CN113570273B

CN113570273B - Spatialization method and system for irrigation farmland statistical data

Info

Publication number: CN113570273B
Application number: CN202110887578.3A
Authority: CN
Inventors: 朱秀芳; 刘莹
Original assignee: Beijing Normal University
Current assignee: Beijing Normal University
Priority date: 2021-08-03
Filing date: 2021-08-03
Publication date: 2023-09-05
Anticipated expiration: 2041-08-03
Also published as: CN113570273A

Abstract

The invention provides a spatialization method and a spatialization system for irrigation farmland statistical data, wherein the method comprises the following steps: acquiring a land coverage utilization diagram of a region to be spatially; covering the area without planting crops in the land coverage utilization map by utilizing the first mark to obtain a cultivated land distribution map of the area to be spatially formed; acquiring meteorological data of a region to be spatially; the meteorological data comprise precipitation, potential evaporation and actual evaporation; and determining an irrigation farmland distribution map of the region to be spatially formed according to the meteorological data and the farmland distribution map. According to the invention, the space of the irrigation farmland statistical data can be realized by processing the land coverage utilization image and the air image data, so that the fine-scale irrigation farmland distribution map is obtained.

Description

Spatialization method and system for irrigation farmland statistical data

Technical Field

The invention relates to the technical field of statistical data spatialization, in particular to a spatialization method and system for irrigation farmland statistical data.

Background

Irrigation is an important land management mode and is an important means for increasing crop yield or reducing negative effects caused by extreme climates such as drought. The extraction of information such as farmland irrigation area, irrigation distribution, irrigation amount and irrigation time is very important for grain safety, economic development, water resource management and the like, wherein the irrigation area and the irrigation distribution are the most basic irrigation information, and can be obtained through irrigation farmland mapping. At present, the irrigation farmland mapping methods can be summarized into two types: irrigation farmland mapping based on remote sensing classification and irrigation farmland mapping based on statistical data spatialization.

The irrigation cultivated land mapping method based on remote sensing classification often selects a plurality of characteristic variables capable of distinguishing the irrigation cultivated land from the rain cultivated land, such as a wave band index (spectral reflectivity, vegetation index, water index and the like), a physical index (starting and stopping time of a growing season, peak value and the like), climate and environmental variables (surface temperature, evaporation and the like) and adopts a supervised or unsupervised classification method to extract the irrigation cultivated land. The core of the statistical data-based spatialization irrigation farmland mapping method is to build a reasonable spatial distribution model, while the core of the spatialization model is to find out a certain or some variables highly related to irrigation, wherein the variables are rasterized data (such as 1 km) with higher spatial resolution than irrigation statistical data, and then build distribution rules based on the variables.

Whether the irrigation farmland drawing is based on statistical data spatialization or the irrigation drawing based on remote sensing classification, selecting proper characteristic variables is an important link. In general, the characteristic variables currently used to reflect the irrigation potential of cultivated lands are largely divided into two categories: the first type of parameter reflects the effect of irrigation mainly by describing the change of soil moisture or crop moisture caused by irrigation; the second category of parameters focuses on irrigation-induced changes in vegetation growth, such as NDVI peaks, etc. Parameters also combine vegetation and non-vegetation factors to characterize the irrigation potential of the farmland. However, most of the characteristic parameters are indirect in terms of the irrigation possibility, and the physical mechanism is not clear enough.

The remote sensing classification-based irrigation farmland mapping method has great dependence on ground real irrigation farmland samples, and the space statistical units (usually administrative unit levels, but cannot distinguish irrigation conditions in administrative units) of the irrigation farmland information data set of the census data have large areas, so that the irrigation space distribution in the space units cannot be reflected. Compared with the classification method, the statistical data spatialization method has the advantages that the principle is clear, the dependence on ground real irrigation cultivated land samples is low, the reproducibility of the method in other areas is high, and the method is favorable for manufacturing large-scale and long-time sequence irrigation cultivated land data products.

Disclosure of Invention

The invention aims to provide a spatialization method and a spatialization system for statistical data of irrigation cultivated land, which can realize spatialization of the statistical data of the irrigation cultivated land so as to obtain a fine-scale distribution map of the irrigation cultivated land.

In order to achieve the above object, the present invention provides the following solutions:

a spatialization method of irrigation farmland statistics, comprising:

acquiring a land coverage utilization diagram of a region to be spatially;

covering the area, which is not planted with crops, in the land coverage utilization map by using a first mark to obtain a cultivated land distribution map of the area to be spatially formed;

Acquiring meteorological data of the region to be spatially; the meteorological data comprise precipitation, potential evaporation and actual evaporation;

determining an irrigation farmland distribution map of the region to be spatially formed according to the meteorological data and the farmland distribution map; the irrigation farmland distribution map is used for measuring the spatialization degree of the irrigation farmland statistical data of the region to be spatialization.

Optionally, the covering the area of the land covering utilization map, where no crop is planted, with the first identifier, to obtain a cultivated land distribution map of the area to be spatially, specifically includes:

dividing the land cover into a plurality of pixels by using a map;

and covering the area without planting crops in each pixel in the land coverage utilization map by using the first mark to obtain a cultivated land distribution map of the area to be spatially formed.

Optionally, determining an irrigation farmland distribution map of the region to be spatially according to the meteorological data and the farmland distribution map specifically includes:

determining the pixels corresponding to the areas planted with crops in the land cover utilization map as cultivated land pixels to obtain a cultivated land pixel set;

calculating the weather drought index of each cultivated land pixel by using a formula CWDI=1-Pre/PET according to the precipitation amount and the potential evaporation amount; the cultivated land pixels are pixels corresponding to areas where crops are planted in the land coverage utilization map; wherein CWDI, pre and PET are respectively the meteorological drought index, rainfall and potential evaporation of the same farmland pixel;

Calculating the agricultural drought index of each cultivated land pixel by using a formula CWSI=1-AET/PET according to the actual evaporation amount and the potential evaporation amount; wherein CWSI, AET and PET are respectively the agricultural drought index, actual evaporation capacity and potential evaporation capacity of the same farmland pixel;

constructing a two-dimensional feature space by taking the meteorological drought index as an abscissa and the agricultural drought index as an ordinate;

determining the upper envelope lines of all the cultivated land pixels in the two-dimensional feature space as rain-raising indication lines;

according to the rain-raising indication line and the agricultural drought index of each cultivated land pixel, a formula is utilizedDetermining an irrigation probability index of each cultivated land pixel; wherein, IPI _i An irrigation probability index for the ith cultivated land pixel; CWSIrainfed (i) is the agricultural drought index corresponding to the same point on the rain-raising indication line as the meteorological drought index of the ith cultivated land pixel, and CWSI (i) is the agricultural drought index of the ith cultivated land pixel;

determining that the cultivated land pixels with the largest irrigation probability index in the cultivated land pixel set are irrigation cultivated land pixels;

judging whether the sum of the areas of all the irrigation cultivated land pixels, in which crops are planted, reaches the actual irrigation cultivated land area or not, and obtaining a first judgment result;

If the first judgment result is yes, covering the area with crops planted in each irrigation cultivated land pixel in the cultivated land distribution map by using a second mark to obtain an irrigation cultivated land distribution map;

and if the second judgment result is negative, removing the irrigation farmland pixels from the farmland pixel set, and returning to the step of determining that the farmland pixel with the largest irrigation probability index in the farmland pixel set is the irrigation farmland pixel.

Optionally, after covering the area with crops planted in each irrigation farmland pixel in the farmland distribution map by using the second identifier to obtain an irrigation farmland distribution map, the method further includes:

determining that the cultivated land pixels which are not covered by the second mark are rain-cultivated land pixels, and covering the area with crops planted in each rain-cultivated land pixel in the irrigation cultivated land distribution map by utilizing a third mark to obtain the rain-cultivated land distribution map.

A spatialization system of irrigation farmland statistics, comprising:

the land cover utilization map acquisition module is used for acquiring a land cover utilization map of the region to be spatially;

the cultivated land distribution map determining module is used for covering the area, which is not planted with crops, of the land coverage utilization map by utilizing the first mark to obtain a cultivated land distribution map of the area to be spatially formed;

The meteorological data acquisition module is used for acquiring meteorological data of the region to be spatially; the meteorological data comprise precipitation, potential evaporation and actual evaporation;

and the irrigation and farmland distribution map determining module is used for determining an irrigation and farmland distribution map of the region to be spatially according to the meteorological data and the farmland distribution map.

Optionally, the tilling area distribution map determining module specifically includes:

the pixel dividing unit is used for dividing the land coverage into a plurality of pixels by using a map;

and the cultivated land distribution map determining unit is used for covering the area, which is not planted with crops, in each pixel in the land coverage utilization map by utilizing the first mark to obtain the cultivated land distribution map of the area to be spatially formed.

Optionally, the irrigation farmland distribution map determining module specifically includes:

a cultivated land pixel set determining unit, configured to determine a pixel corresponding to a region where crops are planted in the land coverage utilization map as a cultivated land pixel, and obtain a cultivated land pixel set;

the meteorological drought index calculation unit is used for calculating the meteorological drought index of each cultivated land pixel according to the precipitation amount and the potential evaporation amount by using a formula CWDI=1-Pre/PET; the cultivated land pixels are pixels corresponding to areas where crops are planted in the land coverage utilization map; wherein CWDI, pre and PET are respectively the meteorological drought index, rainfall and potential evaporation of the same farmland pixel;

The agricultural drought index calculation unit is used for calculating the agricultural drought index of each cultivated land pixel according to the actual evaporation capacity and the potential evaporation capacity by using a formula CWSI=1-AET/PET; wherein CWSI, AET and PET are respectively the agricultural drought index, actual evaporation capacity and potential evaporation capacity of the same farmland pixel;

the two-dimensional feature space construction unit is used for constructing a two-dimensional feature space by taking the meteorological drought index as an abscissa and the agricultural drought index as an ordinate;

the rain-raising indication line determining unit is used for determining that the upper envelope lines of all the cultivated land pixels in the two-dimensional characteristic space are rain-raising indication lines;

the irrigation probability index calculation unit is used for utilizing a formula according to the rain-raising indication line and the agricultural drought index of each cultivated land pixelDetermining an irrigation probability index of each cultivated land pixel; wherein, IPI _i An irrigation probability index for the ith cultivated land pixel; CWSIrainfed (i) is the agricultural drought index corresponding to the same point on the rain-raising indication line as the meteorological drought index of the ith cultivated land pixel, and CWSI (i) is the agricultural drought index of the ith cultivated land pixel;

an irrigation cultivated land pixel determining unit, configured to determine that a cultivated land pixel with the largest irrigation probability index in the cultivated land pixel set is an irrigation cultivated land pixel;

The judging unit is used for judging whether the sum of the areas of all the irrigation cultivated land pixels, in which crops are planted, reaches the actual irrigation cultivated land area or not, and obtaining a judging result; if the judgment result is yes, calling an irrigation farmland distribution map determining unit; if the judging result is negative, removing the irrigation cultivated land pixels from the cultivated land pixel set, and calling the irrigation cultivated land pixel determining unit;

and the irrigation cultivated land distribution map determining unit is used for covering the area, in which crops are planted, in each irrigation cultivated land pixel in the cultivated land distribution map by using a second mark to obtain the irrigation cultivated land distribution map.

Optionally, the irrigation farmland distribution map determining module further includes:

the rain-cultivation farmland distribution map determining unit is used for determining that farmland pixels which are not covered by the second mark are rain-cultivation farmland pixels, and covering areas with crops planted in each rain-cultivation farmland pixel in the irrigation farmland distribution map by utilizing the third mark to obtain the rain-cultivation farmland distribution map.

According to the specific embodiment provided by the invention, the invention discloses the following technical effects:

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions of the prior art, the drawings that are needed in the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of a spatialization method of irrigation farmland statistics provided by an embodiment of the present invention;

FIG. 2 is a schematic representation of the percentage of area harvested in the field irrigated by Blastan in 2017 according to an embodiment of the present invention;

FIG. 3 is a technical roadmap of a spatialization method of statistical data of irrigated land according to an embodiment of the present invention;

FIG. 4 is a diagram of a pixel sample distribution provided by an embodiment of the present invention;

FIG. 5 is a schematic diagram of a two-dimensional feature space provided by an embodiment of the present invention;

fig. 6 is a schematic diagram of determining a rain-raising indicator line according to an embodiment of the present invention; FIG. 6 (a) is a sample distribution diagram of a two-dimensional feature space according to an embodiment of the present invention; FIG. 6 (b) is a diagram showing a sample distribution diagram after the two-dimensional feature space de-discretization according to an embodiment of the present invention; FIG. 6 (c) is a schematic diagram of a rain-keeping indicator line according to an embodiment of the present invention;

FIG. 7 is a graph of the probability index of irrigation in Blas California over 2017 provided by an embodiment of the present invention;

FIG. 8 is a diagram of a distribution diagram of a rain-cultivated land according to an embodiment of the present invention;

fig. 9 is a schematic structural diagram of a spatialization system of statistical data of irrigation farmland according to an embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

In order that the above-recited objects, features and advantages of the present invention will become more readily apparent, a more particular description of the invention will be rendered by reference to the appended drawings and appended detailed description.

Fig. 1 is a flowchart of a spatialization method of statistical data of irrigation farmland according to an embodiment of the present invention, and as shown in fig. 1, the present invention provides a spatialization method of statistical data of irrigation farmland, including:

Step 101: acquiring a land coverage utilization diagram of a region to be spatially;

step 102: covering the area without planting crops in the land coverage utilization map by utilizing the first mark to obtain a cultivated land distribution map of the area to be spatially formed;

step 103: acquiring meteorological data of a region to be spatially; the meteorological data comprise precipitation, potential evaporation and actual evaporation;

step 104: and determining an irrigation farmland distribution map of the region to be spatially formed according to the meteorological data and the farmland distribution map.

Step 102, specifically includes:

dividing a land cover into a plurality of pixels by using a map;

Step 104 specifically includes:

determining pixels corresponding to the areas planted with crops in the land coverage utilization map as cultivated land pixels to obtain a cultivated land pixel set;

calculating the weather drought index of each cultivated land pixel by using a formula CWDI=1-Pre/PET according to the precipitation amount and the potential evaporation amount; the cultivated land pixels are pixels corresponding to areas planted with crops in the land coverage utilization map; wherein CWDI, pre and PET are respectively the meteorological drought index, rainfall and potential evaporation of the same farmland pixel;

Calculating the agricultural drought index of each cultivated land pixel by using a formula CWSI=1-AET/PET according to the actual evaporation capacity and the potential evaporation capacity; wherein CWSI, AET and PET are respectively the agricultural drought index, actual evaporation capacity and potential evaporation capacity of the same farmland pixel;

constructing a two-dimensional feature space by taking a meteorological drought index as an abscissa and an agricultural drought index as an ordinate;

determining the upper envelope lines of all cultivated land pixels in the two-dimensional characteristic space as rain-raising indication lines;

according to the rain-raising indication line and the agricultural drought index of each cultivated land pixel, benefitingBy the formulaDetermining an irrigation probability index of each cultivated land pixel; wherein, IPI _i An irrigation probability index for the ith cultivated land pixel; CWSIrainfed (i) is the agricultural drought index corresponding to the same point on the rain-raising indication line as the meteorological drought index of the ith cultivated land pixel, and CWSI (i) is the agricultural drought index of the ith cultivated land pixel;

if the first judgment result is yes, covering the area with crops planted in each irrigation cultivated land pixel in the cultivated land distribution map by using the second mark to obtain the irrigation cultivated land distribution map;

If the second judgment result is negative, the irrigation cultivated land pixels are removed from the cultivated land pixel set, and the step of determining that the cultivated land pixel with the largest irrigation probability index in the cultivated land pixel set is the irrigation cultivated land pixel is returned.

In addition, the spatialization method of the irrigation farmland statistical data provided by the invention is characterized in that after the second mark is used for covering the area with crops planted in each irrigation farmland pixel in the farmland distribution map, the irrigation farmland distribution map is obtained, and then the spatialization method further comprises the following steps:

determining that the cultivated land pixels which are not covered by the second mark are the rain-cultivation cultivated land pixels, and covering the area with crops planted in each rain-cultivation cultivated land pixel in the irrigation cultivated land distribution map by utilizing the third mark to obtain the rain-cultivation cultivated land distribution map.

Fig. 3 is a technical roadmap of a spatialization method of statistical data of irrigation farmland, provided by an embodiment of the invention, as shown in fig. 3, the spatialization method of statistical data of irrigation farmland provided by the invention comprises the following steps:

1. study area selection

The invention selects the Nebra Ka (Nebra ka) of the United states as a research area, wherein the latitude position of the Nebra Ka is between 40 and 43 degrees N, the longitude position is between 95 and 105 degrees W, 39 percent of land area is used for agricultural production, and the method is one of the main agricultural production areas of the United states. There are over 10 tens of thousands of active irrigation wells in the state, and up to 1 ten thousands of irrigation wells per decade are one of the most densely irrigated areas of the world. The state major crop plants included corn (50% of 2017 harvest area) and soybean (30% of 2017 harvest area), and other minor crops included winter wheat, sorghum, and alfalfa (about 11% of 2017 total harvest area). 4 to 10 months is the growing season of the state main crop. In recent years, students have produced high-resolution irrigation farmland distribution maps in a long time series including the region. The state is selected as a research area to facilitate the utilization of existing irrigation farmland data products, and sample data is selected for comparative analysis of irrigation characteristic variables. Fig. 2 is a schematic diagram of the percentage of area harvested in 2017 in the field irrigated by briska in accordance with an embodiment of the present invention.

2. Data and method

2.1 data description

The data used in the present invention are shown in table 1.

Table 1 data illustrates

As shown in table 1, the data used in the present invention include: data for calculating irrigation probability indexes, data for sample selection and auxiliary data.

2.1.1 irrigation probability index construction data

The data required for the extraction of the rain-raising indication line and the construction of the irrigation probability index are as follows: precipitation, actual vapor deposition, and potential vapor deposition. The detailed description and preprocessing steps of the data are as follows:

(1) Precipitation data

Precipitation data is derived from the PRISM climate research group at oregon state university, which downloads the website: http:// prism. The climate research group collects climate observation data through a wide monitoring network, and develops a space climate data set by combining a complex quality control method with various modeling technologies so as to research short-term and long-term climate modes. The dataset covers a variety of spatiotemporal resolutions since 1895, providing 6 basic climate elements: precipitation, minimum/high temperature, average dew point, maximum/small water vapor pressure deficit.

The spatial resolution of the obtained precipitation data is 0.04 degrees, the spatial reference is the North America geographic coordinate system (GCS NorthAmerican 1983) in 1983, and the precipitation data are stored in a BIL data format. And carrying out projection conversion, nearest neighbor resampling, clipping and batch processing of format conversion on the precipitation data through Arcmap software modeling to obtain precipitation data of Nebulaska, wherein the spatial resolution is 500 meters, the spatial reference is an area projection (Albers Conic EqualArea) such as an Arabian cone, and the format is TIF.

(2) Actual evapotranspiration data and potential evapotranspiration data

Actual and potential evapotranspiration originate from the MOD16A2 sixth version of the data product of MODIS (medium resolution imaging spectrometer, an important instrument for observing global biological and physical processes in the american Earth Observation System (EOS) program). The product is an 8-day synthetic data product generated with 500m spatial resolution, and the download website is as follows: https:// lpdac. Usgs. Gov/products/mod16a2v006/. The product is obtained by calculating based on a Penman-Monteth model by using daily weather re-analysis data and remote sensing data such as vegetation attribute, albedo, land coverage and the like measured by MODIS as input data, and comprises the following data layers: actual evaporation, latent heat flux, and a quality control layer that extends from the input vegetation attribute data, wherein evaporation is calculated using the sum of the eight days of data, and latent heat data is calculated using the eight days of data mean.

This data was downloaded using the USGS (United States Geological Survey, U.S. geological prospecting) aρ EEARS (The Application for Extracting and Exploring Analysis Ready Samples) download tool. The tool can download the MODIS data which are spliced in the appointed area and have the appointed format, and the steps of projection conversion and data splicing are omitted. Since the projection types specified in the tool do not contain the area projection types such as the albert cone, the spatial reference type of the downloaded data is the world-level geographical coordinate system of 1984 edition (GCS WGS 1984). Because the data product only contains a quality control layer of input parameters in the calculation process, the quality of the product cannot be evaluated, and the data cannot be directly used for quality control.

During preprocessing, firstly, 8 days of data are combined into month scale data by utilizing a MATLAB programming program, real evapotranspiration data are obtained by combining scaling factors of the data, then projection conversion, nearest neighbor resampling and batch processing of clipping are respectively carried out on the data by utilizing Arcmap software modeling, and real evapotranspiration data and potential evapotranspiration data of the area projection with the spatial resolution of 500 m and the spatial reference of Arabian cone and the like in Neibutz are obtained. The following pretreatment steps of the MODIS data products are similar to the steps of the product except for quality control, so that only the quality control method is introduced for the following pretreatment steps of the MODIS data products, and the repeated steps are not repeated.

2.1.2 sample selection data

Due to the specificity of irrigation, it is very difficult to obtain the spatial positions of the irrigation and the rain-raising farmland which are absolutely accurate. Thus, the present study selected two existing sets of irrigation farmland distribution maps covering the state of Nebulaska, irrigation facility distribution maps of Nebulaska, and registered irrigation wells for selection of irrigation farmland and rain-fed farmland samples.

AIM-HPA (annual irrigation map of the plateau aquifer, annual Irrigation Maps-High Plains Aquifer) is an irrigation arable dataset covering 30 meters spatial resolution per year in 1984 to 2017 of the american plateau aquifer, with the following download web addresses: https:// www.hydroshare.org/resource/a371fd69d41b4232806d81e17fe4efcb/. The data set is obtained by classifying the data set by Deines and the like by utilizing Landsat (land satellite plan of the national aviation and aerospace agency) images, environment variables and ground live data based on random forest classifiers, filling the blank of the image by applying a Bayesian land coverage updating algorithm after classification, and reducing errors in irrigation time sequences. The data set is a large-scale irrigation farmland data product with the longest time and the highest spatial resolution by 2019, the classification accuracy is 91 percent, and the county irrigation area statistical data provided by NASS (national wind and dampness Association in England) can be captured to be about 85 percent of change.

The MIrAD-US (US irrigation agriculture medium resolution imaging spectrometer dataset The Moderate Resolution Imaging Spectroradiometer Irrigated Agriculture Dataset for the United States) is a US irrigation agriculture dataset developed by the US geological survey earth resource observation and science center and updated synchronously with US agricultural census data, aiming at providing a comprehensive, consistent and timely geospatial dataset for US native irrigation agriculture. The data set is obtained by modeling by using a statistical data spatialization method, and is obtained by distributing irrigation area statistical data to a space grid by taking the statistical data as total quantity constraint and taking a annual NDVI (vegetation coverage index) peak value as a distribution basis. The dataset contains irrigation farmland data with spatial resolution of 250 meters and 1000 meters in 2002, 2007, 2012 and 2017, and the download website is: https:// www.sciencebase.gov/category/item/5 db08e84e4b0b0c58b56e04f. The distribution map of the irrigation facilities in the state of Nebula is provided by advanced land management information technology center (CALMIT, center for Advanced Land Management Information Technologies) of the Lincoln division of Nebula university, and the download website is: https:// calmit. Un. Edu/metadata-2005-nebraska-land-use-center-pivot-irgear-s systems. The data layer provides, in the form of a vector file, a central hub irrigation system and other irrigation system profiles determined from Landsat 5 satellite images over the year 2005 and orthographic images of the agricultural services bureau. The registered irrigation wells in the state of Nebulaska are derived from registered groundwater well data provided by websites of natural resources department of the state of Nebulaska, and download websites are as follows: https:// dnr. The data are stored in the form of vector points, and water wells with irrigation purposes and active states are screened out by using an attribute table to form registered active irrigation water well data.

When the data are used for selecting sample points, only the area projection such as the Arabic cone is needed to be used as the unified space reference, and other preprocessing steps are not needed. The irrigation farmland distribution map AIM-HPA and MIrAD-US can also be applied to irrigation farmland drawing effect evaluation, so that the data needs to be converted into 500 m spatial resolution, and the method comprises the following specific steps: creating grids with spatial resolution of 500 meters covering the study area by using Arcmap, counting the irrigation area percentage of each grid, determining the grids with irrigation area percentage larger than 0 (referring to the existence of irrigation areas in the grids) as irrigation area pixels,

2.1.3 auxiliary data

The auxiliary data in the present invention mainly includes: crop layer, real irrigation area statistics, crop climate data, and inner boulder vector boundaries.

(1) Crop layer

The crop layer used in the present invention is a crop data layer (Cropland Data Layer, CDL) provided by the national agricultural statistics agency of the united states department of agriculture, which downloads the website: https:// nassgeodata. Gmu. The data has a spatial resolution of 30 meters and is widely used for covering crop types.

The original crop layer contains too many crop types, only the main crop is focused in the invention, and thus the preprocessing step of the data comprises: converting the data into an area projection such as an Arabic cone by using Arcmap for spatial reference and converting the data into an area projection such as an Arabic cone by GCS WGS 1984; recoding the crop type by utilizing MATLAB programming, wherein the corresponding conditions of the coded crop type and the pixel value are as follows: corn-1, soybean-2, winter wheat-3, alfalfa-4, sorghum-5, and other crops-6. After the coding is finished, a grid with the spatial resolution of 500 meters covering the research area is created by utilizing Arcmap, the proportion of the crop area in the grid to the whole grid area is counted, and a main crop area percentage distribution map of the research area with the spatial resolution of 500 meters is obtained. The use of the crop area percentages in the present invention is as follows: the pixel with the crop area percentage larger than 0 is extracted as a basis for determining the spatial position of crops when the statistical data are utilized to spatially map irrigation cultivated land, and the pixels participate in the mapping process; in the step of mapping irrigation cultivated land based on the statistics data spatialization, determining the area allocated as the irrigation cultivated land in the pixels by using the crop area percentage; and after the preparation, the irrigation cultivated land distribution map and the rain-cultivation cultivated land distribution map are overlapped to present an irrigation cultivated land density distribution map and a rain-cultivation cultivated land density distribution map.

(2) Real irrigation data

The real irrigation data used in the present invention is real irrigation area data in agricultural census data provided by the national agricultural statistics agency of the united states department of agriculture. The census is performed every five years, and contains real irrigation data in county units, and the download website is as follows: https:// www.nass.usda.gov/Publications/AgCensus/. It should be noted that the actual irrigation area in this census data is recorded only once, regardless of multiple irrigations and multiple crops. The purposes of this data are: in the irrigation farmland mapping process based on the statistical data spatialization, the total irrigation farmland area of each county is provided as a total amount constraint in the spatialization process.

(3) Crop weather data

Crop climate data used in the present invention is derived from the date of crop sowing and harvest released by the U.S. national agricultural statistics agency of the U.S. department of agriculture at 10 months 2010. It downloads the website: https:// downloads. Usda. Library. Cornell. Edu/usda-espis/files/vm 40xr56k/dv13zw65p/w9505297 d/mounting-10-29-2010. Pdf. The report is based on a history of crop growth history of approximately 20 years and expert knowledge to arrive at a date of crop sowing and harvest that corresponds to most years. The report lists the major crops and their climatic information per state. This data is used primarily to determine candidate times for the data to be analyzed in the study based on the time of sowing and harvesting of the major crop in the state of the inner bougaian: 4-10 months. This period covers the growing season of other major crops in the state of the inner placian besides winter wheat, and the winter wheat is not separately extracted and analyzed because the proportion of the winter wheat in the research area is small.

(4) Research area vector

Vector boundary data of the study area is derived from GADM (global administrative division database), which downloads the website as:http://www.gadm.org/。

(5) Meteorological data

The meteorological data of the study area is derived from the national meteorological data set (NHD, national Hydrography Dataset) provided by the united states geological exploration. This data is used primarily to extract the major river in the state of the inner boulder to evaluate the relationship between irrigation probability index and water distribution, with the download website being: https:// www.sciencebase.gov/category/item/5 a96cdb e4b06990606c4d29.

2.2 research methods

2.2.1 sample Point selection

Combining the CDL of the crop layer, the AIM-HPA and MIrAD-US of the existing irrigation farmland data set, the irrigation facility distribution map of Nebulaska and the registered irrigation well distribution map, selecting irrigation farmland sample points and rain-raising farmland sample points in a research area according to the following rules, and taking all sample points as 7: the ratio of 3 is randomly divided into training samples and verification samples.

(1) The selection rules of irrigation farmland sample points are divided into two types, wherein the first type is applicable to the area covered by AIM-HPA irrigation farmland distribution map, and the second type is applicable to the eastern area of Nebulaska which is not covered by AIM-HPA. The first type of selection rule needs to satisfy the following conditions simultaneously: 1) AIM-HPA is irrigation farmland in 2002-2017; 2) MIrAD-US was irrigation farmland in 2002, 2007, 2012 and 2017; 3) Areas with irrigation equipment in 2005. Since the eastern area of the inner plaska, which is not covered by AIM-HPA, belongs to a moist area, the water supply in the area is relatively sufficient, the demand for irrigation is relatively small, and superposition analysis shows that the pixels of the partial area MIrAD-US which are irrigation farmlands in four years 2002, 2007, 2012 and 2017 are relatively small, the second type of selection is changed to simultaneously meet the following conditions: 1) MIrAD-US was irrigation farmland in 2017; 2) Areas with irrigation equipment in 2005; 3) Near the registered active irrigation water well site.

(2) Since AIM-HPA and MIrAD-US irrigation crop distribution maps only contain irrigation crop pixels, extraction of the rain-cultivated land in combination with crop layer CDL is required. For AIM-HPA with 30 m spatial resolution, adopting recoded crop image layer with 30 m resolution to extract rain-raising farmland: the AIM-HPA is used as a non-irrigation farmland image element and the CDL is used as a crop image element to be identified as a rain-raising farmland. For the MIrAD-US with the spatial resolution of 250 meters, the crop layers with the resolution of 250 meters after recoding and mode aggregation are adopted for extracting the rain-cultivated land: the pixel with the MIrAD-US as a non-irrigation cultivated land pixel and the CDL as a crop is identified as a rain-cultivated land. The selection rules of the sample points of the rain-cultivated land are also divided into two types according to whether AIM-HPA is covered or not. The following conditions are simultaneously satisfied in the area of AIM-HPA with coverage: 1) AIM-HPA is a rain-cultivated land in 2002-2017; 2) MIrAD-US was a rainy farmland in 2002, 2007, 2012, and 2017; 3) Areas without irrigation equipment in 2005. The following conditions are simultaneously satisfied in the area not covered by AIM-HPA: 1) MIrAD-US was a rainy farmland in 2017; 2) Areas without irrigation equipment in 2005; 3) Away from the registered active irrigation water well site.

638 irrigation farmland sample points (440 training samples, 198 verification samples) and 496 rain-farming sample points (343 training samples, 153 verification samples) were selected according to the above rule, and the spatial distribution thereof is shown in fig. 4.

2.2.2 irrigation probability index calculation

First, the weather drought index and the agricultural drought index are calculated using precipitation, actual evapotranspiration and potential evapotranspiration. According to the invention, the reference crop evapotranspiration in the crop water deficit index is replaced by the potential evapotranspiration to obtain a new crop water deficit index CWDI to represent the meteorological drought condition of crops, and the crop water stress index CWSI is selected to represent the agricultural drought condition of crops. For both CWDI and CWSI, the larger the value, the more severe the drought, and the smaller the value, the lighter the drought.

And secondly, taking CWDI as an abscissa axis and CWSI as an ordinate axis, and drawing scattered points of all farmland pixels in a two-dimensional characteristic space formed by weather drought and agricultural drought. When meteorological drought conditions are similar, the lighter the agricultural drought conditions, the higher the probability that the cultivated land pixels are irrigated crops; conversely, the more serious the agricultural drought conditions, the higher the likelihood that the cultivated land pixels are rain-fed crops. Therefore, the upper envelope curve of the scattered points in the characteristic space is fitted to serve as a rain-raising indication line. As shown in fig. 5, the rain-raising indication line adopts the following formula:

CWSI _rainfed ＝a×CWDI _top +b, wherein CWSI _rainfed Representing a rain-raising indication line obtained by fitting; cWDI _top Weather drought index values representing scatter points constituting the upper envelope; a and b are the slope and intercept of the upper envelope curve obtained by fitting. After the rain-raising indication lines are extracted, the distance between each cultivated land pixel and the rain-raising indication line in the vertical axis direction is calculated to be used as an irrigation probability index IPI, and the larger the value of the irrigation probability index is, the larger the probability that the pixel is irrigated is.

In addition, it should be noted that some pixel points are located above the fitted rain-raising indication line, and the irrigation probability indexes of the pixel points are directly assigned to 0 and are not calculated any more.

In order to make the value of the irrigation probability index easier to understand, the calculated irrigation probability index value is stretched to between 0 and 1 by a maximum-minimum standardized method. The formula for the max-min normalization is as follows:

IPI _{normalization} ＝(IPI-IPI _min )/(IPI _max -IPI _min )

wherein IPI is _{normalization} For the normalized irrigation probability index value, IPI represents the untreated irrigation probability index value, IPI _max 、IPI _min And the maximum value and the minimum value which can be obtained by the irrigation probability index in the whole research area are respectively.

As shown in fig. 6 (a), there are often outliers in the scatter diagram, and if all the scatter points are used for fitting, the scatter points on the left and right sides will interfere with the fitting effect of the whole envelope, so that the envelope is fitted Before, the scatter diagram was extracted at 0.001 intervals from CWDI, and outliers were removed with CWSI values exceeding the 2-quartile spacing between the upper and lower quartiles as a limit, as shown in fig. 6 (b). After outliers were removed, CWDI was segmented at 0.001 intervals, the maximum CWDI points in each interval were extracted, and fitting of the upper envelope was performed using the scattered points with CWDI between 0.05 and 0.75, resulting in the results shown in fig. 6 (c). 6 (c) the formula of the rain-raising indication line is CWSI _rainfed ＝0.2788×CWDI _top +0.6748。

2.2.3 spatialization of irrigation statistics

The irrigation cultivated land drawing is carried out based on a statistical data spatialization method, wherein the size of an irrigation probability index is used as a space allocation weight to allocate the actual irrigation cultivated land area to cultivated land pixels in sequence. The minimum statistical unit of the actual irrigation area statistics published by the state of the Nebulaska is at the county level, so that the present study distributes irrigation farmland areas in units of county level. In addition, the position of the cultivated land pixels of the present study was determined using the percentage of cultivated land area on a 500 meter spatial resolution grid, and all pixels with a percentage of cultivated land area greater than 0 were considered cultivated land pixels. The specific space allocation rule is as follows:

firstly, marking an cultivated land pixel with the largest irrigation probability index as an irrigation cultivated land pixel in a county, simultaneously calculating the cultivated land area in the current pixel according to the cultivated land area percentage of the current pixel, recording the cultivated land area which is currently determined as the irrigation cultivated land pixel as the drawing irrigation cultivated land area, selecting the pixel with the large irrigation probability index as the irrigation cultivated land pixel if the drawing irrigation cultivated land area of the current pixel is smaller than the irrigation cultivated land area of the statistical data, and repeating the steps until the drawing irrigation cultivated land area of the current county is closest to the irrigation cultivated land area of the statistical data. Then, all counties are distributed according to the same method, and the irrigation cultivated land pixel position of the whole research area can be obtained after distribution is finished. And then marking the positions of the pixels of the rain-cultivated land: the pixels for which the percentage of cultivated land area is greater than 0 but which are not marked as irrigation cultivated land are all marked as rain-fed cultivated land pixels. After marking, obtaining an irrigation cultivated land and rain-cultivated land distribution map with spatial resolution of 500 m in the Blacka state in 2017, and overlapping the irrigation cultivated land and rain-cultivated land distribution map with a cultivated land area percentage layer to obtain an irrigation cultivated land density distribution map and a rain-cultivated land density distribution map in the Blacka state in 2017. The probability index profile of the irrigation in the state of boucian in 2017 is shown in fig. 7; the rain-raising farmland distribution diagram is shown in figure 8.

2.2.4 evaluation of the drawing Effect of irrigation farmland

The method is characterized in that a method similar to the method for selecting training samples during remote sensing classification is adopted for selecting verification samples, and the confusion matrix of the drawing result obtained by calculating statistical data spatialization of the verification samples is utilized for evaluating the overall position accuracy of the drawing result.

3. Results and analysis

3.1 irrigation probability index construction results

According to the obtained rain-keeping indication line of the state of the Blackin 2017, the irrigation probability index is calculated for each cultivated land pixel, and the irrigation probability index is stretched to be between 0 and 1 in a maximum and minimum standardized mode, and the result is shown in figure 7. It can be seen that the irrigation probability index of the state of briska in 2017 as a whole shows a tendency to be high in the east and low in the west, similar to the trend of the distribution of air temperature and precipitation in that state and the gradient of the change in agricultural activity. In addition, areas of higher irrigation probability index are also distributed substantially along the state-dense river network. Preliminary knowledge that the probability of irrigation is greater in the area where the agricultural activities in the state of Nebulaska are dense and close to water sources such as rivers.

3.2 spatial results of irrigation statistics

The evaluation results of the positional accuracy of the irrigation farmland distribution map by the confusion matrix using the verification samples are shown in table 2.

Table 2 irrigation farmland mapping confusion matrix

As can be seen from Table 2, overall accuracy and Kappa coefficient (which means a multi-element discrete method for evaluating classification accuracy and error matrix of a remote sensing image) of the statistical data space mapping method provided by the invention are relatively high.

4. Conclusion:

irrigation has important effects on aspects such as grain safety, water resource management, climate change and the like. The development of irrigation farmland mapping will also provide more data base and practical possibilities for related research. Aiming at the problems of large automatic drawing difficulty of irrigation cultivated lands, weak physical mechanism of characteristic parameters of the existing irrigation cultivated lands and insufficient suitability analysis of drawing methods of different irrigation cultivated lands, the research provides a novel irrigation probability index with a strong physical basis, realizes the manufacture of an irrigation cultivated land distribution map by using a statistical data spatialization method based on the irrigation probability index, and obtains the following main conclusion: (1) The irrigation probability index of the state of the BlackCalifornia in 2017 shows the tendency of high east and low west, is similar to the distribution trend of air temperature and precipitation of the state and the change gradient of agricultural activities, has obvious positive correlation with other irrigation characteristics, and shows that the irrigation probability index can better reflect the actual irrigation condition of the state of the BlackCalifornia. In addition, the irrigation probability index has different capacities of reflecting irrigation conditions under different climates, and can well capture real irrigation conditions in the middle and west regions where the climates are relatively drought and normal, but has reduced capacities of reflecting real irrigation conditions in the eastern regions where the climates are relatively humid. (2) The total precision of the irrigation cultivated land mapping based on the statistical data spatialization is 86.92%, and the Kappa coefficient is 0.74.

Fig. 9 is a schematic structural diagram of a spatialization system of statistical data of irrigation farmland according to an embodiment of the present invention, and as shown in fig. 9, the present invention further provides a spatialization system of statistical data of irrigation farmland, including:

a land cover utilization map acquisition module 901, configured to acquire a land cover utilization map of a region to be spatialized;

a cultivated land distribution map determining module 902, configured to cover, with the first identifier, a region of the land coverage utilization map, where no crop is planted, and obtain a cultivated land distribution map of the region to be spatially formed;

the meteorological data acquisition module 903 is configured to acquire meteorological data of an area to be spatially; the meteorological data comprise precipitation, potential evaporation and actual evaporation;

an irrigation-farmland-profile determination module 904 for determining an irrigation-farmland-profile of the region to be spatialized from the meteorological data and the farmland-profile.

Wherein, the tilling area distribution map determining module 902 specifically includes:

the pixel dividing unit is used for dividing the land coverage by using the map into a plurality of pixels;

The irrigation farmland distribution map determining module 904 specifically includes:

the cultivated land pixel set determining unit is used for determining pixels corresponding to the areas where crops are planted in the land coverage utilization map as cultivated land pixels to obtain a cultivated land pixel set;

the meteorological drought index calculation unit is used for calculating the meteorological drought index of each cultivated land pixel by using a formula CWDI=1-Pre/PET according to the precipitation amount and the potential evaporation amount; the cultivated land pixels are pixels corresponding to areas planted with crops in the land coverage utilization map; wherein CWDI, pre and PET are respectively the meteorological drought index, rainfall and potential evaporation of the same farmland pixel;

the two-dimensional feature space construction unit is used for constructing a two-dimensional feature space by taking a weather drought index as an abscissa and an agricultural drought index as an ordinate;

the judging unit is used for judging whether the sum of the areas of all the irrigation cultivated land pixels, in which crops are planted, reaches the actual irrigation cultivated land area or not, and obtaining a judging result; if the judgment result is yes, calling an irrigation farmland distribution map determining unit; if the judgment result is negative, the irrigation cultivated land pixels are removed from the cultivated land pixel set, and an irrigation cultivated land pixel determining unit is called;

and the irrigation farmland distribution map determining unit is used for covering the area with crops planted in each irrigation farmland pixel in the farmland distribution map by using the second mark to obtain the irrigation farmland distribution map.

In addition, the irrigation farmland map determination module 904 further includes:

the rain-cultivation farmland distribution map determining unit is used for determining that the farmland pixels which are not covered by the second mark are rain-cultivation farmland pixels, and covering the area with crops planted in each rain-cultivation farmland pixel in the irrigation farmland distribution map by utilizing the third mark to obtain the rain-cultivation farmland distribution map.

In the present specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the system disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.

The principles and embodiments of the present invention have been described herein with reference to specific examples, the description of which is intended only to assist in understanding the methods of the present invention and the core ideas thereof; also, it is within the scope of the present invention to be modified by those of ordinary skill in the art in light of the present teachings. In view of the foregoing, this description should not be construed as limiting the invention.

Claims

1. A method for spatialization of irrigation farmland statistics, the method comprising:

acquiring a land coverage utilization diagram of a region to be spatially;

determining an irrigation farmland distribution map of the region to be spatially formed according to the meteorological data and the farmland distribution map;

the determining the irrigation farmland distribution map of the region to be spatially according to the meteorological data and the farmland distribution map specifically comprises the following steps:

if the first judgment result is negative, the irrigation farmland pixels are removed from the farmland pixel set, and the step of determining that the farmland pixel with the largest irrigation probability index in the farmland pixel set is the irrigation farmland pixel is returned.

2. The method for spatialization of irrigation farmland statistics according to claim 1, wherein the covering the land coverage area of the map where no crop is planted with the first identifier, to obtain a farmland distribution map of the area to be spatialized specifically comprises:

dividing the land cover into a plurality of pixels by using a map;

3. The method for spatialization of irrigation farmland statistics according to claim 1, wherein after said covering of the area of the irrigation farmland picture elements in each of the farmland distribution map with the second markers, obtaining an irrigation farmland distribution map, further comprises:

4. A spatialization system of irrigation farmland statistics, characterized in that it comprises:

the irrigation and farmland distribution map determining module is used for determining an irrigation and farmland distribution map of the region to be spatially according to the meteorological data and the farmland distribution map;

the irrigation farmland distribution map determining module specifically comprises:

the irrigation probability index calculation unit is used for utilizing a formula according to the rain-raising indication line and the agricultural drought index of each cultivated land pixel Determining an irrigation probability index of each cultivated land pixel; wherein, IPI _i An irrigation probability index for the ith cultivated land pixel; CWSIrainfed (i) is the agricultural drought index corresponding to the same point on the rain-raising indication line as the meteorological drought index of the ith cultivated land pixel, and CWSI (i) is the agricultural drought index of the ith cultivated land pixel;

5. The spatialization system of irrigation farmland statistics according to claim 4, characterized in that said farmland profile determination module comprises in particular:

6. The spatialization system of irrigation arable land statistics of claim 4, wherein the irrigation arable land profile determination module further comprises: