CN110427995A - A kind of Bayes's soil moisture evaluation method based on multi- source Remote Sensing Data data - Google Patents
A kind of Bayes's soil moisture evaluation method based on multi- source Remote Sensing Data data Download PDFInfo
- Publication number
- CN110427995A CN110427995A CN201910671852.6A CN201910671852A CN110427995A CN 110427995 A CN110427995 A CN 110427995A CN 201910671852 A CN201910671852 A CN 201910671852A CN 110427995 A CN110427995 A CN 110427995A
- Authority
- CN
- China
- Prior art keywords
- data
- soil moisture
- probability
- value
- 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.)
- Granted
Links
- 239000002689 soil Substances 0.000 title claims abstract description 77
- 238000011156 evaluation Methods 0.000 title claims abstract description 9
- 238000000034 method Methods 0.000 claims abstract description 58
- 241000132092 Aster Species 0.000 claims abstract description 24
- 230000009467 reduction Effects 0.000 claims abstract description 23
- 238000009826 distribution Methods 0.000 claims abstract description 20
- 230000003746 surface roughness Effects 0.000 claims abstract description 15
- 238000004458 analytical method Methods 0.000 claims abstract description 10
- 238000010219 correlation analysis Methods 0.000 claims abstract description 6
- 230000007613 environmental effect Effects 0.000 claims description 50
- 238000005259 measurement Methods 0.000 claims description 25
- 238000012417 linear regression Methods 0.000 claims description 10
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 9
- 238000012544 monitoring process Methods 0.000 claims description 5
- 238000005070 sampling Methods 0.000 claims description 5
- 238000012545 processing Methods 0.000 claims description 4
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 claims description 3
- 229910052782 aluminium Inorganic materials 0.000 claims description 3
- 239000004411 aluminium Substances 0.000 claims description 3
- 230000015572 biosynthetic process Effects 0.000 claims description 2
- 238000010606 normalization Methods 0.000 claims description 2
- 238000000611 regression analysis Methods 0.000 claims description 2
- 238000000513 principal component analysis Methods 0.000 abstract description 41
- 241001269238 Data Species 0.000 abstract description 7
- 240000008042 Zea mays Species 0.000 description 42
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 42
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 39
- 235000005822 corn Nutrition 0.000 description 39
- 238000011160 research Methods 0.000 description 17
- 238000011065 in-situ storage Methods 0.000 description 11
- 230000012010 growth Effects 0.000 description 9
- 238000009331 sowing Methods 0.000 description 8
- 230000003698 anagen phase Effects 0.000 description 6
- 238000011835 investigation Methods 0.000 description 6
- 238000010899 nucleation Methods 0.000 description 6
- 230000008859 change Effects 0.000 description 5
- 230000018109 developmental process Effects 0.000 description 4
- 238000005315 distribution function Methods 0.000 description 4
- 239000002420 orchard Substances 0.000 description 4
- 235000016383 Zea mays subsp huehuetenangensis Nutrition 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 235000009973 maize Nutrition 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000009395 breeding Methods 0.000 description 2
- 230000001488 breeding effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 238000003306 harvesting Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000005855 radiation Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 235000017060 Arachis glabrata Nutrition 0.000 description 1
- 244000105624 Arachis hypogaea Species 0.000 description 1
- 235000010777 Arachis hypogaea Nutrition 0.000 description 1
- 235000018262 Arachis monticola Nutrition 0.000 description 1
- 208000003643 Callosities Diseases 0.000 description 1
- 229920000742 Cotton Polymers 0.000 description 1
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 244000299507 Gossypium hirsutum Species 0.000 description 1
- 206010020649 Hyperkeratosis Diseases 0.000 description 1
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 244000131316 Panax pseudoginseng Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000003335 Production assurance Methods 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000002457 bidirectional effect Effects 0.000 description 1
- 239000004927 clay Substances 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000001035 drying Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 238000009313 farming Methods 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000012625 in-situ measurement Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 235000020232 peanut Nutrition 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000004064 recycling Methods 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
- 239000004576 sand Substances 0.000 description 1
- 230000002195 synergetic effect Effects 0.000 description 1
- 238000010189 synthetic method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
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
- G01N33/246—Earth materials for water content
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2415—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
- G06F18/24155—Bayesian classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/251—Fusion techniques of input or preprocessed data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/758—Involving statistics of pixels or of feature values, e.g. histogram matching
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Chemical & Material Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Food Science & Technology (AREA)
- Pathology (AREA)
- Immunology (AREA)
- Biochemistry (AREA)
- Analytical Chemistry (AREA)
- Medicinal Chemistry (AREA)
- Remote Sensing (AREA)
- Geology (AREA)
- General Life Sciences & Earth Sciences (AREA)
- Environmental & Geological Engineering (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Multimedia (AREA)
- Image Processing (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The present invention relates to a kind of Bayes's soil moisture evaluation method based on multi- source Remote Sensing Data data.This method is directed to how to obtain high-precision soil moisture probability-weighted soft data, the calculating that 12 kinds of multi-source datas are weighted probability soft data has been merged for the first time, albedo (A) is obtained including NO emissions reduction FY3-B soil moisture product, using MODIS product, vegetation index NDVI(V) and surface temperature LST(T);Altitude data is obtained using ASTER product;And the gradient, slope aspect, planar curvature, profile curvature, surface roughness, humidity index and relief intensity data are calculated by the altitude data that ASTER product obtains;And probability-weighted soft data is obtained using two kinds of Weight Determinations of multivariate correlation analysis and principal component analysis.The present invention also analyzes the precision analysis of different soft data quantity, and give one's hand sufficient soft data quantity, and the soil moisture spatial distribution for obtaining higher precision plays a significant role.
Description
Technical field
The present invention relates to a kind of soil moisture evaluation method based on multi-source data, especially fusion NO emissions reduction soil moisture
Bayes's soil moisture evaluation method of the multi- source Remote Sensing Data datas such as product data and terrain.
Background technique
Lacking, there are soil moisture (SM) data of high spatial resolution, which to have become, improves Watershed Scale Eco-hydrological Model
One of main bottleneck of accuracy.Bayes's maximum entropy (BME) algorithm is calculated for modeling the estimation of extensive special heterogeneity
Method, and can integrate a plurality of types of data with different accuracy and quality.It theoretically, will be with SM sky using BME algorithm
Between relevant a plurality of types of data integrations SM precision can be improved into SM Spatial outlier.
Soil moisture (SM) is not only hydrological model, Climatic Forecast Models, and mould is estimated in draught monitor model and crop yield
An important factor for key parameter and Global climate change and surface data assimilation of type.Traditional SM in situ measurement can be
Single-point precise measurement SM is not able to satisfy the requirement of extensive dynamic SM monitoring.With the development of satellite remote sensing technology and perfect, open
Many is sent out based on visible/near infrared, the SM monitoring method of thermal infrared and active/passive microwave satellite data, this is but also prison
SM is surveyed to be possibly realized.In particular, passive and active microwave satellite data is only influenced minimum by cloudy day and rainy weather, and it is passive
Microwave data particularly provides the high-sensitivity method of extensive observation SM.Therefore, it is distant to have become land face SM for remote sensing microwave data
The main input for feeling product is with a wide range of applications in global SM monitoring.However, most of quilts in most of regions
Dynamic microwave remote sensing product as was expected accurate (RMSE > 0.06cm3 / cm 3).Due to the essence of passive microwave SM product
Spend lower, real value is very limited.In addition, most of passive microwave SM searching algorithms are all directed to uniform outer surface progress
Optimization.Due to the limitation of sensoring and image-forming principle, the spatial resolution of passive microwave satellite is very low, at tens kilometers
In range.This causes the inside of microwave pixel heterogeneous, this complicates the verifying of SM product.Therefore, there is an urgent need to develop tools
There is the SM product of high spatial resolution.
Summary of the invention
BME algorithm considers the uncertainty of data, and can effectively be merged according to the principle of maximum informational entropy various
Data.The data definition accurately measured is hard data (HD) by BME, and will have probabilistic data definition is soft data
(SD).SD is converted to by probability SD by using probability-distribution function (such as histogram and gauss of distribution function), and is used
HD calibrates the probability SD of SM to obtain distribution function after probability.Finally, estimating SM by using distribution function after probability.This
Invention particular content is as follows:
A kind of Bayes's soil moisture evaluation method based on multi- source Remote Sensing Data data, comprising the following steps:
Step 1 obtains soil moisture soft data and carries out data processing
Specified region, the soil moisture data of scheduled date are obtained using FY3-B soil Reel moisture monitoring product;And to FY3-B soil
Earth moisture data carries out NO emissions reduction processing, is divided 25km resolution ratio FY3-B soil moisture data space by using regression model
Resolution rises to 1km resolution ratio soil moisture data;
(1)
WhereinA 25For 25 km average values of albedo (A),V 25For the 25 km average values of vegetation index NDVI(V),T 25For earth's surface
Temperature LST(T) 25 km average values,mWithnIt is in FY3-B pixel respectivelyiRow and thejThe quantity of 1km pixel in column;A ijIt is
In FY3-BiRow and thejThe albedo value of 1Km pixel in column,V ijIt is in FY3-BiRow and thej1Km pixel in column
NDVI value,T ijIt is in FY3-B iRow and thejThe surface temperature LST value of 1Km pixel in column;
Linear regression model (LRM) are as follows: SM=a1+a2 A+a3 V+a4 T+a5 AV+a6 AT+a7 VT (2)
Wherein SM is soil moisture;a1、a2、a3、a4、a5、a6And a7It is regression coefficient;A is albedo, V is vegetation index, T is
Surface temperature;
Firstly, the A that formula (1) is calculated25、V25And T25Formula (2) are brought into, according to corresponding grid cell size 25km resolution ratio
FY3-B soil moisture data SM25, pass through the regression coefficient in regression analysis computation model;
Secondly, by the albedo (A) of MODIS product 1km resolution ratio, vegetation index NDVI(V) and surface temperature LST(T) bring into
The Soil moisture SM of 1km resolution ratio has accordingly been calculated in model formation (2)1;
Step 2 obtains environmental factor soft data
Albedo (A) is obtained using MODIS product, vegetation index NDVI(V) and surface temperature LST(T);Use ASTER product
Obtain altitude data;And the gradient, slope aspect, planar curvature, section song are calculated by the altitude data that ASTER product obtains
Rate, surface roughness, humidity index and relief intensity data;
Step 3 obtains NO emissions reduction FY3-B soil moisture data SM using step 11Environmental factor soft data is obtained with step 2, is adopted
With histogram form of probability, probability soft data is obtained;
Step 4 determines environmental factor soft data in step 2 using multivariate correlation analysis method and principal component analytical method
Weighted value, and obtain probability-weighted soft data;
Firstly, calculating thejA environmental factor and soil moisture data between quantitative probabilities, be denoted as PSD(EF j ):
(3)
Wherein Ef represents the environmental factor data obtained in step 2;nRepresent SM'snA section;kIndicate the of SMkIt is a
Section,kValue from 1 ton;mRepresent Ef'smA section;It is the of SM in first of Ef intervalkIn a interval
Sampling number;It is the of SMkTotal number of sample points in a section;EfmIt ismThe environmental factor in a section;Each
In section, probability distribution data are constant values, and the size in all sections is identical;
Secondly, calculating wherein thejThe weighted value w of a environmental factor data j Are as follows:
(4)
Wherein,iFor the number of environmental factor in step 2,iFor value from 1 to 11, Rj isjA environmental factor and the FY3B soil water
Related coefficient between divided data; EfjIt isjA environmental factor;NCC(Efj) it isjThe normalization phase relation of a environmental factor
Number, the sum of NCC of all environmental factors are equal to 1, wherein NCC(Efj) value more than or equal to 0.18 environmental factor be leading environment
The factor retains its environmental factor respective weights value, NCC(Efj) environmental factor weighted value of the value less than 0.18 be 0;
Again, the probability-weighted for calculating SM, is denoted as WPSD,
(5)
Step 5, actual measurement obtain soil moisture hard data
In step 1 the same area, phase same date, the pedotheque of 0-5cm depth is collected, and measures soil using aluminium case baking method
Earth moisture content;All parametric cubics are measured, and average value is used for subsequent analysis, are denoted as;
Step 6, using Bayes's maximum entropy method to the soil moisture probability-weighted soft data in step 4And in step 5
Measured dataIt is merged as hard data, estimates soil moisture data, be as a result denoted as;
Step 7 assesses Prediction of Soil Water Content using root-mean-square error (RMSE) and related coefficient (R)As a result standard
Exactness, wherein
(6)
(7)
Wherein,nIt is the total degree that soil moisture is surveyed in step 5,It is in step 5iSecondary actual measurement Soil moisture,It is in step 5nThe average value of secondary actual measurement soil moisture,It is in step 6iSecondary Bayes's maximum entropy
Soil moisture,In step 6nThe average value of secondary Bayes's maximum entropy soil moisture.
Further, probability-weighted soft data number is set as 4 classes, respectively 500,450,400 and 350, obtains
Soil moisture estimation result under four type different weights probability soft data numbers.
This method estimates soil moisture using hard data and soil moisture probability-weighted soft data.This method will be real
Soil moisture data are surveyed as hard data, soft data then has chosen the data such as terrain, it is contemplated that terrain is to the soil water
The influence divided.Higher resolution and higher is estimated the advantage of the invention is that multi- source Remote Sensing Data data is integrated into BME algorithm
The SM of precision.
Detailed description of the invention
Bayes soil moisture evaluation method schematic diagram of the Fig. 1 based on multi- source Remote Sensing Data data;
Fig. 2 survey region is summarized;
The spatial prediction of the SM content in three corn growth stages of the Fig. 3 based on MCA and PCA method;
Fig. 4 is using the estimation SM of BME algorithm and NO emissions reduction FY3-B SM product compared with in-site measurement SM.
Specific embodiment
Object of this investigation is to estimate higher resolution and higher by the way that multi- source Remote Sensing Data data to be integrated into BME algorithm
The SM of precision, as shown in Fig. 1, specific embodiment is as follows for process approach signal:
One, hard data prepares
Survey region (Fig. 2) is located at Hebei province's Hengshui City, is located at northwest China (38 ° 3'00 " N, 115 ° of 27'54 " E).This is
One typical farming region, only a few land cover pattern (exposed soil, corn, orchard and meadow), the soil texture is uniform.The ground
Area belongs to typical continental monsoon climate, and summer temperature is high, abundant rainfall, and winter cold is dry.Therefore, SM content is raw in agricultural
It plays an important role in production, especially in summer.In order to measure SM content, field trial has been carried out in summer.Sample point such as Fig. 2
It is shown.Corn, cotton and peanut are the chief crops in these places.In The Soils are loams, including about 44.6% silt
Mud, 17.5% clay and 37.9% sand.
In summer, corn accounts for nearly the 50% of the research gross area.In our current research, corn is divided into spring maize and summer corn.Spring
Corn seeding is in late April or the first tenday period of a month in May, and last 10 days in August harvest.Summer corn is usually at the beginning of late May or 6 months
Sowing, October harvest.Therefore, experiment is carried out June to September.This period includes the sowing of corn growth, heading stage, flower
Phase and maturity period.
It tests in June 27 (sowing stage, i.e., the 178th day), August 14th (heading stage of corn, i.e., the 226th day) and 9
7 (stage at florescence, i.e., the 249th day) of the moon carries out for 2014.We have selected about 25 kilometers of area × 25 kilometer as key
Sample areas.Sample areas includes SM at, the parameter of each website by 23 samples, soil dielectric constant, and vegetation is aqueous
Amount and leaf area index (LAI).Pedotheque is collected in the depth of 0-5cm to use aluminium case to measure SM content.Measure all ginsengs
Average value three times, and is used for subsequent analysis by number.During this investigation it turned out, actual measurement SM data in situ are divided into two parts.One
The HD for dividing actual measurement SM data in situ to be used as BME analysis.Another is for verifying the estimation SM from BME algorithm.It uses
E5071C Vector Network Analyzer(Keysight Technologies Inc., Santa Rosa, CA, USA)
Soil dielectric constant is measured in the lab.Using comprising 101 needles and total length is more than that the needle plate of 1m measures thick equal Fang Gaodu
With the correlation length of roughness parameter.Use sampling and drying means measurement vegetation water content (VWC).It is planted using LAI-2200C
Object Canopy Analyzer (LI-COR Biosciences, Lincoln, NE, USA) measures LAI five times.Ground data, including SM contain
Amount, soil dielectric constant and vegetation water content, as shown in table 1.)
1. ground data of table abstract
DO | N | SMC (g/cm3) | VWC (kg/m 2 ) | H (cm) | LAI | |||
178th day | Exposed soil/meadow/orchard | 64 | 5.62–48.1 | 0 | 2.3–20.1 | 1.578 | 0.4–3.4 | / |
226th day | Maize Region/orchard | 60 | 9.1–40.9 | 0.48–5.13 | 3.7–25.0 | 1.33 | 0.2–3.0 | 2.89 |
249th day | Maize Region/orchard | 60 | 8.6–39.3 | 1.53–3.90 | 3.4–17.8 | 1.33 | 0.4–2.2 | 3.35 |
(N: sampling number;R: soil dielectric constant;: the average soil weight;VWC: vegetation water content;H: average corn is high
Degree;DO: main object)
Two, multi-source remote sensing soft data obtains
There are many factor for influencing SM retrieval, such as surface temperature, Vegetation canopy water content and roughness.It is traditional to utilize 2-3 kind mostly
Auxiliary data estimates SM, or does not account for terrain influence, or do not account for the soil moisture after NO emissions reduction.Therefore, only
It is difficult using a kind of data to assess SM.Therefore, we are by the FY3-B SM, NDVI-LST-Albedo of 25km resolution ratio
NDVI, Albedo and the LST product of MODIS(MODIS) and ASTER DEM Satellite Product data combine estimation SM space
Distribution.During this investigation it turned out, we used three kinds of MODIS products, i.e. NDVI, Albedo and LST product, resolution ratio 1
km.Advanced spaceborne heat radiation and reflected radiation meter global digital elevation model (ASTER DEM), spatial resolution is up to 30
m.The FY3-B(third wind and cloud B satellite of 25km resolution ratio) SM product, these auxiliary datas are for using BME algorithm development more
High-resolution SM data.
FY3-B SM product is free (download link http://satellite.nsmc.org.cn/PortalSite/
Default.aspx).The product is on the Coriolis satellite of the horizontal and vertical polarization data based on 10.65 GHz data
AMSR-E(Advanced Microwave scanning radiometer-earth observing system) sensor SM searching algorithm exploitation.It is calculated in AMSR-E
In method, Q/H model selected as roughness model estimate the influence of surface roughness.
These three MODIS products are MOD11A1(surface temperature product, LST), MCD43B3(Albedo product) and
MOD13A2(vegetation index product: normalized differential vegetation index product and enhancing vegetation index, NDVI and EVI), they all have sky
Between resolution ratio 1km.MOD11A1 LST product is using the Split-window algorithm of MODIS the 31st and the 32nd wave band with the resolution ratio of 1km
It retrieves (Yang et al.1199).MCD43B3 Albedo product was produced by 16 days anisotropic models, represented MCD43A
The average value of the basic 500m value of Albedo product.MOD13A2 NDVI product by using MODIS specific synthetic method,
Based on production assurance index, from daily, retrieved in the Bidirectional surface reflection of atmospheric correction, removal low quality pixel is simultaneously kept
The NDVI value of high quality.ASTER DEM is based on the detailed observation to NASA a new generation earth observation satellite TERRA.This is current
One of most complete global digital elevation data (DEM), with 30 meters of resolution ratio.During this investigation it turned out, 30 meters of ASTER DEM
Data are re-sampled to 1 kilometer of resolution ratio.
Using ASTER dem data, it is extracted the gradient, slope aspect, planar curvature, profile curvature, surface roughness, humidity and refers to
7 kinds of data (table 2) such as several and relief intensity.This research is generated by LST, Albedo, NDVI, ASTER DEM and by ASTEM DEM
7 class data as environmental factor data, directly affect SM estimation.Utilize the NDVI-LST-Albedo/ in nearly 3 periods
MODIS, FY3-B SM and ASTER dem data: on June 27th, 2014,2014 on August 14, and on September 7th, 2014, i.e., it is beautiful
Three breeding times (the 178th day, the 226th day and the 249th day) of rice.
Seven kinds of terrain datas (gradient, slope aspect, planar curvature, profile curvature, surfaces that 2 ASTER dem data of table generates
Roughness, humidity index and relief intensity)
Title | Algorithm description |
Slope aspect | From each pixel to its maximum direction of fall of adjacent pixel change rate. |
The gradient | The gradient is the second dervative of digital elevation model variation |
Planar curvature | It is perpendicular to the second dervative on the surface in greatest gradient direction. |
Profile curvature | It is second dervative of the curved surface along greatest gradient direction. |
Surface roughness (sr) | ,It is slope, unit: radian.SR is the long-pending ratio with its projected area of earth surface in given area. |
Humidity index (WI) | , As indicates to flow through on surface the catchment area for the per unit contour length that certain is put.It is the gradient. |
Relief intensity | ,(DEMmax)nIt isn×nMaximum value in region;(DEMmin)nIt isn×nMinimum value in region.In our current research,nSetting It is 3. |
Three, the estimation of Bayes's soil moisture and verifying
Some researchers show to depend on SM, NDVI and LST using the development of Microwave Optics/infrared product Synergistic method
Between schematic relationships.Many researchs use MODIS satellite image data with finer than FY3-B original resolution (25 km)
Resolution ratio (1 km) obtain SM information.
In our current research, MODIS albedo (A), NDVI(V) and LST(T) with 25 km FY3-B SM spatial resolution phases
Matching.SM is obtained with the spatial resolution of 1km using linear regression model (LRM).FY3-B SM, MODIS albedo, NDVI and LST
For linear regression analysis, microwave and optics/IR parameter intensity are connected to improve the FY3-B of coarse resolution sky
Between resolution ratio.The resolution ratio of MODIS albedo, NDVI and LST are that 25 km calculating is as follows:
(1)
Wherein A25, V25And T25It is albedo respectively, the 25 km average values of NDVI and LST,mWithnIt is in FY3-B pixel respectivelyiRow and thejThe quantity of 1 km pixel in column.
Linear regression model (LRM) uses as follows:
(2)
Wherein SM is the SM of 25km;A1, a2, a3, a4, a5, a6 and a7 are regression coefficients;Use the time series in April to October
Clear sky data carry out the regression equation in Estimation Study region.Then the regression model of application estimation, including regression coefficient, estimate 25km
The 1 km SM of FY3-B, 1km albedo, NDVI and LST.Although each regression coefficient (i.e. a1, a2, a3, a4, a5, a6 and a7)
Each FY3-B pixel is remained unchanged to fixing the date, but each value changes between not same date, because all variables are at any time
Between change and regression relation change based on these variables.But this method has ignored the importance of SM ground data.
Therefore, it is necessary to merge ground data to obtain the numerical map of the higher precision of SM.
In order to carry out probability statistics to each environmental factor data, using identical discretization method, by each environment because
The range of subdata is also divided intomA section.In conjunction with discrete SM and environmental factor data, by calculate each Interval Discrete environment because
Quantitative probabilities between son and SM data, derive the probability SD of SM.Between obtained SM and environmental factorm×nIt is quantitative
Probabilistic relation is probability SD (i.e. PSD(Ef))。
(3)
Wherein Ef represents one of environmental factor, i.e. NDVI in this research, LST, Albedo, DEM, the gradient, slope aspect, Plane Curved
Rate, profile curvature, surface roughness, humidity index and relief intensity.nRepresent SM'snA section;kIndicate the of SMkIt is a
Section, from 1 ton;mRepresent Ef'smA section;L represents first of section of SM, from 1 tom;It is between first of Ef
Every the of middle SMkSampling number in a interval.It is the of SMkTotal number of sample points in a section;EfmIt ismIt is a
The environmental factor in section.In each section, probability distribution data are constant values, and the size in all sections is identical.This is general
Rate distribution form is known as histogram form of probability.
Since different environmental factors has different influences to SM content, the weighting of the varying environment factor should be specified
Value.Weighted value is normalized using equation 10.Assuming that the quantity of environmental factor is i, then adding for SM is obtained using lower equation
Weigh probability SD(, that is, WPSD).
(5)
Wherein W1 is the weighted value of the first environment factor, and Wj isjThe weighted value of a environmental factor.In this study, using two
Kind of method MCA(multivariate correlation analysis) and PCA(principal component analysis) the method calculating varying environment factor weighted value.Different rings
The normalizated correlation coefficient of the border factor is considered as weighted value.
MCA method is a kind of multiple linear regression model:
The related coefficient normalizing equation of the border factor (R):
(4)
SM is SM content, it represents NO emissions reduction FY3-B SM in this research;R1 is the related coefficient of first environmental factor;
Ef1 is first environmental factor;NCC(Efj) it isjThe normalizated correlation coefficient of a environmental factor.In our current research, environment because
Subdata is auxiliary data, i.e. NDVI-LST-Albedo/MODIS product data, ASTER dem data, the gradient, slope aspect, plane
Curvature, profile curvature, surface roughness, humidity index and relief intensity, totally 11 kinds of auxiliary datas (i.e. i=11).All environment
The sum of NCC of the factor is equal to 1.NCC value is considered the weighted value of environmental factor, to generate probability-weighted SD.
Reduce the number of environmental factor using principal component analysis (PCA) method.In this approach, multivariable factor quilt
Multiple principal components are reduced to, to simplify modeling, and have obtained more steady result.Firstly, being handled by PCA influences soil
11 environmental factor data of moisture recycling, obtain 11 main components. EfiData be not standardized environment in Eq.11 because
Subdata, but the main component after PCA.The NCC value of 11 main components calculates.Then, selection and SM are highly relevant
Dominant factor of several principal components as the different corn growth stages.It is used to obtain finally, the NCC value of these dominant factors is considered as
Obtain the weighted value of probability-weighted SD.
According to the condition probability formula of Bayes's maximum entropy, the prior probability of HD modification SM content can be used.Parameter z quilt
The posterior probability being defined as in the position of prediction x0.z0Posterior probability be expressed as
Wherein h is HD points;M-h is SD points;SM estimation is using special-temporal slices knowledge integration graphic user interface (SEKS-
GUI) software development.This is the software package provided free, for establishing space-time model, and by combining scientific software
Library, Bayes's maximum entropy library and correlation GUI file are predicted and mapping parameters.In conjunction with FY3-BSM retrieval products, MODIS and
The environmental factor data of ASTERDEM product predict SM content using Bayes's maximum entropy method (MEM).Probability SD is using straight
Square figure form of probability, by SM content and environmental factor discretization.The rear PDF of SM is developed in SEKS-GUI software, i.e.,
Utilize the probability-weighted SD of HD data (i.e. scene SM measurement) calibration SM.
In this study, the accuracy of SM prediction result: RMSE(root-mean-square error is assessed using two statistical indicators),
That is formula 6 and R(related coefficient), i.e. formula 7.
(6)
(7)
Wherein,It is actual measurement SM in situ;It is the SM predicted using BME algorithm;nIt is practical survey in situ
Measure the sum of SM.
In the present embodiment, with the related coefficient for calculating FY3-B SM NO emissions reduction and 11 kinds of auxiliary datas, including NDVI-
LST-Albedo/MODIS, ASTER dem data generate the gradient, slope aspect, planar curvature, profile curvature, surface roughness,
Humidity index and relief intensity product.(table 2).
Related coefficient and NCC(normalizated correlation coefficient are calculated in three growth phases of corn), as shown in table 3.In addition to
Except 226th day (heading stage of corn), the strongest correlation between NDVI and SM content.NCC is in heading stage highest
(0.68), minimum (0.427) in sowing time.At the 226th day, the correlation between LST and SM content was most strong (NCC:-0.718).
Therefore, leading factor is different according to growth stage.In sowing stage (the 178th day), LST, NDVI, planar curvature and section are bent
Rate (absolute value >=0.372 of NCC) is the principal element of SM retrieval.In heading stage (the 226th day), LST, NDVI, albedo and
Profile curvature (absolute value >=0.395 of NCC) is leading factor.In florescence (the 249th day), NDVI, profile curvature, surface area
It is principal element with humidity index (absolute value >=0.198 of NCC).
Related coefficient between table 3. NO emissions reduction FY3-B SM and Ef
Ef | R/178th | NCC/178th | R/226th | NCC/226th | R/249th | NCC/249th |
Albedo | / | / | 0.454 | -1.767 | 0.010 | 0.010 |
LST | -0.355 | 2.241 | -0.718 | 2.795 | -0.070 | -0.073 |
NDVI | 0.427 | -2.697 | 0.680 | -2.647 | 0.580 | 0.604 |
DEM | -0.003 | 0.020 | 0.259 | -1.008 | -0.180 | -0.188 |
Surface roughness | 0.006 | -0.035 | 0.023 | -0.090 | -0.010 | -0.010 |
Relief intensity | -0.116 | 0.733 | 0.024 | -0.093 | -0.050 | -0.052 |
Humidity index | 0.092 | -0.582 | -0.269 | 1.046 | 0.190 | 0.198 |
Planar curvature | -0.383 | 2.418 | 0.210 | -0.817 | 0.030 | 0.031 |
The gradient | -0.019 | 0.117 | 0.014 | -0.055 | 0.010 | 0.010 |
Slope aspect | -0.180 | 1.134 | -0.085 | 0.332 | 0.190 | 0.198 |
Profile curvature | 0.372 | -2.349 | -0.395 | 1.536 | 0.260 | 0.271 |
(Ef: environmental factor;R: related coefficient;NCC: normalizated correlation coefficient;178th day: the corn seeding phase;226th day:
Corn heading stage;249th day: corn florescence)
Then, dominant factor of the several main components highly relevant from SM as the different corn growth stages is selected.These masters
The NCC of inducement is considered as the weighted value of these dominant factors, as shown in table 4.From table 4, on the day of the 178th day, PC2,
PC4, PC8 and PC9 are chosen as principal element (NCC >=0.181).The 226th day, PC2, PC6, PC8 and PC9 be chosen as it is leading because
Son, NCC >=0.243.At the 249th day, select PC2, PC4, PC8 and PC10 as dominant factor, and NCC >=0.593(table
4).
Table 4. reduced after PCA FY3-B SM andEfBetween related coefficient
Ef | R/178th | NCC/178th | R/226th | NCC/226th | R/249th | NCC249th |
PC1 | -0.146 | 0.114 | -0.137 | 0.327 | 0.019 | 0.059 |
PC2 | -0.507 | 0.394 | -0.637 | 1.522 | -0.193 | -0.593 |
PC3 | -0.003 | 0.003 | -0.044 | 0.104 | 0.053 | 0.162 |
PC4 | -0.257 | 0.200 | 0.072 | -0.171 | 0.281 | 0.862 |
PC5 | -0.076 | 0.059 | -0.034 | 0.082 | -0.044 | -0.136 |
PC6 | 0.056 | -0.044 | 0.204 | -0.488 | 0.069 | 0.211 |
PC7 | -0.058 | 0.045 | -0.029 | 0.070 | -0.050 | -0.153 |
PC8 | -0.392 | 0.306 | 0.102 | -0.243 | 0.199 | 0.609 |
PC9 | 0.233 | -0.181 | 0.139 | -0.332 | 0.156 | 0.480 |
PC10 | -0.134 | 0.104 | -0.087 | 0.209 | -0.260 | -0.797 |
PC11 | / | / | 0.034 | -0.080 | 0.096 | 0.294 |
(Ef: environmental factor;R: related coefficient;NCC: normalizated correlation coefficient;178th day: the corn seeding phase;226th day:
Corn heading stage;249th day: corn florescence)
In general, using two kinds of data (HD and SD) in BME analysis.HD refers to thering is small error and high-precision certain numbers
According to or true, such as data in situ, there is high-precision historical data.Herein, HD is measured in three field experiments
SM content.HD, i.e. original position SM data are used to estimate the posterior probability density in SEKS-GUI software in the rear class of BME algorithm
Distribution.From the perspective of the theory of knowledge, we not fully depend on certain data or certain facts to the understanding of certain things,
But uncertain information is additionally depended on, such as with the remotely-sensed data of some mistakes, expert opinion and Heuristics.These data
It is known as SD in BME analysis.NO emissions reduction FY3-B SM data and auxiliary data with some errors are considered SD.It is auxiliary
Helping data includes NDVI-LST-Albedo/MODIS, seven seed types that ASTER dem data and ASTER dem data obtain
Data, the i.e. gradient, slope aspect, planar curvature, profile curvature, surface roughness, humidity index and relief intensity.Due to probability SD's
Quantity may influence probability SD, therefore different probability SD points (table 5) are arranged at 4 kinds, and case 1 is 500 soft numbers
Strong point;Case 2 is 450 soft data points;Case 3 is 400 soft data points;Case 4 is 350 soft data points (table 5).
5 HD of table points and probability SD points
(SD: soft data;HD: hard data;178th day: the corn seeding phase;226th day: corn heading stage;249th day: corn
Florescence)
In our current research, MCA and PCA method is respectively used to obtain the weighted value (table 3 and table 4) of the varying environment factor.Therefore, 2
Three growth phases of (two methods calculate weighing value, MCA and PCA) × 3(corn) × 4 cases (table 5), in this study
Have estimated totally 24 probability SD data sets.
SEKS-GUI software has user-friendly interface, and is the library BMElib(BME write with Matlab) building
's.In three corn growth stages (sowing stage, i.e., the 178th day, heading-stage, i.e., the 226th day, florescence, i.e., the 249th day)
SM content distribution figure is as shown in Figure 3.In Fig. 3, preceding four column represent four cases, 5) case 1 to case 4(is shown in Table.Last column
It represents in the 178th day (corn seeding phase), the drop ruler at the 226th day (heading stage of corn) and the 249th day (florescence of corn)
Spend FY3-B SM.In each case, it is shown that use MCA and PCA in three growth phases of corn (sowing, heading and florescence)
6 distribution maps of the SM estimation that method generates.
During this investigation it turned out, being based on two methods (MCA and PCA), the spatial distribution of the SM estimated using BME algorithm is 1
The resolution ratio of kilometer, for calculating the weighted value of environmental factor.MCA and PCA method uses IDL(interactive data language) journey
Sequence is realized.At each pixel region (1 kilometer × 1 kilometer), it will be assumed that the estimated value of SM is uniform.It can from Fig. 3
Out, the BME algorithm at three, this research area corn growth stage, sowing time, heading stage and florescence has estimated the spatial distribution map of SM.In
The bottom of each SM distribution map has some pieces, and for especially case 4. in these blocks, the SM of estimation is not fine.For example, In
The heading stage of corn, i.e., the 249th day, the prediction SM using MCA method are the big yellow block of SM distribution map bottom, especially case
Example 4. is at MCA the 249th day, from 500 to 350 from case 1 to the quantity of case 4(soft data point), the quantity of soft data point compared with
Small, the area of yellow block is larger.On other dates (178 or 226 days), the quantity of soft data is smaller, and the area of block is bigger.Block is raw
At the reason of may be the missing values of existing probability soft data or to lack the value of auxiliary data in the block region.When each
In phase (each column of Fig. 3), the space structure of the estimation SM of 4 cases is similar, and there is similar SM to change for they
Range.At the 178th day, the variation range of the estimation SM value of 4 cases was 0.08 to 0.3 m3/m3, when SM value is less than 0.012
When m3/m3, the negligible amounts of SM pixel.At the 226th day, the variation range of the estimation SM value of case 4 was 0.08 to 0.36 m3
/ m3.At the 249th day, it was 0.12 to 0.26 m3/m3 that the variation range of the estimation SM value of 4 cases is very narrow.Obviously, estimate
SM value larger space variation occur at the 226th day.The 178th day and the 249th day, SM seemed more evenly.178th day and
The reason of 249th day homogeneous SM may be vegetation more evenly.Estimation SM based on MCA method is compared with PCA method, is estimated
The space structure of the PCA based on SMA of meter is more consistent in 4 cases.Between estimation SM from MCA and PCA method
Spatial distribution characteristic shows preliminary realization.Therefore, the quantitative verification work of estimation SM has been carried out in our current research.
Practical SM measurement in situ is divided into two parts.A part of original position actual measurement SM data (88 points in total) are used as hard number
According to.Another (54 points in total) is for verifying the estimation SM from BME algorithm.Three lifes in corn are shown in Fig. 4
Long stage (178,226 and 249) is from the relationship between the actual measurement SM and SM in situ that BME algorithm is estimated.The X-axis of Fig. 4 is former
Position.Measurement result, the Y-axis of Fig. 4 are the estimation SM from BME algorithm.Although there are excessively high estimation and that underestimates ask in Fig. 4
Topic, but most of points are distributed near 1:1 line.
In addition, we pass through each case (case 1, case 2, case 3 and the case for calculating every kind of method (MCA and PCA)
4) RMSE and R value and original position actual measurement SM carries out absolutely verifying work.In three periods (the 178th day, the 226th day and
249 days) in, RMSE and R value is calculated separately using 15 points, 16 points and 23 points of actual measurement in situ.RMSE the and R value of calculating is such as
Shown in table 6.
As can be seen from Table 6, using PCA method, in case 1, the 178th day RMSE value is minimum, be 0.049 m3/
m3., less than 0.079 m3/m3 RMSE value, this is calculated between NO emissions reduction FY3-B SM and in-site measurement SM for it.It is on record
The 178th day of PCA is used in example 1, R value maximum value is 0.639.It is significantly larger than 0.096 R value, this is in NO emissions reduction FY3-B
It is calculated between SM and in-site measurement SM.At the 226th day, the minimum RMSE generated in case 2 and case 3 using MCA method
Value is 0.062m3/m3, less than the 0.091m3/m3 generated by reducing FY3-B SM.It is given birth under case 1 using MCA method
At R value maximum value be 0.53.At the 249th day, the minimum RMSE value that NO emissions reduction FY3-B SM is generated be 0.051 m3/
m3.During this period, the lower RMSE value of the estimation SM from BME algorithm is 0.053 m3/m3, is higher than NO emissions reduction FY3-B SM
0.051 m3/the m3 generated.Generally speaking, we conclude that, using PCA method, the estimation of BME algorithm
The precision of SM is better than NO emissions reduction FY3-B SM, and minimum RMSE value and largest r value are respectively 0.049 m3/m3 and 0.639.Compare
The precision of estimation SM based on MCA and PCA method, it has been found that the 178th day and the 249th day, the estimation SM based on PCA method
RMSE value be less than the RMSE value based on MCA, the R value based on PCA method be higher than be based on MCA method.However, at the 226th day, base
It is greater than the MCA method that is based in the RMSE value of the estimation SM of PCA method, the R value based on PCA method is lower than the R based on MCA method
Value.In addition, estimating the accuracy of SM significantly lower than the 178th day and the 249th day at the 226th day.Reason may be serious vegetation
It influences.Therefore, the 178th day based on PCA method and the 249th day estimation SM are best, other than the 249th day, based on PCA's
Estimate that SM is better than the SM based on MCA.
The NO emissions reduction that table .6 measures SM in situ using RMSE the and R value and reduction FY3-B SM of BME algorithm estimation SM
5. conclusion
BME algorithm is used in SEKS-GUI software, SM content is estimated three corn growth stages, including three in this research
A program, first program: HD and probability SD preparation;Step 2: SM estimates;Third program: estimate the verifying of SM.First
In a program, MCA and PCA are for calculating the weight value of environmental factor.Then, by using NO emissions reduction FY3-B and auxiliary data
(i.e. NDVI-LST-Albedo/MODIS, ASTER DEM, slope, aspect ratio, planar curvature, profile curvature, rough surface
Degree, histogram form of probability) generating probability SD.In three growth phases of corn, the humidity index of 4 cases and
The alleviation amplitude product that ASTER DEM is generated.Finally, obtaining 24 probability SD data sets.During second, BME is used
Algorithm estimates SM content according to the HD data and probability SD that generate first during in SEKS-GUI software.The third side
The SM of method, estimation is assessed for original position SM measurement (RMSE and R, i.e. equation 13-14).
Based on MCA and PCA method, principal element depends on the growth phase of corn in table 3 and 4.In the corn seeding phase,
LST, NDVI, planar curvature and profile curvature are occupied an leading position.In the heading-stage, LST, NDVI, albedo and profile curvature are accounted for
Leading position.In florescence, NDVI and profile curvature are dominant.NDVI is one of the principal element of all growth phases, should be by
It is considered as the important parameter of SM spatial prediction.Secondly, PCA is a kind of effective space dimension reduction method, data can be subtracted from 11 kinds
It is few to arrive only 4 kinds of main components.According to the SM estimated result (Fig. 3 and table 6) for using BME algorithm, the 178th day based on PCA and
249th day estimation SM is optimal.At the 226th day, the estimation SM based on MCA was better than the SM based on PCA method.In addition, I
Draw a conclusion, the SM(estimated using BME algorithm for example, the case 1 based on PCA, RMSE=0.049 m3/m3, R=
0.639) consistent with in-site measurement SM, rather than NO emissions reduction FY3-B SM(RMSE=0.079m3/m3, R=0.096).
In our current research, we are estimated and are estimated SM using BME algorithm using three kinds of different programs.It is walked at first
In rapid, we have prepared the HD and probability SD that estimate SM using BME algorithm in SEKS-GUI software.Number is measured with SM in situ
According to as HD data.Probability SD is respectively by FY3-BSM, NDVI-LST-Albedo/MODIS product of NO emissions reduction, ASTER DEM
Product and the gradient based on MCA (multivariate correlation analysis) and PCA (principal component analysis), slope aspect, planar curvature, section are bent
Rate, surface roughness, humidity index and relief intensity data develop.Wherein, the gradient, slope aspect, planar curvature, section are bent
Rate, surface roughness, humidity index and relief intensity data are from ASTER DEM product.MCA and PCA is for obtaining difference
(i.e. NDVI, LST, Albedo, BME, the gradient, slope aspect, planar curvature, profile curvature, surface roughness, humidity refer to environmental factor
Several and relief intensity) weighted value, this research is using distribution form of the histogram probability distribution as probabilistic standard difference.This point
Cloth form assumes that the probability distribution data in each section are constants, and the size in all sections is all identical.This research uses
Two methods (MCA and PCA) test three breeding times of 4 corns, obtain 24 probabilistic standard difference data collection.Second
During a, estimation space SM is entered data to using HD and probability SD as SEKS-GUI software.On this basis, it utilizes
BME has obtained the mapping graph of 16 SM estimation.During third, the SM estimated using BME algorithm is by using equation 13-
14 live SM measurement is verified.
During this investigation it turned out, since the accuracy for using regression model to reduce FY3-B SM is not high, the drop ruler with error
Degree FY3-B SM generates probability SD(i.e. NDVI-LST-Albedo/MODIS with auxiliary data using only SD),
ASTER DEM and the gradient, aspect ratio, planar curvature, profile curvature, surface roughness and are generated humidity index by ASTER DEM
Relief intensity product.Histogram form of probability is considered as the form of probability SD distribution, wherein assuming probability distribution data
It is steady state value in each section, and the size in all sections is identical.In the generating process of probability-weighted SD, we are used
Two methods of multivariate correlation analysis (MCA) method and principal component analysis (PCA) method determine the weight between auxiliary data.
In abrasion, since probability distribution is influenced by sample points, we used four case 4(1:500 soft datas of case
Point;2:450 soft data point of case;3:400 soft data point of case;4:350 soft data point of case).As certainty value
HD be used to correct the probability of the probability SD of SM to obtain SM after be distributed, that is, obtain the more high accuracy probability of SM.In the research
In, actual measurement SM in situ is considered as HD, is distributed after being used to calibrate probability of the probability soft data to obtain SM.In program 2
In, by using BME algorithm (special-temporal slices knowledge synthesising pattern user interface) in SEKS-GUI, two kinds of weights based on HD
The probability SD for determining method (MCA and PCA) and SM estimates that SM is in 3 corn growth stages at 4 kinds.Software, this is
One software package provided free is currently scientific software library BMElib(Bayes's maximum entropy library) and GUI(graphical user circle
Face) combination.
The above is only the preferred embodiments of the invention, are not intended to limit the invention creation, all in the present invention
Any modifications, equivalent replacements, and improvements etc. done within the spirit and principle of creation, should be included in the guarantor of the invention
Within the scope of shield.
Claims (2)
1. a kind of Bayes's soil moisture evaluation method based on multi- source Remote Sensing Data data, it is characterised in that: this method includes following
Step:
Step 1 obtains soil moisture soft data and carries out data processing
Specified region, the soil moisture data of scheduled date are obtained using FY3-B monitoring soil moisture product;And to FY3-B soil
Earth moisture data carries out NO emissions reduction processing, is divided 25km resolution ratio FY3-B soil moisture data space by using regression model
Resolution rises to 1km resolution ratio soil moisture data;
(1)
WhereinA 25For 25 km average values of albedo (A),V 25For the 25 km average values of vegetation index NDVI(V),T 25For earth's surface
Temperature LST(T) 25 km average values,mWithnIt is in FY3-B pixel respectivelyiRow and thejThe quantity of 1km pixel in column;A ijIt is
In FY3-BiRow and thejThe albedo value of 1Km pixel in column,V ijIt is in FY3-BiRow and thej1Km pixel in column
NDVI value,T ijIt is in FY3-B iRow and thejThe surface temperature LST value of 1Km pixel in column;
Linear regression model (LRM) are as follows: SM=a1+a2 A+a3 V+a4 T+a5 AV+a6 AT+a7 VT(2)
Wherein SM is soil moisture;a1、a2、a3、a4、a5、a6And a7It is regression coefficient;A is albedo, V is that vegetation index, T are ground
Table temperature;
Firstly, the A that formula (1) is calculated25、V25And T25Formula (2) are brought into, according to corresponding grid cell size 25km resolution ratio
FY3-B soil moisture data SM25, pass through the regression coefficient in regression analysis computation model;
Secondly, by the albedo (A) of MODIS product 1km resolution ratio, vegetation index NDVI(V) and surface temperature LST(T) bring into
The Soil moisture SM of 1km resolution ratio has accordingly been calculated in model formation (2)1;
Step 2 obtains environmental factor soft data
Albedo (A) is obtained using MODIS product, vegetation index NDVI(V) and surface temperature LST(T);Use ASTER product
Obtain altitude data;And the gradient, slope aspect, planar curvature, section song are calculated by the altitude data that ASTER product obtains
Rate, surface roughness, humidity index and relief intensity data;
Step 3 obtains NO emissions reduction FY3-B soil moisture data SM using step 11Environmental factor soft data is obtained with step 2, is adopted
With histogram form of probability, probability soft data is obtained;
Step 4 determines environmental factor soft data in step 2 using multivariate correlation analysis method and principal component analytical method
Weighted value, and obtain probability-weighted soft data;
Firstly, calculating thejA environmental factor and soil moisture data between quantitative probabilities, be denoted asP SD (EF j ):
(3)
Wherein Ef represents the environmental factor data obtained in step 2;nRepresent SM'snA section;kIndicate the of SMkIt is a
Section,kValue from 1 ton;mRepresent Ef'smA section;It is the of SM in first of Ef intervalkIn a interval
Sampling number;It is the of SMkTotal number of sample points in a section;EfmIt ismThe environmental factor in a section;Each
In section, probability distribution data are constant values, and the size in all sections is identical;
Secondly, calculating wherein thejThe weighted value w of a environmental factor data j Are as follows:
(4)
Wherein,iFor the number of environmental factor in step 2,iFor value from 1 to 11, Rj isjA environmental factor and the FY3B soil water
Related coefficient between divided data; EfjIt isjA environmental factor;NCC(Efj) it isjThe normalization phase relation of a environmental factor
Number, the sum of NCC of all environmental factors are equal to 1, wherein NCC(Efj) value more than or equal to 0.18 environmental factor be leading environment
The factor retains its environmental factor respective weights value, NCC(Efj) environmental factor weighted value of the value less than 0.18 be 0;
Again, the probability-weighted for calculating SM, is denoted as WPSD,
(5)
Step 5, actual measurement obtain soil moisture hard data
In step 1 the same area, phase same date, the pedotheque of 0-5cm depth is collected, and measures soil using aluminium box baking method
Earth moisture content;All parametric cubics are measured, and average value is used for subsequent analysis, are denoted as;
Step 6, using Bayes's maximum entropy method to the soil moisture probability-weighted soft data in step 4And in step 5
Measured dataIt is merged as hard data, estimates soil moisture data, be as a result denoted as;
Step 7 assesses Prediction of Soil Water Content using root-mean-square error (RMSE) and related coefficient (R)As a result accurate
Degree, wherein
(6)
(7)
Wherein,nIt is the total degree that soil moisture is surveyed in step 5,It is in step 5iSecondary actual measurement Soil moisture,It is in step 5nThe average value of secondary actual measurement soil moisture,It is in step 6iSecondary Bayes's maximum entropy
Soil moisture,In step 6nThe average value of secondary Bayes's maximum entropy soil moisture.
2. a kind of Bayes's soil moisture evaluation method based on multi- source Remote Sensing Data data according to claim 1, feature
It is, wherein probability-weighted soft data number is set as 4 classes, respectively 500,450,400 and 350 in step 6, obtains
Soil moisture estimation result under four type different weights probability soft data numbers.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910671852.6A CN110427995B (en) | 2019-07-24 | 2019-07-24 | Bayesian soil moisture estimation method based on multi-source remote sensing data |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910671852.6A CN110427995B (en) | 2019-07-24 | 2019-07-24 | Bayesian soil moisture estimation method based on multi-source remote sensing data |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110427995A true CN110427995A (en) | 2019-11-08 |
CN110427995B CN110427995B (en) | 2023-11-17 |
Family
ID=68412139
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910671852.6A Active CN110427995B (en) | 2019-07-24 | 2019-07-24 | Bayesian soil moisture estimation method based on multi-source remote sensing data |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110427995B (en) |
Cited By (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111089827A (en) * | 2020-01-07 | 2020-05-01 | 华南农业大学 | Measuring system and measuring method for water penetration of citrus garden soil |
CN111651411A (en) * | 2020-04-21 | 2020-09-11 | 成都信息工程大学 | Complex terrain remote sensing soil moisture product downscaling method |
CN111742329A (en) * | 2020-05-15 | 2020-10-02 | 安徽中科智能感知产业技术研究院有限责任公司 | Mining typical ground object dynamic monitoring method and platform based on multi-source remote sensing data fusion and deep neural network |
CN111861222A (en) * | 2020-07-22 | 2020-10-30 | 中国水利水电科学研究院 | Method for acquiring farmland and grassland roughness facing regional scale wind erosion |
CN112528555A (en) * | 2020-11-23 | 2021-03-19 | 中国科学院空天信息创新研究院 | Soil moisture map drawing method and device |
CN112560570A (en) * | 2020-09-29 | 2021-03-26 | 中国科学院大学 | High-resolution earth surface temperature estimation method based on cooperative downscaling and data fusion |
CN113033262A (en) * | 2019-12-25 | 2021-06-25 | 中移(成都)信息通信科技有限公司 | Model training method and crop yield estimation method |
CN113419046A (en) * | 2021-06-17 | 2021-09-21 | 北京大学 | Improved soil humidity product bivariate fusion method |
CN113763409A (en) * | 2021-08-27 | 2021-12-07 | 南通大学 | Gauss filtering-based multi-threshold segmentation method for high-dimensional brain nuclear magnetic image |
CN114461983A (en) * | 2022-02-09 | 2022-05-10 | 中国科学院、水利部成都山地灾害与环境研究所 | Satellite precipitation product space downscaling method based on water balance principle |
CN115204479A (en) * | 2022-07-01 | 2022-10-18 | 南京信息工程大学 | Crop survival area prediction method and system based on multi-source remote sensing data |
CN117235607A (en) * | 2023-11-13 | 2023-12-15 | 天津市农业发展服务中心 | Soil moisture content real-time monitoring system and monitoring method |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105550423A (en) * | 2015-12-09 | 2016-05-04 | 浙江大学 | CMORPH satellite precipitation data downscaling method based on Fuzzy-OLS (Ordinary Least Squares) |
CN106019408A (en) * | 2016-05-10 | 2016-10-12 | 浙江大学 | Multi-source-remote-sensing-data-based high-resolution-ratio satellite remote-sensing estimation method |
CN106446444A (en) * | 2016-10-14 | 2017-02-22 | 中国科学院遥感与数字地球研究所 | Soil moisture spatial predication research based on Bayes maximum entropy and priori knowledge |
CN106483147A (en) * | 2016-10-14 | 2017-03-08 | 中国科学院遥感与数字地球研究所 | The long-term sequence passive microwave soil moisture accuracy improvements research worked in coordination with based on MODIS and measured data |
US20170122889A1 (en) * | 2014-06-18 | 2017-05-04 | Texas Tech University System | Portable Apparatus for Soil Chemical Characterization |
CN108169161A (en) * | 2017-12-12 | 2018-06-15 | 武汉大学 | A kind of corn planting regional soil humidity appraisal procedure based on modified MODIS indexes |
CN108268735A (en) * | 2018-01-29 | 2018-07-10 | 浙江大学 | Soil moisture NO emissions reduction method based on multi-source remote sensing satellite fused data |
CN109002604A (en) * | 2018-07-12 | 2018-12-14 | 山东省农业科学院科技信息研究所 | A kind of soil moisture content prediction technique based on Bayes's maximum entropy |
CN109460789A (en) * | 2018-11-07 | 2019-03-12 | 中国农业科学院农田灌溉研究所 | A kind of soil moisture fusion method based on Bayes's maximum entropy |
-
2019
- 2019-07-24 CN CN201910671852.6A patent/CN110427995B/en active Active
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170122889A1 (en) * | 2014-06-18 | 2017-05-04 | Texas Tech University System | Portable Apparatus for Soil Chemical Characterization |
CN105550423A (en) * | 2015-12-09 | 2016-05-04 | 浙江大学 | CMORPH satellite precipitation data downscaling method based on Fuzzy-OLS (Ordinary Least Squares) |
CN106019408A (en) * | 2016-05-10 | 2016-10-12 | 浙江大学 | Multi-source-remote-sensing-data-based high-resolution-ratio satellite remote-sensing estimation method |
CN106446444A (en) * | 2016-10-14 | 2017-02-22 | 中国科学院遥感与数字地球研究所 | Soil moisture spatial predication research based on Bayes maximum entropy and priori knowledge |
CN106483147A (en) * | 2016-10-14 | 2017-03-08 | 中国科学院遥感与数字地球研究所 | The long-term sequence passive microwave soil moisture accuracy improvements research worked in coordination with based on MODIS and measured data |
CN108169161A (en) * | 2017-12-12 | 2018-06-15 | 武汉大学 | A kind of corn planting regional soil humidity appraisal procedure based on modified MODIS indexes |
CN108268735A (en) * | 2018-01-29 | 2018-07-10 | 浙江大学 | Soil moisture NO emissions reduction method based on multi-source remote sensing satellite fused data |
CN109002604A (en) * | 2018-07-12 | 2018-12-14 | 山东省农业科学院科技信息研究所 | A kind of soil moisture content prediction technique based on Bayes's maximum entropy |
CN109460789A (en) * | 2018-11-07 | 2019-03-12 | 中国农业科学院农田灌溉研究所 | A kind of soil moisture fusion method based on Bayes's maximum entropy |
Non-Patent Citations (2)
Title |
---|
王春梅 等: "被动微波土壤水分产品真实性检验研究进展", 《浙江农业学报》 * |
王春梅 等: "被动微波土壤水分产品真实性检验研究进展", 《浙江农业学报》, vol. 31, no. 05, 31 May 2019 (2019-05-31), pages 846 - 854 * |
Cited By (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113033262B (en) * | 2019-12-25 | 2022-08-05 | 中移(成都)信息通信科技有限公司 | Model training method and crop yield estimation method |
CN113033262A (en) * | 2019-12-25 | 2021-06-25 | 中移(成都)信息通信科技有限公司 | Model training method and crop yield estimation method |
CN111089827A (en) * | 2020-01-07 | 2020-05-01 | 华南农业大学 | Measuring system and measuring method for water penetration of citrus garden soil |
CN111651411A (en) * | 2020-04-21 | 2020-09-11 | 成都信息工程大学 | Complex terrain remote sensing soil moisture product downscaling method |
CN111651411B (en) * | 2020-04-21 | 2023-03-10 | 成都信息工程大学 | Complex terrain remote sensing soil moisture product downscaling method |
CN111742329A (en) * | 2020-05-15 | 2020-10-02 | 安徽中科智能感知产业技术研究院有限责任公司 | Mining typical ground object dynamic monitoring method and platform based on multi-source remote sensing data fusion and deep neural network |
CN111742329B (en) * | 2020-05-15 | 2023-09-12 | 安徽中科智能感知科技股份有限公司 | Mining typical feature dynamic monitoring method and platform based on multi-source remote sensing data fusion and deep neural network |
CN111861222A (en) * | 2020-07-22 | 2020-10-30 | 中国水利水电科学研究院 | Method for acquiring farmland and grassland roughness facing regional scale wind erosion |
CN111861222B (en) * | 2020-07-22 | 2023-11-14 | 中国水利水电科学研究院 | Method for obtaining roughness of cultivated land and grassland facing regional scale wind erosion |
CN112560570B (en) * | 2020-09-29 | 2023-07-25 | 中国科学院大学 | High-resolution earth surface temperature estimation method for collaborative downscaling and data fusion |
CN112560570A (en) * | 2020-09-29 | 2021-03-26 | 中国科学院大学 | High-resolution earth surface temperature estimation method based on cooperative downscaling and data fusion |
CN112528555B (en) * | 2020-11-23 | 2023-02-21 | 中国科学院空天信息创新研究院 | Soil moisture map drawing method and device |
CN112528555A (en) * | 2020-11-23 | 2021-03-19 | 中国科学院空天信息创新研究院 | Soil moisture map drawing method and device |
CN113419046B (en) * | 2021-06-17 | 2022-02-18 | 北京大学 | Improved soil humidity product bivariate fusion method |
CN113419046A (en) * | 2021-06-17 | 2021-09-21 | 北京大学 | Improved soil humidity product bivariate fusion method |
CN113763409A (en) * | 2021-08-27 | 2021-12-07 | 南通大学 | Gauss filtering-based multi-threshold segmentation method for high-dimensional brain nuclear magnetic image |
CN114461983A (en) * | 2022-02-09 | 2022-05-10 | 中国科学院、水利部成都山地灾害与环境研究所 | Satellite precipitation product space downscaling method based on water balance principle |
CN115204479A (en) * | 2022-07-01 | 2022-10-18 | 南京信息工程大学 | Crop survival area prediction method and system based on multi-source remote sensing data |
CN117235607A (en) * | 2023-11-13 | 2023-12-15 | 天津市农业发展服务中心 | Soil moisture content real-time monitoring system and monitoring method |
CN117235607B (en) * | 2023-11-13 | 2024-02-13 | 天津市农业发展服务中心 | Soil moisture content real-time monitoring system and monitoring method |
Also Published As
Publication number | Publication date |
---|---|
CN110427995B (en) | 2023-11-17 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110427995A (en) | A kind of Bayes's soil moisture evaluation method based on multi- source Remote Sensing Data data | |
Betbeder et al. | Assimilation of LAI and dry biomass data from optical and SAR images into an agro-meteorological model to estimate soybean yield | |
Huang et al. | Assimilating a synthetic Kalman filter leaf area index series into the WOFOST model to improve regional winter wheat yield estimation | |
AU2016244067B2 (en) | Forecasting national crop yield during the growing season | |
Huang et al. | Jointly assimilating MODIS LAI and ET products into the SWAP model for winter wheat yield estimation | |
Wang et al. | Predictive mapping of soil total nitrogen at a regional scale: A comparison between geographically weighted regression and cokriging | |
Casa et al. | Forcing a wheat crop model with LAI data to access agronomic variables: Evaluation of the impact of model and LAI uncertainties and comparison with an empirical approach | |
Wendroth et al. | Predicting yield of barley across a landscape: a state-space modeling approach | |
Montzka et al. | Modelling the water balance of a mesoscale catchment basin using remotely sensed land cover data | |
Tian et al. | Estimating zero-plane displacement height and aerodynamic roughness length using synthesis of LiDAR and SPOT-5 data | |
CN108647401B (en) | Watershed nitrogen and phosphorus pollution assessment method based on space remote sensing technology | |
WO2016160384A1 (en) | Forecasting national crop yield during the growing season | |
Yin et al. | Estimation of grassland height based on the random forest algorithm and remote sensing in the Tibetan Plateau | |
US20210223433A1 (en) | Determination of location-specific weather information for agronomic decision support | |
CN110222870A (en) | Assimilate the Regional Fall Wheat yield estimation method of satellite fluorescence data and crop growth model | |
CN109325433A (en) | Introduce the black soil region soybean biomass multi-temporal remote sensing inversion method of terrain factor | |
Zhang et al. | Winter wheat identification by integrating spectral and temporal information derived from multi-resolution remote sensing data | |
Liu et al. | Winter wheat yield estimation based on assimilated Sentinel-2 images with the CERES-Wheat model | |
CN108205718A (en) | Production method and system are surveyed in a kind of cereal crops sampling | |
Cui et al. | Estimating maize yield in the black soil region of Northeast China using land surface data assimilation: integrating a crop model and remote sensing | |
Matthews et al. | A field parcel-oriented approach to evaluate the crop cover-management factor and time-distributed erosion risk in Europe | |
Garg et al. | Digital mapping of soil landscape parameters | |
Zhang et al. | Estimation of rice yield from a C-band radar remote sensing image by integrating a physical scattering model and an optimization algorithm | |
Wang et al. | The distributed CERES-Maize model with crop parameters determined through data assimilation assists in regional irrigation schedule optimization | |
Cheng et al. | ChinaCropSM1 km: a fine 1 km daily soil moisture dataset for dryland wheat and maize across China during 1993–2018 |
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 | ||
TA01 | Transfer of patent application right |
Effective date of registration: 20230717 Address after: 100190 No. 19 West North Fourth Ring Road, Haidian District, Beijing Applicant after: Aerospace Information Research Institute,Chinese Academy of Sciences Address before: No. a 20, Datun Road, Chaoyang District, Beijing 100101 Applicant before: Institute of Remote Sensing and Digital Earth, Chinese Academy of Sciences |
|
TA01 | Transfer of patent application right | ||
GR01 | Patent grant | ||
GR01 | Patent grant |