Method for extracting water from irrigated land in arid and semi-arid areas

By acquiring and analyzing remote sensing and environmental data in arid and semi-arid areas, using random forest model classification, the problem of irrigated land and dry land data extraction is solved, and high-precision data of cultivated land type is provided to support more scientific water conservancy irrigation plans.

CN114581787BActive Publication Date: 2024-11-26AEROSPACE INFORMATION RES INST CAS
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210088243.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-01-25
Publication Date
2024-11-26
Estimated Expiration
2042-01-25

AI Technical Summary

Technical Problem

In arid and semi-arid areas, it is difficult for the existing technology to extract the data on arable land type of irrigated land and dry land with high precision, affecting the formulation of water conservancy irrigation plans and agricultural output.

Method used

By obtaining remote sensing data, environmental data and sample point data of arid and semi-arid areas, normalized vegetation index, water index, temperature data and precipitation data are selected as classification characteristics, and a random forest model is used to classify irrigated land and dry land.

Benefits of technology

It has achieved high-precision irrigated land extraction, providing a detailed data foundation for irrigation and water conservancy projects, and improving the pertinence and scientific nature of water resource management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114581787B_ABST
    Figure CN114581787B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of cultivated land technology, and provides a method for extracting irrigated land in arid and semi-arid areas. The method for extracting irrigated land in arid and semi-arid areas comprises: obtaining basic data of arid and semi-arid areas; selecting classification features of irrigated land and dry land in arid and semi-arid areas based on the basic data; and extracting irrigated land based on the classification features. The method for extracting irrigated land in arid and semi-arid areas provided by the present invention can well extract irrigated land in arid and semi-arid areas, obtain high-precision cultivated land type data, and provide a detailed data basis for irrigation and water conservancy projects.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of cultivated land, and in particular to a method for extracting irrigated land in arid and semi-arid areas. Background Art

[0002] In recent years, most studies have been conducted on irrigated farmland and dryland, and most of the irrigated farmland is irrigated paddy fields, not irrigated land. In the arid and semi-arid regions of northwest my country, due to terrain restrictions, the farmland is relatively fragmented, there are few large tracts of irrigated land, and there is no targeted study on the spatiotemporal changes of irrigated land and dryland. However, these fragmented irrigated lands and their changes have a huge impact on food production and the ecological environment in areas with low agricultural output and scarce water resources, and their existence cannot be ignored. Only by clarifying the quantity and spatial distribution structure of local irrigated land and dryland can we formulate water conservancy and irrigation plans in a targeted and scientific manner. Summary of the invention

[0003] The present invention provides a method for extracting irrigated land in arid and semi-arid areas, which is used to obtain high-precision cultivated land type data and provide a detailed data basis for irrigation and water conservancy projects. The present invention provides a method for extracting irrigated land in arid and semi-arid areas, comprising: obtaining basic data of arid and semi-arid areas; selecting classification features of irrigated land and dry land in arid and semi-arid areas based on the basic data; and extracting the irrigated land based on the classification features.

[0004] According to a method for extracting irrigated land in an arid or semi-arid region provided by the present invention, the step of obtaining basic data of the arid or semi-arid region further comprises: obtaining remote sensing data, environmental data and sample point data of the arid or semi-arid region.

[0005] According to a method for extracting irrigated land in an arid and semi-arid area provided by the present invention, the step of obtaining remote sensing data of the arid and semi-arid area further includes: obtaining a normalized vegetation index and a water body index during the crop growing season based on the remote sensing data; the step of obtaining environmental data of the arid and semi-arid area further includes: obtaining temperature data and precipitation data of multiple preset time periods.

[0006] According to a method for extracting irrigated land in arid and semi-arid areas provided by the present invention, the step of selecting classification characteristics of irrigated land and dry land in arid and semi-arid areas based on the basic data further includes: selecting the normalized vegetation index and the water body index in the crop growing season as the first classification characteristics of irrigated land and dry land.

[0007] According to a method for extracting irrigated land in arid and semi-arid areas provided by the present invention, the step of selecting the normalized vegetation index and the water body index in the crop growing season as the first classification features of irrigated land and dry land further includes: selecting the normalized vegetation index and the water body index in July and August in the crop growing season as the second classification features of irrigated land and dry land.

[0008] According to a method for extracting irrigated land in arid and semi-arid areas provided by the present invention, the step of selecting classification features of irrigated land and dry land in arid and semi-arid areas based on the basic data also includes: selecting a first ratio of the normalized vegetation index to precipitation data, the negation of the second ratio of the normalized vegetation index to temperature data, a third ratio of the water body index to precipitation data, and the negation of the fourth ratio of the water body index to temperature data as the third classification feature.

[0009] According to a method for extracting irrigated land in arid and semi-arid areas provided by the present invention, the step of extracting the irrigated land based on the classification features further includes: inputting the second classification feature, the first classification feature and the third classification feature, and the second classification feature and the third classification feature into a random forest model respectively to obtain classification information of multiple irrigated lands and dry lands.

[0010] According to a method for extracting irrigated land in arid and semi-arid areas provided by the present invention, the method further includes: verifying the first classification feature and the second classification feature to obtain a preferred time feature.

[0011] A method for extracting irrigated land in arid and semi-arid areas according to the present invention also includes: verifying the third classification feature to obtain an optimal preset time length.

[0012] The method for extracting irrigated land in arid and semi-arid areas provided by the present invention can well extract irrigated land in arid and semi-arid areas, obtain high-precision cultivated land type data, and provide a detailed data basis for irrigation and water conservancy projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0014] Figure 1 It is a flow chart of the method for extracting irrigated land in arid and semi-arid areas provided by the present invention;

[0015] Figure 2It is a technical flow chart of the method for extracting irrigated land in arid and semi-arid areas provided by the present invention; DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0017] The features of the terms "first" and "second" in the specification and claims of the present invention may include one or more of the features explicitly or implicitly. In the description of the present invention, unless otherwise specified, "plurality" means two or more.

[0018] Combine the following Figure 1 and Figure 2 The present invention describes the method for extracting water from irrigated land in arid and semi-arid regions.

[0019] like Figure 1 As shown, in one embodiment of the present invention, a method for extracting irrigated land in arid and semi-arid regions specifically comprises the following steps:

[0020] Step 01: Obtain basic data on arid and semi-arid areas.

[0021] Specifically, the northern part of the Loess Plateau is a typical area of ​​arid and semi-arid areas, and the unique topography forms a distribution of cultivated land in this area, with both plains distributed in patches and scattered terraces and slopes in mountainous areas. From a spatial perspective, the precipitation in the northwest of the study area is generally lower than that in the southeast, so there is a greater demand for water supplementation and more irrigated dry land, so this area is selected as a typical case area. Obtaining basic data for this area includes obtaining remote sensing data, environmental data, and sample point data for this area.

[0022] Step 02: Select the classification characteristics of irrigated land and dry land in arid and semi-arid areas based on basic data.

[0023] Specifically, since water is the biggest constraint on the growth of local crops in the arid and semi-arid regions of Northwest China, the difference in soil moisture content is mainly reflected in the growth of vegetation. Therefore, the vegetation growth rather than soil moisture (which is not easily captured by remote sensing because it is blocked by vegetation and changes too quickly) is used to distinguish whether irrigation is required. Among the vegetation indices, the Normalized Difference Vegetation Index (NDVI) is considered to be the most commonly used and stable indicator of vegetation growth, while the MDWI is the index that best reflects the moisture content of vegetation. These two indices can be selected as classification features.

[0024] Furthermore, dryland irrigation completely depends on precipitation, and crop growth is closely related to precipitation, while irrigated land can be irrigated by other means. At the same time, temperature also has a serious impact on the growth of dryland crops. The inverse of the ratio of the normalized vegetation index to temperature (NDVI / -T), the inverse of the ratio of the water index to temperature (MDWI / -T), the ratio of the normalized vegetation index to precipitation (NDVI / P), and the ratio of the water index to precipitation (MDWI / P) can also be selected as classification features.

[0025] Step 03: Extract irrigated land based on classification features.

[0026] Specifically, the selected classification features and the irrigated land and dry land sample data are input into the random forest model to obtain the classification of dry land and irrigated land. Specifically, the normalized vegetation index and water index are input into the random forest model to obtain the classification results of irrigated land and dry land; the normalized vegetation index, water index, the ratio of the normalized vegetation index to temperature (NDVI / -T), the ratio of the water index to temperature (MDWI / -T), the ratio of the normalized vegetation index to precipitation (NDVI / P), and the ratio of the water index to precipitation (MDWI / P) are input into the random forest model as classification feature factors to obtain the classification results of irrigated land and dry land. In this embodiment, when the proportion of irrigated land in an administrative unit reaches more than 95%, the cultivated land characteristics of the area can be determined to be irrigated land; when the proportion of dry land reaches more than 95%, the cultivated land characteristics of the area can be determined to be dry land.

[0027] The method for extracting irrigated land in arid and semi-arid areas provided by the present invention can well extract irrigated land in arid and semi-arid areas, obtain high-precision cultivated land type data, and provide a detailed data basis for irrigation and water conservancy projects.

[0028] like Figure 2 As shown, in one embodiment of the present invention, the step of acquiring basic data of arid and semi-arid areas further includes: acquiring remote sensing data, environmental data and sample point data of the arid and semi-arid areas.

[0029] Specifically, the normalized vegetation index and water body index can be extracted from the remote sensing data. In this embodiment, the environmental data refers to temperature data and precipitation data, and the sample point data refers to irrigated land and dry land sample point data.

[0030] Furthermore, in one embodiment of the present invention, the step of obtaining remote sensing data of arid and semi-arid regions further includes: obtaining a normalized vegetation index and a water index during a crop growing season based on remote sensing data. The step of obtaining environmental data of arid and semi-arid regions further includes: obtaining temperature data and precipitation data for multiple preset time periods.

[0031] Specifically, in this embodiment, the crop growing season refers to April to September, the first half of each month is divided into one period, and the second half of each month is divided into one period. Obtaining the normalized vegetation index and water body index of the crop growing season means obtaining the normalized vegetation index and water body index of each crop period as basic data.

[0032] Acquiring temperature data and precipitation data for multiple preset time periods refers to: acquiring temperature data and precipitation data for a certain time period before remote sensing imaging, optionally, such as acquiring temperature data and precipitation data for 15 days before imaging, temperature data and precipitation data for 30 days, temperature data and precipitation data for 60 days, and temperature data and precipitation data from sowing time to shooting time as basic data.

[0033] It can be understood that the preset durations listed above can be set according to specific regional characteristics and are not limited to the durations listed in this embodiment.

[0034] In one embodiment of the present invention, the step of selecting classification features of irrigated land and dry land in arid and semi-arid areas based on basic data further includes: selecting the normalized vegetation index and water body index during the crop growing season as the first classification features of irrigated land and dry land.

[0035] Specifically, the normalized difference vegetation index and water body index in each period of the crop growing season, i.e., April to September, were used as the first classification features of irrigated land and dry land.

[0036] Furthermore, the normalized vegetation index and water body index in July and August during the crop growing season were selected as the second classification characteristics of irrigated land and dry land. Specifically, in July and August, the water and heat conditions are the best, and crops grow fastest. Most crops enter the maturity period in August and reach the peak of the normalized vegetation index. Throughout the growing season, the impact of long-term drought stress on vegetation growth and vegetation water content is concentrated in July and August. Starting from September, some crops enter the harvest stage, and the normalized vegetation index of crops has a downward trend.

[0037] Furthermore, the step of selecting classification characteristics of irrigated land and dry land in arid and semi-arid areas based on basic data also includes: selecting a first ratio of the normalized vegetation index to precipitation data, the negation of the second ratio of the normalized vegetation index to temperature data, a third ratio of the water body index to precipitation data, and the negation of the fourth ratio of the water body index to temperature as the third classification feature.

[0038] Specifically, dryland irrigation depends entirely on precipitation, so the normalized vegetation index and water index are positively correlated with precipitation in the arid and semi-arid regions of Northwest China, while this relationship is not obvious in irrigated land. On the one hand, the normalized vegetation index and water index of dryland are relatively low, because of the stress of insufficient water, so the ratio of the normalized vegetation index to precipitation and the ratio of the water index to precipitation are low. On the other hand, the utilization efficiency of precipitation in irrigated land is higher than that in dry land, so the ratio of the normalized vegetation index to precipitation and the ratio of the water index to precipitation can distinguish irrigated land from dry land.

[0039] In addition, the higher the temperature, the stronger the evaporation. The soil moisture content and vegetation moisture content will be affected. Dry land is under drought stress, and the higher the temperature, the worse the moisture condition and the worse the growth. However, irrigated land is not under drought stress, so this relationship is not obvious, and the higher the temperature, the better the growth because of its better heat. Therefore, the inverse of the ratio of the normalized vegetation index and the water index to the temperature can be used to distinguish between irrigated land and dry land.

[0040] In one embodiment of the present invention, the second classification feature, the first classification feature and the third classification feature, and the second classification feature and the third classification feature are input into a random forest model to obtain classification information of multiple irrigated lands and dry lands.

[0041] Specifically, by inputting the normalized vegetation index and water body index of the growing season into the random forest model, the irrigated land and dry land in the sample area can be classified; by inputting the normalized vegetation index and water body index of July and August into the random forest model, the irrigated land and dry land in the sample area can also be classified; by inputting the normalized vegetation index, water body index, the first ratio, the second ratio, the third ratio and the fourth ratio of the growing season into the random forest model, the irrigated land and dry land in the sample area can be classified; by inputting the normalized vegetation index, water body index, the first ratio, the second ratio, the third ratio and the fourth ratio of July and August into the random forest model, the irrigated land and dry land in the sample area can also be classified.

[0042] The first classification feature and the second classification feature are verified through the confusion matrix to obtain the preferred time features as July and August. The third classification feature is verified through the confusion matrix to obtain the optimal temperature and precipitation data for 30 days before shooting in multiple preset time periods.

[0043] Table 1 is the classification method information, as shown in the following table:

[0044] Table 1 Classification method information

[0045]

[0046]

[0047] As can be seen from the above table, only the normalized vegetation index and water body index of the crop growing season (10 periods) are input into the random forest model to obtain good results for the classification of irrigated land and dry land, with an overall accuracy and kappa coefficient of 0.824 and 0.632 respectively. After the normalized vegetation index, water body index, and the first ratio, second ratio, third ratio, and fourth ratio of the growing season are input into the random forest model, the classification accuracy of irrigated land and dry land will be improved, and the overall accuracy and kappa coefficient will reach 0.857 and 0.697 respectively. In the screening and classification of each key period, the normalized vegetation index and water body index from the end of June to the beginning of August are used to try classification, and the overall accuracy of 0.847 and the kappa coefficient of 0.684 are obtained. Good results, which are close to the classification accuracy of the growing season. Then, the three months of June / July / August are classified separately. It can be seen that the accuracy of July and August is higher, which is the most distinguishing characteristic period. The classification results of irrigated land and dry land were obtained by using the four-period data of July and August and the classification of the last half of July and the first half of August (two periods), and the normalized vegetation index, water body index, first ratio, second ratio, third ratio and fourth ratio were all input into the random forest model. The overall accuracy and Kappa coefficient of the classification results of irrigated land and dry land were high, which were 0.851, 0.687 and 0.860, 0.710 respectively.

[0048] In the comparison of various classification methods, the classification results with the first ratio, second ratio, third ratio and fourth ratio were all more accurate than those without them. The overall accuracy was increased by 4.03% on average, while the Kappa coefficient was increased by 8.54% on average, indicating that the classification accuracy of irrigated land and dry land was improved after adding environmental data, and the user accuracy and producer accuracy were more balanced.

[0049] In general, the classification results of the three types of classification, namely, from late June to early August, July / August (4 periods), and late July and early August (2 periods), are good, and are close to the classification accuracy of the growing season. However, the classification accuracy of only using the normalized vegetation index and water body index during this period is worse than that of the growing season, but after introducing the first ratio, the second ratio, the third ratio, and the fourth ratio, the classification accuracy is close to that of the growing season. It can be seen that adding precipitation data and temperature data can use fewer periods and still maintain good classification results.

[0050] The method for extracting irrigated land in arid and semi-arid areas provided by an embodiment of the present invention uses the normalized vegetation index, the water body index, the first ratio of the normalized vegetation index to precipitation data, the second ratio of the normalized vegetation index to temperature data, the third ratio of the water body index to precipitation data, and the fourth ratio of the water body index to temperature data as classification features. It is possible to use fewer classification factors to classify dry land and irrigated land in arid and semi-arid areas, thereby improving the extraction efficiency and classification accuracy of irrigated land.

[0051] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for extracting irrigated land in arid and semi-arid areas, characterized in that: include: Obtain basic data on arid and semi-arid regions; Selecting classification characteristics of irrigated land and dry land in arid and semi-arid regions based on the basic data; Extracting the irrigated land based on the classification features; The step of obtaining basic data of arid and semi-arid regions further comprises: obtaining remote sensing data, environmental data and sample point data of arid and semi-arid regions; The step of acquiring remote sensing data of arid and semi-arid regions further includes: acquiring a normalized vegetation index and a water body index during a crop growing season based on the remote sensing data; The step of obtaining environmental data of arid and semi-arid regions further includes: obtaining temperature data and precipitation data of a plurality of preset time periods; The step of selecting classification features of irrigated land and dry land in arid and semi-arid regions based on the basic data further includes: selecting the normalized difference vegetation index and the water body index in the crop growing season as the first classification features of irrigated land and dry land; The step of selecting the normalized vegetation index and the water body index in the crop growing season as the first classification feature of irrigated land and dry land further includes: selecting the normalized vegetation index and the water body index in July and August in the crop growing season as the second classification feature of irrigated land and dry land; The step of selecting the classification features of irrigated land and dry land in arid and semi-arid areas based on the basic data further includes: selecting a first ratio of the normalized vegetation index to the precipitation data, the inverse of a second ratio of the normalized vegetation index to the temperature data, a third ratio of the water body index to the precipitation data, and the inverse of a fourth ratio of the water body index to the temperature data as the third classification feature; The step of extracting the irrigated land based on the classification feature further includes: inputting the second classification feature, the first classification feature and the third classification feature, and the second classification feature and the third classification feature into a random forest model respectively to obtain classification information of multiple irrigated lands and dry lands.

2. The method for extracting irrigated land in arid and semi-arid areas according to claim 1, characterized in that: Also includes: The first classification feature and the second classification feature are verified to obtain a preferred time feature.

3. The method for extracting irrigated land in arid and semi-arid areas according to claim 1, characterized in that: Also includes: The third classification feature is verified to obtain an optimal preset duration.