D3, and extreme drought D4; The drought level classification and precipitation threshold based on SPI are as follows: 1) Mild drought D1, -1.3 < SPI ≤ -0.8, and the precipitation threshold is P D1 ; 2) Moderate drought D2, -1.6 < SPI ≤ -1.3, and the precipitation threshold is P D2 ; 3) Severe drought D3, -2.0 < SPI ≤ -1.6, and the precipitation threshold is P D3 ; 4) Severe drought D4, SPI≤-2.0, precipitation threshold is P D4 .

Images