CN111505625B

CN111505625B - Active and passive combined microwave remote sensing detection method for ice and snow internal state distribution

Info

Publication number: CN111505625B
Application number: CN202010249594.5A
Authority: CN
Inventors: 董晓龙; 白东锦; 朱迪
Original assignee: National Space Science Center of CAS
Current assignee: National Space Science Center of CAS
Priority date: 2020-04-01
Filing date: 2020-04-01
Publication date: 2022-05-03
Anticipated expiration: 2040-04-01
Also published as: CN111505625A

Abstract

The invention discloses an active and passive combined microwave remote sensing detection method for ice and snow internal state distribution, which comprises the following steps: acquiring an observation value of the change of the ice and snow echo intensity along with time measured by an active ultra-wideband surface penetration radar; converting the observation value of the ice and snow echo intensity along with the change of time into the observation value of the ice and snow echo intensity along with the change of the ice and snow inner depth; according to an observation value of ice and snow echo intensity along with the change of ice and snow internal depth, determining dielectric characteristic vertical distribution information corresponding to an ice and snow internal layered structure and an internal reflecting layer; obtaining an ice and snow radiation brightness temperature value measured by a passive ultra-wideband microwave radiometer; establishing an ice and snow radiation brightness temperature forward model and an echo intensity forward model based on the vertical distribution of the ice and snow internal layered structure and the corresponding dielectric characteristics of the internal reflecting layer; and obtaining the distribution data of the ice and snow internal state by adopting a statistical regression inversion algorithm or a physical inversion algorithm. The method changes the current situation that the state distribution in the ice and snow is predicted and estimated by adopting the hypothesis model.

Description

Active and passive combined microwave remote sensing detection method for ice and snow internal state distribution

Technical Field

The invention relates to the field of microwave remote sensing, in particular to an active and passive combined microwave remote sensing detection method for ice and snow internal state distribution.

Background

The physical properties of the ice and snow interior and the substrate determine the material balance of the ice and snow, wherein the state distribution of the ice and snow interior such as temperature, density, freezing and thawing state and the like is very important, and the state distribution needs to be effectively estimated and measured for reasonably predicting the future ice and snow evolution process and the influence thereof on the climate. However, due to the limitation of environment and technical implementation cost, field measurement of ice and snow internal properties is difficult to perform, so that actual observation data of ice and snow internal state distribution is very deficient and is often predicted through an empirical model, and output results of different models are often greatly different, so that the reliability of prediction cannot be guaranteed. Therefore, it is very significant to develop a method for detecting the state distribution of the internal temperature, density, freezing and thawing state of the ice and snow.

In the state distribution, the ice and snow internal temperature distribution means an ice and snow internal physical temperature that changes with depth. The ice and snow internal temperature distribution is influenced by the surface accumulation rate, the surface temperature, the geothermal flux, the ice thickness and the heat exchange and stress action in ice, is very important for understanding ice and snow dynamics and thermodynamics, and is one of basic parameters in the energy and substance exchange analysis of ice and snow and the external environment and the future climate prediction. Currently, data on the internal temperature of ice and snow mainly come from field measurements of drill holes and ice cores. However, due to the limitation of the region and the technical implementation cost, the observation of the temperature distribution inside the ice and snow is extremely sparse, and temperature measurement covering the entire ice and snow area cannot be provided. For the requirements of researching ice and snow evolution and material balance, information of temperature distribution in ice and snow in a wide range needs to be covered, and remote sensing is one of the most suitable means. Currently, remote sensing research on inversion of temperature inside ice and snow is mainly based on an UWBRAD ultra-wideband software-defined microwave radiometer developed by the association of state university in ohio and italy, which utilizes the response of the radiated bright temperature of different frequency channels to temperature distributions at different depths to perform inversion of the temperature distributions. However, after the measured data is analyzed, it is found that the ice and snow emission is affected by a series of factors including a layered structure, random fluctuation of density distribution, an internal significant reflection layer effect and the like, and the factors cannot be effectively restricted only by using the passive microwave radiometer, so that the capability of inverting the ice and snow internal temperature distribution by using the measurement result of the passive microwave radiometer is severely limited. Since the random fluctuation of the density distribution and the effect of the internal significant reflection layer are both dielectric characteristics corresponding to the internal reflection layer of ice and snow, in the following description of the present invention, the random fluctuation of the density distribution and the effect of the internal significant reflection layer are referred to as dielectric characteristics corresponding to the internal layered structure and the internal reflection layer.

The ice and snow internal density distribution means an ice and snow internal density varying with a depth. The density distribution inside the snow and ice is subjected to the effects of snowfall events, thermal insulation changes, monsoon events and snow pressing processes, and forms a form of average density distribution plus random fluctuation. Information on the density distribution of the interior of ice and snow is important for predicting the contribution of ice and snow to sea level. At present, the data of the density distribution inside the ice and snow mainly come from the field measurement of the drill hole and the ice core. Therefore, the distribution is extremely sparse due to the limitation of the observation area, and observation covering the entire ice and snow area cannot be provided. For predicting the contribution of ice and snow to sea level, the measured ice and snow volume change is converted into mass change, the conversion process needs space coverage and time change information of ice and snow internal density distribution, and remote sensing is one of the most suitable means. At present, remote sensing research on ice and snow internal density inversion is mainly based on ground-based single-transmission multi-reception or multi-transmission multi-reception ground surface penetrating radar, interference processing is carried out on observed values of echo intensities observed by different receiving antennas along with time change, a relation between an echo interference phase corresponding to reflection position depth and electromagnetic wave propagation speed of corresponding position depth is established, and then ice and snow density depth distribution is inverted according to the relation between the electromagnetic wave propagation speed and density. This method requires that distinguishable differences in echo intensities between the transmit-receive paths of the multiple groups of transmit-receive antennas change with time, so as to extract a more accurate interference phase, and thus requires that the distance differences between the different groups of transmit-receive antennas are relatively large, which is relatively easy to implement for a foundation platform, but is difficult to expand for aerial and space platforms, so that the space coverage is also limited.

The ice and snow internal freezing and thawing state is specially used for replacing the ablation phenomenon of an ice and snow substrate, the substrate ablation is directly related to geothermal flux, the distribution of the geothermal flux is different, so that the thermal balance state of an ice bed and the ice and snow substrate can also change along with the space position, the higher geothermal flux can cause the ice and snow substrate to melt, even generate an ice lake, thereby generating a lubricating effect on the substrate, reducing friction, changing a stress structure, accelerating the reduction of ice and snow mass at the corresponding position and the movement speed to the ice and snow edge, and accelerating the contribution to the sea level. At present, the judgment of the freezing and thawing state in the ice and snow mainly depends on a surface penetration radar, the reflectivity of the ice and snow substrate is quantitatively analyzed through the echo intensity of the ice and snow substrate, and the freezing and thawing state is judged according to the characteristics of the reflectivity of the ice and snow substrate. Quantitative analysis of the reflectivity of the ice and snow substrate needs to accurately estimate the attenuation in the ice and snow, and this requires estimation and prediction of state distribution parameters determining the attenuation such as the temperature in the ice and snow by other observation means or empirical models, and the uncertainty is usually very high, so that the freeze-thaw state cannot be accurately judged. On the other hand, in the previous observation results of bright temperature of the area of the iced lake of the south pole Vostok in the SMOS task of the European Bureau, it is found that the iced lake corresponds to a cold target in the observation of the radiated bright temperature, so that the radiated bright temperature at the corresponding position is obviously lower than the radiated bright temperature value of the surrounding frozen substrate area. Therefore, the passive radiation bright temperature observation value can reflect the freezing and thawing state in the ice and snow to a certain extent. However, due to a plurality of undetermined factors in the radiation transmission process, the phenomenon that the radiation brightness temperature is low can only be used as an indication tendency, and cannot be used as a judgment index of the freeze-thaw state.

In summary, the current remote sensing detection means for the internal properties of ice and snow mainly comprises a ground grouping transceiver surface penetration radar, an airborne multi-band surface penetration radar and an airborne ultra-wideband microwave radiometer. The ground grouping transceiver surface penetrates through radar for observing the density distribution in the ice and snow. The airborne multiband surface penetration radar is used for describing an ice and snow internal reflection layer and ice and snow substrate terrain. The airborne ultra-wideband microwave radiometer measures radiation bright temperature values of different frequency channels, and the temperature distribution in the ice and snow is depicted through the response of the radiation bright temperatures of the different frequency channels to the temperature distribution in different depths.

However, the active and passive remote sensing detection means proposed at present can not realize the effective inversion of the state distribution of the internal temperature, density, freezing and thawing state and the like of ice and snow in the space range, for example, the ground grouping transceiver surface penetration radar is difficult to expand to the air and the space platform, so that the observation of the density distribution in a large range can not be realized, and the scientific target requirement can not be met. The airborne active multi-band surface penetration radar can only reflect the position of a reflecting layer in ice and snow and the size of corresponding echo intensity, and can only estimate the parameters of the random fluctuation change part of density for density distribution; quantitative analysis of dielectric attenuation is required for temperature distribution and freeze-thaw state inside ice and snow, and dielectric attenuation cannot be estimated by means of active radar echo intensity alone. The inversion of the ultra-wideband microwave radiometer is limited by the prediction capability of a forward model of the ice and snow radiation bright temperature on an actual bright temperature observation value, the forward model of the radiation bright temperature is established based on the vertical distribution of dielectric characteristics corresponding to an ice and snow layered structure and an internal reflection layer, the vertical distribution of the dielectric characteristics corresponding to the layered structure and the internal reflection layer can only be estimated through a priori model and hypothesis, and the output results of different models can have great difference.

Disclosure of Invention

The invention aims to overcome the defects of the prior art and provides an active and passive combined microwave remote sensing detection method for ice and snow internal state distribution.

In order to achieve the purpose, the invention provides an active and passive combined microwave remote sensing detection method for ice and snow internal state distribution, which comprises the following steps:

acquiring an observation value of the change of the ice and snow echo intensity along with time measured by an active ultra-wideband surface penetration radar;

obtaining an ice and snow radiation brightness temperature value measured by a passive ultra-wideband microwave radiometer;

inputting an observation value of the ice and snow echo intensity along with the change of time and an ice and snow radiation brightness temperature value into a trained ice and snow state model to obtain ice and snow internal state distribution data; the ice and snow state model is obtained by adopting a statistical regression inversion algorithm.

As an improvement of the method, the observation value of the change of the ice and snow echo intensity measured by the active ultra-wideband surface penetration radar along with the time is obtained; the method specifically comprises the following steps:

measuring in a detection frequency band by using an active ultra-wideband surface penetration radar to obtain ice and snow echo signals;

performing surface clutter suppression on the ice and snow echo signals;

and converting the ice and snow echo signals after surface impurity suppression into a time sequence of ice and snow echo intensity to obtain an observation value of ice and snow echo intensity along with time change.

As an improvement of the method, the ice and snow radiation brightness temperature value measured by the passive ultra-wideband microwave radiometer is obtained; the method specifically comprises the following steps:

measuring in a detection frequency band by using a passive ultra-wideband microwave radiometer to obtain a measured value;

converting the measured value into an ice and snow radiation brightness temperature value of each channel through system calibration and an antenna directional pattern;

and evaluating and processing each channel to obtain the radiation brightness temperature value of the ice and snow.

As an improvement of the above method, the method further comprises: training the ice and snow state model; the method specifically comprises the following steps:

acquiring distribution data of ice and snow internal temperature, density and freeze-thaw state, and performing space-time matching on the distribution data, an observation value of ice and snow echo intensity changing along with time and an ice and snow radiation bright temperature value to obtain a matching data set of the observation value of radiation bright temperature and the observation value of ice and snow echo intensity changing along with time and the ice and snow internal temperature, density and freeze-thaw state distribution data;

randomly selecting part of data from the matched data set as a training set, and using the rest of data as a test set; obtaining an independent variable radiation bright temperature observation value, an observation value of the echo intensity changing along with time and an ice and snow state model of dependent variable ice and snow internal temperature, density and freeze-thaw state distribution data by utilizing a training set;

and testing the established ice and snow state model by using a test set, and evaluating the performance of the ice and snow state model to obtain an optimal ice and snow state model.

The invention provides an active and passive combined microwave remote sensing detection method for ice and snow internal state distribution, which comprises the following steps:

converting the observation value of the ice and snow echo intensity along with the change of time into the observation value of the ice and snow echo intensity along with the change of the ice and snow inner depth;

according to an observation value of ice and snow echo intensity along with the change of ice and snow internal depth, determining the vertical distribution information of dielectric characteristics corresponding to an ice and snow internal layered structure and an internal reflection layer;

establishing an ice and snow radiation bright temperature forward model and an echo intensity forward model containing ice and snow internal temperature, density and freeze-thaw state based on the ice and snow internal layered structure and the corresponding dielectric characteristic vertical distribution of the internal reflecting layer;

setting ice and snow internal temperature, density and freeze-thaw state distribution as parameters to be inverted, establishing a cost function, and inputting the parameters to be inverted into an ice and snow radiation brightness temperature forward model to obtain a simulated brightness temperature value; inputting the parameters to be inverted into an echo intensity forward model to obtain an echo intensity analog value; and inputting the simulated brightness temperature value, the echo intensity simulated value, the ice and snow radiation brightness temperature value and the observation value of the ice and snow echo intensity changing along with the ice and snow internal depth into a cost function, adjusting the parameters to be inverted until the result of the cost function is within a set threshold range, and obtaining ice and snow internal temperature, density and freeze-thaw state distribution corresponding to the parameters to be inverted, namely ice and snow internal state distribution data.

performing surface clutter suppression on the ice and snow echo signals;

As an improvement of the above method, the observation value of the ice and snow echo intensity changing along with time is converted into the observation value of the ice and snow echo intensity changing along with the ice and snow inner depth; the method specifically comprises the following steps:

according to the observation value of the ice and snow echo intensity changing along with the time, the propagation time of the electromagnetic wave corresponding to each observation echo intensity in the ice and snow medium is determined, and the depth position z in the ice and snow medium corresponding to the ice and snow echo intensity observation value is calculated according to the following formula_echoSo as to obtain an observation value of the ice and snow echo intensity along with the change of the ice and snow inner depth;

z_echo＝v_ice·t_echo/2

wherein, t_echoThe round trip time v of the electromagnetic wave corresponding to the observed value of the ice and snow echo intensity propagating in the ice and snow medium_icePropagation speed of electromagnetic wave in ice and snow medium:

wherein c is light velocity, epsilon'_iceIs the real part of the dielectric constant of ice and snow.

As an improvement of the method, the vertical distribution information of the dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflecting layer is determined according to the observed value of the ice and snow echo intensity along with the change of the ice and snow internal depth; the method specifically comprises the following steps:

determining the position of the ice and snow internal reflecting layer and the echo intensity value of the corresponding position according to the ice and snow echo intensity;

determining a reflection mechanism corresponding to the reflection layer according to the echo characteristics; wherein the reflection mechanism comprises a density fluctuation reflection mechanism, a conductivity mutation reflection mechanism and a reflection mechanism caused by impurities;

determining the position of an internal reflecting layer and the echo intensity of which the density fluctuation is a reflecting mechanism, and carrying out empirical estimation on parameters of random fluctuation in ice and snow density distribution according to the position of the internal reflecting layer and the echo intensity;

determining the position of an internal reflecting layer with the conductivity mutation as a reflecting mechanism and the echo intensity, wherein the position of the internal reflecting layer is the depth of the conductivity mutation layer, and empirically estimating the size of the conductivity mutation layer relative to the background conductivity according to the size of the echo intensity corresponding to the reflecting position relative to the echo intensity corresponding to the adjacent non-reflecting position;

determining the position of an internal reflecting layer with impurities as a reflecting mechanism and the echo intensity, wherein the position of the internal reflecting layer is the depth of the impurity layer, and empirically estimating the dielectric parameter of the impurity layer relative to the dielectric parameter of a background medium according to the echo intensity corresponding to the reflecting position relative to the echo intensity corresponding to the adjacent non-reflecting position;

and obtaining the dielectric characteristic vertical distribution information corresponding to the ice and snow internal layered structure and the internal reflecting layer according to the random fluctuation parameters in the ice and snow density distribution, the size of the conductivity mutation layer relative to the background conductivity and the size of the dielectric parameter of the impurity layer relative to the background dielectric parameter.

converting the measured value into an ice and snow radiation brightness temperature value of each channel through system calibration and an antenna directional diagram;

evaluating and processing each channel to obtain the radiation brightness temperature value of the ice and snow;

and obtaining the ice and snow radiation brightness temperature value measured by the passive ultra-wideband microwave radiometer.

Compared with the prior art, the invention has the advantages that:

1. the invention provides an active and passive combined microwave remote sensing detection method for ice and snow internal state distribution, which changes the current situation that the ice and snow internal state distribution in the current space range can only be predicted and estimated by adopting a hypothesis model.

2. According to the invention, an ice and snow internal layered structure obtained by active remote sensing echo analysis and dielectric characteristic vertical distribution information corresponding to an internal reflection layer are used for constructing an ice and snow radiation brightness temperature forward model and an echo intensity forward model which are more in line with reality, so that the uncertainty of inversion of ice and snow internal state distribution by using radiation brightness temperature and echo intensity observation values is reduced.

3. According to the invention, the density random fluctuation variation part parameters can be estimated through the observation value of the ice and snow echo intensity along with the time variation detected by the active surface penetration radar, and the rest density distribution parameters can be used as parameters to be inverted through the observed values of the radiation brightness and the echo intensity, so that the defects that the density distribution whole parameters can not be effectively estimated by using the active multi-band surface penetration radar alone, and the density distribution inversion mode of the surface penetration radar of the active multi-group receiving and transmitting antenna device is difficult to expand to the air and a space platform are overcome.

4. According to the method, the inversion of the freezing and thawing state in the ice and snow is carried out according to the relation between the freezing and thawing state in the ice and snow and the radiation brightness of the low-frequency channel and the intensity of the ice and snow echo, so that the defect that quantitative analysis of the reflectivity of an ice and snow substrate cannot be accurately carried out due to the fact that a surface penetration radar is used alone is overcome, and the uncertainty of the judgment of the freezing and thawing state in the ice and snow is reduced.

5. The active and passive combined microwave remote sensing detection load can be carried on a surface, air and space platform, so that observation of different ranges and resolutions of state distribution such as internal temperature, density, freeze-thaw state and the like of ice and snow is realized, and the limitation that the current observation of the internal state distribution of ice and snow mainly depends on the observation range of remote sensing of a drill hole, an ice core or a foundation platform is overcome.

Drawings

FIG. 1 is a flow chart of an active and passive combined microwave remote sensing detection method for ice and snow internal state distribution according to the present invention;

FIG. 2 is a specific flowchart of obtaining vertical distribution information of dielectric characteristics corresponding to an internal layered structure of ice and snow and an internal reflection layer by using an active remote sensing observation value analysis in the active and passive combined microwave remote sensing detection method for internal state distribution of ice and snow provided by the present invention;

fig. 3 is a specific flowchart of a statistical regression inversion algorithm established in the active and passive microwave remote sensing method for detecting ice and snow internal state distribution according to embodiment 1 of the present invention;

fig. 4 is a specific flowchart of establishing a physical inversion algorithm in the active and passive microwave remote sensing method for ice and snow internal state distribution according to embodiment 2 of the present invention.

Detailed Description

The invention aims to overcome the defects that the accuracy of an ice and snow radiation brightness temperature forward model is limited due to the fact that the vertical distribution of dielectric characteristics corresponding to an ice and snow layered structure and an internal reflection layer cannot be restrained by singly utilizing passive microwave remote sensing observation, so that the radiation brightness temperature value measured by utilizing the passive microwave remote sensing cannot effectively invert the state distribution such as the internal temperature, density and freeze-thaw state of ice and snow, and the defects that all parameters of the internal density distribution of ice and snow and the attenuation of ice and snow media cannot be effectively estimated by singly utilizing the active microwave remote sensing observation, so that the observation value of the echo intensity measured by utilizing the active microwave remote sensing along with the change of time cannot effectively restrain and judge the state distribution such as the internal temperature, density and freeze-thaw state of ice and snow, and provides a driving and passive combined microwave remote sensing detection method for the internal state distribution of ice and snow. The active ultra-wideband surface penetration radar can effectively describe an internal layered structure from the ice and snow surface to the ice bottom, and the observed value of the change of the detected echo intensity along with time can be used for analyzing and obtaining the information of the vertical distribution of dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflecting layer, so that a more practical radiation bright-temperature forward model and an echo intensity forward model can be established according to the information of the vertical distribution of the dielectric characteristics corresponding to the layered structure and the internal reflecting layer; the method comprises the steps that a radiation bright temperature forward model comprises state distribution of ice and snow internal temperature, density, a freeze-thaw state and the like, an inversion algorithm is established according to the relation between the state distribution of the ice and snow internal temperature, density, the freeze-thaw state and the like and ice and snow radiation bright temperature and ice and snow echo intensity, parameters to be inverted of the state distribution of the ice and snow internal temperature, density, the freeze-thaw state and the like can be solved based on an observation value and a radiation bright temperature observation value of echo intensity changing along with time, and then inversion results of the state distribution of the ice and snow internal temperature, density, the freeze-thaw state and the like are obtained according to a corresponding model. Therefore, the method can realize effective observation and inversion of state parameter distribution such as ice and snow internal temperature, density, freeze-thaw state and the like. In addition, the designed active and passive combined microwave detection amount has no strict requirement on the detection platform, namely the realization installation of the equipment load and the detection mode do not change essentially due to the change of the detection platform, and different platform selections only affect the space range and the resolution ratio of observation, so the invention can be carried on the surface, the air or the space platform according to the specific scientific target detection requirement, realize the observation of the distribution of the ice and snow internal states with different space ranges and resolutions, and overcome the defect that the remote sensing mode only carried on a foundation platform or the local limited observation range of field measurement can only be carried.

Step 1) adopting an active and passive combined detection system, wherein an ultra-wideband surface penetration radar is adopted for active detection, and an ultra-wideband microwave radiometer is adopted for passive detection;

step 2) determining detection frequency, acquiring an observation value of the echo intensity along with time change through an actively detected surface penetration radar, and acquiring a radiation brightness and temperature observation value of the detection frequency through a passively detected microwave radiometer;

step 3) based on the observation value of the echo intensity along with the change of time obtained in the step 2), obtaining an ice and snow internal layered structure and the vertical distribution of dielectric characteristics corresponding to an internal reflection layer, and establishing an ice and snow radiation bright temperature forward model and an echo intensity forward model which contain the state distribution of ice and snow internal temperature, density, freeze-thaw state and the like;

and 4) establishing an inversion algorithm according to the relation between the state distribution of the ice and snow internal temperature, density, freeze-thaw state and the like and the ice and snow radiation bright temperature and echo intensity, and performing inversion to obtain the ice and snow internal state distribution based on the observation value of the echo intensity changing along with time and the radiation bright temperature observation value in the step 2).

The method steps are explained below.

In the step 1), the active microwave detector is an ultra-wideband surface penetration radar, and the passive microwave detector is an ultra-wideband microwave radiometer and is carried on a surface, air or space platform.

In step 2), determining the detection frequency specifically includes:

and 2-1) determining an active detection frequency range, wherein the detection capability and the appropriate resolution ratio are required for an internal layered structure from the ice and snow surface to the ice bottom. The echo detection of a high frequency band is used for ice and snow surface observation and fine structure description in a snow layer; the echo detection of a higher frequency band is used for describing the internal structure of ice and snow within a depth range of tens of meters; echo detection of a lower frequency band is used for describing an internal structure of ice and snow within a depth range of hundreds of meters; the low-frequency-band echo detection is used for describing the internal structure of ice and snow within a depth range of thousands of meters;

step 2-2) determining a passive detection frequency band, wherein the passive detection frequency band is required to have sensitivity to radiation at thousands of meters of different depths in the ice and snow; the detected ultra-wideband radiation brightness-temperature spectrum can be used for restraining internal state distribution;

and 2-3) selecting the optimal detection frequency combination as the detection frequency from the detection frequency bands selected in the step 2-1) and the step 2-2) through sensitivity and correlation analysis.

In the step 2), acquiring an observation value of the change of the ice and snow echo intensity of the required detection frequency along with time through an active ultra-wideband surface penetration radar specifically comprises:

step 2-4) surface clutter suppression is carried out on echo signals received by the active radar, and the suppression method which can be adopted comprises the methods of forward-orbit synthetic aperture processing or cross-orbit interference processing and the like;

and 2-5) converting the echo signals after the surface clutter suppression into a time sequence of echo intensity.

In step 2), acquiring a radiation brightness temperature observation value of a required detection frequency through a passive ultra-wideband microwave radiometer specifically comprises:

step 2-6), converting the electric signal received by the passive microwave radiometer into the brightness temperature of ice and snow radiation through system calibration and an antenna directional diagram;

and 2-7) evaluating and processing the radiated brightness temperature of each channel, such as processing the channel possibly having radio frequency interference by an RFI radio frequency interference elimination algorithm.

In step 3), obtaining the vertical distribution of the dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflection layer from the observation value of the echo intensity changing along with time specifically comprises:

and 3-1) converting the time sequence of the radar echo intensity into an observation value of the echo intensity changing along with the depth position in the ice and snow. The concrete mode is as follows: determining the propagation speed of electromagnetic wave in ice and snow according to the refractive index determined by the dielectric constant of the ice and snow medium

Wherein c is light velocity, epsilon'_iceIs the real part of the dielectric constant of ice and snow, v_iceIs the propagation speed of electromagnetic waves in the ice and snow medium.

Further, the propagation time of the electromagnetic wave in the ice and snow medium, which is generated according to each observation echo intensity, is determined according to the observation value of the ice and snow echo intensity changing along with the time, and the propagation speed of the electromagnetic wave in the ice and snow medium is multiplied to obtain the depth position corresponding to the ice and snow echo intensity observation value, which is expressed as the depth position corresponding to the ice and snow echo intensity observation value

z_echo＝v_ice·t_echo/2

Wherein, t_echoThe round trip time z of the electromagnetic wave corresponding to the observed value of the ice and snow echo intensity propagating in the ice and snow medium_echoDepth location in ice and snow medium corresponding to ice and snow echo intensity observed value

The echo intensity time sequence of the steps 2-4) and 2-5) can be converted into an observation value of the echo intensity changing along with the position of the depth in the ice and snow.

Step 3-2) analyzing the relation between the observation value of the echo intensity changing along with the depth position and the vertical distribution of dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflecting layer, wherein the specific mode is as follows:

(1) and analyzing the reflection mechanism of the reflected echo inside the radar. The reflection mechanism of the reflected echo inside the ice and snow generally includes: dielectric discontinuities caused by density contrast due to density fluctuations; dielectric discontinuity caused by abrupt conductivity changes; dielectric discontinuity caused by impurities.

For dielectric discontinuity caused by ice and snow density fluctuation, the intensity contrast caused by the density fluctuation determines the intensity of the echo at the reflection position;

for dielectric discontinuity caused by abrupt conductivity change, the size of the conductivity peak relative to the background medium conductivity determines the size of the echo intensity at the reflection position relative to the echo intensity at the adjacent non-reflection position;

for impurity induced dielectric discontinuities, the magnitude of the impurity layer dielectric parameter relative to the background dielectric parameter determines the magnitude of the echo intensity at the reflection site relative to the echo intensity at the adjacent non-reflection site.

(2) And determining a corresponding reflection mechanism according to the intensity characteristics of the reflected echoes. And analyzing the reflection echo characteristics corresponding to different reflection mechanisms through the internal structure data of the ice and snow and the active remote sensing observation value measured on site. The density fluctuation as the reflection mechanism is characterized in that the intensity of the generated reflection echo is relatively dense corresponding to the reflection interface, and the reflection interface is usually not present in deep ice.

(3) And estimating the vertical distribution of the dielectric characteristics corresponding to the internal reflection layer according to the intensity of the reflected echo. And analyzing the empirical relationship between the reflected echo intensities corresponding to different reflection mechanisms and corresponding dielectric characteristics through the internal structure data of the ice and snow and the active remote sensing observation values measured on site.

And 3-3) deducing the vertical distribution of the dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflecting layer according to the echo intensity depth distribution obtained in the step 3-1) based on the relationship between the echo intensity and the vertical distribution of the dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflecting layer in the step 3-2). The concrete mode is as follows:

(1) determining the position of the reflecting layer and the echo intensity value of the corresponding position according to the echo intensity;

(2) determining a reflection mechanism corresponding to the reflection layer according to the echo intensity characteristics;

(3) determining an empirical model of density distribution and undetermined parameters for the reflected echo with density fluctuation as a reflection mechanism, estimating the density contrast size according to the reflected echo intensity based on the empirical relationship between the reflected echo intensity corresponding to the density fluctuation reflection mechanism in the step 3-2) and corresponding dielectric characteristics, and further estimating the parameters describing density random fluctuation in the density model according to the density contrast size;

(4) for the reflected echo with the conductivity mutation as a reflection mechanism, estimating the size of the conductivity mutation layer relative to the background conductivity according to the intensity of the reflected echo based on the empirical relationship between the intensity of the reflected echo corresponding to the conductivity mutation reflection mechanism and the corresponding dielectric characteristics in the step 3-2);

(5) for the reflection echo with the impurity as a reflection mechanism, estimating the size of the dielectric parameter of the impurity layer relative to the dielectric parameter of the background medium according to the reflection echo intensity based on the empirical relationship between the reflection echo intensity corresponding to the impurity reflection mechanism in the step 3-2) and the corresponding dielectric characteristic;

(6) and (5) integrating the steps (1) to (5) to obtain the result that the dielectric characteristics of the ice and snow internal layered structure correspond to the internal reflecting layer in vertical distribution.

And 3-4) establishing an ice and snow radiation bright temperature forward model and an echo intensity forward model which have the same layered structure and the dielectric characteristic vertical distribution corresponding to the internal reflecting layer and comprise the state distribution of ice and snow internal temperature, density, freeze-thaw state and the like according to the ice and snow internal layered structure and the dielectric characteristic vertical distribution information corresponding to the internal reflecting layer obtained in the steps 3-1) -3).

In step 4), if a statistical regression inversion algorithm, including but not limited to a neural network inversion algorithm, is used, the method specifically includes:

step 4-1) acquiring state distribution data of ice and snow internal temperature, density, freeze-thaw state and the like serving as true values, and then performing space-time matching on the state distribution data of ice and snow internal temperature, density, freeze-thaw state and the like and an observation value of echo intensity changing along with time and a radiation bright temperature observation value to obtain a matching data set of the radiation bright temperature observation value and the observation value of echo intensity changing along with time and the state distribution data of ice and snow internal temperature, density, freeze-thaw state and the like;

step 4-2) randomly selecting part of data as a training set A, and using the rest of data as a test set B; obtaining an independent variable radiation bright temperature observation value, an observation value of the echo intensity along with the time change and an ice and snow state model of state distribution data of a dependent variable such as ice and snow internal temperature, density and freeze-thaw state by using the training set A;

step 4-3) testing the established ice and snow state model by using the test set B, evaluating the performance of the model and obtaining an optimal ice and snow state model;

and 4-4) inputting the real-time radiation bright temperature observation value and the observation value of the echo intensity changing along with time, which are obtained in the step 2), into a trained ice and snow state model to obtain state distribution such as ice and snow internal temperature, density, freeze-thaw state and the like, namely ice and snow internal state distribution data.

In step 4), if a physical inversion algorithm is adopted, the method specifically comprises the following steps:

step 4-1) determining an empirical model of state distribution such as ice and snow internal temperature, density, freeze-thaw state and the like and corresponding parameters to be inverted;

step 4-2) establishing a cost function, adjusting to-be-inverted parameters of state distribution such as ice and snow internal temperature, density and freeze-thaw state, substituting the to-be-inverted parameters into an observation value which is established based on time-space matching and contains the ice and snow radiation bright temperature forward model and the echo intensity forward model of the state distribution parameters such as ice and snow internal temperature, density and freeze-thaw state, calculating corresponding radiation bright temperature simulation values and echo intensity simulation values, substituting the simulation values and actual observation values into the cost function, and searching to-be-inverted parameter solutions of state distribution such as ice and snow internal temperature, density and freeze-thaw state, wherein the cost function results meet a set threshold range;

and 4-3) substituting the solution of the parameters to be inverted of the state distribution such as the ice and snow internal temperature, density, freeze-thaw state and the like obtained by inversion into the corresponding model to obtain the inversion result of the state distribution such as the ice and snow internal temperature, density, freeze-thaw state and the like.

The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and examples.

Example 1

With reference to fig. 1, the present embodiment can be specifically described as follows:

step 101: and detecting the state distribution such as the temperature, the density and the freeze-thaw state in the ice and snow according to the relationship between the state distribution such as the temperature, the density and the freeze-thaw state in the ice and snow and the relationship between the radiation brightness temperature value observed by the ultra-wideband microwave radiometer and the echo intensity observed by the ultra-wideband surface penetration radar.

The radiation brightness temperature values of different frequency bands of ice and snow are measured by using a passive ultra-wideband microwave radiometer, and the ice and snow echo strength values of different depths are measured by using an active ultra-wideband surface penetration radar. The radiation brightness temperatures of different frequency bands reflect the total radiation quantity from different depths of ice and snow, and are related to the internal state distribution of different depths; the echo intensities at different depths are the combined action result of the reflectivity and the attenuation of ice and snow on the reflectivity, and are determined by the state distribution in the ice and snow, so that the detection of the state distribution such as the temperature, the density, the freeze-thaw state and the like in the ice and snow can be realized by using active and passive combined microwave remote sensing observation.

Step 102: and detecting the vertical distribution information of the dielectric characteristics corresponding to the internal layered structure from the ice and snow surface to the ice bottom and the internal reflecting layer according to the relationship between the vertical distribution information of the dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflecting layer and the echo intensity detected by the ultra-wideband radar.

The observation value of the change of the ice and snow echo intensity along with the time is measured by using an active ultra-wideband surface penetration radar, and the echo intensity is the combined action result of the reflectivity and the ice and snow attenuation on the reflectivity, so that the reflection mechanism corresponding to the internal reflection layer can be determined through echo intensity characteristic analysis, and the empirical estimation of the corresponding dielectric characteristics is carried out. The radar penetration depth of different frequency bands is different, the penetration depth of a relatively high frequency band is shallow, the radar is suitable for observation of a shallow ice and snow layered structure, and the penetration depth of a relatively low frequency band is deep, and the radar is suitable for observation of a deep ice and snow layered structure. Therefore, the active ultra-wideband surface penetration radar is utilized, and the detection of the vertical distribution information of the internal layered structure from the ice and snow surface to the ice bottom and the dielectric characteristics corresponding to the internal reflection layer can be realized.

Step 103: based on the passive ultra-wideband microwave radiometer and the active ultra-wideband surface penetration radar detection principle in the steps 101 and 102, an active and passive frequency band sensitive to temperature distribution, density distribution and freeze-thaw state and an active frequency band capable of describing the vertical distribution of the internal layered structure from the ice and snow surface to the ice bottom and the dielectric characteristics corresponding to the internal reflection layer are selected.

(1) According to the passive detection principle of the step 101, the radiation brightness temperatures observed by different frequency channels in the ultra-wideband microwave radiometer correspond to the total radiation amount from different depths of ice and snow, and the temperature distribution, density distribution and freeze-thaw state information in the ice and snow can be restrained by the brightness temperature observation values of the different frequency channels;

(2) the description of the layered structure of the ice and snow inner part requires the radar to have the detection capability of the reflecting layer of the ice and snow inner part with different depths and have proper resolution according to the active detection principle of the step 101 and the step 102. The observation of topographic information on the surface of ice and snow and fine structures in the snow layer requires a high-frequency-band-based ultra-wideband radar; the internal reflection layer of the ice and snow with the depth of tens of meters on the shallow layer is usually on the centimeter and sub-centimeter level, so the observation resolution of the internal layering of the ice and snow on the shallow layer at least reaches the centimeter level, and the observation of the shallow layer of ten meters is required to be based on the ultra-wideband radar of a higher frequency band; the layering interval of the ice and snow internal reflection layer with the depth of hundreds of meters is usually larger than that of a shallow layer, and the frequency band required by observation of the ice and snow internal reflection layer with the depth of hundreds of meters is lower in combination with the penetration requirement; the layering interval of the internal reflecting layer of the ice and snow with the depth of thousands of meters is usually large, and the frequency band required by observation of the internal reflecting layer of the ice and snow with the depth of thousands of meters is lower in combination with the penetration requirement. And by combining the observation value of the change of the ultra-wideband surface penetrating radar echo intensity along with the time, the vertical distribution information of the internal layered structure from the ice and snow surface to the ice bottom and the dielectric characteristics corresponding to the internal reflecting layer can be analyzed and obtained.

(3) Through sensitivity and correlation analysis, the optimal detection frequency combination is selected in the sensitive frequency bands selected in the steps (1) and (2), and the detection of the temperature distribution, the density distribution and the freeze-thaw state in the ice and snow is realized.

Step 104: and determining an active and passive combined microwave remote sensing detection scheme of state distribution such as the temperature, density, freezing and thawing state and the like in the ice and snow according to application requirements and detection principles in the step 101 and the step 102.

(1) According to the detection principle in the step 101, the required passive microwave radiometer is an ultra-wideband microwave radiometer, and an ice and snow radiation brightness temperature observation value is obtained;

(2) according to the detection principle in the step 102, the required active radar is an ultra-wideband surface penetration radar, and an observation value of the ice and snow echo intensity along with the change of time is obtained;

(3) according to the observation requirements of state distribution such as the temperature, density and freeze-thaw state in the ice and snow and the selection of a carrying platform, determining the spatial resolution and the vertical resolution of the required active and passive microwave detectors;

(4) and (3) carrying out system design of the active and passive microwave detectors, and determining various system parameters such as antenna aperture, polarization mode, radiation sensitivity, pulse width, radar echo signal-to-noise ratio, bandwidth and the like according to the platform operation mode.

Step 105: acquiring a radiation brightness and temperature observation value of a required detection frequency channel through a passive ultra-wideband microwave radiometer, wherein the step further comprises the following processing method:

(1) through system calibration and combination of an antenna directional diagram, an electric signal received by the passive microwave radiometer is converted into the brightness temperature of ice and snow radiation.

(2) And evaluating and processing the radiated brightness temperature of each channel, such as processing an RFI radio frequency interference elimination algorithm aiming at the channel possibly having radio frequency interference.

Step 106: the method comprises the steps of acquiring ice and snow echo signals of required detection frequency through an active ultra-wideband surface penetration radar, carrying out surface clutter suppression on the echo signals, and converting the echo signals into a time sequence of echo intensity. As shown in fig. 2, this step further includes the following processing methods:

(1) the method comprises the following steps of performing surface clutter suppression on echo signals received by an active radar, wherein the suppression method comprises the methods of forward-orbit synthetic aperture processing or cross-orbit interference processing and the like;

(2) converting the echo signals after the surface clutter suppression into a time sequence of echo intensity;

step 107: in order to obtain state distribution such as ice and snow internal temperature, density, freeze-thaw state and the like from the observed value of the radiation bright temperature and the observed value of the echo intensity along with the change of time, a statistical regression inversion algorithm is established.

As shown in FIG. 3, the statistical regression inversion algorithm is established as follows:

(1) acquiring state distribution data of ice and snow internal temperature, density, freeze-thaw state and the like serving as true values, and then performing space-time matching on the state distribution data of the ice and snow internal temperature, density, freeze-thaw state and the like and an observation value of echo intensity changing along with time and a radiation bright temperature observation value to obtain a matching data set of the radiation bright temperature observation value and the observation value of echo intensity changing along with time and the state distribution data of the ice and snow internal temperature, density, freeze-thaw state and the like;

(2) randomly selecting part of data as a training set A, and using the rest of data as a test set B; obtaining an independent variable radiation bright temperature observation value, an observation value of the echo intensity along with the time change and an ice and snow state model of state distribution data of a dependent variable such as ice and snow internal temperature, density and freeze-thaw state by using the training set A;

(3) testing the established ice and snow state model by using the test set B, evaluating the performance of the model and obtaining an optimal ice and snow state model;

(4) and (4) inputting the real-time radiation bright-temperature observation values and the observation values of the echo intensity changing along with time, which are obtained in the steps 105 and 106, into a trained ice and snow state model to obtain state distribution of the ice and snow internal temperature, density, freeze-thaw state and the like.

Example 2

The specific steps of this example are as follows:

steps 101 to 105 are the same as in example 1.

Step 106: the method comprises the steps of obtaining an ice and snow echo signal of a required detection frequency through an active ultra-wideband surface penetration radar, carrying out surface clutter suppression on the echo signal, converting the echo signal into a time sequence of echo intensity, determining the propagation speed of electromagnetic waves in an ice and snow medium by combining the refractive index of the ice and snow medium, and obtaining an observation value of the echo intensity along with the change of depth. And acquiring the dielectric characteristic vertical distribution information of the ice and snow internal layered structure corresponding to the internal reflection layer through the depth distribution of the echo intensity.

As shown in fig. 2, this step further includes the following processing methods:

(3) determining the propagation velocity of electromagnetic waves in the ice and snow according to the refractive index determined by the dielectric constant of the ice and snow medium

z_echo＝v_ice·t_echo/2

Wherein, t_echoThe round trip time z of the electromagnetic wave corresponding to the observed value of the ice and snow echo intensity propagating in the ice and snow medium_echoAnd the depth position in the ice and snow medium corresponding to the ice and snow echo intensity observed value.

And (3) converting the observation value of the echo intensity changing along with time obtained in the steps (1) and (2) into an observation value of the echo intensity changing along with the depth position in the ice and snow.

(4) And determining the position of the ice and snow internal reflecting layer and a corresponding echo intensity value according to the observation value of the echo intensity along with the change of the ice and snow internal depth position.

(5) Determining the position of an internal reflecting layer with impurities as a reflecting mechanism and the echo intensity, considering the position of the corresponding internal reflecting layer as the depth of the impurity layer, and performing empirical estimation on the dielectric parameter of the impurity layer relative to the dielectric parameter of a background medium according to the echo intensity corresponding to the reflecting position relative to the echo intensity corresponding to the adjacent non-reflecting position;

(6) determining the position of an internal reflecting layer and the echo intensity of a reflecting mechanism, wherein the conductivity mutation is determined to be the position of the internal reflecting layer, the position of the corresponding internal reflecting layer is regarded as the depth of the conductivity mutation layer, and the empirical estimation is carried out on the conductivity of the conductivity mutation layer relative to the background medium according to the echo intensity corresponding to the reflecting position relative to the echo intensity corresponding to the adjacent non-reflecting position;

(7) and determining the position and the echo intensity of an internal reflecting layer with density fluctuation as a reflecting mechanism, and carrying out empirical estimation on parameters of a random fluctuation part in ice and snow density distribution according to the position and the echo intensity value of the corresponding internal reflecting layer.

The concrete mode is as follows:

first, the depth distribution of the reflectivity of the reflection mechanism, which is the density fluctuation, is deduced from the depth distribution of the echo intensities.

Further, the parameters of the random fluctuation portion in the density distribution are estimated from the characteristics of the reflectance distribution. Taking an exponential density distribution model with a segmented snow cover pressing characteristic as an example, the density distribution is in the form of:

where ρ is_i＝916.7kg/m³Is the density of pure ice, ρ_sIs the surface density of ice and snow, ρ_c1Is the first density of turning points, p_c2The second inflection point density. L is a radical of an alcohol₁、L₂、L₃Attenuation length of exponential average density, z, over three sections of regions divided for two turning points respectively_c1And z_c2Are the depth positions corresponding to the two turning points respectively. N (sigma)_ρ) Expressed as σ_ρIs a normal distribution of standard deviation, σ_ρ1、σ_ρ2、σ_ρ3Respectively, the standard deviation, alpha, of the randomly varying part of the density over the three regions_ρ1、α_ρ2、α_ρ3The attenuation lengths of the randomly varying density portions in the three regions are shown respectively.

At this time, the position depth corresponding to the first turning point of the reflectivity distribution is the position depth z of the first density turning point_c1The position depth corresponding to the second turning point of the reflectivity distribution is the position depth z of the second density turning point_c2. Further based on the relationship between reflectivity and density contrast when density fluctuations are the reflection mechanism, i.e.

Wherein R is_layerIs the reflection coefficient of the reflection layer, d is the thickness of the reflection layer, lambda is the radar operating wavelength, and Δ ρ is the density contrast in g/cm³；

Calculating the variation of density contrast with depth according to the reflectivity distribution, and respectively calculating the density random fluctuation variation parameter sigma on each segment area according to the variation of the density contrast depth on the segment area_ρ1、σ_ρ2、σ_ρ3And alpha_ρ1、α_ρ2、α_ρ3Carrying out estimation;

step 107: based on the vertical distribution of dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflecting layer obtained by echo intensity analysis, an ice and snow radiation brightness temperature forward model and an echo intensity forward model which have the same layered structure and the same dielectric characteristics corresponding to the internal reflecting layer and comprise state distributions of ice and snow internal temperature, density, freeze-thaw state and the like are established.

Step 108: in order to obtain state distribution such as ice and snow internal temperature, density, freeze-thaw state and the like from the observed value of the radiation bright temperature and the observed value of the echo intensity along with the change of time, a physical inversion algorithm is established.

As shown in fig. 4, the physical inversion algorithm is established as follows:

(1) determining an empirical model of state distribution such as ice and snow internal temperature, density, freeze-thaw state and the like and corresponding parameters to be inverted;

(2) establishing a cost function, adjusting to-be-inverted parameters of state distribution such as ice and snow internal temperature, density and freeze-thaw state, substituting the to-be-inverted parameters into an observation value which is established based on time-space matching and contains the ice and snow radiation bright temperature forward model and the echo intensity forward model of the state distribution parameters such as ice and snow internal temperature, density and freeze-thaw state, calculating corresponding radiation bright temperature analog value and echo intensity analog value, substituting the analog value and actual observation value into the cost function, and searching to-be-inverted parameter solutions of the state distribution such as ice and snow internal temperature, density and freeze-thaw state, wherein the cost function results meet a set threshold range; the cost function may take the following form, but is not limited thereto.

An example of a cost function is as follows. Setting parameters to be inverted of ice and snow internal temperature as geothermal flux G and annual average accumulation rate M, and setting parameters to be inverted of density as surface density rho_iAnd density turning point density ρ_cFreezing and thawing stateThe parameter to be inverted is the ice bottom reflectivity R_bed. On-site measurement data of the existing partial area can be used as a geothermal flux and an average annual accumulation rate prior value which respectively correspond to G^priorAnd M^priorAnd a priori value of density turning point

Establishing a cost function for the above parameter estimation to be inverted can be expressed as

Wherein the content of the first and second substances,

as an observed value of the radiant light temperature as a function of frequency,

a radiation brightness temperature analog value calculated for the passive forward model;

as an observation of the echo intensity as a function of depth,

an echo intensity analog value calculated for the active forward model; g^priorAnd M^priorPrior values for geothermal flux and annual average accumulation rate, respectively;

is a prior value of density at a density turning point; sigma_TBIs the standard deviation of the actually measured radiation brightness temperature,

for standard deviation of measured echo intensity, σ_GAnd σ_MRespectively the standard deviation of the prior geothermal flux and the prior annual average accumulation rate;

is the standard deviation of the prior density turning point density; n is a radical of_fThe number of frequencies of the radiation brightness temperature observation considered in the cost function; n is a radical of_dIs the number of depths of the echo intensity observation considered in the cost function.

(3) The ice and snow internal temperature, density and freeze-thaw state distribution corresponding to the parameters to be inverted are obtained at this time, and are ice and snow internal state distribution data.

In summary, embodiments 1 and 2 provide an active and passive combined microwave remote sensing method for ice and snow internal state distribution. The observation values of the radiation brightness temperature of different frequency channels of the passive ultra-wideband microwave radiometer and the observation values of the echo intensity of different depths of the active ultra-wideband surface penetrating radar are related to the state distribution of the temperature, the density, the freeze-thaw state and the like in the ice and snow; a radiation bright temperature and echo intensity forward model constructed by obtaining the vertical distribution information of dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflection layer based on the ice and snow echo intensity observation value analysis is closer to reality. And inversion of state distribution parameters such as ice and snow internal temperature, density, freeze-thaw state and the like can be performed by using the radiation bright temperature observation values of different frequency channels and the observation value of the echo intensity changing along with time. Therefore, state distribution information such as ice and snow internal temperature, density, freeze-thaw state and the like can be obtained through active and passive combined microwave remote sensing detection and inversion according to the ice and snow radiation bright temperature observation value and the observation value of echo intensity changing along with time.

The invention has the technical effects that:

1. the method can observe the radiation bright temperature of different frequency bands of ice and snow through a passive ultra-wideband radiometer, and observe the echo intensity of different depths of ice and snow through an active ultra-wideband surface penetration radar; the method comprises the steps that the vertical distribution information of dielectric characteristics corresponding to an ice and snow internal layered structure and an internal reflection layer is obtained through the analysis of an observation value of echo intensity along with the change of time, a radiation brightness temperature forward model and an echo intensity forward model which are more in line with the reality are established according to the vertical distribution information of the dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflection layer, and the uncertainty of state distribution inversion of ice and snow internal temperature, density, freeze-thaw state and the like through the measurement of the echo intensity by a radiation brightness temperature and ultra-wide band surface penetration radar measured by an ultra-wide band microwave radiometer is reduced;

2. the method can acquire the temperature distribution information in the ice and snow from the ice and snow surface to the ice bottom, maintains the sensitivity to the radiation of the low-frequency channel of the ultra-wide band microwave radiometer to the depth of kilometers, maintains the sensitivity to the deep echo intensity through the low-frequency band of the ultra-wide band surface penetration radar, so that the deep ice and snow temperature can be inverted, and maintains the sensitivity to the radiation of the high-frequency channel and the echo intensity of the shallow layer, so that the shallow ice and snow temperature can be inverted. The estimation capability of the deep ice and snow temperature is very beneficial to the analysis of the ice and snow substrate state and energy matter exchange;

3. the method can acquire density distribution information in ice and snow, and estimates density random fluctuation change parameters in density distribution through an observation value of the echo intensity changing along with time in a density fluctuation depth range by a surface penetration radar. And the rest density distribution parameters are inverted by the radiation brightness temperature value measured by the passive ultra-wideband microwave radiometer and the echo intensity measured by the active ultra-wideband surface penetration radar. The estimation capability of ice and snow density distribution is very important for predicting the contribution of ice and snow to sea level;

4. the method can acquire the information of the freezing and thawing state in the ice and snow, judge the position of the under-ice lake through the ultra-wideband microwave radiometer, and judge the freezing and thawing state in the ice and snow according to the relationship between the freezing and thawing state in the ice and snow, the echo intensity of the ice and snow and the radiation brightness temperature value of the low-frequency channel. The estimation capability of the freezing and thawing state in the ice and snow is very beneficial to the prediction of the geothermal flow distribution;

5. the method can be used for describing the complete internal layered structure from the ice and snow surface to the ice bottom and the corresponding dielectric characteristic vertical distribution information of the internal reflecting layer. Observing a fine structure in the snow layer and topographic and elevation information on the surface of the ice and snow in a high-frequency section of the ultra-wideband surface penetration radar; observing an ice and snow internal reflecting layer within a logarithmic ten-meter depth range at a higher frequency range; observing an ice and snow internal reflecting layer within a depth range of hundreds of meters at a lower frequency range; and observing the ice and snow internal reflecting layer within a kilometer depth range at a low frequency range. And the ultra-wideband surface penetration radar can judge the positions of water layers and water cavities in ice and snow, and depends on the strong reflection characteristics of the water layers in the ice and snow and the extremely weak echo under the water layers, even the phenomenon that the echo of an ice bed disappears. Thereby completely depicting the vertical distribution information of the internal layered structure from the ice and snow surface to the ice bottom and the corresponding dielectric characteristics of the internal reflection layer;

6. the method can obtain the coverage of the ice and snow area in a larger range and the observation with higher resolution. The remote sensing of ice and snow areas in different ranges can be effectively carried out through the surface, air and space platforms, the resolution of the surface and air platforms is higher than that of the space platform, for example, the spatial resolution of a radiated bright and warm product of a microwave radiometer carried by an SMOS task of the European space administration is about 50km, the spatial resolution of the radiated bright and warm product of the air platform can be better than 10km, and the spatial resolution of the lower air platform can be better than 1 km. Therefore, the device can be carried on different platforms according to the requirements of the coverage range and the resolution ratio of the required ice and snow area, the detection of the state distribution of the ice and snow internal temperature, density, freezing and thawing state and the like with the required resolution ratio can be realized, the coverage of the ice and snow in different ranges can be realized, and the device is very beneficial to different scientific target requirements;

7. according to the method, the vertical distribution information of the dielectric characteristics corresponding to the ice and snow internal layered structure and the internal reflection layer can be obtained through the active ultra-wideband surface penetration radar remote sensing observation value, a more accurate radiation brightness and temperature forward model is constructed, and the defect that the vertical distribution of the dielectric characteristics corresponding to the layered structure and the internal reflection layer cannot be restrained by singly using a passive microwave radiometer is overcome. The method can estimate the parameters of the random fluctuation change part of the density through the remote sensing observation value of the active ultra-wideband surface penetration radar, and carry out inversion on other parameters in ice and snow density distribution through the ice and snow radiation bright temperature observation value and the echo intensity observation value, thereby overcoming the defect that a plurality of groups of ground transmitting and receiving surface penetration radars are difficult to expand to an air and space platform and the defect that the random fluctuation change parameters of the density can only be restricted by singly using the active ultra-wideband surface penetration radar. The method can judge the freezing and thawing state of the interior of the ice and snow through the echo intensity observation value obtained by active remote sensing and the low-frequency-band radiation bright temperature observation value obtained by passive remote sensing, and overcomes the defect that the reflectivity of the ice and snow substrate cannot be accurately estimated because the ice and snow attenuation cannot be effectively restricted by singly using the active remote sensing observation value. And estimating the state distribution of the internal temperature, density, freezing-thawing state and the like of the ice and snow in different spatial ranges and resolutions.

Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention and are not limited. Although the present invention has been described in detail with reference to the embodiments, it will be understood by those skilled in the art that various changes may be made and equivalents may be substituted without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. An active and passive combined microwave remote sensing detection method for ice and snow internal state distribution comprises the following steps:

acquiring an observation value of the change of the ice and snow echo intensity along with time, which is measured by an active ultra-wideband surface penetration radar;

inputting an observation value of the ice and snow echo intensity along with the change of time and an ice and snow radiation brightness temperature value into a trained ice and snow state model to obtain ice and snow internal state distribution data; the ice and snow state model is obtained by adopting a statistical regression inversion algorithm; the ice and snow internal state distribution data comprise ice and snow internal temperature, density and freeze-thaw state;

training the ice and snow state model; the method specifically comprises the following steps:

2. The method for detecting the state distribution in the ice and snow by remote sensing of the active and passive combination of the microwaves according to claim 1, wherein the observation value of the change of the ice and snow echo intensity with time measured by an active ultra-wideband surface penetration radar is obtained; the method specifically comprises the following steps:

performing surface clutter suppression on the ice and snow echo signals;

3. The method for detecting the state distribution in the ice and snow by remote sensing of the active and passive combination of the microwaves according to claim 2, wherein the method is characterized in that the ice and snow radiation brightness temperature value measured by a passive ultra-wideband microwave radiometer is obtained; the method specifically comprises the following steps:

4. An active and passive combined microwave remote sensing detection method for ice and snow internal state distribution comprises the following steps:

5. The method for detecting the state distribution in the ice and snow by remote sensing of the active and passive combination of the microwaves according to claim 4, wherein the observation value of the change of the ice and snow echo intensity with time measured by an active ultra-wideband surface penetration radar is obtained; the method specifically comprises the following steps:

performing surface clutter suppression on the ice and snow echo signals;

6. The method for detecting the state distribution in the ice and snow according to claim 5, wherein the observation value of the ice and snow echo intensity changing with time is converted into the observation value of the ice and snow echo intensity changing with the ice and snow inner depth; the method specifically comprises the following steps:

z_echo＝v_ice·t_echo/2

wherein c is light speed, epsilon'_iceIs the real part of the dielectric constant of ice and snow.

7. The method for detecting the state distribution inside the ice and snow according to claim 6, wherein the vertical distribution information of the dielectric characteristics corresponding to the layered structure inside the ice and snow and the internal reflecting layer is determined according to the observation value that the echo intensity of the ice and snow changes with the depth inside the ice and snow; the method specifically comprises the following steps:

8. The method for detecting the state distribution in the ice and snow by remote sensing of the active and passive combination of the microwaves according to claim 7, wherein the method is characterized in that the ice and snow radiation brightness temperature value measured by a passive ultra-wideband microwave radiometer is obtained; the method specifically comprises the following steps:

and acquiring the ice and snow radiation brightness temperature value measured by the passive ultra-wideband microwave radiometer.