CN109342697A - Based on random forest-normal stabilizing pile soil organic carbon prediction technique - Google Patents
Based on random forest-normal stabilizing pile soil organic carbon prediction technique Download PDFInfo
- Publication number
- CN109342697A CN109342697A CN201811491043.9A CN201811491043A CN109342697A CN 109342697 A CN109342697 A CN 109342697A CN 201811491043 A CN201811491043 A CN 201811491043A CN 109342697 A CN109342697 A CN 109342697A
- Authority
- CN
- China
- Prior art keywords
- soil organic
- organic carbon
- random forest
- soil
- environmental factor
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/24—Earth materials
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Pure & Applied Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Environmental & Geological Engineering (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Probability & Statistics with Applications (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Operations Research (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Geology (AREA)
- Remote Sensing (AREA)
- Algebra (AREA)
- Evolutionary Biology (AREA)
- Food Science & Technology (AREA)
- Medicinal Chemistry (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to be based on random forest-normal stabilizing pile soil organic carbon prediction technique.First, pedotheque organic carbon is measured；Secondly, environmental factor is extracted and screens environmental factor relevant to soil organic carbon；Finally, the spatial distribution of soil organic carbon is predicted based on random forest-Ordinary Kriging Interpolation model and correlative environmental factors.The method of the present invention realizes the spatial prediction of soil organic carbon by means of the relationship between soil organic matter and environmental factor (landform, weather, vegetation).
Description
Technical field
The present invention relates to remote sensing technique application fields, and it is organic to be based particularly on random forest-normal stabilizing pile soil
Carbon content prediction technique.
Background technique
Due to the presence of regional soil attribute spatial and temporal variation characteristic, so that traditional sampling and drafting method are difficult accurate characterization
The spatial distribution characteristic and continuity of soil organic matter, it is difficult to obtain high precision soil organic carbon information.In recent years, with remote sensing
The fast development of information technology, it is different that digitlization soil cartography (digital soilmapping, DSM) becomes accurate acquisition space
The effective means of the big regional soil distribution of organic of matter.DSM is mainly using Soil-landscape model as theoretical basis at present, i.e.,
The environmental factor closely related with Zinc fractions is obtained by " 3S " technology and computer technology, is simulated using mathematical model
Quantitative relationship between soil and environmental factor, and this relationship is extrapolated to area of space, realize the space of soil organic matter
Forecast of distribution.Have numerous statistical models and be applied to soil organic matter forecasting research, these models mostly use greatly conventional statistics
Or geo-statistic method.With the continuous development of artificial intelligence technology, more and more scholars carry out soil using machine learning model
The research of relationship between earth attribute and its environmental factor, random forest (randomforest, RF) model is powerful with its in the middle
Nonlinear fitting ability becomes one of prediction effective method of soil organic matter spatial distribution.However, learning mould with other machines
Type is the same, and Random Forest model only considers the relationship between soil organic carbon and environmental factor, has ignored the space of variable
Autocorrelation, and then influence soil organic matter precision of prediction.Thus, the spatial autocorrelation effect between variable how is effectively treated
It is of great significance to the accurate prediction for realizing soil organic carbon.
Summary of the invention
The purpose of the present invention is to provide be based on random forest-normal stabilizing pile soil organic carbon prediction side
Method realizes that the space of soil organic matter is pre- by means of the relationship between soil organic matter and environmental factor (landform, weather, vegetation)
It surveys, realizes the accurate prediction of soil organic matter and the distribution of related soil attribute space.
To achieve the above object, the technical scheme is that being had based on random forest-normal stabilizing pile soil
Machine carbon content prediction technique, includes the following steps:
Step S1, pedotheque organic carbon is measured；
Step S2, environmental factor is extracted and screens environmental factor relevant to soil organic matter；
Step S3, random forest-Ordinary Kriging Interpolation model and correlative environmental factors are based on to the sky of soil organic carbon
Between be distributed and predicted.
Further, the step S1 specific implementation are as follows: with laying sample, pedotheque and correspondence are geographical with acquiring sample
Coordinate measures soil organic carbon using potassium bichromate-Outside Heating Method.
Further, the specific implementation steps are as follows by the step S2:
Step S21, it is based on remotely-sensed data and weather basic data, extracts environmental factor related with soil organic matter, packet
Include terrain factor, Vegetation factors, climatic factor；
Step S22, precision of prediction of the error OOB error to soil organic matter outside the bag to reduce Random Forest model
It impacts, judges whether the environmental factor retains by rejecting the monotonicity of the OOB error after environmental factor one by one, if
OOB error increase then retains the environmental factor, on the contrary then reject.
Further, in step S21, the terrain factor includes height above sea level, the gradient；The Vegetation factors include that normalization is planted
By index, green normalized differential vegetation index, soil can adjust vegetation index, amendment type soil adjusts vegetation index, canopy structure
Insensitive vegetation index, triangle vegetation index；The climatic factor includes average annual rainfall, average annual temperature.
Further, the specific implementation steps are as follows by the step S3:
Step S31, modeling sample collection and test samples collection are divided into all soil sampling data；
Step S32, modeling sample collection data are based on, are simulated between soil organic matter and environmental factor using Random Forest model
Mathematical relationship, and make spatial prediction distribution map；
Step S33, the residual values of sampling point are obtained according to soil organic carbon measured value and Random Forest model predicted value,
Ordinary Kriging Interpolation interpolation is carried out to the residual values of soil organic carbon；
Step S34, by the soil organic carbon predicted value based on Random Forest model and based on common lattice league (unit of length) method
Residual error estimated value carries out space and adds operation；
Step S35, each sampling point environmental factor data based on test samples collection, using random forest-Ordinary Kriging Interpolation mould
Type predicts soil organic carbon, and is compared with actual measurement soil organic carbon, determines and is based on common gram of random forest-
The validity of league (unit of length) model prediction soil organic matter；
Step S36, soil is adopted using random forest-ordinary Kriging for the environmental factor in non-sampled area
Survey region where sample carries out the prediction of soil organic matter spatial distribution.
Further, in the step S32, the mathematical relationship between soil organic matter and environmental factor is as follows:
In formula:For spatially i-th point of the predicted value of the soil organic carbon based on Random Forest model；
a1,a2,…,anFor environmental factor；Mathematical relationship of the f between soil organic carbon and environmental factor.
Further, in the step S33, soil organic carbon measured value and Random Forest model predicted value obtain sample
The residual values formula of point and Ordinary Kriging Interpolation interpolation formula is carried out respectively such as formula (2), (3) to the residual values of soil organic carbon
It is shown:
In formula: r (xi) it is soil organic carbon spatially i-th point of residual values；C(xi) contain for soil organic matter
Measure spatially i-th point of measured value；Spatially for the soil organic carbon based on Ordinary Kriging Interpolation model
I-th point of residual error estimated value；N is the quantity of known eyeball；wiFor spatially i-th point of weighted value；
Convolution (2), (3) obtain in the step S34, the soil organic carbon prediction based on Random Forest model
Value carries out space with the residual error estimated value based on common lattice league (unit of length) method and adds the formula of operation as follows:
In formula,For based on random forest-normal stabilizing pile soil organic carbon spatially i-th
The predicted value of a point.
Compared to the prior art, the invention has the following advantages: the method for the present invention, by means of soil organic matter and ring
Relationship between the border factor (landform, weather, vegetation), the relationship based on Random Forest model simulation soil organic matter and environmental factor
And the space distribution situation of normal stabilizing pile simulation residual values, it realizes the spatial prediction of soil organic matter, solves single machine
Device learning model fails to consider the spatial autocorrelation effect of variable, is soil organic matter and the essence that related soil attribute space is distributed
Quasi- prediction provides a kind of technology.
Detailed description of the invention
Fig. 1 is that area's research area geographical location and field soil sampling division figure are studied in the embodiment of the present invention.
Fig. 2 is not consider the autocorrelative soil organic matter spatial distribution map of the variable space in the embodiment of the present invention.
Fig. 3 is the Semivariance model figure of soil organic matter residual values in the embodiment of the present invention.
Fig. 4 is soil organic matter residual values spatial distribution map in the embodiment of the present invention.
Fig. 5 is that (left figure is based on random gloomy for soil organic matter measured value and predicted value regression analysis in the embodiment of the present invention
It is that woods model obtains as a result, right figure is the result obtained based on random forest-Ordinary Kriging Interpolation model).
Fig. 6 is to be distributed in the embodiment of the present invention based on random forest-normal stabilizing pile soil organic matter spatial prediction
Figure.
Specific embodiment
With reference to the accompanying drawing, technical solution of the present invention is specifically described.
The present invention provides random forest-normal stabilizing pile soil organic carbon prediction technique is based on, including such as
Lower step:
Step S1, pedotheque organic carbon is measured；
Step S2, environmental factor is extracted and screens environmental factor relevant to soil organic matter；
Step S3, random forest-Ordinary Kriging Interpolation model and correlative environmental factors are based on to the sky of soil organic carbon
Between be distributed and predicted.
Specific implementation process of the invention is illustrated below in conjunction with specific embodiment.
Embodiment:
The specific research area of selection is present embodiments provided to be illustrated.Research area's overview: research area is located at Fujian governor
The river Ting Xian Tian Town is located in Wuyi Mountains (25 ° of 33 ' 48 ' N of N~25 °, 116 ° of 18 ' 31 ' E of E~116 °), and the gross area is
296km2, wherein mountainous region area 213km2, main advantage tree species have masson pine (Pinus massoniana) and China fir
(Cunninghamia lanceolata).Area category middle subtropical zone monsoon climate, 17.5 DEG C~18.8 DEG C of average temperature of the whole year, year
Average rainfall 1700mm, for topography and geomorphology based on low mountains and hills, soil types is southern typical red soil mound based on red soil
Ling Qu.Due to historical reasons, area mountainous region natural vegetation is studied by serious damage, severe water and soil erosion becomes Southern Red Soil water
One of the region of soil flow mistake most serious.In recent years by closing hillsides to facilitate afforestation, the water-and-soil conservation measures such as low yield forest rebuilding, make to study area
Vegetation obtains certain recovery, creates advantage for the storage of topsoil organic carbon.
(1) soil organic matter measures
In January, 2015 carries out collecting soil sample work in research area, is taken into account with superiority and inferiority as principle, in research area setting 59
A standard site (Fig. 1), while using 400 handhold GPS of Mai Zhelun Neptune acquisition sample coordinate data.It is adopted in each sample ground
The soil about 500g for collecting 0~20cm depth is packed into polybag, and takes back laboratory and allow its natural air drying.Soil sample natural air drying in bag
Afterwards, part soil sample is taken to be ground up, sieved, the measurement for soil organic carbon.Using potassium bichromate-Outside Heating Method measurement soil
Earth organic carbon content.
(2) environmental factor is extracted and is screened
First, it is based on remotely-sensed data and weather basic data, extracts environmental factor related with soil organic matter.
1. using data:
Remotely-sensed data: river Tian Town transit time is January 21, orbit number 120-042, multi light spectrum hands space in 2015
The Landsat-8OLI image data that resolution ratio is 30 meters.River Tian Town digital elevation model (digital elevation
Model, DEM) data, spatial resolution 30m.
Meteorological data: the rainfall of Fujian Province Changting County river Tian Town offer, temperature record.
2. environmental factor is extracted:
Based on soil genesis theory, obtaining in terms of landform, vegetation, weather three influences soil organic carbon
Environmental factor, wherein terrain factor is mainly obtained by DEM；Vegetation index is the important parameter for characterizing soil organic carbon,
6 vegetation index indexs in close relations with soil organic matter are extracted by remote sensing image；Weather is to cause regional soil organic
An important factor for carbon content spatial variability, the 14 small towns rainfall equal every year and average annual temperature provided by Changting County weather bureau
The research area climatic factor data that spatial resolution is 30m are obtained by inverse distance weighted interpolation.Varying environment Factor Source is shown in
Table 1.
The description of 1 environmental factor of table and source
Note: B1、B2、B3、B4The reflectivity of blue wave band, green light band, red spectral band and near infrared band is respectively represented,
Corresponding to OLI image data the 2nd, 3,4,5 wave bands.
Second, the outer error condition of bag between environmental factor is analyzed, environmental factor is screened.
1. Random Forest model parameter setting
The basic ideas of Random Forest model are: 1) being taken out from original training set by bootstrap (bootstrap) sampling
K sample is taken, and the sample size of each sample is in the same size with original training set；2) decision tree is utilized respectively to each
Sample is modeled, and k modeling result is obtained；3) it according to the modeling result of all decision trees, is finally predicted by ballot
As a result.The key problem of Random Forest model modeling is determining ntree (quantity of decision tree in forest) and mtry (split vertexes
The stochastic variable number at place), it ntree is set as 1000, mtry is set as 3 in the present invention.
2. environmental factor is screened
The environmental factor for influencing soil organic matter has very much, and it is different that different environmental factors influences importance degree to it
Sample.Error (out ofbag error, OOB error) is to soil organic matter outside bag in order to reduce Random Forest model
Precision of prediction impacts, this research is lower to importance degree and generates accumulated error to the prediction of soil organic carbon
The factor is rejected, and mainly judges whether the factor retains by rejecting the monotonicity of OOB error after the factor one by one, if
OOB error increase then retains the factor, on the contrary then reject, to realize the environmental factor screening process of Random Forest model.
Table 2 be each environmental factor gradually rejects after RF model bag outside error condition, there it can be seen that by environmental factor SAVI,
After SIPI and MAT is rejected, the outer error of the bag of model decreases, and error outside the model bag after other environmental factors are rejected
It is bigger than error outside the model bag before rejecting.Therefore, select 7 environment such as ELE, S, NDVI, GNDVI, MSAVI, TVI and P because
The participation factor of the son as random forest regression model.
The screening of 2 environmental factor of table
Note: ALL is that all environmental factors participate in Random Forest model.
(3) soil organic matter spatial distribution is predicted
Spatial prediction is carried out to soil organic matter based on random forest-Ordinary Kriging Interpolation model and correlative environmental factors.
First, sampled data is divided into modeling collection and inspection set, the present invention will using reservation sample cross check system
It randomly selects 44 in data to 59 samples to be used to establish prediction model as modeling collection, using remaining 15 as inspection set
For verifying model, Fig. 1 is shown in specific distribution.
Second, modeling sample data are based on, soil organic matter is realized by the random forest software package in R software
Mathematical relationship (formula 1) between content and environmental factor, while using 9.3 software of ArcGIS to the prediction knot of soil organic matter
Fruit carries out spatial distribution drawing.As a result such as Fig. 2.
In formula:For spatially i-th point of the predicted value of the soil organic carbon based on Random Forest model；
a1,a2,…,anFor environmental factor；Mathematical relationship of the f between soil organic carbon and environmental factor.
Third obtains the residual values (formula of sampling point according to soil organic carbon measured value and Random Forest model predicted value
2) semi-variance function analysis (Fig. 3), while using GS+9.0 software to residual values is carried out, obtains optimal function model parameter, and
Ordinary Kriging Interpolation is carried out to the residual values of soil organic carbon in 9.3 software of ArcGIS based on optimal function model parameter
Interpolation (formula 3, Fig. 4).
In formula: r (xi) it is soil organic carbon spatially i-th point of residual values；C(xi) contain for soil organic matter
Measure spatially i-th point of measured value；Spatially for the soil organic carbon based on Ordinary Kriging Interpolation model
I-th point of residual error estimated value；N is the quantity of known eyeball；wiFor spatially i-th point of weighted value.
4th, it will be based on random with the raster symbol-base function (raster calculator) in 9.3 software of ArcGIS
The soil organic carbon predicted value of forest model carries out space with the residual error estimated value based on common lattice league (unit of length) method and adds operation, i.e.,
It obtains based on random forest-normal stabilizing pile soil organic matter spatial prediction result (formula 4).
In formula,For based on random forest-normal stabilizing pile soil organic carbon spatially i-th
The predicted value of a point.
5th, using the coefficient of determination (R2), root-mean-square error (RMSE), average relative error (MAE) carry out evaluation model
Accuracy, specific formula for calculation are as follows:
In formula: yiFor i-th of sampling point soil organic matter measured value；For i-th of sampling point soil organic matter predicted value；For
The average value of soil organic matter measured value.
It obtains having based on Random Forest model and random forest-Ordinary Kriging Interpolation model soil using 15 inspection sets
Machine carbon predicted value, by being compared with corresponding soil organic matter measured value, while to soil organic matter measured value and predicted value
Regression analysis is carried out, and draws 1:1 relation line, compares the precision of prediction of 2 kinds of models, the results are shown in Table 3, Fig. 5.Random forest-is general
The soil organic matter predicted value of logical Krieger model and the relationship of measured value are as follows: y=0.6980x+0.4877, corresponding decision system
Number R2It is 0.58, improves 15% than Random Forest model, the goodness of fit of the tropic and 1:1 relation line is preferable, precision of prediction
It is higher.In addition, random forest-Ordinary Kriging Interpolation model RMSE and MAE decrease compared to Random Forest model.With
Random Forest model is compared, and random forest-Ordinary Kriging Interpolation model precision of prediction is higher, and has different journeys in terms of error
The reduction of degree, thus the spatial prediction of soil organic carbon can be carried out with the model.
3 RF model of table and RF-OK model prediction accuracy
6th, for the environmental factor in non-sampled area, using random forest-ordinary Kriging, to soil sampling institute
Survey region carry out the prediction of soil organic matter spatial distribution, as a result see Fig. 6.
The above are preferred embodiments of the present invention, all any changes made according to the technical solution of the present invention, and generated function is made
When with range without departing from technical solution of the present invention, all belong to the scope of protection of the present invention.
Claims (7)
1. being based on random forest-normal stabilizing pile soil organic carbon prediction technique, which is characterized in that including as follows
Step:
Step S1, pedotheque organic carbon is measured；
Step S2, environmental factor is extracted and screens environmental factor relevant to soil organic matter；
Step S3, the space of soil organic carbon is divided based on random forest-Ordinary Kriging Interpolation model and correlative environmental factors
Cloth is predicted.
2. the method according to claim 1, wherein the step S1 specific implementation are as follows: with laying sample, adopt
With collecting sample pedotheque and corresponding geographical coordinate, measure soil organic carbon using potassium bichromate-Outside Heating Method.
3. the method according to claim 1, wherein the specific implementation steps are as follows by the step S2:
Step S21, it is based on remotely-sensed data and weather basic data, extracts environmental factor related with soil organic matter, including ground
The shape factor, Vegetation factors, climatic factor；
Step S22, error OOB error causes the precision of prediction of soil organic matter outside the bag to reduce Random Forest model
It influences, judges whether the environmental factor retains by rejecting the monotonicity of the OOB error after environmental factor one by one, if OOB
Error increase then retains the environmental factor, on the contrary then reject.
4. according to the method described in claim 3, it is characterized in that, the terrain factor includes height above sea level, the gradient in step S21；
The Vegetation factors include normalized differential vegetation index, green normalized differential vegetation index, the adjustable vegetation index of soil, amendment type soil
Earth adjusts vegetation index, the insensitive vegetation index of canopy structure, triangle vegetation index；The climatic factor includes average annual rainfall
Amount, average annual temperature.
5. the method according to claim 1, wherein the specific implementation steps are as follows by the step S3:
Step S31, modeling sample collection and test samples collection are divided into all soil sampling data；
Step S32, modeling sample collection data are based on, the number between soil organic matter and environmental factor is simulated using Random Forest model
Relationship, and make spatial prediction distribution map；
Step S33, the residual values of sampling point are obtained according to soil organic carbon measured value and Random Forest model predicted value, to soil
The residual values of earth organic carbon content carry out Ordinary Kriging Interpolation interpolation；
Step S34, by the soil organic carbon predicted value based on Random Forest model and the residual error based on common lattice league (unit of length) method
Estimated value carries out space and adds operation；
Step S35, each sampling point environmental factor data based on test samples collection, it is pre- using random forest-Ordinary Kriging Interpolation model
Soil organic matter is surveyed, and is compared with actual measurement soil organic matter, determines and is based on random forest-Ordinary Kriging Interpolation model prediction
The validity of soil organic matter；
Step S36, for the environmental factor in non-sampled area, using random forest-ordinary Kriging, to soil sampling institute
Survey region carry out the prediction of soil organic matter spatial distribution.
6. according to the method described in claim 5, it is characterized in that, in the step S32, between soil organic matter and environmental factor
Mathematical relationship it is as follows:
In formula:For spatially i-th point of the predicted value of the soil organic carbon based on Random Forest model；a1,
a2,…,anFor environmental factor；Mathematical relationship of the f between soil organic carbon and environmental factor.
7. according to the method described in claim 5, it is characterized in that, in the step S33, soil organic carbon measured value and
Random Forest model predicted value obtains the residual values formula of sampling point and carries out in common gram to the residual values of soil organic carbon
Lattice interpolation formula is respectively as shown in formula (2), (3):
In formula: r (xi) it is soil organic carbon spatially i-th point of residual values；C(xi) it is that soil organic carbon exists
Spatially i-th point of measured value；It is spatially i-th of the soil organic carbon based on Ordinary Kriging Interpolation model
The residual error estimated value of point；N is the quantity of known eyeball；wiFor spatially i-th point of weighted value；
Convolution (2), (3) obtain in the step S34, the soil organic carbon predicted value based on Random Forest model with
Residual error estimated value based on common lattice league (unit of length) method carries out space and adds the formula of operation as follows:
In formula,For based on spatially i-th point of random forest-normal stabilizing pile soil organic carbon
Predicted value.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811491043.9A CN109342697A (en) | 2018-12-07 | 2018-12-07 | Based on random forest-normal stabilizing pile soil organic carbon prediction technique |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811491043.9A CN109342697A (en) | 2018-12-07 | 2018-12-07 | Based on random forest-normal stabilizing pile soil organic carbon prediction technique |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109342697A true CN109342697A (en) | 2019-02-15 |
Family
ID=65297487
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811491043.9A Pending CN109342697A (en) | 2018-12-07 | 2018-12-07 | Based on random forest-normal stabilizing pile soil organic carbon prediction technique |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109342697A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110118994A (en) * | 2019-05-24 | 2019-08-13 | 同济大学 | A kind of nonmarine source rock quantitative forecasting technique based on seismic inversion and machine learning |
| CN110276160A (en) * | 2019-07-02 | 2019-09-24 | 四川农业大学 | A kind of region of no relief soil organic matter three-dimensional spatial distribution analogy method |
| CN114511239A (en) * | 2022-02-18 | 2022-05-17 | 中国农业科学院农业信息研究所 | Method and device for dividing soil heavy metal pollution risk area, electronic equipment and medium |
| CN114740180A (en) * | 2022-04-07 | 2022-07-12 | 中山大学 | Soil organic carbon estimation method and device based on multi-source remote sensing data |
| CN115266612A (en) * | 2022-07-27 | 2022-11-01 | 福建农林大学 | Southern hilly area cultivated land available phosphorus drawing method based on high-resolution environment variable |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108801934A (en) * | 2018-04-10 | 2018-11-13 | 安徽师范大学 | A kind of modeling method of soil organic carbon EO-1 hyperion prediction model |
-
2018
- 2018-12-07 CN CN201811491043.9A patent/CN109342697A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108801934A (en) * | 2018-04-10 | 2018-11-13 | 安徽师范大学 | A kind of modeling method of soil organic carbon EO-1 hyperion prediction model |
Non-Patent Citations (4)
| Title |
|---|
| 杜佩颖等: "平原丘陵过渡区土壤有机质空间变异及其影响因素", 《土壤学报》 * |
| 杨煜岑等: "基于随机森林的农耕区土壤有机质空间分布预测", 《浙江农业学报》 * |
| 王茵茵等: "基于多分辨率遥感数据与随机森林算法的土壤有机质预测研究", 《土壤学报》 * |
| 齐雁冰等: "基于遥感与随机森林算法的陕西省土壤有机质空间预测", 《自然资源学报》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110118994A (en) * | 2019-05-24 | 2019-08-13 | 同济大学 | A kind of nonmarine source rock quantitative forecasting technique based on seismic inversion and machine learning |
| CN110118994B (en) * | 2019-05-24 | 2020-10-02 | 同济大学 | Continental facies hydrocarbon source rock quantitative prediction method based on seismic inversion and machine learning |
| CN110276160A (en) * | 2019-07-02 | 2019-09-24 | 四川农业大学 | A kind of region of no relief soil organic matter three-dimensional spatial distribution analogy method |
| CN114511239A (en) * | 2022-02-18 | 2022-05-17 | 中国农业科学院农业信息研究所 | Method and device for dividing soil heavy metal pollution risk area, electronic equipment and medium |
| CN114511239B (en) * | 2022-02-18 | 2023-04-18 | 中国农业科学院农业信息研究所 | Method and device for dividing soil heavy metal pollution risk area, electronic equipment and medium |
| CN114740180A (en) * | 2022-04-07 | 2022-07-12 | 中山大学 | Soil organic carbon estimation method and device based on multi-source remote sensing data |
| CN115266612A (en) * | 2022-07-27 | 2022-11-01 | 福建农林大学 | Southern hilly area cultivated land available phosphorus drawing method based on high-resolution environment variable |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109342697A (en) | Based on random forest-normal stabilizing pile soil organic carbon prediction technique | |
| Tadesse et al. | Evaluating a satellite-based seasonal evapotranspiration product and identifying its relationship with other satellite-derived products and crop yield: A case study for Ethiopia | |
| Nelson et al. | Estimating Siberian timber volume using MODIS and ICESat/GLAS | |
| CN103364781B (en) | Based on the grainfield ground reference point screening technique of remotely-sensed data and Geographic Information System | |
| CN107527014A (en) | Crops planting area RS statistics scheme of sample survey design method at county level | |
| Pfeifer et al. | Leaf area index for biomes of the Eastern Arc Mountains: Landsat and SPOT observations along precipitation and altitude gradients | |
| CN104408258A (en) | Large-scale soil organic carbon spatial distribution simulation method involving environmental factors | |
| CN108984803A (en) | A kind of method and system of crop yield spatialization | |
| CN105809148A (en) | Crop drought recognition and risk evaluation method based on remote sensing time-space-spectrum fusion | |
| CN107330413A (en) | A kind of mother drug plants recognition methods based on remote sensing technology | |
| CN111667183A (en) | Method and system for monitoring cultivated land quality | |
| CN110378926A (en) | A kind of aboveground vegetation ecological water evaluation method based on airborne LiDAR and Sentinel-2A data | |
| CN108205718A (en) | Production method and system are surveyed in a kind of cereal crops sampling | |
| Girma | Potential impact of climate and land use changes on the water resources of the Upper Blue Nile Basin | |
| Kılıç et al. | An integrated framework to identify and map gullies in a Mediterranean region of Turkey | |
| Zheng et al. | Review of regional carbon counting methods for the Chinese major ecological engineering programs | |
| Schaepman et al. | River floodplain vegetation scenario development using imaging spectroscopy derived products as input variables in a dynamic vegetation model | |
| CN115861793A (en) | Regional ecological safety pattern construction method based on minimum cumulative resistance model | |
| Pisek et al. | Assessment of a global leaf area index product from SPOT-4 VEGETATION data over selected sites in Canada | |
| Ludwig et al. | Using GIS to predict concentrated flow erosion in cultivated catchments | |
| Tao et al. | The Integrated Precipitation and 20 Hydrology Experiment. Part I: Quality High-Resolution Landscape Attributes 21 Datasets | |
| Duncan | Modelling the impact of land cover and forestry change on the hydrological characteristics of the Two Streams sub-catchment, Natal Midlands | |
| Lee et al. | Enhanced Watershed Modeling by Incorporating Remotely Sensed Evapotranspiration and Leaf Area Index | |
| Lapides et al. | Inclusion of bedrock vadose zone in dynamic global vegetation models is key for simulating vegetation structure and function | |
| Mkaya | Evaluation of the impact of land use change on catchment hydrology: The case of Wundanyi River catchment in Taita Hills, Kenya |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190215 |
|
| RJ01 | Rejection of invention patent application after publication |