A method for optimizing area of regional forest, farmland and grassland in a semi-arid sandy area and proportion of the area

Abstract

The application discloses a kind of semi-arid sandy area regional forest field grass sand area and its proportion optimization method, comprising: the relationship between soil moisture deep leakage and rainfall, vegetation coverage under different land use types is established;The estimation model of the total amount of soil moisture deep leakage of the sand area to be optimized is constructed, according to the land area of forest field grass sand and the amount of soil moisture deep leakage, the total amount of soil moisture deep leakage of the sand area to be optimized is estimated every year;The linear model of the total amount of soil moisture deep leakage of the sand area to be optimized is constructed, under the constraint condition of ensuring the water resource safety of the sand area to be optimized, the optimization area and its proportion of forest field grass sand in the sand area to be optimized are inversed by applying machine learning regression model.The application solves the problem that, in the current northern semi-arid sandy area management, the relationship between regional water resource utilization and ecological restoration of main land use types such as forest, grass, sand and field is not considered as a whole, and the stability of sand-fixing vegetation and the safety of regional water resources cannot be ensured.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of sand land protection and management, and particularly relates to a method for optimizing the area and proportion of forest, farmland, grassland and sand land in a semi-arid sand area, which is suitable for optimizing the area proportion of forest, farmland, grassland and sand land in a semi-arid sand area in the north of China. BACKGROUND

[0002] In the vast desert area in the north of China, water resource is a core factor for determining the regional vegetation construction, ecological restoration and sustainable development of social economy. Under the premise of fully considering the restriction of regional water resource, how to optimize the area proportion of forest, farmland, grassland and sand land in a sand area to improve the balance between the utilization and supply of regional water resource has important theoretical value and practical significance for realizing the management and restoration of ecological environment in the sand area.

[0003] In the past desert theoretical research and management practice, due to the restriction of multiple factors such as theoretical cognition, industry isolation and technical means, the forestry, agriculture and animal husbandry, water conservancy and environment departments mainly focus on the research and practice of industry or local problems, and do not comprehensively consider the relationship between the regional water resource restriction and ecological environment management by taking into account the main land use types such as forest land, grassland, bare sand and farmland. With the rapid increase of sand area management, development and population in recent years, the area proportion of farmland with high water consumption increases rapidly, while the area proportion of bare sand with high deep water infiltration decreases sharply. In addition, the negative influence of climate change on water resource input, which leads to the destruction of the balance between the utilization and supply of regional water resource, and actual problems such as the obvious decrease of underground water level, the recession of artificial sand-fixing vegetation and the activation of fixed dunes. Therefore, under the premise of fully considering the balance of water resource supply in the sand area, how to determine the optimization configuration method of the area proportion of forest, farmland, grassland and sand land in a small area in the sand area has become a key problem for realizing the ecological restoration and water resource safety in the sand area. SUMMARY

[0004] The application provides a method for optimizing the area and proportion of forest, farmland, grassland and sand land in a semi-arid sand area, so as to solve the problem that the relationship between the regional water resource utilization and ecological environment management of the main land use types such as forest land, grassland, bare sand and farmland is not considered comprehensively in the current sand area management, and the stability of sand-fixing vegetation and water resource safety cannot be ensured.

[0005] The method for optimizing the area and proportion of forest, farmland, grassland and sand land in a semi-arid sand area provided by the application has the characteristics that the optimization method comprises the following steps.

[0006] Step one, determining the deep water infiltration of soil in the forest land, farmland, grassland and bare sand land in the sand area to be optimized, and establishing the relationship between the deep water infiltration of soil and rainfall and vegetation coverage under different land use types;

[0007] Step two, construct an estimation model of the total amount of soil moisture deep percolation in the sand area to be optimized, estimate the total amount of soil moisture deep percolation in the sand area to be optimized every year according to the area and soil moisture deep percolation of forest land, farmland, grassland and bare sand land;

[0008] Step three, construct a linear model of the total amount of soil moisture deep percolation in the sand area to be optimized, based on the estimated value of the total amount of soil moisture deep percolation in the sand area to be optimized every year, apply machine learning regression model to inverse the optimized area and its proportion of forest land, farmland, grassland and bare sand land in the sand area to be optimized under the constraint condition of ensuring the safety of water resources in the sand area to be optimized.

[0009] Further, the expression of the relationship between the soil moisture deep percolation and the rainfall and vegetation coverage under different land types is:

[0010] Y i =a1×X1+a2×X2+a0 (1)

[0011] In the formula: Y i is the soil moisture deep percolation of different land types, mm; i is the different utilization land types of forest land, farmland, grassland and bare sand land; X1 is the vegetation coverage, %; X2 is the rainfall, mm; a0, a1 and a2 are fitting parameters.

[0012] Further, the expression of the estimation model is:

[0013] D t =b1Y1+b2Y2+b3Y3+b4Y4 (2)

[0014] In the formula: D t is the total amount of soil moisture deep percolation in the sand area to be optimized every year, mm; Y1, Y2, Y3 and Y4 are the soil moisture deep percolation of forest land, farmland, grassland and bare sand land, respectively, mm; b1, b2, b3 and b4 are the area of forest land, farmland, grassland and bare sand land in the sand area to be optimized, respectively, km 2 .

[0015] Further, the expression of the linear model is:

[0016]

[0017] Wherein: and are the area weight vector and the percolation characteristic vector of forest land, farmland, grassland and bare sand land, respectively:

[0018]

[0019]

[0020] In the formula: ω 11 , ω 21 , ω 31 , ω 41 are the areas of forest land, farmland, grassland and bare sand land in the sand area to be optimized in the first year, respectively, km 2 ; ω 1n , ω 2n , ω 3n , ω 4n are the areas of forest land, farmland, grassland and bare sand land in the sand area to be optimized in the nth year, respectively, km 2 ; Y t11 , Y t21 , Y t31 , Y t41 are the unit area infiltration amounts of forest land, farmland, grassland and bare sand land in the sand area to be optimized in the first year, respectively, mm; Y t1n , Y t2n , Y t3n , Y t4n are the unit area infiltration amounts of forest land, farmland, grassland and bare sand land in the sand area to be optimized in the nth year, respectively, mm; the unit area infiltration amount is obtained by formula (1).

[0021] Further, the constraint condition in the linear model is to ensure the minimum infiltration amount of the sand area to be optimized according to the minimum annual rainfall.

[0022] Further, in the application of the machine learning regression model inversion process, the estimated value of the total amount of soil moisture deep infiltration of the sand area to be optimized each year is used as much as possible.

[0023] Compared with the prior art, the sand area forest land grass sand area and the optimization method of the proportion thereof provided by the present application are suitable for the optimization and configuration of the areas and proportions of forest land, farmland, grassland and sand in the semi-arid and semi-humid sand area in the north, are based on the long-term research and accumulation of deep soil moisture infiltration and groundwater recharge of sand, and are combined with the low coverage sand control theory and technology, and creatively propose the natural constraint rule of considering the area proportion of main types such as forest land, farmland, grassland and sand in the sand area and the balance of water resource supply, to determine the optimization area and proportion of main land use types in the sand area, and finally realize the win-win of water resource safety and ecological restoration in the sand area. BRIEF DESCRIPTION OF DRAWINGS

[0024] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following will briefly introduce the drawings needed to be used in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only exemplary, and other drawings can be obtained by the provided drawings without creative labor for those skilled in the art.

[0025] Figure 1 The figure shows the process of the area optimization method of the forest, farmland, grassland and sand area in the semi-arid sand region in the embodiments of the present application. DETAILED DESCRIPTION

[0026] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0027] At present, in the sand region management planning and practice in China, although the areas of the main land use types such as forest, farmland, grassland and sand are overall planned and designed, they do not consider the water transmission and conversion law between natural precipitation, soil moisture and groundwater system, and do not involve the quantitative relationship between water resource consumption and groundwater recharge of various land use types, which cannot ensure the stability of sand-fixing vegetation and water resource safety, and cannot realize the comprehensive goal of regional ecological, social and economic sustainable development.

[0028] And the deep soil water infiltration refers to all the water fluxes below the soil root layer reaching the groundwater level, which is the main way to recharge groundwater, and links the soil water dynamics below the root zone in the aeration zone and the land hydrological cycle, and plays an important hydrological and ecological benefit in different regions.

[0029] The present application uses the deep soil water infiltration (200cm) in the growth season to represent the index of the water resource utilization and balance relationship under the conditions of the main land use types such as forest, farmland, grassland and sand, to quantify the relationship between rainfall, soil moisture and groundwater recharge under each land use type.

[0030] The present application discloses an area optimization method of forest, farmland, grassland and sand area in a semi-arid sand region and the proportion thereof, which comprises the following steps.

[0031] Step one, determining the deep soil water infiltration of the forest land, farmland, grassland and bare sand land in the sand region to be optimized, and establishing the relationship between the deep soil water infiltration and the rainfall and vegetation coverage under different land use types, which is expressed as: Y = aX + b

[0032] Y i = a1 x X1 + a2 x X2 + a0 (1)

[0033] In the formula, Y i is the soil water deep seepage of different land use types, mm; i is the forest land, farmland, grassland and bare sand land of different land types; X1 is the vegetation coverage, %; X2 is the rainfall, mm; a0, a1 and a2 are fitting parameters. The to-be-optimized sand area region can be considered as a township, county, town and the like. In addition, X1 and X2 in formula (1) are parameters for different land use types in the same year, and X1 and X2 in different years and different land types are different. The fitting parameters a0, a1 and a2 of formula (1) can be obtained by using a binary linear regression analysis method through statistics of rainfall, vegetation coverage and deep seepage data in the to-be-optimized region for many years.

[0034] According to the definition of soil deep seepage, the soil water seeps downward in the soil layer below the plant root layer. In the sand land, it is generally considered that the deep seepage water cannot be directly absorbed and utilized by the plant root system, and it is also difficult to rise and evaporate due to capillary action, so it can only seep downward to the underground saturated layer, thereby achieving the effect of recharging underground water. Therefore, according to the characteristics of soil hydrology of the sand land, the soil deep seepage is considered to be equivalent to the underground water recharge. In addition, the relationship between natural rainfall, vegetation coverage of different land use types and deep seepage given by formula (1) can reflect the quantitative relationship between rainfall, soil water and underground water recharge under different land use types.

[0035] Step two, constructing an estimation model of the total amount of soil water deep seepage of the to-be-optimized sand area region, estimating the total amount of soil water deep seepage of the to-be-optimized sand area region each year according to the area and soil water deep seepage of the forest land, farmland, grassland and bare sand land.

[0036] The expression of the estimation model is:

[0037] D t = b1 Y1 + b2 Y2 + b3 Y3 + b4 Y4 (2)

[0038] In the formula, D t is the total amount of soil water deep seepage of the to-be-optimized sand area region each year, mm; Y1, Y2, Y3 and Y4 are the soil water deep seepage of the forest land, farmland, grassland and bare sand land, respectively, mm; b1, b2, b3 and b4 are the area of the forest land, farmland, grassland and bare sand land in the to-be-optimized sand area region, respectively, km 2 . It should be noted that D tThe estimated value of the total deep soil water infiltration amount of a certain year, wherein the year is not marked with parameters, and different years of related parameters Y and b can be replaced to obtain the estimated value of the total deep soil water infiltration amount of the corresponding year.

[0039] Step three, constructing a linear model of the total deep soil water infiltration amount of the sand region to be optimized, under the constraint condition of ensuring the water resource safety of the sand region to be optimized, based on the estimated value of the total deep soil water infiltration amount of the sand region to be optimized each year, applying a machine learning regression model to inverse the optimized area and its proportion of the forest land, farmland, grassland and bare sand land in the sand region to be optimized.

[0040] The expression of the linear model is:

[0041]

[0042] Among them: and are the area weight vector and the infiltration characteristic vector of the forest land, farmland, grassland and bare sand land, respectively:

[0043]

[0044]

[0045] In the formula: ω 11 , ω 11 , ω 11 , ω 11 are the area of the forest land, farmland, grassland and bare sand land in the sand region to be optimized in the first year, respectively, km 2 ; ω 1n , ω 1n , ω 1n , ω 1n are the area of the forest land, farmland, grassland and bare sand land in the sand region to be optimized in the nth year, respectively, km 2 ; Y t11 , Y t21 , Y t31 , Y t41 are the unit area infiltration amount of the forest land, farmland, grassland and bare sand land in the sand region to be optimized in the first year, respectively, mm; Y t1n , Y t2n , Y t3n , Y t4n are the unit area infiltration amount of the forest land, farmland, grassland and bare sand land in the sand region to be optimized in the nth year, respectively, mm; wherein the unit area infiltration amount can be obtained by formula (1), that is, the deep soil water infiltration amount in formula (1) is the deep soil water infiltration amount per unit area.

[0046] The constraint condition in the linear model is to ensure the minimum leakage of the to-be-optimized sand area region according to the minimum annual rainfall. In the process of applying the machine learning regression model inversion, as many years as possible are used to estimate the total amount of deep soil moisture leakage of the to-be-optimized sand area region every year, such as the deep soil moisture leakage of the to-be-optimized sand area region in the past 40 years, so as to improve the inversion accuracy.

[0047] In the optimization method of the above area and its proportion, based on the systematic understanding of the change rule of deep soil infiltration of the main land types such as forest land, grassland, farmland and bare sand land in the semi-arid and semi-humid sand area in the early stage, the quantitative relationship between soil infiltration and rainfall and vegetation coverage under the condition of each type can be established. Taking the sand area region as the research object, according to the relationship among rainfall, vegetation coverage and infiltration, the deep soil moisture leakage per unit area in the small area in the past nearly 40 years and the estimated value of the total deep soil moisture leakage in the to-be-optimized area every year can be estimated, and the optimization area of forest, field, grass and sand in the region can be determined by applying the machine learning regression model inversion, and then the proportion of the area of different land use types (the proportion of each type area can be easily realized by computer by calculating the area of each land use type, so the calculation of the proportion is not repeated) is obtained, which provides a basis for realizing the integrated protection and systematic management of forest, field, grass and sand in the semi-arid sand area.

[0048] The scheme of the present application is guided by the concept of "respecting nature, adapting to nature and protecting nature", based on the long-term research and accumulation of deep soil moisture infiltration and groundwater recharge of sand land, and combined with the low coverage sand control theory and technology, creatively proposes the natural constraint rule of considering the area proportion of main types such as forest, field, grass and sand in a small area and the balance of water resource supply, to determine the optimization area and proportion of main land use types in a small area, and finally realize the win-win of water resource safety and ecological restoration in the sand area.

[0049] Implementation case

[0050] A method for optimizing the area and proportion of forest, field, grass and sand in a semi-arid sand area region - taking Wushenqi Galutu Town in the Mu Us Sand Land as an example.

[0051] Wushenqi Galutu Town is the seat of Wushenqi Government in Ordos City, Inner Mongolia Autonomous Region, located in the hinterland of Mu Us Sand Land, with a regional area of about 2328.6km 2 ; The population is 989,000, and the main industries are characteristic planting, animal husbandry, breeding and other industries.

[0052] According to statistics, the forest land of Garutou Town covers an area of 50.76 km2, accounting for 2.18% of the total area; the farmland covers an area of 19.13 km2, accounting for 0.82% of the total area; the grassland covers an area of 1016.86 km2, accounting for 43.67% of the total area; the sandy land covers an area of 1010.26 km2, accounting for 43.38% of the total area; and other land covers an area of 231.62 km2, accounting for 9.95%. That is, the four types of land use, i.e. forest, farmland, grassland and sandy land, account for 90.05% of the total area, so it is an ideal test area for the balanced development of regional water resources and various types of land use. Taking Garutou as an example, the optimization method of the area and proportion of forest, farmland, grassland and sandy land in a small area is studied as follows:

[0053] An optimization method of the area and proportion of forest, farmland, grassland and sandy land in a semi-arid sandy area, comprising the following steps:

[0054] (1) First, obtain the relevant data of rainfall, coverage and infiltration of each year under the land use types of forest, farmland, grassland and sandy land in the Mu Us Sandy Land through consulting relevant literature and other methods, and then apply the binary linear regression analysis method to obtain the parameters a0, a1 and a2 listed in formula 1, so as to establish the quantitative relationship between soil infiltration and rainfall under different land use types of forest land, farmland, grassland and bare sandy land;

[0055] Y i =-98.675-248.056×X1+0.724×X2(R 2 =0.643;P<0.05) (1)

[0056] In the formula: Y i is the deep soil water infiltration of different land types (mm); X1 is the vegetation coverage (%); X2 is the rainfall (mm); P is the significance level of statistical test, P < 0.05 indicates that formula (1) is statistically valid and reliable.

[0057] In this embodiment, the annual rainfall data is from the China Meteorological Data Network (http: / / data.cma.cn / ). The vegetation coverage of forest, farmland, grassland and sandy land in each year is calculated according to the formula on page 24 of “Response of Vegetation Coverage Change in Mu Us Sandy Land to Climate Change” (Feng Ying, 2015, Master's Degree Thesis of North Forest), and the specific formula is as follows:

[0058] A 1i =-0.001*B i +0.63

[0059] A 2i =0.01*B i +0.58

[0060] A 3i = 0.01 * B i + 0.4

[0061] A 4i = 0.004 * B i + 0.19

[0062] In the formula: A 1i , A 2i , A 3i and A 4i are the vegetation cover of forest, farmland, grassland and sand in the i-th year, respectively; B i refers to the year, and 1 is set in 1980, and then increases by 1 every year until 2022.

[0063] (2) Construct an estimated model of soil water infiltration in the sand area to be optimized, calculate the estimated value of the total annual deep infiltration of soil water in Galutu Town from 1981 to 2022 according to the area of the four main land types of forest, farmland, grassland and sand and the deep soil water infiltration obtained in step (1), and the results are shown in Table 1.

[0064] Table 1 Estimated value of total annual deep infiltration of soil water in Galutu Town from 1981 to 2022 (unit: 10 3 m 3 )

[0065]

[0066] Note: The infiltration amount is negative, which means that due to the low rainfall in the year (generally less than 300 mm), the soil layer in the sand area not only has no deep infiltration, but also consumes the water storage in the soil.

[0067] (3) On the premise of ensuring deep infiltration of soil in Galutu Town (annual precipitation is generally greater than 300 mm), that is, excluding data with estimated total infiltration less than 0, based on the unit area infiltration obtained in step (1) and the annual deep annual infiltration estimated value obtained in step (2), the optimized area of forest, farmland, grassland and sand in Galutu Town in the past 42 years is estimated by applying a machine learning regression model, and the proportion of the optimized area is 14.798%, 0.037%, 53.002% and 32.164%, respectively.

[0068] The above examples are only exemplary embodiments of the present application and are not intended to limit the present application, and the protection scope of the present application is defined by the claims. Those skilled in the art can make various modifications or equivalent replacements to the present application within the spirit and protection scope of the present application, and such modifications or equivalent replacements shall also be considered to fall within the protection scope of the present application.

Claims

1. A method for optimizing the area of semi-arid sandy region forest, farmland and grassland and the proportion of each area, characterized in that: The optimization method comprises: Step one, determining the deep soil water seepage of forest land, farmland, grassland and bare sand land in the sand area to be optimized, and establishing the relationship between the deep soil water seepage and rainfall and vegetation coverage under different land use types, and the relationship expression is: Y i = a1xX1+ a2xX2+ a0 (1); where: Y i is the deep soil water infiltration of different land types; i is the different land use types of forest land, farmland, grassland and bare sand land; X1 is the vegetation coverage; X2 is the rainfall; a0, a1 and a2 are fitting parameters; Step two, constructing an estimation model of the total deep soil water seepage of the sand area to be optimized, and estimating the total deep soil water seepage of the sand area to be optimized each year according to the area and deep soil water seepage of forest land, farmland, grassland and bare sand land; The expression of the estimation model is: D t = b1Y1+ b2Y2+ b3Y3+ b4Y4 (2); In the formula: D t is the total annual deep soil moisture infiltration of the sand area region to be optimized; Y1, Y2, Y3 and Y4 are the deep soil water seepage of forest land, farmland, grassland and bare sand land respectively; b1, b2, b3 and b4 are the areas of forest land, farmland, grassland and bare sand land in the sand area to be optimized respectively; Step three, constructing a linear model of the total deep soil water seepage of the sand area to be optimized, and based on the estimated value of the total deep soil water seepage of the sand area to be optimized each year, the optimized area and proportion of forest land, farmland, grassland and bare sand land in the sand area to be optimized are inversed by applying a machine learning regression model under the constraint condition of ensuring water resource safety of the sand area to be optimized; The expression of the linear model is: （3）； where: and are the area weight vectors and the leakage characteristic vectors for forest land, farmland, grassland, and bare sand land, respectively: （4）； （5）； In the formula: ω 11 , ω 21 , ω 31 , ω 41 are the areas of forest land, farmland, grassland and bare sand land in the sand area to be optimized in the first year, respectively. ω 1n , ω 2n , ω 3n , ω 4n are the areas of forest land, farmland, grassland and bare sand land in the sand area to be optimized in the nth year, respectively. Y t11 , Y t21 , Y t31 , Y t41 The unit area infiltration amounts of the forest land, farmland, grassland and bare sand land in the sand area to be optimized in the first year are in turn Y t1n , Y t2n , Y t3n , Y t4n The unit area infiltration of each type of land in the sand area to be optimized in the nth year, in sequence, is woodland, farmland, grassland, and bare sand land; the unit area infiltration is obtained by formula (1).

2. The optimization method of the area and proportion of forest land, farmland, grassland and bare sand land in a semi-arid sand area region according to claim 1, characterized in that: The constraint condition in the linear model is to ensure the minimum seepage of the sand area to be optimized according to the minimum annual rainfall.

3. The optimization method of the area and proportion of forest land, farmland, grassland and bare sand land in a semi-arid sand area region according to claim 1, characterized in that: In the process of applying the machine learning regression model to inversion, as many estimated values of the total deep soil water seepage of the sand area to be optimized each year as possible are used.

Images