CN106872466B - A kind of lake ice freeze thawing monitoring method and system based on dynamic Decomposition of Mixed Pixels method - Google Patents
A kind of lake ice freeze thawing monitoring method and system based on dynamic Decomposition of Mixed Pixels method Download PDFInfo
- Publication number
- CN106872466B CN106872466B CN201611266615.4A CN201611266615A CN106872466B CN 106872466 B CN106872466 B CN 106872466B CN 201611266615 A CN201611266615 A CN 201611266615A CN 106872466 B CN106872466 B CN 106872466B
- Authority
- CN
- China
- Prior art keywords
- lake
- pixel
- bright temperature
- temperature value
- mixed
- 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
- 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
- G01N21/84—Systems specially adapted for particular applications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30181—Earth observation
- G06T2207/30184—Infrastructure
Landscapes
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Geophysics And Detection Of Objects (AREA)
- Radiation Pyrometers (AREA)
- Radar Systems Or Details Thereof (AREA)
Abstract
The present invention provides a kind of lake ice freeze thawing monitoring method and system based on dynamic Decomposition of Mixed Pixels method.The method includes S1, are based on lake particular optical image and passive microwave remote sensing data, obtain the bright temperature value of lake mixed pixel and the bright temperature value of loke shore Pure pixel；S2 is based on lake particular optical image and the bright temperature value of lake mixed pixel, the area ratio in lake and loke shore in the lake mixed pixel is obtained by Overlap Analysis；S3, based on lake and loke shore area ratio in the bright temperature value of the lake mixed pixel, the bright temperature value of the loke shore Pure pixel and the lake mixed pixel, the bright temperature value of lake Pure pixel of the lake mixed pixel is obtained, using linear spectral unmixing method to monitor lake ice freeze thawing phenomenon.The present invention monitors the sub-pixed mapping lake freeze thawing that passive microwave remote sensing data is applied to high Asia, obtains the bright temperature information in lake in sub-pixed mapping, realizes high Asia passive microwave remote sensing sub-pixed mapping lake ice freeze thawing monitoring.
Description
Technical field
The present invention relates to lake ice freeze thawing monitoring technical fields, are based on dynamic Decomposition of Mixed Pixels more particularly, to one kind
The lake ice freeze thawing monitoring method and system of method.
Background technique
Based on Qinghai-Tibet Platean, by the mountain ranges such as the Himalayas, Kun Lun Mountain, Hengduanshan Mountains In China, the Qilian mountains and Tianshan Mountains and plateau
The high Asia region of area composition, it is global High aititude that for altitude ranges 2000 between 8844m, mean sea level, which is about 4046m,
Lake area the most intensive, domestic lake spread all over the place, and area is greater than 1km2, and there are about 1210.
Lake ice parameter is one of important sensitive factor of whole world change, and the weather and environment of high Asia are to the region earth
System model has important influence, is the hot spot region of the most sensitive region of Global climate change and current research, high
Mountain lake is very sensitive for climate change, especially the time of high mountain lake freeze thawing and duration, is often expected to use for recording
In announcement regional climate and environmental change feature.
This area's both high and cold physical geographic environment feature, causes the area uninhabited, lake monitoring station lacks, and makes
It obtains lake freeze thawing parameter to be difficult to obtain from ground monitoring, especially period of history lake ice does not have to observe data substantially, therefore, usually
Need to carry out the acquisition of freeze thawing state by history remote sensing observations data and existing observation data.Wherein it is usually used in monitoring lake jelly
The optical remote sensing data melted mainly have AVHRR (Advanced Very High Resolution Radiometer), MODIS
(Moderate Resolution Imaging Spectroradiometer) etc., common passive microwave remote sensing data will have
SSMI/S (Special Sensor Microwave Imager/Sounder), AMSR-E (Advanced Microwave
Scanning Radiometer-Earth Observing System)、FY3MWRI(Microwave Radiation
) and AMSR2 (The Advanced Microwave Scanning Radiometer 2) etc. Imager.
High Asia atmosphere convection is active, often has cloud cover, affects when optical sensor passes by and effectively obtain lake
Information is melted in frost, so that there are the serious problems of shortage of data for optical image.Such as Kropacek (2013) utilizes MODIS optics
When 59 lake freezing-thawing conditions of product surveillance Qinghai-Tibet Platean, for the considerations of polluting to cloud, use is selected to close through maximum accumulated snow
At eight days Snow Products of MYD10A2 monitor lake freeze thawing, but select the data monitoring lake freeze thawing situation there is at least 8
Its above even 16 days monitoring error.
And microwave radiance transfer remotely-sensed data have round-the-clock observing capacity, influenced by sexual intercourse weather conditions it is less, over the ground
It is strong to observe space-time expending, and has the detection of ice water phase transformation compared with high sensitivity, particularly suitable for ice and snow freeze thawing study on monitoring,
It is widely used in the monitoring of land table permafrost change, snow-cover variation monitoring and sea ice variation monitoring.But since passive microwave radiates
Resolution of scoring is coarse, and mixed pixel effect is serious, it is caused to be greatly limited in lake freeze thawing variation monitoring, and does not have
A large amount of lake freeze thawing situations can be monitored as optical data by having.
At present high Asia region using the lake of passive microwave remote sensing data study on monitoring be concentrated mainly on Qinghai Lake with
And the large-size lakes such as Nam Co, or even also only have 35 large-size lakes in the lake that entire Northern Hemisphere region is properly used for monitoring.
And high Asia region lake is densely distributed, clearly and to understand area's weather and environmental change band in depth from the angle of lake freeze thawing
The influence and response come, it is to be understood that the freeze thawing situation in the more lakes in the area.Although passive microwave remote sensing data mixed pixel is imitated
Should be obvious, but in optical remote sensing and ground observation the lake freeze thawing in available high Asia region related data, it is passive to make
Microwave remote sensing data become a kind of valuable material of monitoring lake freeze thawing.
Generally for more finer earth's surface monitoring feature information are obtained from the passive microwave image of coarse resolution, often
There are following several solutions, such as pixel analysis technology, remote sensing image NO emissions reduction technology and the increasing for combining multi- source Remote Sensing Data data
Strong resolution technique.Area of lake is often relatively small compared with the pixel resolution of passive microwave remote sensing data；Lake region
Information is recorded only in the even single pixel of limited several pixels (when more than one microwave pixel of area of lake wants small).And
Satellite sensor and earth's surface lake position are in a kind of state of relative motion, and sensor revisits capture record lake region information
Pixel center point coordinate always change, the area of lake for causing single pixel to be captured also changes therewith, leads to tradition
The method of the resolution ratio enhancing such as passive microwave be difficult to be applicable in.
Aggregate analysis shows that in global change research due, high Asia is especially Qinghai-Tibet, and number is melted in river and lake frost
It is lacked according to comparing；China only has the lake ice routine observation data of Qinghai Lake and less Lake Namco in recent years at present, therefore goes through
It is very important for the acquisition of the lake ice freeze thawing data in history period.Microwave radiance transfer remotely-sensed data can be used for recalling history
The observation of period lake ice freeze thawing, however the problem is that, current a variety of methods may be implemented to carry out Big Lakess and monitor, but
It is in the case where small lakes, signal often occurs smudgy, differentiates the problems such as improper.Therefore need to fully consider data
The exploitation of availability methodology, and the problem of break through its intrinsic spatial resolution, to provide lake ice freeze thawing for more lakes
Parameter information.
Summary of the invention
The present invention provide it is a kind of overcome the above problem or at least be partially solved the above problem based on dynamic mix picture
The lake ice freeze thawing monitoring method and system of first decomposition method.
According to an aspect of the present invention, a kind of lake ice freeze thawing monitoring side based on dynamic Decomposition of Mixed Pixels method is provided
Method, comprising:
S1 is based on lake particular optical image and passive microwave remote sensing data, obtains the bright temperature value of lake mixed pixel and lake
The bright temperature value of bank Pure pixel；
S2, is based on lake particular optical image and the bright temperature value of lake mixed pixel, obtained by Overlap Analysis described in
The area ratio in lake and loke shore in the mixed pixel of lake；
S3 is based on the bright temperature value of the lake mixed pixel, the bright temperature value of the loke shore Pure pixel and the lake mixing picture
Lake and loke shore area ratio, the pure picture in lake of the lake mixed pixel is obtained using linear spectral unmixing method in member
The bright temperature value of member, to monitor lake ice freeze thawing phenomenon.
According to another aspect of the present invention, a kind of lake ice freeze thawing monitoring based on dynamic Decomposition of Mixed Pixels method is also provided
System, comprising:
Preliminary bright temperature data obtains module, for being based on lake particular optical image and passive microwave remote sensing data, obtains
The bright temperature value of lake mixed pixel and the bright temperature value of loke shore Pure pixel；
Lake and land percentage module, for being based on lake particular optical image and the bright temperature of lake mixed pixel
Value, the area ratio in lake and loke shore in the lake mixed pixel is obtained by Overlap Analysis；
Dynamic is accurate to obtain module, for being based on the bright temperature value of the lake mixed pixel, the bright temperature of loke shore Pure pixel
It is mixed to obtain the lake using linear spectral unmixing method for lake and loke shore area ratio in value and the lake mixed pixel
The bright temperature value of lake Pure pixel of pixel is closed, to monitor lake ice freeze thawing phenomenon.
A kind of lake ice freeze thawing monitoring method and system based on dynamic Decomposition of Mixed Pixels method that the application proposes, in conjunction with lake
Particular optical image and passive microwave remote sensing data are moored, the ratio value in lake and loke shore in passive microwave pixel is dynamically obtained,
Microwave radiance transfer dynamic mixed pixel linear decomposition is carried out, to obtain the bright temperature value of lake Pure pixel；The lake is pure
The bright temperature value of pixel can be used for monitoring lake ice freeze thawing phenomenon.
Herein described method, which can be realized, to be applied to commonly used in the passive microwave remote sensing data of large scale earth's surface monitoring
The sub-pixed mapping grade lake freeze thawing of high Asia monitors, and obtains the bright temperature information in lake in the sub-pixed mapping, it can be achieved that high Asia
Area's passive microwave remote sensing sub-pixed mapping lake ice freeze thawing monitoring.
Detailed description of the invention
Fig. 1 is a kind of lake ice freeze thawing monitoring method flow chart based on dynamic Decomposition of Mixed Pixels method of the present invention；
Fig. 2 is rectangle dynamic buffering schematic diagram of the present invention；
Fig. 3 is that a kind of lake ice freeze thawing based on dynamic Decomposition of Mixed Pixels method of the present invention monitors system schematic；
Fig. 4 a is the bright temperature schematic diagram in Qinghai Lake long-term sequence of the present invention lake；
Fig. 4 b is the bright temperature schematic diagram in Hoh Xil Lake long-term sequence of the present invention lake；
Fig. 4 c is that the present invention reaches the then bright temperature schematic diagram in wrong long-term sequence lake；
Fig. 4 d is the bright temperature schematic diagram in the lake Sai Ku long-term sequence lake of the present invention；
Fig. 4 e is the bright temperature schematic diagram in Ba Mu mistake long-term sequence lake of the present invention；
Fig. 4 f is visitd for the present invention and is invited the bright temperature schematic diagram in cloth mistake long-term sequence lake；
Fig. 4 g is the bright temperature schematic diagram in hole mistake long-term sequence lake of the present invention；
Fig. 5 a is the present invention to two kinds of bright temperature contrast schematic diagrams in information enhancement method lake up to then mistake；
Fig. 5 b is the present invention to visiing two kinds of bright temperature contrast schematic diagrams in information enhancement method lake for inviting cloth mistake.
Specific embodiment
With reference to the accompanying drawings and examples, specific embodiments of the present invention will be described in further detail.Implement below
Example is not intended to limit the scope of the invention for illustrating the present invention.
The region that the present invention is embodied is high Asia, and the high Asia refers to that Central Asia is with Qinghai-Tibet Platean
The High aititude region at center.The present invention chooses apparent six lakes of high Asia area discrepancy as implementation lake ice jelly
Melt the object of monitoring, they be respectively Hoh Xil Lake, up to then mistake, Sai Kuhu, Ba Mu are wrong, visit and invite that cloth is wrong and hole is wrong.It is spaceborne
The average area of lake that passive microwave sensors A MSR-E pixel captures this six lakes accounts for pixel area ratio from big to small
Respectively 0.57,0.55,0.41,0.37,0.28 and 0.22, such as table 1.
Table 1
The present invention is when being embodied there are four types of used data, comprising: AMSR-E passive microwave remote sensing data,
Landsat TM image data enhances the AMSR-E data and cloudless MODIS data day by day of resolution ratio.Described in wherein selecting
AMSR-E passive microwave remote sensing data extracts lake and the bright temperature value in loke shore region；The Landsat TM image data is selected to calculate
The changeable weight in lake and lake；Select it is described enhancing resolution ratio AMSR-E data extract the bright temperature value in lake, and with mix picture
The bright temperature value that first decomposition method is extracted carries out effect comparison；And select the cloudless MODIS data day by day as extracting lake
The verify data of freeze thawing parameter.
The AMSR-E passive microwave remote sensing data is obtained by microwave radiance transfer spoke meter observation data observation.It is described passive
Microwave spoke meter has 12 imager channels of horizontal polarization and vertical polarization, and frequency is respectively 6.9GHz, 10.65GHz,
18.7GHz, 23.8GHz, 36.5GHz, 89.0GHz, due to ice and water, the emissivity under the channel 18.7GHz differs greatly (Che
Deng 2009；Tao etc., 2014), therefore the present invention is mainly based on the V polarization data of 18.7GHz, and wherein 18.7GHz's is original
The resolution ratio 27*16km of data.
The Landsat TM image data is obtained by Landsat-5 moonscope.The Landsat-5 satellite is in 1984
Successful launch carries TM (Thematic Mapper) sensor of 30*30m spatial resolution to operational failure in 2011.
Landsat-5 satellite is during operation persistently to surface observation, and it is comprehensive to obtain a large amount of earth's surface overlay areas, the history of high quality
Data, for resource management, ECOLOGICAL ENVIRONMENTAL MONITORING, geologic survey etc. provides data resource abundant.The present invention utilizes
Landsat-5 data and the AMSR-E pixel that passes by are laid out analysis, calculate loke shore area ratio in lake in AMSR-E pixel.
The AMSR-E data of the enhancing resolution ratio are by Brigham Young University, the U.S. (Brigham Young
University, BYU) microwave remote sensing earth team (Microwave Earth Remote Sensing group, MERS)
Using AMSR-E data combination scatterometer image according to image reconstruction algorithm (Long etc., 1998；Early etc., 2001) generate
Enhance the image of resolution ratio.The AMSR-E data of the enhancing resolution ratio are different in the spatial resolution of different frequency, wherein
The resolution ratio in the channel 6GHz, 10GHz, 18GHz is that the resolution ratio in the channel 12.5*12.5km, 23GHz and 36GHz is 7.5*
7.5km the resolution ratio in the channel 89GHz is 2.5*2.5km.The present invention selects the polarized enhancing of globular projection 18.7GHz V to differentiate
The AMSR-E data pick-up of rate is up to then mistake data compared with visit the bright temperature in lake for inviting cloth mistake as effect.
Day by day the cloudless MODIS data be Qinghai-Tibet Platean MODIS day by day cloudless Snow Cover Area data set (Qiu Yubao etc.,
2016) according to the product cloudless day by day of MOD10A1 and MYD10A1 exploitation in a text.The Qinghai-Tibet Platean MODIS cloudless product day by day
Snow face volume data concentrates the terrain classification that these four Regional Sustainable Development of The Qinghai tibet Plateau are not known including land, accumulated snow, lake water and lake ice lake.
Using the Qinghai-Tibet Platean MODIS, the lake water of cloudless Snow Cover Area data set and lake ice situation of change verify base of the present invention day by day
In dynamic Decomposition of Mixed Pixels method lake ice freeze thawing monitoring method lake freezing-thawing condition to monitoring result.
As shown in Figure 1, a kind of lake ice freeze thawing monitoring method based on dynamic Decomposition of Mixed Pixels method of the present invention, comprising:
S1 is based on lake particular optical image and passive microwave remote sensing data, obtains the bright temperature value of lake mixed pixel and lake
The bright temperature value of bank Pure pixel；
S2, is based on lake particular optical image and the bright temperature value of lake mixed pixel, obtained by Overlap Analysis described in
The area ratio in lake and loke shore in the mixed pixel of lake；
S3 is based on the bright temperature value of the lake mixed pixel, the bright temperature value of the loke shore Pure pixel and the lake mixing picture
Lake and loke shore area ratio, the pure picture in lake of the lake mixed pixel is obtained using linear spectral unmixing method in member
The bright temperature value of member, to monitor lake ice freeze thawing phenomenon.
Passive microwave remote sensing data described in the present embodiment is AMSR-E passive microwave remote sensing data；The particular optical image
For high-resolution optical image, specially Landsat TM image data.
The spatial resolution of passive microwave remote sensing data is low, and single pixel can capture several different types of topographical features
Information, mixed pixel effect are serious.Passive microwave sensor during obtaining earth's surface information also by it is some other because
Element influences, the calibration calibration including sensor, irrelevant signal caused by polarization effect etc. of energy radiation, and from it is outer
The influence of space energy radiation and earth atmosphere.
The present embodiment selects AMSR-E L2A bright temperature data, and the AMSR-E L2A bright temperature data provides via data flat
Platform has carried out the related basic handling such as radiation calibration and information smoothing.The land of the single pixel of the AMSR-E L2A bright temperature data
The bright temperature value of table, by the area weight integral representation of the true bright temperature value of each atural object for being distributed in the pixel, formula is as follows:
Formula (1) indicates the bright temperature TB ' (x that passive microwave sensor receives0,yo) it is by detecting center (x0,yo)
The bright temperature TB (x, y) of each atural object in pixel is sought by area weight integral,For the antenna gain of sensor.
The present invention is analyzed using the passive microwave remote sensing data, in conjunction with the lake particular optical image to described
Lake in the pixel of passive microwave remote sensing data carries out the positioning of position and coordinate, it is first determined the bright temperature value of lake mixed pixel
With the bright temperature value of loke shore Pure pixel；Lake and loke shore ratio in the lake mixed pixel are further determined that, to obtain lake
The bright temperature value of Pure pixel.The bright temperature value of lake Pure pixel is to monitor the most important data of lake ice freeze thawing phenomenon.
As an optional embodiment, the S1 further comprises:
S1.1 obtains the edge in lake using edge detection method from the lake particular optical image；Based on the lake
Particular optical image and the pixel position comparative analysis of the passive microwave remote sensing data and the edge in the lake are moored, institute is obtained
State the lake center of passive microwave remote sensing data；
S1.2, from the passive microwave remote sensing data, the nearest bright temperature of microwave pixel of lake center described in selected distance
Value is the bright temperature value of lake mixed pixel；From the passive microwave remote sensing data, lake center described in selected distance is nearest
Multiple loke shore Pure pixels the bright temperature value of microwave pixel, after taking mean value, be the bright temperature value of loke shore Pure pixel.
For passive microwave remote sensing data since image resolution ratio is too low, the present embodiment utilizes high-resolution Landsat
The position in corresponding lake in TM image data, is compared point with the position in the lake of the pixel of the passive microwave remote sensing data
Analysis, determines lake center in the pixel of the passive microwave remote sensing data.
The lake region earth's surface information detected when satellite passes by, it will usually it is recorded centrally in neighbouring several pixels,
And the latitude and longitude coordinates of microwave pixel centre coordinate point fine difference can all occur when revisiting every time；And passive microwave remote sensing number
According to resolution ratio it is low, therefore with high-resolution lake particular optical image be comparison, be compared analysis.Make described passive micro-
Lake and loke shore pixel in the pixel of wave remotely-sensed data and the lake in high-resolution lake particular optical image and loke shore
Image positional relationship be overlapped as far as possible unanimously；As long as having been determined in the lake in high-resolution lake particular optical image in this way
Heart point, so that it may the central point that lake in the pixel of the passive microwave remote sensing data is obtained according to corresponding positional relationship, into
And obtain the coordinate of lake center point, i.e., the lake center in the pixel of the described passive microwave remote sensing data.
After determining lake center, using one or more microwave pixels where the lake as the bright temperature of lake mixed pixel
It is worth, includes the bright temperature value in pure lake and the bright temperature value of pure loke shore in the bright temperature value of lake mixed pixel.
It chooses the pixel remote enough from lake center and obtains the bright temperature value of loke shore Pure pixel.Because of the limited area in lake,
As long as so choosing pixel remote enough is obtained with pure loke shore, to obtain the bright temperature value of loke shore Pure pixel.
As an optional embodiment, the S2 further comprises:
S2.1 is based on passive microwave remote sensing data, obtains lake mixed pixel；In the lake of the lake mixed pixel
The rectangular area of the spatial resolution of particular sensor specific frequency is generated centered on the heart；
S2.2 is laid out analysis based on the rectangular area and the lake particular optical image, decomposites the lake
Moor the ratio in lake and loke shore in mixed pixel.
Passive microwave remote sensing data selected by the present embodiment is observed obtained by data observation by microwave radiance transfer spoke meter, and
Satellite is in the state constantly moved with the earth in high air to surface scanning process, therefore during sensor revisits, it sees every time
Minor shifts can all occur for pixel position in the passive microwave remote sensing data measured, and the lake captured/loke shore area is ceaselessly
It changes, other than area is far smaller than the lake of pixel resolution.Therefore the observation data for needing repeatedly to pass by satellite
Comprehensive analysis is carried out, and is laid out analysis in conjunction with high-resolution Landsat TM image data, determines the lake mixing
The ratio in lake and loke shore in pixel.
Specifically, the bright temperature value of lake Pure pixel described in the S3 meets:
TBsensor=aTBlake+bTBland (2)
Wherein, TBsensorFor the bright temperature of lake mixed pixel of the single microwave pixel of the bright temperature value of lake mixed pixel
Value；
TBlakeFor the bright temperature value of lake Pure pixel, indicate that lake region is bright in the bright temperature value of lake mixed pixel
Temperature value；
TBlandFor the bright temperature value of loke shore Pure pixel, indicate that lake mixed pixel bright temperature value Mid-continent domain is bright
Temperature value；
A is the area weight of lake region in the bright temperature value of lake mixed pixel；
B is the area weight in lake mixed pixel bright temperature value Mid-continent domain；
Wherein, the ratio of a and b is the ratio in lake and loke shore in the lake mixed pixel.
The present invention is based on area of lake to carry out linear spectral unmixing, wherein assuming what passive microwave radiometer was captured
Land table information is homogeneous, and the earth's surface land information obtained is only related with the area of atural object, ignores the lesser factor of other influences.
The bright temperature of the bright temperature in lake region and land area in formula (2) all includes the influence of the factors such as atmosphere, and lake face
Product weight a and loke shore area weight b passes through high-resolution Landsat TM optical image and is calculated.
As an optional embodiment, the S1.2 further comprises: choosing 1 centered on the lake center
The bright temperature value of microwave pixel is the bright temperature value of lake mixed pixel；
The bright temperature value of microwave pixel of lake center described in selected distance nearest 2-3 pure loke shore pixels and do it is average after
For the bright temperature value of loke shore Pure pixel.
Specifically, the coordinate of the lake mixed pixel meets:
Wherein, (X, Y) is the coordinate of the lake center, (Xi,Yi) be the passive microwave remote sensing data any pixel
The coordinate of central point, (Xmin,Ymin) it is spaceborne bright temperature data coordinate corresponding to the bright temperature value of lake mixed pixel.
As an optional embodiment, the S2.1 further comprises:
S2.1.1 generates the first border circular areas and the second border circular areas respectively centered on the lake center；
S2.1.2 obtains the minimum outsourcing rectangle of first border circular areas and second border circular areas, is based on isogonism
Horizontal axis cyclotomy column projection coordinate system characteristic obtains the specific microwave frequency pixel resolution with the passive microwave remote sensing data
The rectangular area of the same size.
The size of first border circular areas and the size of second border circular areas are according to the passive microwave remote sensing number
According to a Pixel size difference and choose different sizes；Whether which kind of size, the present embodiment pass through first circle
What the isogonism horizontal axis cyclotomy column projection coordinate system characteristic of the minimum outsourcing rectangle of shape region and second border circular areas obtained
The rectangular area is equal with a Pixel size of the passive microwave remote sensing data.
Specifically, a spatial resolution value of the radius of first border circular areas for specific microwave frequency, described second
The radius of border circular areas is another spatial resolution value of specific microwave frequency.
The present embodiment uses passive microwave sensors A MSR-E pixel, and the pixel size of the AMSR-E pixel is 27*16
Kilometer, a spatial resolution value are 16 kilometers, and another spatial resolution value is 27 kilometers, therefore described first is round
The radius in region is 8 kilometers, and the radius of second border circular areas is 13.5 kilometers.
Since the area of lake that passive microwave sensors A MSR-E pixel is captured is not stop variation, in order to accurately find out
The weight of lake and loke shore in each pixel, the present embodiment are delayed centered on the AMSR-E pixel point closest from lake center
The rectangular area of a 27*16km is gone out, the lake lake that each satellite revisits pixel is then calculated by Overlap Analysis dynamic
Bank area weight.Finally the ratio of the lake found out and loke shore is substituted into formula (2), decomposites lake region in each pixel
Bright temperature value.
The present embodiment first buffers into two border circular areas using any in WGS84 projection coordinate system:
First border circular areas indicates are as follows: (Xi-a)2+(Yi-b)2=82
Second border circular areas indicates are as follows: (Xi-a)2+(Yi-b)2=13.52
The smallest outsourcing rectangle for seeking two circles again obtains then according to isogonism horizontal axis cyclotomy column projection coordinate system characteristic
27*16km rectangle IJKL is obtained, as shown in Figure 2.Finally calculate the rectangle IJKL that passes by daily of satellite, and with it is high-resolution
Landsat TM optical image Overlap Analysis, dynamic Decomposition go out the ratio in loke shore and lake.
As shown in figure 3, the present invention also provides a kind of, the lake ice freeze thawing based on dynamic Decomposition of Mixed Pixels method monitors system, packet
It includes:
Preliminary bright temperature data obtains module, for being based on lake particular optical image and passive microwave remote sensing data, obtains
The bright temperature value of lake mixed pixel and the bright temperature value of loke shore Pure pixel；
Lake and land percentage module, for being based on lake particular optical image and the bright temperature of lake mixed pixel
Value, the area ratio in lake and loke shore in the lake mixed pixel is obtained by Overlap Analysis；
Dynamic is accurate to obtain module, for being based on the bright temperature value of the lake mixed pixel, the bright temperature of loke shore Pure pixel
It is mixed to obtain the lake using linear spectral unmixing method for lake and loke shore area ratio in value and the lake mixed pixel
The bright temperature value of lake Pure pixel of pixel is closed, to monitor lake ice freeze thawing phenomenon.
The present invention a kind of lake ice freeze thawing monitoring method and system based on dynamic Decomposition of Mixed Pixels method is achieved beneficial
Effect can be obtained by following comparative analysis.
(1) Contrast on effect of mixed pixel dynamic Decomposition result
Please referring to Fig. 4 a- Fig. 4 g, Fig. 4 a is the bright temperature schematic diagram in Qinghai Lake long-term sequence of the present invention lake, and Fig. 4 b is this hair
The bright temperature schematic diagram in bright Hoh Xil Lake long-term sequence lake, Fig. 4 c are that the present invention illustrates up to the then bright temperature in wrong long-term sequence lake
Figure, Fig. 4 d are the bright temperature schematic diagram in the lake Sai Ku long-term sequence lake of the present invention, and Fig. 4 e is Ba Mu mistake long-term sequence lake of the present invention
Bright temperature schematic diagram is moored, Fig. 4 f is visitd for the present invention and invited the bright temperature schematic diagram in cloth mistake long-term sequence lake, and Fig. 4 g is that hole mistake of the present invention is long
The bright temperature schematic diagram in time series lake.
Using Decomposition of Mixed Pixels method to Hoh Xil Lake, up to then mistake, Sai Kuhu, Ba Mu are wrong, visit and invite that cloth is wrong and hole is wrong
The bright temperature in the lake region of (wherein Qinghai Lake is served only for comparing) carries out resolution process at six.It can be seen that utilization from Fig. 4 a- Fig. 4 g
The bright temperature curve of lake long-term sequence that mixed pixel linear dynamic decomposition method obtains temperature curve brighter than original mixed pixel
The real features of lake freeze thawing variation can more be protruded.Wherein effect the most significantly invites cloth wrong for visiing for Fig. 4 f, utilizes AMSR-E
Original bright temperature data cannot detect the bright temperature variation characteristic of the lake freeze thawing substantially, but handle by mixed pixel linear decomposition
Afterwards, the bright temperature curve in long-term sequence lake is similar with the bright temperature curvilinear motion feature in large-size lake Qinghai, and discomposing effect is good.
But from the point of view of the pixel analysis of the hole Fig. 4 g mistake treated bright temperature curve, effect is not it is obvious that having more
Noise, its lake freeze thawing variation characteristic effect trend is also not too much obvious after Fourier filtering is handled.In Fig. 4 a- Fig. 4 g
Can be seen that mixed pixel linear decomposition method when the lake object area detected is about AMSR-E pixel 0.3, discomposing effect
Already close to the limit.
(2) enhancing resolution ratio bright temperature data compares analysis
In order to further analyze the practicability of lake Decomposition of Mixed Pixels method, the MERS team agglomeration of U.S. BYU is quoted
Resolution ratio is that the AMSR-E 18.7GHz V polarization data of the enhancing resolution ratio of 12.5*12.5km compares and analyzes.
Selection, which reaches then mistake and visits, invites wrong two lakes of cloth example as a comparison, and Fig. 5 a is of the invention to two kinds of letters for reaching then mistake
Cease the bright temperature contrast schematic diagram in Enhancement Method lake；Fig. 5 b is the present invention to visiing two kinds of bright temperature in information enhancement method lake for inviting cloth mistake
Contrast schematic diagram.It can from the bright temperature curve in lake that is then wrong and visiing the AMSR-E data of enhancing resolution ratio for inviting wrong two lakes of cloth is reached
Find out, the AMSR-E data for enhancing resolution ratio highlight not apparent humidification for lake freeze thawing characteristic signal.From
It is seen up to the then bright temperature curve in wrong lake, although the AMSR-E data for enhancing resolution ratio can be from the bright temperature for generally distinguishing the lake reluctantly
Freeze thawing characteristic information, but its data noise is more；And from the point of view of smaller lake visits and invites cloth mistake, enhance the AMSR-E number of resolution ratio
Its lake freeze thawing characteristic signal is shown according to that cannot embody.It is supervised by comparing discovery Decomposition of Mixed Pixels method for lake freeze thawing
It surveys in application and is more applicable in than combining the method for other image reconstructions enhancing AMSR-E resolution ratio.
(3) Contrast on effect that Decomposition of Mixed Pixels freeze thawing differentiates
After carrying out Decomposition of Mixed Pixels processing to the bright temperature of lake region pixel, using TIMESAT (Time-series of
Satellite sensor data) software extraction lake freeze thawing parameter.TIMESAT software systems are initially to be applied to processing length
The NDVI data of time series, to monitor the cyclical growth variation of plant[25][26].Have in view of TIMESAT software systems fast
The ability of the long-term sequence data of speed processing seasonal variety and the bright temperature changed in conjunction with passive microwave data in lake freeze thawing
Lake bright temperature data is embedded into the software algorithm by periodic law, combines simple visually correct to mention using the software systems
Take lake freeze thawing parameter.
After the lake freeze thawing parameter extraction of passive microwave data comes out, with Qinghai-Tibet MODIS cloudless snow surface day by day
Volume data collection is as verify data, and up to then mistake, lake and Hoh Xil Lake are matched as verifying lake in library for selection.By to Qinghai-Tibet Platean
Day by day cloudless Snow Cover Area data set, exposure mask are cut MODIS, and the lake the MODIS ice/water area change for obtaining long-term sequence is bent
Line, by extracting freeze thawing parameter in lake using coverage of water 90% and 20% as threshold value.It is verified, it is extracted using AMSR-E data
Lake freeze thawing result and the lake extracted of the cloudless data of MODIS freeze to melt with lake the Pearson correlation coefficients of result and distinguish
For 0.968 and 0.987.
The application is based on mixed pixel dynamic linear decomposition principle, based on based on high-resolution optical image, invents
A kind of dynamic mixed pixel decomposition method of lake ice freeze thawing microwave information can be applied to the satellite-borne microwave sensing of middle low resolution
Device, such as AMSR-E greatly expand the application power of passive microwave mixed pixel, it is made to can be used for middle small lakes freeze thawing parameter
It extracts, carries out the contrast verification of lake ice freeze thawing monitoring, the Pearson of freeze thawing time by cloudless MODIS snow ice data day by day
Related coefficient is more than 0.96, it is believed that the method precision that paper is proposed is higher.
Analyzed by Comparative result available: warp is compared with the AMSR-E bright temperature data of enhancing resolution ratio, using mixing
Pixel dynamic Decomposition method is in terms of lake freeze thawing information enhancement than combining scatterometer image according to the increasing of image reconstruction algorithm
Strong result is more applicable.Mixed pixel dynamic Decomposition method can be generally applicable to area of lake and account for single 30% or more the pixel of AMSR-E
Lake, apply in lake pixel scale.
When lake between 20-30%, the speciality itself and loke shore surrounding geographical environment by lake shape and lake are special
Sign influences, and discomposing effect has certain uncertainty；Pass through the mixed pixel dynamic Decomposition method analysis to SSMIS data, knot
Fruit shows that this method can be applied in different historical datas and the now data of development higher resolution.This method is equally
It is applicable to the freeze thawing monitoring of other water in cold regions body surface types, such as bay sea ice freeze thawing monitoring or elongated river
River ice monitoring, such as yellow river ice flood monitoring and warning.
Passive microwave remote sensing data is in the case where more stars continuously acquire data at present, and the method for the invention can dash forward
The spatial resolution precision of broken data itself limits, and obtains the freeze thawing information of more lake ice to the maximum extent, can also be for when next
In generation, more high-resolution passive microwave data provided the monitoring thinking of enhancing freeze thawing information.
Finally, the present processes are only preferable embodiment, it is not intended to limit the scope of the present invention.It is all
Within the spirit and principles in the present invention, any modification, equivalent replacement, improvement and so on should be included in protection of the invention
Within the scope of.
Claims (7)
1. a kind of lake ice freeze thawing monitoring method based on dynamic Decomposition of Mixed Pixels method characterized by comprising
S1 is based on lake particular optical image and passive microwave remote sensing data, obtains the bright temperature value of lake mixed pixel and loke shore is pure
The bright temperature value of net pixel；
S2 is based on lake particular optical image and the bright temperature value of lake mixed pixel, obtains the lake by Overlap Analysis
The area ratio in lake and loke shore in mixed pixel；
S3, based in the bright temperature value of the lake mixed pixel, the bright temperature value of the loke shore Pure pixel and the lake mixed pixel
Lake and loke shore area ratio, the lake Pure pixel for obtaining the lake mixed pixel using linear spectral unmixing method are bright
Temperature value, to monitor lake ice freeze thawing phenomenon；
Wherein, the S1 further comprises:
S1.1 obtains the edge in lake using edge detection method from the lake particular optical image；It is special based on the lake
Determine the pixel position comparative analysis of optical image and the passive microwave remote sensing data and the edge in the lake, obtains the quilt
The lake center of dynamic microwave remote sensing data；
S1.2, from the passive microwave remote sensing data, the nearest bright temperature value of microwave pixel of lake center described in selected distance is
The bright temperature value of lake mixed pixel；From the passive microwave remote sensing data, nearest more of lake center described in selected distance
The bright temperature value of microwave pixel of a loke shore Pure pixel is the bright temperature value of loke shore Pure pixel after taking mean value；
Wherein, the S2 further comprises:
S2.1 is based on passive microwave remote sensing data, obtains lake mixed pixel；Lake center with the lake mixed pixel is
It is centrally generated the rectangular area of the spatial resolution of particular sensor specific frequency；
S2.2 is laid out analysis based on the rectangular area and the lake particular optical image, it is mixed to decomposite the lake
Close the ratio in lake and loke shore in pixel.
2. the method as described in claim 1, which is characterized in that the bright temperature value of lake Pure pixel described in the S3 meets:
TBsensor=aTBlake+bTBland
Wherein, TBsensorFor the bright temperature value of lake mixed pixel of the single microwave pixel of the bright temperature value of lake mixed pixel；
TBlakeFor the bright temperature value of lake Pure pixel, the bright temperature value in lake region in the bright temperature value of lake mixed pixel is indicated；
TBlandFor the bright temperature value of loke shore Pure pixel, the bright temperature value in lake mixed pixel bright temperature value Mid-continent domain is indicated；
A is the area percentage weight of lake region in the bright temperature value of lake mixed pixel；
B is the area percentage weight in lake mixed pixel bright temperature value Mid-continent domain；
Wherein, the ratio of a and b is the area ratio in lake and loke shore in the lake mixed pixel.
3. the method as described in claim 1, which is characterized in that the S1.2 further comprises: choosing with the lake center
Centered on 1 bright temperature value of microwave pixel be the bright temperature value of lake mixed pixel；
The bright temperature value of microwave pixel of lake center described in selected distance nearest 2-3 pure loke shore pixels and do it is average after for institute
State the bright temperature value of loke shore Pure pixel.
4. method as claimed in claim 1 or 3, which is characterized in that the coordinate of the lake mixed pixel meets:
Wherein, (X, Y) is the coordinate of the lake center, (Xi,Yi) be the passive microwave remote sensing data any pixel center
The coordinate of point, (Xmin,Ymin) it is spaceborne bright temperature data coordinate corresponding to the bright temperature value of lake mixed pixel.
5. the method as described in claim 1, which is characterized in that the S2.1 further comprises:
S2.1.1 generates the first border circular areas and the second border circular areas respectively centered on the lake center；
S2.1.2 obtains the minimum outsourcing rectangle of first border circular areas and second border circular areas, is based on isogonism horizontal axis
Cyclotomy column projection coordinate system characteristic obtains the specific microwave frequency pixel resolution size with the passive microwave remote sensing data
The consistent rectangular area.
6. method as claimed in claim 5, which is characterized in that the radius of first border circular areas is specific microwave frequency
One spatial resolution value, the radius of second border circular areas are another spatial resolution value of specific microwave frequency.
7. a kind of lake ice freeze thawing based on dynamic Decomposition of Mixed Pixels method monitors system characterized by comprising
Preliminary bright temperature data obtains module, for being based on lake particular optical image and passive microwave remote sensing data, obtains lake
The bright temperature value of mixed pixel and the bright temperature value of loke shore Pure pixel；
Lake and land percentage module are led to for being based on lake particular optical image and the bright temperature value of lake mixed pixel
Cross the area ratio that Overlap Analysis obtains lake and loke shore in the lake mixed pixel；
Dynamic is accurate to obtain module, for based on the bright temperature value of the lake mixed pixel, the bright temperature value of the loke shore Pure pixel and
Lake and loke shore area ratio in the lake mixed pixel obtain the lake mixing picture using linear spectral unmixing method
The bright temperature value of lake Pure pixel of member, to monitor lake ice freeze thawing phenomenon.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611266615.4A CN106872466B (en) | 2016-12-31 | 2016-12-31 | A kind of lake ice freeze thawing monitoring method and system based on dynamic Decomposition of Mixed Pixels method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611266615.4A CN106872466B (en) | 2016-12-31 | 2016-12-31 | A kind of lake ice freeze thawing monitoring method and system based on dynamic Decomposition of Mixed Pixels method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106872466A CN106872466A (en) | 2017-06-20 |
| CN106872466B true CN106872466B (en) | 2019-08-13 |
Family
ID=59164702
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611266615.4A Active CN106872466B (en) | 2016-12-31 | 2016-12-31 | A kind of lake ice freeze thawing monitoring method and system based on dynamic Decomposition of Mixed Pixels method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106872466B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107749144B (en) * | 2017-09-28 | 2019-11-22 | 成都理工大学 | A kind of flood level method for early warning of ice-lake breach and its application |
| CN110569733B (en) * | 2019-08-09 | 2022-02-01 | 中国科学院南京地理与湖泊研究所 | Lake long time sequence continuous water area change reconstruction method based on remote sensing big data platform |
| CN113252183A (en) * | 2021-07-06 | 2021-08-13 | 河南工业大学 | Processing method of 89GHz data for Antarctic ice cover surface snow melting detection |
| CN113610708B (en) * | 2021-07-28 | 2023-11-17 | 国家卫星气象中心(国家空间天气监测预警中心) | Imaging method and device for passive microwave remote sensing flood information |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101963664A (en) * | 2010-09-28 | 2011-02-02 | 中国科学院东北地理与农业生态研究所 | Microwave remote sensing pixel element decomposing method based on land and water living beings classifying information |
| CN104198052A (en) * | 2014-09-25 | 2014-12-10 | 国家卫星海洋应用中心 | Acquisition method for sea ice concentration on basis of ocean No. II satellite scanning microwave radiometer |
-
2016
- 2016-12-31 CN CN201611266615.4A patent/CN106872466B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101963664A (en) * | 2010-09-28 | 2011-02-02 | 中国科学院东北地理与农业生态研究所 | Microwave remote sensing pixel element decomposing method based on land and water living beings classifying information |
| CN104198052A (en) * | 2014-09-25 | 2014-12-10 | 国家卫星海洋应用中心 | Acquisition method for sea ice concentration on basis of ocean No. II satellite scanning microwave radiometer |
Non-Patent Citations (1)
| Title |
|---|
| Developing a Global Data Record of Daily Landscape Freeze/Thaw Status Using Satellite Passive Microwave Remote Sensing;Youngwook Kim等;《IEEE Transactions on Geoscience and Remote Sensing》;20110331;第49卷(第3期);第949-960页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106872466A (en) | 2017-06-20 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106872466B (en) | A kind of lake ice freeze thawing monitoring method and system based on dynamic Decomposition of Mixed Pixels method | |
| Derksen et al. | Development of a tundra-specific snow water equivalent retrieval algorithm for satellite passive microwave data | |
| Liang et al. | Time-series snowmelt detection over the Antarctic using Sentinel-1 SAR images on Google Earth Engine | |
| Brooker et al. | Mapping the activity and evolution of retrogressive thaw slumps by tasselled cap trend analysis of a Landsat satellite image stack | |
| Mubea et al. | Monitoring land-use change in Nakuru (Kenya) using multi-sensor satellite data | |
| Caruso et al. | Oil spills and slicks imaged by synthetic aperture radar | |
| Kim et al. | Detection of the Hebei Spirit oil spill on SAR imagery and its temporal evolution in a coastal region of the Yellow Sea | |
| Lara et al. | Automated detection of thermoerosion in permafrost ecosystems using temporally dense Landsat image stacks | |
| Almar et al. | On the use of the Radon transform to estimate longshore currents from video imagery | |
| Kwok | Declassified high-resolution visible imagery for Arctic sea ice investigations: An overview | |
| Zeng et al. | Examination of POLDER/PARASOL and MODIS/Aqua cloud fractions and properties representativeness | |
| Catalán et al. | Microwave backscattering from surf zone waves | |
| Tian et al. | An experiment for oil spill recognition using RADARSAT-2 image | |
| Giles et al. | Semi-automated feature-tracking of East Antarctic sea ice from Envisat ASAR imagery | |
| Díaz Méndez et al. | Wind and wave observations off the south Pacific Coast of Mexico using TerraSAR-X imagery | |
| Ranjbar et al. | Remote sensing of a high-Arctic, local dust event over Lake Hazen (Ellesmere Island, Nunavut, Canada) | |
| Schwarz et al. | Detection of windthrow in mountainous regions with different remote sensing data and classification methods | |
| Chavez Jr et al. | Monitoring dust storms and mapping landscape vulnerability to wind erosion using satellite and ground-based digital images | |
| Zhou et al. | Identifying soil freeze/thaw states using scattering and coherence time series of high-resolution C-band synthetic aperture radar in the Qinghai-Tibet Plateau | |
| Zhang et al. | Detection of power transmission tower from SAR image based on the fusion method of CFAR and EF feature | |
| Banks | Assessing RADARSAT-2 polarimetric SAR for mapping Shoreline Cleanup and Assessment Technique (SCAT) classes in the Canadian Arctic | |
| Jishad et al. | Thermal front detection using satellite-derived sea surface temperature in the northern indian ocean: Evaluation of gradient-based and histogram-based methods | |
| Pulvirenti et al. | Integration of SAR intensity and coherence data to improve flood mapping | |
| Yang et al. | Fusion of Landsat 8 Operational Land Imager and Geostationary Ocean Color Imager for hourly monitoring surface morphology of lake ice with high resolution in Chagan Lake of Northeast China | |
| Huh et al. | Remote sensing of turbid coastal and estuarine waters: a method of multispectral water-type analysis |
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 |