CN115292616B - Vegetation blue sky albedo estimation method and device based on spectrum invariant theory - Google Patents
Vegetation blue sky albedo estimation method and device based on spectrum invariant theory Download PDFInfo
- Publication number
- CN115292616B CN115292616B CN202210770366.1A CN202210770366A CN115292616B CN 115292616 B CN115292616 B CN 115292616B CN 202210770366 A CN202210770366 A CN 202210770366A CN 115292616 B CN115292616 B CN 115292616B
- Authority
- CN
- China
- Prior art keywords
- vegetation
- data
- sentinel
- albedo
- reflectivity
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
- G01N21/17—Systems in which incident light is modified in accordance with the properties of the material investigated
- G01N21/55—Specular reflectivity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2455—Query execution
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Quality & Reliability (AREA)
- Software Systems (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Image Processing (AREA)
Abstract
The invention provides a vegetation blue sky albedo estimation method and device based on a spectrum invariant theory, wherein the method comprises the following steps: acquiring Sentinel-2 data and static meteorological satellite data of a target area; extracting clear empty Sentinel-2 data in the Sentinel-2 data; determining the single scattering albedo of the blade and the soil background reflectivity based on clear sky Sentinel-2 data; calculating vegetation growth rhythms of a target area, and determining a time sequence of a daily vegetation leaf area index by combining with Sentinel-2 data; determining an hour-by-hour sky scattered light proportion based on the stationary meteorological satellite data; and acquiring the vegetation blue sky albedo of the target area based on the blade single scattering albedo, the soil background reflectivity, the time sequence of the day-by-day vegetation leaf area index and the sky scattered light proportion hour by hour. The invention gets rid of the dependence on multi-angle satellite observation data for vegetation albedo estimation, and can provide ten-meter-level and hour-by-hour continuous vegetation albedo data under different vegetation types and different weather conditions.
Description
Technical Field
The invention relates to the technical field of remote sensing image recognition, in particular to a vegetation blue sky albedo estimation method and device based on a spectrum invariant theory.
Background
The Albedo of the earth's surface short wave band refers to the ratio of the reflected radiation energy to the incident energy of all earth's surfaces in the hemispherical space of the solar short wave band (0.3-3.0 μm), and is a key parameter widely applied to global change researches such as the earth's surface energy balance, medium-long term weather prediction, global climate model (Global Climate Model, GCM) and the like.
Vegetation is an important component of surface coverage, accounting for 29% of the land surface to atmosphere interface (Graetz, 1990). Compared with other land object types such as bare soil, water body and the like, the space-time dynamic of vegetation albedo is more remarkable. In the space dimension, vegetation types are various, the heterogeneity of spatial distribution is strong, and different vegetation types and albedo difference are obvious. When vegetation type and vegetation coverage change, a complex series of changes in surface energy balance will occur, in which albedo plays a key role (Charney et al, 1975). On the time dimension, on one hand, the vegetation albedo presents a fluctuation change rule in the year along with the periodical change of the vegetation growth rhythm; on the other hand, global vegetation cover shows a significant greenish trend (zhueal, 2016) in nearly 30 years under the interference of external environment, such as global warming, human activity interference, etc., and this greenish trend has significantly affected albedo of the earth surface, and will also lead to redistribution of earth surface heat and moisture, ultimately affecting the energy balance and moisture circulation of the earth (Zeng et al, 2017,2018;Li et al, 2018).
Remote sensing is the only means to obtain long-term, large-scale earth surface albedo time series. With the remarkable increase of medium-high resolution satellite data resources, medium-high resolution albedo remote sensing estimation is a current research hotspot. However, the existing medium-high resolution albedo remote sensing algorithm cannot meet the requirement of dynamic continuous monitoring of vegetation coverage areas due to limitation of revisit periods and observation angles, and a targeted albedo algorithm with high space-time resolution is urgently needed.
Disclosure of Invention
In order to overcome the defect that the existing albedo estimation algorithm cannot meet the requirements of vegetation space-time rapid change monitoring, the inventor performs intensive research, provides a Sentinel-2 hour-by-hour vegetation albedo estimation A-P algorithm based on a spectrum invariance theory, and performs accuracy verification of the algorithm by combining ground actual measurement data.
The technical content of the invention comprises:
a vegetation blue sky albedo estimation method based on spectrum invariant theory, the method comprising:
acquiring Sentinel-2 data and static meteorological satellite data of a target area;
extracting clear empty Sentinel-2 data in the Sentinel-2 data, and determining single scattering albedo of the blade and the soil background reflectivity based on the clear empty Sentinel-2 data;
calculating vegetation growth rhythms of the target area, and determining a daily vegetation leaf area index time sequence by combining the Sentinel-2 data;
determining an hour-by-hour sky scattered light proportion based on the stationary meteorological satellite data;
and acquiring the vegetation blue sky albedo of the target area based on the blade single scattering albedo, the soil background reflectivity, the time series of the daily vegetation leaf area indexes and the hourly sky scattered light proportion.
Further, the extracting the clear empty Sentinel-2 data in the Sentinel-2 data comprises the following steps:
acquiring Sentinel-2 data with cloud coverage smaller than a set value from the Sentinel-2 data;
preprocessing the Sentinel-2 data smaller than a set value to obtain clear sky Sentinel-2 data, wherein the preprocessing comprises the following steps: radiometric calibration, geometric correction and atmospheric correction of images.
Further, the determining the single scattering albedo of the blade based on clear sky Sentinel-2 data in the Sentinel-2 data comprises:
aiming at the change of the single scattering albedo of the blade of the sunny Sentinel-2 data in the visible light wave band, a lookup table with chlorophyll values as lookup conditions is established based on a PROSAIL model so as to determine the single scattering albedo of the blade in the visible light wave band;
based on the clear sky Sentinel-2 data, establishing a lookup table from infrared soil reflectivity, vegetation coverage and pure vegetation canopy near infrared reflectivity index to the blade single scattering albedo so as to determine the blade single scattering albedo of a near infrared band;
obtaining the single scattering albedo of the blade in the short wave infrared band according to the equivalent water thickness of the blade and the single scattering albedo of the blade in the near infrared band;
and synthesizing the single scattering albedo of the blade in the visible light band, the near infrared band and the short wave infrared band to obtain the single scattering albedo of the blade.
Further, the establishing a lookup table from infrared soil reflectivity, vegetation coverage, pure vegetation canopy near infrared reflectivity index to the blade single scattering albedo based on the clear sky Sentinel-2 data includes:
estimating vegetation contribution degree of a vegetation covered part in a pixel based on the data of the clear sky Sentinel-2, wherein the pixel comprises the vegetation covered part and a soil part;
acquiring vegetation coverage in the pixels;
calculating a near infrared reflectivity index of a pure vegetation canopy according to the vegetation coverage and the vegetation contribution degree;
when the correlation between the pure vegetation canopy near infrared reflectivity index and the blade single scattering albedo is established by means of a PROSAIL model, the distribution and the fixed value of parameters are simulated to determine the influence of each parameter on the blade single scattering albedo, so that a lookup table from the infrared soil reflectivity, vegetation coverage and pure vegetation canopy near infrared reflectivity index to the blade single scattering albedo is obtained.
Further, the calculating vegetation growth rhythms of the target area includes:
obtaining the maximum value NDVI of the original normalized difference vegetation index in one year max Minimum NDVI of the initial normalized difference vegetation index at the up-take stage min ；
Extracting a dynamic threshold tau in a growing season based on the variation amplitude of the original normalized difference vegetation index;
according to the maximum value NDVI max Said minimum value NDVI min Calculating a turning green period t according to the dynamic threshold tau;
calculating the withered and yellow period t';
and obtaining the vegetation growth rhythm according to the turning green period t and the withered and yellow period t'.
Further, the determining a time series of day-to-day vegetation leaf area indexes by combining the Sentinel-2 data comprises:
carrying out outlier rejection on the Sentinel-2 data by utilizing cloud mask data;
acquiring an original normalized difference vegetation index time sequence based on the Sentinel-2 data with the outlier removed, and reconstructing the original normalized difference vegetation index time sequence by combining the vegetation growth rhythm to acquire a normalized difference vegetation index time sequence in a growing season and a normalized difference vegetation index time sequence in a non-growing season;
setting a daily vegetation leaf area index according to the normalized difference vegetation index time sequence of the non-growing season;
aiming at the normalized difference vegetation index time sequence of the growing season, inversion of the daily vegetation leaf area index is carried out by utilizing a LUT method based on a BRDF unified model;
comprehensively growing Ji Yufei to obtain a discrete time sequence of the daily vegetation leaf area index;
and reconstructing the discrete daily vegetation leaf area index time sequence by adopting a time sequence reconstruction algorithm to obtain a daily LAI time sequence.
Further, the determining an hour-by-hour sky scattered light proportion based on the stationary meteorological satellite data includes:
estimating the downstream direct radiation and the downstream total radiation data of M kilometer resolution based on the stationary meteorological satellite data;
and estimating the sky scattered light proportion of each hour according to the downlink direct radiation and the downlink total radiation data so as to obtain the sky scattered light proportion of each hour.
Further, the obtaining vegetation blue sky albedo of the target area based on the blade single scattering albedo, the soil background reflectivity, the time series of the daily vegetation leaf area index and the hourly sky scattered light proportion includes:
setting the probability P that photons collide with the blades again after being scattered by the blades, and calculating the probability a that photons entering in a single direction are absorbed by vegetation after one or more collisions with the vegetation by combining the single scattering albedo c_dir Probability a that photons incident in hemispherical direction are absorbed by vegetation after one or more collisions with the vegetation c_diff ；
Setting a function G describing canopy leaf inclination angle distribution and an aggregation index xi, and calculating interception probability i of canopy on unidirectional incident photons on each day by combining the time sequence of the daily vegetation leaf area index 0 ；
Calculating the interception probability i 0 Integrating in hemispherical space to obtain the interception probability of incident photons in hemispherical direction by the canopy
Setting probability q of scattered photons falling out of the canopy after collision with the blade d Calculating vegetation scattering reflectivity of photons uniformly incident in all directions by combining the single scattering albedo and the probability P
Based on the soil background reflectivity, the vegetation scattering reflectivitySaid interception probability->Respectively calculating the probability P that photons are not absorbed after being subjected to any collision between soil and vegetation sr Probability of absorption P with photons after undergoing any number of collisions between soil and vegetation sa
According to the proportion of the scattered light of the hourly sky and the interception probabilityi 0 The interception probability i 0 Said probability a c_dir Said probability a c_diff The vegetation scattering reflectivitySaid probability P sr Obtaining the absorption of vegetation parts under different wavelengths;
according to the soil background reflectivity, the hourly sky scattered light proportion and the interception probability i 0 The interception probability i 0 The vegetation scattering reflectivitySaid probability P sa Obtaining the absorption of soil parts under different wavelengths;
and obtaining the vegetation blue sky albedo according to the absorption of the plant parts under the different wavelengths and the absorption of the soil parts under the different wavelengths.
A storage medium having a computer program stored therein, wherein the computer program is arranged to perform any of the methods described above when run.
An electronic device comprising a memory having a computer program stored therein and a processor arranged to run the computer program to perform any of the methods described above.
Compared with the prior art, the invention has the beneficial effects that:
1. the vegetation albedo estimation gets rid of the dependence on multi-angle satellite observation data;
2. can provide vegetation albedo data of different vegetation types, ten meters under different weather conditions and continuous from hour to hour.
Drawings
FIG. 1 shows a flowchart of a Sentinel-2 hour-by-hour vegetation albedo remote sensing estimation "A-P" algorithm based on a spectrum invariant theory.
Fig. 2 shows a day-by-day LAI reconstruction flow diagram.
FIG. 3 shows a flowchart for inverting LAI based on BRDF unified model.
Fig. 4 shows estimating hourly vegetation albedo based on reconstructed LAI data.
Fig. 5 shows the results of the "a-P" algorithm accuracy verification based on ground measured data.
Detailed Description
The invention is further described in detail below by means of the figures and examples. The features and advantages of the present invention will become more apparent from the description.
The word "exemplary" is used herein to mean "serving as an example, embodiment, or illustration. Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments. Although various aspects of the embodiments are illustrated in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
Referring to fig. 1, the vegetation blue sky albedo estimation method based on the spectrum invariant theory of the invention comprises the following steps:
step 1: determining parameters required by a model based on clear sky Sentinel-2 data: blade single scattering albedo and soil background reflectivity.
The sunny Sentinel-2 data provided by the invention can meet the requirements that the image is less influenced by cloud (the cloud coverage is less than 5%), and the data preprocessing processes such as radiometric calibration, geometric correction, atmospheric correction and the like of the image are required to be completed, so that the reflectivity data of the atmosphere bottom are obtained, and then the input parameters required by the model are acquired.
(1) Blade single scattering albedo (Single Scattering Albedo, SSA)
Blade SSA can be obtained by measuring the reflectivity and transmissivity of the blade through a ground spectrometer, but SSA of the blade with certain discrete points can only be obtained, and the blade SSA can not be suitable for estimating the albedo of vegetation in a large range based on remote sensing satellite data.
The reflection and transmission characteristics of vegetation blades in different wave bands are related to physicochemical parameters of the blades, and the blade SSA of each wave band of Sentinel-2 obtained according to a PROSPECT model changes along with chlorophyll (Cab), dry matter content (Cm) and equivalent water thickness (Cw): in the visible light band (B2, B3, B4), the blade SSA is mainly affected by Cab; in the near infrared band (B8 a), the blade SSA is mainly affected by Cm; in the short-wave infrared band (B11, B12), the blade SSA is mainly affected by Cm and Cw. Thus, the SSA for each band of the blade may be obtained by constructing a look-up table of blade physicochemical parameters and SSA for each band of the blade.
The Sentinel-2 official trains simulation data acquired based on a PROSAIL model (foundation et al, 1996;Jacquemoud and Baret,1990;Jacquemoud et al, 2009) by using a neural network method, providing a series of physicochemical parameter products of vegetation canopy: vegetation coverage (Fractional Vegetation Cover, FVC), canopy Cab, canopy Cw, etc. (Weiss et al, 2016); thus, in the visible light band, blade SSA can be determined based on a least squares algorithm by building a look-up table from blade Cab to blade SSA. Due to the lack of Cm data, the blade SSA of the near-infrared band and the short-wave infrared band cannot be determined by directly constructing a lookup table of physical and chemical parameters, and the single scattering albedo of the near-infrared blade needs to be obtained by constructing other indexes.
The invention provides an NIR-CR index for estimating vegetation reflectivity in the near infrared band. The NIR-CR index originates from the NIRvH2 index (Zeng et al 2021), which uses the difference in the red-near infrared band reflectance characteristics of soil and vegetation blades to decompose the reflectivity of the near infrared band canopy into a contribution of vegetation reflectivity and a contribution of soil reflectivity. The specific principle is as follows.
a) Absorption and reflection characteristics of blade spectrum in red-near infrared band
The reflectance spectrum curve of the blade is simulated by using a PROSPECT model, and the blade spectrum is found to show strong absorption characteristics in a red light wave band (650-680 nm), because chlorophyll has strong absorption effect in the wave band, and the reflectance (< 0.1) of the blade is very small; whereas in the near infrared band (780-860 nm) a plateau of high reflectivity (> 0.4) is present, at which time the chlorophyll absorption substantially disappears except for the absorption of the dry matter of the leaf, and in the transition between these two phases, the so-called red-edge band, the chlorophyll absorption goes through the process from strongest to weakest, the fluctuation of the chlorophyll affects the beginning and ending wavelengths of the red-edge band, but does not affect the magnitude of the near infrared plateau reflectivity value.
b) Reflection characteristics of red-near infrared band vegetation canopy and soil spectrum
The reflectance curve of the vegetation canopy was simulated based on the PROSAIL model, which shows relatively high reflectance in the red band during periods of low vegetation coverage (LAI < 3), while also showing a slow increase in reflectance during the plateau in the near infrared band, unlike the reflectance spectrum curve of the vegetation blades in the red-near infrared portion. In general, in the red to near infrared band (650-900 nm), the soil reflectivity spectrum curve shows a slow and steady rising trend along with the increase of wavelength, which can approximate to a linear growth process, so that the increase of the vegetation canopy reflectivity in the red and near infrared bands is from the influence of the soil background reflectivity; and when the vegetation LAI is increased from 0.5 to 5, the influence of the soil background reflectivity is gradually reduced, and when the vegetation LAI is increased to a certain degree (LAI > 5), the reflectivity spectrum curve of the vegetation canopy gradually approaches to the characteristics of the reflectivity spectrum curve of the blade: red light exhibits low reflectivity and near infrared exhibits a high reflectivity plateau.
The contribution of vegetation to the red-band canopy reflectivity is not more than 0.02 (Zeng et al, 2021) due to the influence of strong chlorophyll absorption, the leaf scattering rate of the red-band is very small, and the multiple scattering effect caused by the vegetation canopy structure, so that the canopy reflectivity of the red-band can be considered to basically provide information of the soil background reflectivity, and the reflectivity of the soil in the near-infrared band can be obtained as long as the slope of the approximately linear increase of the reflectivity of the soil in the red-near-infrared part is known, and the near-infrared reflectivity of the canopy is decomposed into the soil contribution and the vegetation contribution. Thus, a specific estimation principle of the NIRvH2 index can be expressed as formula (2) (Zeng et al, 2021).
NIRvH2＝BRF(NIR)-BRF(R)-k·(λ NIR -λ R ) (2)
Wherein BRF (NIR) is the reflectivity of the pixel in the near infrared portion, BRF (R) is the reflectivity of the pixel in the red band, and k is the slope of the soil spectrum curve in the red-near infrared band，λ NIR And lambda (lambda) R Wavelengths in the red and near infrared bands, respectively.
In combination with the band setting of the Sentinel-2MSI sensor, the NIRvH2 index can be estimated based on the Sentinel-2 data. The method selects a B4 wave band (center wavelength: 665 nm) of Sentinel-2 as a red light wave band, selects a B7 (center wavelength: 782.8 nm) wave band and a B8a (center wavelength: 842 nm) wave band as two near infrared wave bands in a platform phase, estimates the slope k of the linear growth of soil in the red-near infrared wave band by utilizing the difference of the reflectivities of the B7 and the B8a wave bands, and estimates the NIRvH2 index based on a formula (2).
The estimated NIRvH2 index represents the extent of vegetation contribution of vegetation covered part in the current pixel, the contribution of vegetation and soil to the whole pixel is formed by a certain linear weighted combination, and the weight coefficient is vegetation coverage FVC, so the invention provides an index-NIR-CR describing the near infrared reflectivity of a pure vegetation canopy, and the conversion relation between the NIR-CR and the NIRvH2 is shown as a formula (3).
NIR-CR＝NIRvH2/FVC (3)
Wherein NIR-CR describes the contribution to the pixel reflectivity when the pure vegetation is covered under a black soil background (i.e. the reflectivity of the soil background is 0), and NIRvH2 is the contribution to the pixel reflectivity by the vegetation canopy when the vegetation components and the soil components are mixed according to a certain proportion; FVC is the coverage of vegetation within a pel.
The NIR-CR index reflects the contribution of vegetation components, which are affected by the comprehensive incidence and observation geometry, vegetation canopy structure, blade SSA and the like, to pixel reflectivity without the influence of soil background, and the correlation between the NIR-CR index and the vegetation blade SSA is established by means of a PROSAIL model to obtain the SSA of the near infrared single blade. And simulating the reflectivity of the canopy by using a PROSAIL model for further analysis, wherein the numerical distribution or the fixed value of each parameter is simulated, and the specific set value of each parameter is shown in a table. The absorption coefficients of the components do not have any significant influence on the single scattering albedo of each band, and thus are set to a fixed value at the time of simulation. Soil background reflectivity the spectrum of the dry soil and the wet soil built-in by the PROSAIL model is adopted, and the soil background reflectivity is changed by changing the mixing ratio (rsoil) of the dry soil and the wet soil. In the simulation, the structural parameter (N) of the blade is set to be constant at 1.5, and the reflectivity and the transmissivity of the blade are approximately equal. Meanwhile, assuming that vegetation is a spherical crown, the leaf inclination distribution satisfies that the G function is equal to 0.5.
Table 1 PROSAIL model input parameters and list of numerical distribution ranges thereof
The simulation result of PROSAIL shows that the correlation between the soil background reflectivity and the near infrared SSA of the blade still has a certain influence, and the NIR-CR index and the near infrared SSA of the blade show one-to-one correspondence under the condition that the approximate range of the soil background reflectivity is determined. This is because the NIR-CR index is only as much as possible removed from the effects of soil single scattering and partial vegetation-soil multiple scattering, and thus the estimated NIR-CR index still contains some soil information; therefore, when constructing a lookup table from NIR-CR index to near-infrared leaf SSA based on PROSAIL model, it is necessary to consider near-infrared soil reflectivity, vegetation coverage FVC, and NIR-CR index as lookup conditions, and estimate the near-infrared leaf SSA pixel by pixel using a least squares algorithm.
After the near infrared blade SSA is determined, the blade SSA corresponding to the visible light wave band and the short wave infrared wave band of the Sentinel-2 can be determined by combining the Cab value and the Cw value provided by SNAP and constructing a lookup table based on a PROSAIL model.
(2) Reflectivity of soil background
Direction of vegetation albedo estimation model required soil-hemispherical reflectivityAnd hemisphere-hemisphere reflectivity->Both of these parameters describe the absorption and reflection capabilities of the soil background for direct and scattered radiation reaching the soil, respectively. However, since Sentinel-2 can only provide single angle reflectivity data, determining soil background reflectivity based on Sentinel-2 data requires two assumptions: (1) The soil background has a lambertian character, under this assumption the direction of the soil-hemispherical reflectivity +.>And hemisphere-hemisphere reflectivity->Equal; (2) In the same region, the soil texture will not change for a short period of time (e.g., one year) without the action of external conditions (precipitation or manual irrigation), i.e., the soil's dichroic reflectance characteristics will remain relatively constant. Based on the two assumptions above, the reflectivity of the soil background can be determined from Sentinel-2 images of bare soil periods (without ice and snow effects) in early spring or autumn winter seasons of the study area.
Research results of He et al (2019) show that the statistics of reflectivity of each wave band of the soil according to vegetation types is of practical application significance due to feedback effect of soil water-precipitation (Eltahir et al, 1998) and interaction mechanism of vegetation and soil water (Liu et al, 2010). Therefore, the invention respectively calculates bare soil reflectivities of 6 wave bands of B2, B3, B4, B8a, B11 and B12 corresponding to Sentinel-2 data according to a vegetation space distribution diagram. And according to the normalized difference vegetation index (Normalized Difference Vegetation Index, NDVI) <0.3, the judgment standard of the bare soil pixels is adopted.
Step 2: and filling the missing data of the cloud sky by combining the vegetation growth rhythm and a time sequence reconstruction algorithm, and determining model parameters: time series of Leaf Area Index (LAI) of day-to-day vegetation.
The LAI is an important parameter describing the vegetation canopy structure, is used for determining the interception probability, re-collision probability and the like of the vegetation canopy on photons, is used as an important input parameter of a vegetation albedo model, and is very important for estimating the vegetation albedo. Under a sunny condition, obtaining vegetation LAI data by adopting a LAI inversion algorithm based on a unified model based on the surface reflectivity data provided by Sentinel-2; however, due to the influence of the revisit period of the Sentinel-2 satellite and the cloud pollution problem (including the cloud coverage area and the cloud shadow area) under the cloud conditions, the daily surface reflectivity data cannot be obtained, and the difficulty of accurately estimating the daily vegetation LAI and further estimating the vegetation albedo with high time resolution is high.
The invention proposes a complete daily LAI time sequence reconstruction method by combining a periodical law of vegetation growth and a time sequence reconstruction algorithm, and the LAI data of discrete dates estimated based on a unified model is reconstructed into a daily LAI time sequence, and a specific flow can be divided into the following four parts (figure 2).
(1) And removing abnormal values by using the cloud mask data. Cloud detection is the most critical step in the satellite remote sensing image preprocessing process, otherwise, the pixels affected by the cloud (comprising cloud coverage pixels and cloud shadow region pixels) can increase the estimation error of the subsequent LAI. The influence of the cloud pollution problem on the reflectivity of the pixel comprises: the reflectivity of the pixels affected by the thick cloud is obviously increased in all wave bands, so that saturation is generated when inverting the LAI based on the BRDF unified model; the pixel affected by the thin cloud can cause the reflectivity of each wave band to be improved to a certain extent, so that the estimated LAI time sequence has a false high value; the pixel in the cloud shadow area is mainly irradiated by scattered radiation because solar radiation is shielded by the cloud layer, and the atmospheric scattering in the cloud shadow is strong in shorter wavelength energy (such as visible light wave band) and the scattered radiation energy is relatively weak in longer wavelength (such as near infrared/short wave infrared wave band), so that the shadow pixel is darker than the surrounding environment, the reflectivity data of the pixel is lower, and the estimated LAI value is lower. Aiming at the LAI estimation error problem caused by the cloud pollution, a certain cloud mask file is needed to carry out mask processing on an original Sentinel-2 reflectivity file.
The European air office official provides Sen2Cor software to carry out atmosphere correction on Sentinel-2 images, and provides a scene classification chart and a cloud probability label, and the cloud detection algorithm can detect pixels polluted by thick clouds and coiled clouds on the images to a certain extent, but has relatively weak detection capability on thin clouds and cloud shadow areas.
The Sentinel Hub research team developed a single scene pixel based cloud detector. The principle of operation of this cloud detector is a machine learning method, with training and validation samples from the Hollstein et al hand-marked data set of about 640 ten thousand hand-marked pixels from 108 global and annual Sentinel-2 scenes, which pixels belong to one of the following six classes: clear sky, clouds, shadows, snow, curly clouds or bodies of water (terrestrial). The cloud detection algorithm based on the single scene does not depend on the neighborhood of the pixel cloud probability calculation, but allocates a probability of being covered by cloud to each image pixel according to the spectral response of satellites to the pixel, so that cloud mask data adopted by the cloud detection algorithm based on the single scene is a cloud probability file, the value distribution of each pixel is 0-100,0 indicates that the current pixel is not polluted by cloud, and 100 indicates that the current pixel is polluted by thick cloud. In the range of the research area, the probability of the pixel cloud polluted by thick clouds is approximately 80-100%, the probability of the pixel cloud polluted by thin clouds is distributed at 30-70%, and the probability of the pixel cloud in the cloud image area is approximately distributed at 10-30%. Therefore, the cloud probability of 10 is selected as the threshold for detecting whether the pixel is polluted by the cloud, the pixel reflectivity with the cloud pollution probability less than or equal to 10 is used for LAI inversion, and the pixel with the cloud probability more than 10 does not participate in LAI inversion.
(2) And extracting vegetation growing season based on the reconstructed NDVI data. This step is also done to reject outliers of partial LAI estimates. In a certain area, under the influence of altitude, snow and ice are often covered in non-growing seasons, the reflectivity of each wave band of the pixels covered by the snow is at a high value, and a least square-based lookup table algorithm is adopted when the LAI is inverted by using a unified model, so that false high values of the LAI in the non-growing seasons often occur. In order to solve the problem of false high value in non-growing season, the invention firstly utilizes a time sequence reconstruction algorithm to reconstruct an original NDVI time sequence to obtain day-to-day NDVI data; then combine to moveState thresholding (Groten et al, 2002;et al 2004) extract the growth season range (vegetation turning period and yellow period) pixel by pixel. The dynamic threshold method is simple and easy to use, is widely applied to vegetation weather extraction research based on remote sensing data, the estimated principle is shown as formula (4), and the method selects 40% of the variation range of NDVI as the dynamic threshold of the extracted growing season by combining the previous research (such as everlasting, 2014, hou Meiting, 2012, song Chunqiao, 2012) so as to obtain the vegetation green-turning period of a certain area approximately in the middle and late 4 months.
NDVI(t)＝(NDVI max -NDVI min )×40％ (4)
In the formula, NDVI max Is the maximum value of vegetation NDVI in one year; NDVI min Is the minimum of the NDVI ramp-up phase (1-183 days); t is the vegetation growth season beginning period (turning green period) in one year. The same method determines the end of vegetation growing season (dry period) in the NDVI decrease phase (184-365 days).
After the growth season range of a certain region pixel by pixel is determined, the LAI value of the image in the non-growth season is set to be 0.1, and the LAI inversion is carried out on the image in the growth season by using a BRDF unified model.
(3) And estimating the LAI value in the growing season range based on an algorithm for inverting the LAI by the BRDF unified model. The method for inverting the LAI based on the vegetation BRDF unified model (Xu et al, 2017) is to construct a lookup Table of pixel reflectivities and LAIs of three bands of green, red and near infrared by using the vegetation BRDF unified model, and then to search for the pixel-by-pixel LAIs based on the lookup Table (Look-up Table, LUT) (Ma et al, 2018). Based on a vegetation canopy BRDF unified model, performing LAI inversion by using an LUT method is divided into two parts: 1) A lookup table is established. And according to the BRDF unified model, inputting parameters required by the model, obtaining the reflectivity corresponding to the incident zenith angle and azimuth angle of the sun and the sensor which are time-synchronized with the acquisition of satellite images under different LAIs of different vegetation, and obtaining a lookup table for LAI inversion. 2) Inversion is based on a look-up table. And acquiring the pixel reflectivity from the satellite image, and obtaining the LAI corresponding to the minimum reflectivity phase difference by using a least square method. A specific flow for inverting LAI based on Sentinel-2 reflectivity data is shown in fig. 3.
(4) And reconstructing the LAI time sequence day by day all the year by day by adopting a time sequence reconstruction algorithm. In a certain area, the time resolution of the Sentinel-2 data is 2-3 days due to the overlapping of a plurality of strips of satellite orbits, but clouds and precipitation increase in a growing season (7-8 months), and the problem of data loss caused by cloud pollution is serious, so that the LAI data which is acquired in the third step and has only certain discrete dates in one year is required to be reconstructed by adopting a time sequence reconstruction algorithm.
The smooth method proposed by Garcia (2010) was used by Xiao et al (2015) to reconstruct reflectivity data of remote sensing satellite images. The Garcia smoothing method is a punishment least squares regression method (Penalized Least Square Regression Based on Three-Dimensional Discrete Cosine Transform, DCT-PLS) based on discrete cosine transform, and the invention adopts the same method to reconstruct the LAI daily time sequence.
Step 3: determining model parameters in combination with the stationary meteorological satellite data: proportion of sky scattered light from hour to hour.
Sky scattered light ratio describes the proportion of scattered radiation in atmospheric downstream radiation to total downstream radiation. Since the earth's surface has a dichroic reflection characteristic for light incident in different directions, this characteristic results in the earth's surface having different reflectivities for direct radiation incident in a single direction and scattered radiation incident in hemispherical space. Therefore, when estimating the vegetation blue sky albedo, the sky scattered light proportion needs to be used as input data, and the reflection characteristics of direct radiation and scattered radiation in the incident radiation are respectively considered.
In general, sky diffuse light ratio is affected differently by sun geometry, atmospheric conditions, weather conditions, etc. Under the clear sky condition, the sky cloud quantity is very little, the atmospheric condition is relatively stable, and the sky scattered light proportion is only influenced by the sun geometry condition; however, under the cloud sky condition, the sky scattered light proportion is greatly changed and unpredictable due to the influence of cloud dynamic change, so that the invention introduces 5 km-resolution downlink direct radiation and downlink total radiation data (Letu et al, 2020,2021) estimated based on sunflower No. 8 static meteorological satellite data, estimates the sky scattered light proportion from hour to hour, and finally obtains the blue sky albedo of vegetation.
Step 4: and carrying the determined model parameters into a model, and estimating the albedo of the vegetation blue sky.
The basic formula for estimating the vegetation blue sky albedo is shown as a formula (1), the calculation formula is improved to a certain extent on the basis of the existing research (Peng et al, 2015), and the albedo estimation model is expanded to be suitable for albedo estimation under different vegetation types and different weather conditions.
Where A (λ) is the albedo of the blue sky of the vegetation, av (λ) is the absorption of the vegetation part, and as (λ) is the absorption of the soil part.
The names of the other variables are as shown in Table 2.
TABLE 2 model related physical quantity
/>
Accuracy verification of the invention
According to the invention, the measured data of 4 ground observation sites in a certain area are selected as the accuracy verification data of an A-P algorithm. The specific information of the four sites is shown in table 1. Since the time resolution of four site observations is 1 min/30 min, and the study was aimed at validating the accuracy of hourly vegetation albedo data, it was necessary to integrate the ground measurement data into an hourly time resolution. And (3) selecting downlink radiation and uplink radiation data within a half hour range before and after the whole point to obtain an average value (for example, the average value of 9:30-10:30 represents 10-point radiation data), and finally obtaining the local time 8:00-17: the 00 hour ground measurement albedo data is compared with the vegetation albedo data estimated by the "A-P" algorithm.
Table 1 ground verification site information for a region
Fig. 4 is a graph comparing the albedo of vegetation blue sky with ground measured data on an hour-by-hour basis in a growing season of four ground stations based on the estimation of "a-P", fig. 5 is a result of accuracy verification based on ground measured data, and overall, the model estimation value and the ground measurement value have a higher correlation, especially for two stations of pinus sylvestris and larch, and the points on the scatter diagram are basically concentrated in 1: near line 1, the degree of dispersion is low, a small overestimation exists at the grassland site, and a certain estimation error exists at the highest value and the lowest value at the corn site.
From the specific accuracy statistics results, the correlation coefficient of the pinus sylvestris and larch stations is 0.55, and the correlation coefficient of the farmland stations is 0.72, which shows that the estimated hourly vegetation albedo can better reflect the change rule of the vegetation albedo along with time; in terms of model estimation errors, the MAE of the four stations is within 0.03, wherein the MAE of the pinus sylvestris and larch is about 0.01, the RMSE of the grasslands and corn stations is about 0.03, and the RMSE of the pinus sylvestris and larch is less than 0.015. This estimation accuracy is superior to most current algorithms for inverting earth albedo based on medium-high resolution satellite data (Li et al, 2018; shinai et al, 2014).
Claims (9)
1. A vegetation blue sky albedo estimation method based on spectrum invariant theory, the method comprising:
acquiring Sentinel-2 data and static meteorological satellite data of a target area;
extracting clear empty Sentinel-2 data in the Sentinel-2 data, and determining single scattering albedo of the blade and the soil background reflectivity based on the clear empty Sentinel-2 data;
calculating vegetation growth rhythms of the target area, and determining a daily vegetation leaf area index time sequence by combining the Sentinel-2 data;
determining an hour-by-hour sky scattered light proportion based on the stationary meteorological satellite data;
acquiring vegetation blue sky albedo of the target area based on the blade single scattering albedo, the soil background reflectivity, the time series of the daily vegetation leaf area index and the hourly sky scattered light proportion; the obtaining the vegetation blue sky albedo of the target area based on the blade single scattering albedo, the soil background reflectivity, the daily vegetation leaf area index time sequence and the hourly sky scattered light proportion comprises the following steps:
setting the probability P that photons collide with the blades again after being scattered by the blades, and calculating the probability a that photons entering in a single direction are absorbed by vegetation after being collided with vegetation elements for one or more times by combining the single scattering albedo of the blades c_dir Probability a that photons incident in hemispherical directions are absorbed by vegetation after one or more collisions with vegetation elements c_diff ；
Setting a function G describing canopy leaf inclination angle distribution and an aggregation index xi, and calculating interception probability i of canopy on unidirectional incident photons on each day by combining the time sequence of the daily vegetation leaf area index 0 ；
Calculating the interception probability i 0 Integrating in hemispherical space to obtain the interception probability of incident photons in hemispherical direction by the canopy
Setting probability q of scattered photons falling out of the canopy after collision with the blade d Calculating vegetation scattering reflectivity of photons uniformly incident in all directions by combining the single scattering albedo of the blade and the probability P
Based on the soil background reflectivity, the vegetation scattering reflectivitySaid interception probability->Respectively calculating the probability P that photons are not absorbed after being subjected to any collision between soil and vegetation sr Probability of absorption P with photons after undergoing any number of collisions between soil and vegetation sa ；
According to the proportion of scattered light of the hourly sky, the interception probability i 0 Said probability a c_dir Said probability a c_diff The vegetation scattering reflectivitySaid probability P sr Obtaining the absorption of vegetation parts under different wavelengths;
according to the soil background reflectivity, the hourly sky scattered light proportion and the interception probability i 0 The vegetation scattering reflectivitySaid probability P sa Obtaining the absorption of soil parts under different wavelengths;
and acquiring the vegetation blue sky albedo according to the absorption of the vegetation part under the different wavelengths and the absorption of the soil part under the different wavelengths.
2. The method of claim 1, wherein said extracting clear empty Sentinel-2 data from said Sentinel-2 data comprises:
acquiring Sentinel-2 data with cloud coverage smaller than a set value from the Sentinel-2 data;
preprocessing the Sentinel-2 data smaller than a set value to obtain clear sky Sentinel-2 data, wherein the preprocessing comprises the following steps: radiometric calibration, geometric correction and atmospheric correction of images.
3. The method of claim 1, wherein the determining a leaf single scatter albedo based on clear sky Sentinel-2 data in the Sentinel-2 data comprises:
aiming at the change of the single scattering albedo of the blade of the sunny Sentinel-2 data in the visible light wave band, a lookup table with chlorophyll values as lookup conditions is established based on a PROSAIL model so as to determine the single scattering albedo of the blade in the visible light wave band;
based on the clear sky Sentinel-2 data, establishing a lookup table from infrared soil reflectivity, vegetation coverage and pure vegetation canopy near infrared reflectivity index to the blade single scattering albedo so as to determine the blade single scattering albedo of a near infrared band;
obtaining the single scattering albedo of the blade in the short wave infrared band according to the equivalent water thickness of the blade and the single scattering albedo of the blade in the near infrared band;
and synthesizing the single scattering albedo of the blade in the visible light band, the near infrared band and the short wave infrared band to obtain the single scattering albedo of the blade.
4. The method of claim 3, wherein the building a lookup table from infrared soil reflectivity, vegetation coverage, pure vegetation canopy near infrared reflectivity index to the blade single scatter albedo based on the clear sky Sentinel-2 data comprises:
estimating vegetation contribution degree of a vegetation covered part in a pixel based on the data of the clear sky Sentinel-2, wherein the pixel comprises the vegetation covered part and a soil part;
acquiring vegetation coverage in the pixels;
calculating a near infrared reflectivity index of a pure vegetation canopy according to the vegetation coverage and the vegetation contribution degree;
when the correlation between the pure vegetation canopy near infrared reflectivity index and the blade single scattering albedo is established by means of a PROSAIL model, the distribution and the fixed value of parameters are simulated to determine the influence of each parameter on the blade single scattering albedo, so that a lookup table from the infrared soil reflectivity, vegetation coverage and pure vegetation canopy near infrared reflectivity index to the blade single scattering albedo is obtained.
5. The method of claim 1, wherein the calculating a vegetation growth rhythm of the target area comprises:
obtaining the maximum value NDVI of the original normalized difference vegetation index in one year max Minimum NDVI of the initial normalized difference vegetation index at the up-take stage min ；
Extracting a dynamic threshold tau in a growing season based on the variation amplitude of the original normalized difference vegetation index;
according to the maximum value NDVI max Said minimum value NDVI min Calculating a turning green period t according to the dynamic threshold tau;
calculating the withered and yellow period t ′ ；
According to the turning green period t and the withered and yellow period t ′ Obtaining the vegetation growth rhythm.
6. The method of claim 1, wherein the determining a time series of day-to-day vegetation leaf area indices in combination with the Sentinel-2 data comprises:
carrying out outlier rejection on the Sentinel-2 data by utilizing cloud mask data;
acquiring an original normalized difference vegetation index time sequence based on the Sentinel-2 data with the outlier removed, and reconstructing the original normalized difference vegetation index time sequence by combining the vegetation growth rhythm to acquire a normalized difference vegetation index time sequence in a growing season and a normalized difference vegetation index time sequence in a non-growing season;
setting a daily vegetation leaf area index according to the normalized difference vegetation index time sequence of the non-growing season;
aiming at the normalized difference vegetation index time sequence of the growing season, inversion of the daily vegetation leaf area index is carried out by utilizing a LUT method based on a BRDF unified model;
comprehensively growing Ji Yufei to obtain a discrete time sequence of the daily vegetation leaf area index;
and reconstructing the discrete daily vegetation leaf area index time sequence by adopting a time sequence reconstruction algorithm to obtain a daily LAI time sequence.
7. The method of claim 1, wherein the determining an hour-by-hour sky-scattered-light ratio based on the stationary meteorological satellite data comprises:
estimating the downstream direct radiation and the downstream total radiation data of M kilometer resolution based on the stationary meteorological satellite data;
and estimating the sky scattered light proportion of each hour according to the downlink direct radiation and the downlink total radiation data so as to obtain the sky scattered light proportion of each hour.
8. A storage medium having a computer program stored therein, wherein the computer program is arranged to perform the method of any of claims 1-7 when run.
9. An electronic device comprising a memory having a computer program stored therein and a processor arranged to run the computer program to perform the method of any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210770366.1A CN115292616B (en) | 2022-06-30 | 2022-06-30 | Vegetation blue sky albedo estimation method and device based on spectrum invariant theory |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210770366.1A CN115292616B (en) | 2022-06-30 | 2022-06-30 | Vegetation blue sky albedo estimation method and device based on spectrum invariant theory |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115292616A CN115292616A (en) | 2022-11-04 |
| CN115292616B true CN115292616B (en) | 2023-07-18 |
Family
ID=83823257
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210770366.1A Active CN115292616B (en) | 2022-06-30 | 2022-06-30 | Vegetation blue sky albedo estimation method and device based on spectrum invariant theory |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115292616B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116305875B (en) * | 2023-02-17 | 2023-08-29 | 中国科学院地理科学与资源研究所 | Data processing method and device for leaf area index in numerical weather forecast mode |
| CN116308866B (en) * | 2023-05-23 | 2023-07-28 | 武汉大学 | Rice ear biomass estimation method and system based on canopy reflection spectrum |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020165671A1 (en) * | 2019-02-11 | 2020-08-20 | Università Degli Studi Di Palermo | Method for monitoring vegetation ground covers |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105806784B (en) * | 2016-03-17 | 2018-08-14 | 北京大学 | A kind of vegetation two contributes the acquisition methods of item to sexual reflex Multiple Scattering |
| KR20180129443A (en) * | 2017-05-26 | 2018-12-05 | 이화여자대학교 산학협력단 | Apparatus and method for determining snow albedo by reflecting structural features of vegetation after end of growth |
| CN108896514B (en) * | 2018-06-05 | 2020-11-06 | 桂林航天工业学院 | Method for quantitatively describing simulation influence of blade mirror reflection on canopy reflectivity |
| CN110472525B (en) * | 2019-07-26 | 2021-05-07 | 浙江工业大学 | Noise detection method for time series remote sensing vegetation index |
| CN114663786B (en) * | 2022-03-21 | 2024-03-22 | 南京林业大学 | Stand radiation flux calculating method based on point cloud data and computer graphics |
-
2022
- 2022-06-30 CN CN202210770366.1A patent/CN115292616B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020165671A1 (en) * | 2019-02-11 | 2020-08-20 | Università Degli Studi Di Palermo | Method for monitoring vegetation ground covers |
Non-Patent Citations (1)
| Title |
|---|
| 几种典型地表类型反照率时序变化特征及其参数化研究;薛华柱;张国东;周红敏;王锦地;万华伟;;北京师范大学学报(自然科学版);第55卷(第02期);第272-283页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115292616A (en) | 2022-11-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Köhler et al. | Assessing the potential of sun-induced fluorescence and the canopy scattering coefficient to track large-scale vegetation dynamics in Amazon forests | |
| CN115292616B (en) | Vegetation blue sky albedo estimation method and device based on spectrum invariant theory | |
| Lyu et al. | A new method for grassland degradation monitoring by vegetation species composition using hyperspectral remote sensing | |
| Delegido et al. | Estimating chlorophyll content of crops from hyperspectral data using a normalized area over reflectance curve (NAOC) | |
| Chen et al. | Multiple-scattering scheme useful for geometric optical modeling | |
| Clevers | Imaging spectrometry in agriculture-plant vitality and yield indicators | |
| CN106501186B (en) | A kind of soil moisture content product NO emissions reduction method | |
| Zhao et al. | Effect of vertical distribution of crop structure and biochemical parameters of winter wheat on canopy reflectance characteristics and spectral indices | |
| Dong et al. | A modified version of the kernel-driven model for correcting the diffuse light of ground multi-angular measurements | |
| Abuelgasim et al. | Modeling bidirectional radiance measurements collected by the advanced solid-state array spectroradiometer (ASAS) over Oregon transect conifer forests | |
| Hanan et al. | A method to estimate instantaneous and daily intercepted photosynthetically active radiation using a hemispherical sensor | |
| Hu et al. | Surface albedos and angle-corrected NDVI from AVHRR observations of South America | |
| Nilson et al. | Determination of needle area indices of coniferous forest canopies in the NOPEX region by ground-based optical measurements and satellite images | |
| Xue et al. | One-dimensional variational retrieval of temperature and humidity profiles from the FY4A GIIRS | |
| Gupta et al. | Top-of-the-atmosphere shortwave flux estimation from satellite observations: an empirical neural network approach applied with data from the A-train constellation | |
| He et al. | Direct estimation of land surface albedo from simultaneous MISR data | |
| Meng et al. | Hot dark spot index method based on multi-angular remote sensing for leaf area index retrieval | |
| Cui et al. | An empirical anisotropy correction model for estimating land surface albedo for radiation budget studies | |
| Cierniewski et al. | Annual dynamics of shortwave radiation as consequence of smoothing of previously plowed and harrowed soils in Poland | |
| Rengarajan et al. | Modeling and simulation of deciduous forest canopy and its anisotropic reflectance properties using the Digital Image and Remote Sensing Image Generation (DIRSIG) tool | |
| Chen | Remote sensing of leaf area index of vegetation covers | |
| Armstrong et al. | Spatial variability of mean daily estimates of actual evaporation from remotely sensed imagery and surface reference data | |
| Leblon et al. | A semi‐empirical model to estimate the biomass production of forest canopies from spectral variables Part 1: Relationship between spectral variables and light interception efficiency | |
| Lind et al. | The spatio-temporal relationship between rainfall and vegetation development in Burkina Faso | |
| Hu et al. | BOST: A canopy reflectance model suitable for both continuous and discontinuous canopies over sloping terrains |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |