Natural grassland grazing duration prediction method and device based on remote sensing technology

Abstract

This application discloses a method and device for predicting the grazing duration of natural grasslands based on remote sensing technology. The prediction method includes obtaining the daily forage intake based on the existing aboveground biomass and potential aboveground biomass of the grassland; and obtaining the grazing duration based on the daily forage intake and the total amount of edible forage within a preset future period. The method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in this application can effectively predict the grazing duration of grasslands by accurately calculating the daily forage intake and the total amount of edible forage within a preset future period. Through scientific assessment of the existing and potential aboveground biomass of the grassland, it can ensure the rational utilization of forage resources, avoid grassland degradation caused by overgrazing, and maximize the productive potential of the grassland.

Description

Technical Field

[0001] This application relates to the field of ecological remote sensing, and in particular to a method and device for predicting the grazing duration of natural grasslands based on remote sensing technology. Background Technology

[0002] Grasslands, as a national strategic resource, play a vital role in maintaining ecological security and promoting national economic development. Implementing a grassland-livestock balance management system and rationally regulating grassland utilization intensity can protect grassland ecological health. The core of this system lies in accurately understanding the matching relationship between grassland edible forage production and livestock demand, which requires a precise understanding of grassland natural growth, existing biomass, and actual livestock consumption. However, traditional methods struggle to obtain near-real-time, accurate data on existing and potential aboveground biomass when faced with grazing disturbances, thus failing to determine grazing duration and consequently hindering the effective implementation of the grassland-livestock balance management system. Summary of the Invention

[0003] The purpose of this application is to provide a method and device for predicting the grazing time of natural grasslands based on remote sensing technology. This method can effectively predict the grazing time of natural grasslands, thereby promoting the implementation of a grass-livestock balance management system that uses grazing time as a lever.

[0004] To achieve the above objectives, embodiments of this application provide a method for predicting the grazing duration of natural grasslands based on remote sensing technology, including:

[0005] Daily forage intake was obtained based on the existing aboveground biomass and potential aboveground biomass of grassland. The existing aboveground biomass was obtained by inputting the leaf area index dataset, normalized vegetation index, and remotely sensed surface reflectance into the existing biomass remote sensing inversion model. The potential aboveground biomass was obtained by inputting digital elevation model data, slope topographic factor data, slope aspect topographic factor data, total precipitation, and 2-meter ground temperature data into the potential biomass remote sensing estimation model.

[0006] The grazing duration is determined based on the daily forage intake and the total amount of edible forage within a predetermined time period. The total amount of edible forage within the predetermined time period is obtained based on the rational utilization rate of natural grassland and the potential aboveground biomass of grassland within the predetermined time period. The potential aboveground biomass of grassland within the predetermined time period is obtained by inputting digital elevation model data, slope topographic factor data, slope aspect topographic factor data, total precipitation within the predetermined time period, and the predicted 2-meter ground temperature within the predetermined time period into the potential biomass remote sensing estimation model.

[0007] Optionally, the training process for the existing biomass remote sensing inversion model is as follows:

[0008] Load the preprocessed first training dataset. The first training dataset includes first feature data and first label data. The first feature data includes leaf area index dataset, remote sensing surface reflectance and normalized vegetation index. The first label data includes measured values ​​of aboveground biomass.

[0009] The random forest model was trained using the first training dataset to obtain an existing biomass remote sensing inversion model, which includes:

[0010] Preset the number of decision trees to be created in the random forest model;

[0011] For each decision tree, a predetermined number of features are randomly selected from all features to form the first feature subset considered during the training of the decision tree;

[0012] Use the bootstrap method to draw a first subset of samples with replacement from the first training dataset, which is the same size as the first training dataset.

[0013] Train a decision tree using the first feature subset and the first sample subset;

[0014] Repeat the above process until a predetermined number of decision trees are constructed, thus obtaining the existing biomass remote sensing inversion model.

[0015] Optionally, the leaf area index dataset and surface reflectance are obtained by image stitching, resampling, and reprojection processing of data collected by a medium-resolution imaging spectrometer; the normalized vegetation index is calculated using remote sensing reflectance in the red band and the near-infrared band.

[0016] The measured value of the aboveground biomass is:

[0017]

[0018] In the formula, AGB represents the measured aboveground biomass, n represents the total number of quadrats, and AGB i This represents the aboveground biomass dry weight of the i-th grassland population in a certain sample plot.

[0019] Optionally, the training process for the potential biomass remote sensing estimation model is as follows:

[0020] Load the preprocessed second training dataset, which includes second feature data and second label data. The second feature data includes digital elevation model data, slope topographic factor data, aspect topographic factor data, total precipitation and ground 2-meter temperature data. The second label data includes the existing aboveground biomass of unused grassland and hayfields.

[0021] The random forest model was trained using the second training dataset to obtain a potential biomass remote sensing estimation model, which includes:

[0022] Preset the number of decision trees to be created in the random forest model;

[0023] For each decision tree, a predetermined number of features are randomly selected from all features to form a second feature subset considered during the training of the decision tree;

[0024] Use the bootstrap method to draw a second subset of samples with replacement from the second training dataset, which is the same size as the second training dataset.

[0025] Train the decision tree using the second feature subset and the second sample subset;

[0026] Repeat the above process until a predetermined number of decision trees are constructed, thus obtaining a remote sensing estimation model for potential biomass.

[0027] Alternatively, the process of obtaining the second feature data is as follows:

[0028] Based on the original digital elevation model data, slope topographic factor data and aspect topographic factor data are calculated. The total daily precipitation is obtained by accumulating the total hourly precipitation data. The average daily temperature is obtained by averaging the 2-meter ground temperature data within each day.

[0029] The slope topographic factor data, aspect topographic factor data, daily total precipitation and daily average temperature were transformed and resampled to obtain digital elevation model data with 500-meter spatial resolution and latitude and longitude projection, slope topographic factor data, aspect topographic factor data, total precipitation and hourly ground temperature data at 2 meters.

[0030] Alternatively, the process of obtaining the second tag data is as follows:

[0031] Principal component transformation was performed on the time series data of normalized vegetation index, and the top 5 characteristic bands with the highest characteristics in the principal component analysis results were retained.

[0032] Based on the time series data of the first 5 feature bands, a multi-scale segmentation algorithm is used to identify grassland fences and obtain grassland fence boundary data.

[0033] Based on grassland fence boundary data, a mowing recognition algorithm is used to identify unused grassland and haymaking areas;

[0034] Based on the existing aboveground biomass of grassland output by the existing biomass remote sensing inversion model, the existing aboveground biomass of unused grassland and hayfield is obtained.

[0035] Optionally, the process for obtaining the total precipitation and the predicted ground temperature at 2 meters within the future preset time period is as follows:

[0036] By interpolating the weekly average precipitation and the weekly average ground temperature at 2 meters, the predicted daily average temperature and daily average ground temperature at 2 meters are obtained for the future preset time period.

[0037] The daily average temperature and the predicted daily average ground temperature at 2 meters are scaled for the future preset time period. The initial spatial resolution of 0.25° is converted to a spatial resolution of 500 meters, resulting in a spatial resolution of 500 meters, latitude and longitude projection, the total precipitation for each day in the future preset time period, and the predicted ground temperature at 2 meters.

[0038] This application also provides a device for predicting the grazing time of natural grassland, including:

[0039] The calculation module is configured to obtain the daily forage intake based on the existing aboveground biomass and the potential aboveground biomass of the grassland. The existing aboveground biomass is obtained by inputting the leaf area index dataset, normalized vegetation index, and remotely sensed surface reflectance into the existing biomass remote sensing inversion model. The potential aboveground biomass is obtained by inputting digital elevation model data, slope topographic factor data, slope aspect topographic factor data, total precipitation, and 2-meter ground temperature data into the potential biomass remote sensing estimation model.

[0040] The duration prediction module is configured to determine the grazing duration based on the daily forage intake and the total amount of edible forage within a preset future time period. The total amount of edible forage within the preset future time period is obtained based on the rational utilization rate of natural grassland and the potential aboveground biomass of grassland within the preset future time period. The potential aboveground biomass of grassland within the preset future time period is obtained by inputting digital elevation model data, slope topographic factor data, slope aspect topographic factor data, total precipitation within the preset future time period, and the predicted 2-meter ground temperature within the preset future time period into the potential biomass remote sensing estimation model.

[0041] This application also provides an electronic device, including a processor and a memory, wherein the memory stores an executable program, and the processor executes the executable program to perform the steps of the method as described in any of the preceding claims.

[0042] This application also provides a storage medium carrying one or more computer programs, which, when executed by a processor, implement the steps of any of the methods described above.

[0043] The method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in this application embodiment accurately calculates the daily forage intake of natural grasslands and the total amount of edible forage within a preset future period, and can effectively predict the grazing duration of grasslands.

[0044] The method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in this application supports the rational utilization and management of forage resources by scientifically assessing the existing and potential aboveground biomass of the grassland. This avoids grassland degradation caused by overgrazing and maximizes the productive potential of the grassland. Based on actual conditions such as the grassland's leaf area index and normalized difference vegetation index, as well as environmental factors such as topography, precipitation, and temperature, this method predicts the grazing duration, promoting a healthy cycle in the grassland ecosystem, helping to maintain grassland ecological balance, reduce soil erosion, and preserve biodiversity. Utilizing advanced remote sensing technology and data analysis models, this method provides a reliable tool for ranch managers to make data-driven management decisions. Attached Figure Description

[0045] Figure 1 A flowchart of a method for predicting the grazing duration of natural grasslands based on remote sensing technology, provided in an embodiment of this application;

[0046] Figure 2 A rendering showing the predicted grazing time in a certain area (pastoral and grassland areas) in September 2024, provided for an embodiment of this application.

[0047] Figure 3 A schematic diagram of the structure of the natural grassland grazing duration prediction device provided in the embodiments of this application;

[0048] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0049] Various embodiments and features of this application are described herein with reference to the accompanying drawings. It should be understood that various modifications can be made to the embodiments described herein. Therefore, the foregoing description should not be considered limiting, but merely as examples of embodiments. Other modifications within the scope and spirit of this application will be apparent to those skilled in the art.

[0050] The method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in this application accurately predicts the existing and potential biomass of large-scale grasslands using remote sensing technology. By combining meteorological data and historical grazing patterns, it achieves near real-time prediction of suitable grazing duration. The application of this method will provide strong support for implementing more flexible and precise grassland-livestock balance management, which will not only help protect and restore grassland ecosystems, but also promote the harmonious development of pastoral socio-economic conditions.

[0051] The following description, in conjunction with the accompanying drawings, details the method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in the embodiments of this application. Figure 1 One of the flowcharts for the method of predicting the grazing duration of natural grassland based on remote sensing technology provided in the embodiments of this application is shown below. Figure 1 As shown, the method includes the following steps:

[0052] S100. The daily forage intake is obtained based on the existing aboveground biomass and potential aboveground biomass of the grassland.

[0053] Specifically, daily forage intake = potential aboveground biomass of grassland - existing aboveground biomass of grassland.

[0054] The existing aboveground biomass of grassland is obtained by inputting the leaf area index dataset, remote sensing surface reflectance, and NDVI (Normalized Difference Vegetation Index) into the existing biomass remote sensing inversion model.

[0055] The potential aboveground biomass of grassland is obtained by inputting digital elevation model data, slope topographic factor data, aspect topographic factor data, total precipitation and ground temperature data at 2 meters into the potential biomass remote sensing estimation model.

[0056] In this embodiment, the existing biomass remote sensing inversion model is obtained by training a random forest model using a first training dataset. The first training dataset includes first feature data and first label data. The first feature data includes leaf area index data, remotely sensed surface reflectance, and normalized difference vegetation index (NDVI). The first label data includes measured aboveground biomass values.

[0057] Specifically, the process of obtaining the leaf area index dataset and the surface reflectance dataset is as follows:

[0058] Obtain the LAI (Leaf Area Index) product MOD15A3H and the reflectance product MOD43A4 from MODIS (Moderate Resolution Imaging Spectroradiometer).

[0059] Image stitching, resampling, and reprojection were performed on LAI product MOD15A3H and reflectance product MOD43A4 respectively to obtain MODIS LAI data with 500-meter spatial resolution and latitude and longitude projection of the target area and the surface reflectance on MODIS b band (i.e., blue light band to near-infrared band).

[0060] MODIS is a multifunctional medium-resolution imaging spectrometer that provides abundant data on the Earth's surface and atmosphere, and is widely used in various fields such as vegetation monitoring, surface temperature monitoring, ocean monitoring, atmospheric monitoring, fire monitoring, and snow and ice monitoring. Through various levels of data products, MODIS supports a wide range of scientific research and environmental monitoring work.

[0061] MODIS provides various levels of data products, including Level 0, Level 1, Level 2, Level 3, and Level 4 products. Level 0 products represent unprocessed raw data; Level 1 products represent data that has undergone preliminary processing, including geolocation and radiometric correction; Level 2 products represent data products that have undergone further processing, such as surface reflectance, surface temperature, and vegetation indices; Level 3 products represent data products synthesized from Level 2 products onto a unified spatiotemporal grid, typically used for long-term trend analysis; and Level 4 products represent model-based integrated products.

[0062] MODIS has 36 spectral bands, covering a wide wavelength range from visible light to thermal infrared. The spatial resolution varies across different bands, ranging from 250 meters (visible band) to 1000 meters (thermal infrared band). With a scan width of 2330 kilometers, MODIS can cover almost the entire Earth's surface daily. The orbital design of the Terra and Aqua satellites allows them to cover the same area multiple times a day, providing high temporal resolution data. Because MODIS data undergoes rigorous calibration and verification, its accuracy and consistency are ensured.

[0063] It's important to note that spatial resolution determines the actual ground area represented by each pixel in a remote sensing image. High resolution means each pixel represents a smaller ground area, and vice versa. For a spatial resolution of 500 meters, each pixel represents a ground area of ​​500 meters by 500 meters. In other words, for an image with a spatial resolution of 500 meters, each pixel in the image actually represents a square area of ​​500 meters by 500 meters.

[0064] Latitude and longitude projection is a method of mapping the three-dimensional coordinates (longitude and latitude) of the Earth's surface onto a two-dimensional plane. Its characteristic is that it directly uses the Earth's longitude and latitude values ​​to represent geographical locations.

[0065] Specifically, NDVI is a commonly used remote sensing index used to assess the greenness and health of surface vegetation. NDVI is calculated using reflectance in the red and near-infrared bands. The calculation process for NDVI is as follows:

[0066]

[0067] In equation (1), NIR represents the reflectivity of the near-infrared band, and Red represents the reflectivity of the red band.

[0068] In this embodiment, aboveground biomass can be investigated through quadrat setup, plot selection, and aboveground biomass measurement. Based on grassland field survey data, the measured aboveground biomass values ​​are obtained as follows:

[0069]

[0070] In formula (2), AGB represents the measured value of aboveground biomass, and its unit is g / m³. 2 n represents the total number of sample plots, AGB i This represents the aboveground biomass dry weight of the i-th grassland population in a given plot, expressed in g / m³. 2 .

[0071] In a specific embodiment, the process of training the random forest model using the first training dataset to obtain the existing biomass remote sensing inversion model is as follows:

[0072] Load the preprocessed first training dataset, which includes first feature data and first label data. The first feature data includes leaf area index dataset, remote sensing surface reflectance and normalized vegetation index, and the first label data includes measured values ​​of aboveground biomass.

[0073] The random forest model was trained using the first training dataset to obtain an existing biomass remote sensing inversion model, which includes:

[0074] Preset the number of decision trees to be created in the random forest model;

[0075] For each decision tree, a predetermined number of features are randomly selected from all features to form the first feature subset considered during the training of the decision tree;

[0076] Use the bootstrap method to draw a first subset of samples with replacement from the first training dataset, which is the same size as the first training dataset.

[0077] Train a decision tree using the first feature subset and the first sample subset;

[0078] Repeat the above process until a predetermined number of decision trees are constructed, thus obtaining the existing biomass remote sensing inversion model.

[0079] During training, the first feature data can be divided into a training set and a validation set in a 7:3 ratio. The performance of the existing biomass remote sensing inversion model is evaluated on the validation set, and the model parameters are adjusted based on the validation results to establish the final existing biomass remote sensing inversion model. Specifically, existing aboveground biomass can be expressed as an equation with the following input parameters:

[0080] AGB c =f c (LAI,NDVI,ρ b (3)

[0081] In equation (3), AGB c f represents the existing aboveground biomass of the grassland. c () represents a random forest model for retrieving existing aboveground biomass, ρ b This represents the remotely sensed surface reflectance on the MODIS b-band.

[0082] In the embodiments of this application, the potential aboveground biomass of grassland refers to the aboveground biomass that a natural grassland can achieve in an unused state under specific hydrothermal and topographical conditions.

[0083] The potential biomass remote sensing estimation model was obtained by training the random forest model using a second training dataset. The second training dataset includes second feature data and second label data. The second feature data includes DEM (Digital Elevation Model) data, slope topographic factor data, aspect topographic factor data, total precipitation, and ground temperature at 2 meters. The second label data includes the existing aboveground biomass of unused grasslands and hayfields.

[0084] Specifically, a DEM is a digital model representing the topography of the Earth's surface, storing elevation information of the surface in a grid format. The value of each grid cell (pixel) represents the elevation of that location.

[0085] Slope topographic data represents the angle or gradient of the ground relative to the horizontal plane. It can be expressed as a percentage or in degrees. For example, a 10% slope means that for every 100 meters traveled horizontally, the elevation rises by 10 meters; in degrees, this is approximately 5.71°. Aspect topographic data refers to the direction of maximum descent at a point on the ground. It is typically measured clockwise from north, ranging from 0° (true north) to 360° (back to true north). For example, an east-facing aspect is 90°, and a south-facing aspect is 180°. Both slope and aspect topographic data can be calculated from DEM data.

[0086] The total daily precipitation for the target region is obtained by summing up the hourly total precipitation data from ERA5. ERA5 is the latest global reanalysis product from the ECMWF (European Centre for Medium-Range Weather Forecasts), providing atmospheric, land, and ocean data from 1977 to the present. Compared to ECMWF's previous reanalysis product, ERA-Interim, ERA5 offers improved spatial and temporal resolution and includes more observational data, thus providing more accurate historical climate information.

[0087] The daily average temperature of the target area is calculated by averaging the ground air temperature data at 2 meters above the ground within each day in ERA5.

[0088] By using ENVI (Environment for Visualizing Images) software to transform and resample DEM data, slope topographic factor data, aspect topographic factor data, daily total precipitation and daily average temperature of the target area, we can obtain DEM data, slope topographic factor data, aspect topographic factor data, total precipitation and 2-meter ground temperature data of the target area with a spatial resolution of 500 meters and latitude and longitude projection.

[0089] Specifically, the process of acquiring existing aboveground biomass from unused grasslands and hayfields is as follows:

[0090] Principal component transformation was performed on the time series data of normalized vegetation index, and the top 5 characteristic bands with the highest characteristics in the principal component analysis results were retained.

[0091] Based on the time series data of the first 5 feature bands, a multi-scale segmentation algorithm is used to identify grassland fences and obtain grassland fence boundary data.

[0092] Based on grassland fence boundary data, a mowing recognition algorithm is used to identify unused grassland and haymaking areas;

[0093] Based on the existing aboveground biomass of grassland obtained by formula (3), the existing aboveground biomass of unused grassland and existing aboveground biomass of hayfield are obtained, which are the training labels for the potential biomass remote sensing estimation model.

[0094] In a specific embodiment, the process of training the random forest model using the second training dataset to obtain the potential biomass remote sensing estimation model is as follows:

[0095] Load the preprocessed second training dataset, which includes second feature data and second label data. The second feature data includes digital elevation model data, slope topographic factor data, aspect topographic factor data, total precipitation and ground 2-meter temperature data. The second label data includes the existing aboveground biomass of unused grassland and hayfields.

[0096] The random forest model was trained using the second training dataset to obtain a potential biomass remote sensing estimation model, which includes:

[0097] Preset the number of decision trees to be created in the random forest model;

[0098] For each decision tree, a predetermined number of features are randomly selected from all features to form a second feature subset considered during the training of the decision tree;

[0099] Use the bootstrap method to draw a second subset of samples with replacement from the second training dataset, which is the same size as the second training dataset.

[0100] Train the decision tree using the second feature subset and the second sample subset;

[0101] Repeat the above process until a predetermined number of decision trees are constructed, thus obtaining a remote sensing estimation model for potential biomass.

[0102] During training, DEM data with 500-meter spatial resolution and latitude / longitude projection, slope topographic factor data, aspect topographic factor data, total precipitation, and 2-meter ground temperature data were used as model input parameters. These input parameters were divided into training and validation sets in a 7:3 ratio. Potential aboveground biomass of unused grassland and hayfields was used as the model output parameter. The performance of the potential biomass remote sensing estimation model was evaluated on the validation set, and the model parameters were adjusted based on the validation results to establish the final potential biomass remote sensing estimation model. Specifically, potential aboveground biomass can be expressed as an equation for the following input parameters:

[0103] AGB p =f p (DEM, slope, aspect, tp) t1…tn ,T2m ti…tn (4)

[0104] In equation (4), AGB p f represents the potential aboveground biomass of grassland. p () represents a random forest model predicting potential aboveground biomass, slope represents slope topographic data, aspect represents slope aspect topographic data, and tp t1…tn T2m represents the total precipitation during the period from time t1 to time tn. t1…tnThis represents the ground temperature at a depth of 2 meters during the time interval from time t1 to time tn.

[0105] S200: Based on the daily forage intake and the total amount of edible forage within a preset future period, the grazing time is obtained.

[0106] Specifically, the grazing period is calculated as the total amount of edible forage within a predetermined future period divided by the daily forage intake.

[0107] The total amount of edible forage within the future preset period is obtained based on the reasonable utilization rate of natural grassland and the potential aboveground biomass of grassland within the future preset period. The potential aboveground biomass of grassland within the future preset period is obtained by inputting digital elevation model data, slope topographic factor data, slope aspect topographic factor data, total precipitation within the future preset period, and the predicted ground temperature at 2 meters within the future preset period into the potential biomass remote sensing estimation model.

[0108] In this embodiment of the application, the process of obtaining the total precipitation within a future preset time period and the predicted ground temperature at 2 meters within the future preset time period is as follows:

[0109] By interpolating the weekly average precipitation and the weekly average ground temperature at 2 meters, the predicted daily average temperature and daily average ground temperature at 2 meters are obtained for the future preset time period.

[0110] The daily average temperature and the predicted daily average ground temperature at 2 meters within a future preset time period are scaled to convert the initial 0.25° spatial resolution to a 500-meter spatial resolution, resulting in a 500-meter spatial resolution, the total daily precipitation, and the predicted ground temperature at 2 meters within the future preset time period.

[0111] In one specific embodiment, the weekly average precipitation and weekly average 2-meter ground temperature extracted from the meteorological forecast data of ERA5 IFS (Integrated Forecasting System) are interpolated over time to obtain the daily average temperature and daily average 2-meter ground temperature forecast values ​​for the next 30 days.

[0112] Using ENVI software, the daily average temperature and the daily average 2-meter ground temperature forecast for the next 30 days were scaled to convert the initial 0.25° spatial resolution to 500 meters spatial resolution, resulting in the target area's 500-meter spatial resolution, latitude and longitude projection, daily total precipitation for the next 30 days, and the predicted 2-meter ground temperature.

[0113] The above step S200 will be described in detail below using a specific embodiment.

[0114] In this embodiment of the application, in order to describe the amount of grassland vegetation biomass consumed by grazing activities, the difference between the potential aboveground biomass and the existing aboveground biomass of the grassland is defined as the forage intake, i.e.

[0115] GC = AGB p -AGB c (5)

[0116] In equation (5), GC represents pasture intake, and AGB represents... p AGB represents the potential aboveground biomass of grassland. c This indicates the current aboveground biomass of the grassland.

[0117] For the total forage intake GC during the time period from time t1 to time t2 t1-t2 Prediction can be made using the difference between potential aboveground biomass and existing aboveground biomass over this period, i.e.

[0118] GC t1-t2 =AGB p_t2 -AGB c_t2 -AGB p_t1 +AGB c_t1 (6)

[0119] In equation (6), GC t1-t2 AGB represents the total amount of forage consumed during the time period from t1 to t2. p_t2 AGB represents the potential aboveground biomass of grassland at time t2. c_t2 AGB represents the existing aboveground biomass of the grassland at time t2. p_t1 AGB represents the potential aboveground biomass of grassland at time t1. c_t1 This represents the existing aboveground biomass of the grassland at time t1.

[0120] Daily forage intake (DGC) can be expressed as the total amount of forage consumed (GC) during the time period from t1 to t2. t1-t2 The ratio of the duration from time t1 to time t2, i.e.

[0121]

[0122] Assuming that the daily forage intake remains constant during the time period t1-t3, then the total forage intake GC during the time period t2-t3 is... t2-t3 for:

[0123]

[0124] In this embodiment of the application, without considering the compensatory growth of natural grassland, the existing aboveground biomass AGB of the grassland at future time t3 is... c_t3 This can be considered as the existing aboveground biomass (AGB) of the grassland at time t2. c_t2The increase in aboveground biomass ΔAGB during the t2-t3 time period t2_t3 The sum of

[0125] AGB c_t3 =AGB c_t2 +ΔAGB t2_t3 =AGB c_t2 +AGB p_t3 -AGB p_t2 (9)

[0126] Based on grassland type data for the target area, and in accordance with the People's Republic of China agricultural industry standard "Calculation of Reasonable Carrying Capacity of Natural Grassland" (NY / T 635-2015), a reasonable utilization rate p is set for different grassland types. For example, the reasonable utilization rate for year-round grazing of typical meadow grasslands is 50-55%, and the reasonable utilization rate for year-round grazing of sandy grasslands is 20-30%. Here, the reasonable utilization rate p is taken as the average of the range of reasonable utilization rates for year-round grazing in the standard.

[0127] Combining the natural grassland utilization rate p and the existing aboveground biomass AGB at future time t3 c_t3 The total edible forage (TF) during the time period t2-t3 can be obtained. t2-t3 :

[0128] TF t2-t3 =p·AGB c_t3 =p(AGB) p_t3 -AGB p_t2 +AGB c_t2 (10)

[0129] In this embodiment of the application, if the total forage intake GC during the time period from t2 to t3 is... t2-t3 Greater than the total edible forage amount TF during this period t2-t3 In order to ensure the ecological health of natural grasslands and the orderly management of grass-livestock balance, it is necessary to limit the grazing time during this period.

[0130] The grazing duration t that the total amount of edible forage can support during the period from t2 to t3. g This is the ratio of the total amount of edible forage to the daily intake during that period.

[0131]

[0132] For example, using a geographic location as a unit, a map is drawn showing the predicted grazing time in September 2024 for a certain area (pastoral and grassland area) to which the geographic location belongs. Figure 2 This shows the number of days suitable for grazing within the next 30 days (September 1st to September 30th) predicted in August 2024. From... Figure 2It can be seen that this application can effectively predict the number of grazing days in the future in vast pastoral and grassland areas.

[0133] The method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in this application can effectively predict the grazing duration of grasslands by accurately calculating the daily forage intake of natural grasslands and the total amount of edible forage within a preset future period.

[0134] The method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in this application can ensure the rational use of pasture resources, avoid grassland degradation caused by overgrazing, and maximize the production potential of grasslands by scientifically assessing the existing aboveground biomass and potential aboveground biomass of grasslands.

[0135] The method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in this application predicts the grazing duration based on the actual conditions of the grassland, such as leaf area index and normalized vegetation index, as well as environmental factors such as topography, precipitation, and temperature. This method can promote the healthy cycle of the grassland ecosystem, help maintain the ecological balance of the grassland, reduce soil erosion, and preserve biodiversity.

[0136] The method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in this application utilizes advanced remote sensing technology and data analysis models to offer ranch managers a reliable tool for making data-driven management decisions. This not only improves ranch management efficiency but also helps adapt to the uncertainties brought about by climate change, enhancing the flexibility and resilience of agricultural production.

[0137] The method for predicting the grazing duration of natural grasslands based on remote sensing technology provided in this application can effectively control grazing intensity, help mitigate greenhouse gas emissions, reduce the impact on the surrounding natural environment, and is in line with the concept of green development.

[0138] Based on the same inventive concept, such as Figure 3 As shown in the illustration, this application also provides a device for predicting the grazing duration of natural grassland, comprising:

[0139] The calculation module is configured to obtain the daily forage intake based on the existing aboveground biomass and the potential aboveground biomass of the grassland. Specifically, the daily forage intake = potential aboveground biomass - existing aboveground biomass. The existing aboveground biomass is obtained by inputting leaf area index data, normalized difference vegetation index (NDVI) data, and surface reflectance into the existing biomass remote sensing inversion model. The potential aboveground biomass is obtained by inputting digital elevation model data, slope topographic factor data, aspect topographic factor data, total precipitation, and 2-meter ground temperature data into the potential biomass remote sensing estimation model.

[0140] In this embodiment, the existing biomass remote sensing inversion model is obtained by training a random forest model using a first training dataset. The first training dataset includes first feature data and first label data. The first feature data includes leaf area index data, remotely sensed surface reflectance, and normalized difference vegetation index (NDVI). The first label data includes measured aboveground biomass values.

[0141] In this embodiment, the potential biomass remote sensing estimation model is obtained by training a random forest model using a second training dataset. The second training dataset includes second feature data and second label data. The second feature data includes DEM (Digital Elevation Model) data, slope topographic factor data, aspect topographic factor data, total precipitation, and ground temperature at 2 meters. The second label data includes the existing aboveground biomass of unused grasslands and hayfields.

[0142] The grazing duration prediction module is configured to determine the grazing duration based on daily forage intake and the total amount of edible forage within a preset future time period. Specifically, the grazing duration = total amount of edible forage within the preset future time period / daily forage intake. The total amount of edible forage within the preset future time period is obtained based on the rational utilization rate of natural grassland and the potential aboveground biomass of grassland within the preset future time period. The potential aboveground biomass is obtained by inputting digital elevation model data, slope topographic factor data, aspect topographic factor data, total precipitation within the preset future time period, and the predicted 2-meter ground temperature within the preset future time period into the potential biomass remote sensing estimation model.

[0143] In this embodiment of the application, the process of obtaining the total precipitation within a future preset time period and the ground temperature at 2 meters within the future preset time period is as follows:

[0144] By interpolating the weekly average precipitation and the weekly average ground temperature at 2 meters, the predicted daily average temperature and daily average ground temperature at 2 meters are obtained for the future preset time period.

[0145] The daily average temperature and the predicted daily average ground temperature at 2 meters are scaled for the future preset time period. The initial spatial resolution of 0.25° is converted to a spatial resolution of 500 meters, resulting in a spatial resolution of 500 meters, latitude and longitude projection, the total precipitation for each day in the future preset time period, and the predicted ground temperature at 2 meters.

[0146] Based on the same concept, embodiments of this application also provide an electronic device, such as... Figure 4 As shown, it includes a processor and a memory, the memory storing an executable program, and the processor executing the executable program to perform the steps of the method described above.

[0147] This application also provides a storage medium that carries one or more computer programs, which, when executed by a processor, implement the steps of the method described above.

[0148] The storage medium in this embodiment may be included in an electronic device / system; or it may exist independently and not assembled into an electronic device / system. The storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of this application.

[0149] It should be understood that in the embodiments of this application, the processor may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0150] It should also be understood that the memory mentioned in the embodiments of this application can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. Non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of RAM are available, such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchlink DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0151] It should be noted that when the processor is a general-purpose processor, DSP, ASIC, FPGA, or other programmable logic device, discrete gate or transistor logic device, or discrete hardware component, the memory (storage module) is integrated into the processor.

[0152] It should be noted that the memories described herein are intended to include, but are not limited to, these and any other suitable types of memories.

[0153] It should also be understood that the first, second, third, fourth and various numerical designations used herein are merely for descriptive convenience and are not intended to limit the scope of this application.

[0154] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0155] In implementation, each step of the above method can be completed by integrated logic circuits in the processor's hardware or by instructions in software. The steps of the method disclosed in the embodiments of this application can be directly implemented by a hardware processor, or by a combination of hardware and software modules in the processor. The software modules can reside in random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, registers, or other mature storage media in the art. This storage medium is located in memory, and the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, detailed descriptions are omitted here.

[0156] In the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0157] Those skilled in the art will recognize that the various illustrative logical blocks (ILBs) and steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this application.

[0158] In the several embodiments provided in this application, it should be understood that the disclosed data detection methods, apparatuses, electronic devices, storage media, and chips can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units (or modules) is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual couplings, direct couplings, or communication connections may be through some interfaces; indirect couplings or communication connections between devices or units may be electrical, mechanical, or other forms.

[0159] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0160] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0161] In the above embodiments, implementation can be achieved entirely or partially through software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented entirely or partially in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state drive), etc.

[0162] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for predicting the grazing duration of natural grasslands based on remote sensing technology, characterized in that, include: Daily forage intake was obtained based on the existing aboveground biomass and potential aboveground biomass of grassland. The existing aboveground biomass was obtained by inputting the leaf area index dataset, normalized vegetation index, and remotely sensed surface reflectance into the existing biomass remote sensing inversion model. The potential aboveground biomass was obtained by inputting digital elevation model data, slope topographic factor data, slope aspect topographic factor data, total precipitation, and 2-meter ground temperature data into the potential biomass remote sensing estimation model. The grazing duration is determined based on the daily forage intake and the total amount of edible forage within a predetermined time period. The total amount of edible forage within the predetermined time period is obtained based on the rational utilization rate of natural grassland and the potential aboveground biomass of grassland within the predetermined time period. The potential aboveground biomass of grassland within the predetermined time period is obtained by inputting digital elevation model data, slope topographic factor data, slope aspect topographic factor data, total precipitation within the predetermined time period, and the predicted 2-meter ground temperature within the predetermined time period into the potential biomass remote sensing estimation model. in: Daily forage intake is the difference between the potential aboveground biomass of the grassland and the existing aboveground biomass of the grassland. The grazing period is the total amount of edible forage within a future preset time period divided by the daily forage intake.

2. The method according to claim 1, characterized in that, The training process for the existing biomass remote sensing inversion model is as follows: Load the preprocessed first training dataset. The first training dataset includes first feature data and first label data. The first feature data includes leaf area index dataset, remote sensing surface reflectance and normalized vegetation index. The first label data includes measured values ​​of aboveground biomass. The random forest model was trained using the first training dataset to obtain an existing biomass remote sensing inversion model, which includes: Preset the number of decision trees to be created in the random forest model; For each decision tree, a preset number of features are randomly selected from all features to form the first feature subset considered during the training of the decision tree; Use the bootstrap method to draw a first subset of samples with replacement from the first training dataset, which is the same size as the first training dataset. Train a decision tree using the first feature subset and the first sample subset; Repeat the above process until a predetermined number of decision trees are constructed, thus obtaining the existing biomass remote sensing inversion model.

3. The method according to claim 2, characterized in that, The leaf area index dataset and surface reflectance were obtained by image stitching, resampling, and reprojection of data collected by a medium-resolution imaging spectrometer; the normalized vegetation index was calculated using remote sensing reflectance in the red band and the near-infrared band. The measured value of the aboveground biomass is: ， In the formula, AGB This represents the measured value of aboveground biomass. n This indicates the total number of quadrats. AGB i Indicates the first sample i Aboveground biomass dry weight of a grassland population.

4. The method according to claim 1, characterized in that, The training process for the potential biomass remote sensing estimation model is as follows: Load the preprocessed second training dataset, which includes second feature data and second label data. The second feature data includes digital elevation model data, slope topographic factor data, aspect topographic factor data, total precipitation and ground 2-meter temperature data. The second label data includes the existing aboveground biomass of unused grassland and hayfields. The random forest model was trained using the second training dataset to obtain a potential biomass remote sensing estimation model, which includes: Preset the number of decision trees to be created in the random forest model; For each decision tree, a predetermined number of features are randomly selected from all features to form a second feature subset considered during the training of the decision tree; Use the bootstrap method to draw a second subset of samples with replacement from the second training dataset, which is the same size as the second training dataset. Train the decision tree using the second feature subset and the second sample subset; Repeat the above process until a predetermined number of decision trees are constructed, thus obtaining a remote sensing estimation model for potential biomass.

5. The method according to claim 4, characterized in that, The process of obtaining the second feature data is as follows: Based on the original digital elevation model data, slope topographic factor data and aspect topographic factor data are calculated. The total daily precipitation is obtained by accumulating the total hourly precipitation data. The average daily temperature is obtained by averaging the 2-meter ground temperature data within each day. The slope topographic factor data, aspect topographic factor data, daily total precipitation and daily average temperature were transformed and resampled to obtain digital elevation model data with 500-meter spatial resolution and latitude and longitude projection, slope topographic factor data, aspect topographic factor data, total precipitation and hourly ground temperature data at 2 meters.

6. The method according to claim 4, characterized in that, The process of obtaining the second tag data is as follows: Principal component transformation was performed on the time series data of normalized vegetation index, and the top 5 characteristic bands with the highest characteristics in the principal component analysis results were retained. Based on the time series data of the first 5 feature bands, a multi-scale segmentation algorithm is used to identify grassland fences and obtain grassland fence boundary data. Based on grassland fence boundary data, a mowing recognition algorithm is used to identify unused grassland and haymaking areas; Based on the existing aboveground biomass of grassland output by the existing biomass remote sensing inversion model, the existing aboveground biomass of unused grassland and hayfield is obtained.

7. The method according to claim 1, characterized in that, The process for obtaining the total precipitation and the predicted ground temperature at 2 meters within the future preset time period is as follows: By interpolating the weekly average precipitation and the weekly average ground temperature at 2 meters, the predicted daily average temperature and daily average ground temperature at 2 meters are obtained for the future preset time period. The daily average temperature and the predicted daily average ground temperature at 2 meters within a future preset time period are scaled to convert the initial 0.25° spatial resolution to a 500-meter spatial resolution, resulting in a 500-meter spatial resolution, the total daily precipitation, and the predicted ground temperature at 2 meters within the future preset time period.

8. A device for predicting the grazing time of natural grassland, characterized in that, include: The calculation module is configured to obtain the daily forage intake based on the existing aboveground biomass and the potential aboveground biomass of the grassland. The existing aboveground biomass is obtained by inputting the leaf area index dataset, normalized vegetation index, and remotely sensed surface reflectance into the existing biomass remote sensing inversion model. The potential aboveground biomass is obtained by inputting digital elevation model data, slope topographic factor data, slope aspect topographic factor data, total precipitation, and 2-meter ground temperature data into the potential biomass remote sensing estimation model. The duration prediction module is configured to determine the grazing duration based on the daily forage intake and the total amount of edible forage within a preset future time period. The total amount of edible forage within the preset future time period is obtained based on the rational utilization rate of natural grassland and the potential aboveground biomass of grassland within the preset future time period. The potential aboveground biomass of grassland within the preset future time period is obtained by inputting digital elevation model data, slope topographic factor data, slope aspect topographic factor data, total precipitation within the preset future time period, and the predicted 2-meter ground temperature within the preset future time period into the potential biomass remote sensing estimation model. in: Daily forage intake is the difference between the potential aboveground biomass of the grassland and the existing aboveground biomass of the grassland. The grazing period is the total amount of edible forage within a future preset time period divided by the daily forage intake.

9. An electronic device, characterized in that, It includes a processor and a memory, wherein the memory stores an executable program, and the processor executes the executable program to perform the steps of the method as claimed in any one of claims 1 to 7.

10. A storage medium, characterized in that, The storage medium carries one or more computer programs, which, when executed by a processor, implement the steps of the method as described in any one of claims 1 to 7.

Images