Method for determining carbon dioxide column concentration based on lidar cloud echo signal

Abstract

The present invention discloses a method for determining the columnar concentration of carbon dioxide based on a lidar cloud echo signal, belonging to the field of atmospheric laser remote sensing technology. The method can effectively extract cloud echo signals from echo signals containing reflections from multiple hard targets and calculate the columnar concentration of carbon dioxide above the cloud. The method includes the following steps: preprocessing the original path-integrated differential absorption lidar (IPDA lidar) data to extract auxiliary data such as time, position, payload attitude angle, and velocity; extracting signal features from the signal data after removing background noise, and determining the signal peak position using a screening method based on the median absolute deviation; using the signal peak position to solve the hard target's height from sea level; calculating the distance difference between each echo, matching signal points using a height threshold, and then performing consistency processing on the integral range of the matched echo signals and solving the integral value; filtering the matched signals using statistical features and a height threshold to obtain valid cloud echo signals; solving the differential optical depth (DAOD) using the cloud's online and offline echo signals; calculating the integral weight function (IWF) using temperature, humidity, and pressure data combined with Doppler shift correction; and finally calculating the columnar concentration of carbon dioxide above the cloud (XCO2).

Description

Technical Field

[0001] The present invention belongs to the technical field of atmospheric laser remote sensing, and in particular relates to a method for determining carbon dioxide column concentration based on an IPDA laser radar cloud echo signal. Background Art

[0002] The 2016 Paris Agreement set a long-term global atmospheric temperature target of limiting global average warming to well below 2°C relative to pre-industrial levels, with the goal of limiting warming to 1.5°C by the end of the 21st century. To meet the needs of countries for effective carbon emissions and accounting management, lidar detection based on the principle of path-integrated differential absorption has been proposed, building upon existing technologies. Researchers have also begun developing IPDA lidar systems and conducting research on CO2 concentration inversion.

[0003] Because lasers can penetrate clouds, IPDA lidar can simultaneously obtain echo signals from both the ground and clouds. Existing data processing studies often use the maximum value method to extract signal peaks, which only captures the strongest echo signals. With global cloud coverage reaching up to 60%, effectively extracting and separating cloud echo signals can improve data utilization. Furthermore, the lack of continuity between the cloud layer and the ground, coupled with the increased complexity of cloud echo signals, requires targeted processing and analysis. In this paper, we propose an effective method for extracting cloud echo signals and develop a high-precision inversion method for columnar CO2 concentrations derived from cloud echoes. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to effectively extract the cloud echo signal detected by the IPDA lidar, distinguish multiple layers of clouds and improve data utilization, and use the cloud echo signal to invert the columnar concentration of carbon dioxide above the cloud.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] Step 1. Use the path-integrated differential absorption lidar system to obtain the echo signals of the ground and clouds;

[0007] Step 2: Preprocessing the echo signals of the ground and cloud layers, including:

[0008] Extract auxiliary data segments to obtain time information, position information, payload flight information, as well as monitoring signals and echo signals;

[0009] Step 3. Signal denoising: Remove the background noise of the online and offline signal segments respectively, and perform local weighted quadratic regression smoothing denoising to obtain the denoised signal;

[0010] Step 4. Extracting Signal Features: All minimum points of the denoised signal are used as a dataset of signal points to be screened. This dataset is screened using an outlier screening method based on the median absolute deviation. The peak points of all signals are extracted. The signal points of the online and offline echo signal segments are matched. The integration ranges of the matched echo signals are processed for consistency to obtain the integral value. The peak position and voltage, integration interval and voltage of the signal are obtained. The pre-trigger data points of the online and offline signal segments are respectively taken as noise data.

[0011] Step 5. Calculate the height of the target from the sea level based on the peak position of the signal;

[0012] Step 6. Use statistical features and height thresholds to filter the matched signals and obtain valid cloud echo signals;

[0013] Step 7. Calculate the differential absorption optical depth (DAOD) based on the voltage integration values of the cloud's online and offline echo signals;

[0014] Step 8: Calculate the integral weight function (IWF) using temperature, humidity, and pressure meteorological data combined with Doppler shift correction;

[0015] Step 9: Based on the differential absorption optical depth obtained in step 7 and the integral weight function obtained in step 8, calculate the carbon dioxide column concentration (XCO2) above the cloud and perform multi-point averaging of the carbon dioxide column concentration according to different resolution requirements.

[0016] Preferably, the outlier screening method based on the median absolute deviation used in the signal extraction in step 4 is to screen out the points in the signal point sample that differ from the median by more than three times the converted median absolute deviation, namely, the echo signal points.

[0017] The median absolute deviation (MAD) is defined as:

[0018] MAD=median(|X i -meadian(X i )|)i=1,...,n

[0019] Among them, median(X) represents the median of the data, X i Represents the data point. When MAD is used as a consistent estimator of the standard deviation σ, σ=k·MAD, k is the proportional factor constant. For normally distributed data, k≈1.4826. That is, the echo signal point satisfies:

[0020] X i >MAD+3·σ

[0021] X i <MAD-3·σ

[0022] The integration range consistency processing described in step 4 means that the peak positions of the corresponding online and offline echo signals and the number of signal points taken before and after the peaks are consistent.

[0023] The height H (m) of the hard target from sea level described in step 5 is calculated by the following formula:

[0024] H=hD·cosγ p ·cosγ R

[0025] Where, h is the flight height of the payload (m), γ p is the pitch angle, γ R is the roll angle, and D is the distance from the payload to the hard target (m), which is given by the following formula:

[0026] D=0.5×(d on +d off )

[0027] Among them, d on and d off are the distances (m) from the payload to the hard target center of mass calculated using the online pulse and offline pulse, respectively, and are given by:

[0028]

[0029] Where c = 3 × 10 8 m / s is the speed of light, id on / off The peak position of the Online / Offline echo signal, id 0 on / off The peak position of the online / offline monitoring signal. The data sampling rate is F s MS / s, so the interval between two adjacent signal acquisition points is

[0030] Preferably, the height threshold in step 6 is that the difference between the hard target height H from sea level and the elevation at the corresponding longitude and latitude in the digital elevation model (DEM) is greater than 300m.

[0031] Preferably, the differential absorption optical depth (DAOD) in step 7 is calculated by the following formula:

[0032]

[0033] Among them, (P on ) i and (P off )i are respectively the power of the i-th point on the echo signal selected online and offline after the integration range consistency processing according to claim 2, (P 0 on ) i and (P 0 off ) i are respectively the power of the i-th point on the monitoring signal selected online and offline after the integration range consistency processing in claim 2.

[0034] Preferably, the integral weight function (IWF) in step 8 is calculated by the following formula:

[0035]

[0036] Where NA is Avogadro's constant, P is atmospheric pressure, T is atmospheric temperature, Δν is the frequency shift caused by the Doppler effect, σ is the absorption cross section of carbon dioxide molecules at r, and R i is the ideal gas constant, is the H2O dry air mixture ratio.

[0037] As the preferred step 9, the carbon dioxide column concentration (XCO2) is calculated by the following formula:

[0038]

[0039] The present invention has the following beneficial effects:

[0040] 1) The outlier screening and extraction method of the median absolute deviation is used to effectively separate the cloud echo signals from multiple echo signals of the same set of monitoring signals, thereby improving data utilization;

[0041] 2) Combined with IPDA lidar detection data, the carbon dioxide column concentration above the cloud is inverted. When there are multiple layers of clouds under the same path, the carbon dioxide column concentration of the cloud layer structure is obtained to make the data more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 is a flow chart of an embodiment of a method for determining the columnar concentration of carbon dioxide based on a lidar cloud echo signal according to the present invention;

[0043] Figure 2 2 is a schematic diagram of echo signal height calculation according to an embodiment of the present invention. DETAILED DESCRIPTION

[0044] The embodiments of the present invention are described in detail below with reference to the accompanying drawings. However, the present invention can be implemented in many different ways as defined and covered by the claims.

[0045] See also Figure 1 The method of determining the carbon dioxide column concentration based on the laser radar cloud echo signal of the present invention comprises the following steps:

[0046] Step 1. Use a 1572nm path-integrated differential absorption lidar system to obtain echo signals from the ground and clouds;

[0047] Step 2. Preprocess the signals collected by the 1572nm path-integrated differential absorption lidar system to extract auxiliary data segments containing time, longitude and latitude, flight attitude angles (yaw angle, pitch angle, roll angle), flight altitude, and flight speed (three-dimensional speed), as well as monitoring signal and echo signal data segments containing online and offline wavelengths.

[0048] Step 3. Signal denoising: Remove the background noise of the online and offline signal segments respectively, and perform local weighted quadratic regression smoothing denoising to obtain the denoised signal;

[0049] Step 4. Extraction of signal features: All the minimum points of the denoised signal are used as the data set of signal points to be screened. This data set is screened using the outlier screening method based on the median absolute deviation to extract the peak points of all signals. The median absolute deviation (MAD) is defined as:

[0050] MAD=median(|X i -meadian(X i )|)i=1,...,n

[0051] Among them, median(X) represents the median of the data, X i The method is to select the points in the signal point sample whose difference from the median is more than three times the absolute deviation of the converted median, which is the echo signal point, that is, satisfying:

[0052] X i >MAD+3·σ

[0053] X i <MAD-3·σ

[0054] When MAD is used as a consistent estimator of the standard deviation σ, σ = k·MAD, where k is the proportional factor constant. For normally distributed data, k≈1.4826.

[0055] The signal points of the online and offline echo signal segments are then matched. The integration range of the matched echo signals is then processed for consistency, and the integral value is calculated. Consistency here means that the peak positions and the number of signal points before and after the peaks of the corresponding online and offline echo signals are consistent. Pre-trigger data points of the online and offline signal segments are used as noise.

[0056] Step 5. Calculate the height H (m) of the hard target from the sea level using the extracted signal peak position, see Figure 2 . It is calculated by the following formula:

[0057] H=hD·cosγ p ·cosγ R

[0058] Among them, h is the flight altitude of the aircraft (m), γ p is the pitch angle, γ R is the roll angle, and D is the distance from the payload to the hard target (m), which is given by the following formula:

[0059] D=0.5×(d on +d off )

[0060] Among them, d on and d off are the distances (m) from the load to the target center of mass calculated using the online pulse and the offline pulse, respectively, and are given by the following formula:

[0061]

[0062] Where c = 3 × 10 8 m / s is the speed of light, id on / off The peak position of the Online / Offline echo signal, id 0 on / off The peak position of the signal is monitored for Online / Offline. The data sampling rate is 125MS / s, so the interval between two adjacent signal acquisition points is

[0063] Step 6. Use statistical features and height thresholds to filter the matched signals to obtain valid cloud echo signals; the height threshold is that the difference between the hard target height H from sea level and the elevation at the corresponding latitude and longitude in the digital elevation model (DEM) is greater than 300m.

[0064] Step 7. Calculate the differential absorption optical depth (DAOD) based on the voltage integration values of the cloud's online and offline echo signals. The differential absorption optical depth (DAOD) is calculated using the following formula:

[0065]

[0066] Among them, (P on ) i and (P off ) i are respectively the power of the i-th point on the echo signal selected online and offline after the integration range consistency processing according to claim 2, (P 0 on ) i and (P 0 off ) i are respectively the power of the i-th point on the monitoring signal selected online and offline after the integration range consistency processing in claim 2.

[0067] Step 8: Calculate the integral weight function (IWF) using the temperature, humidity, and pressure meteorological data combined with Doppler shift correction. The temperature, humidity, and pressure data can be found and used based on actual conditions. The integral weight function (IWF) is calculated as follows:

[0068]

[0069] Where NA is Avogadro's constant, P is atmospheric pressure, T is atmospheric temperature, Δν is the frequency shift caused by the Doppler effect, σ is the absorption cross section of carbon dioxide molecules at r, and R i is the ideal gas constant, is the H2O dry air mixture ratio.

[0070] Step 9: Based on the differential absorption optical depth obtained in step 7 and the integral weight function obtained in step 8, calculate the carbon dioxide column concentration (XCO2) above the cloud and perform multi-point averaging of the carbon dioxide column concentration according to different resolution requirements. The carbon dioxide column concentration (XCO2) is calculated by the following formula:

[0071]

[0072] Experiments have shown that this method, through its use of a discrete value screening method based on the median absolute deviation, extracts signals, avoiding the loss of valid signals when multiple echo signals coexist and the inaccurate extraction of valid signals at low signal-to-noise ratios. This method achieves higher data utilization than the minimum value method. The method can obtain the columnar CO2 concentration from multi-layer cloud structures to the load. Compared with in-situ measurements, the results show consistent trends in CO2 column concentration variation across regions and altitudes. The effective and accurate acquisition of CO2 concentration above the cloud increases CO2 concentration data coverage and facilitates multidimensional analysis of CO2 source and sink distribution.

Claims

1. A method for determining the columnar concentration of carbon dioxide based on a lidar cloud echo signal, characterized in that: The following steps are involved: Step 1. Use the path-integrated differential absorption lidar system to obtain the echo signals of the ground and clouds; Step 2: Preprocessing the echo signals of the ground and cloud layers, including: Extract auxiliary data segments to obtain time information, position information, payload flight information, as well as monitoring signals and echo signals; Step 3. Signal denoising: Remove the background noise of the online and offline signal segments respectively, and perform local weighted quadratic regression smoothing denoising to obtain the denoised signal; Step 4. Extracting Signal Features: All minimum points of the denoised signal are used as a dataset of signal points to be screened. This dataset is screened using an outlier screening method based on the median absolute deviation. The peak points of all signals are extracted. The signal points of the online and offline echo signal segments are matched. The integration ranges of the matched echo signals are processed for consistency to obtain the integral value. The peak position and voltage, integration interval and voltage of the signal are obtained. The pre-trigger data points of the online and offline signal segments are respectively taken as noise data. Step 5. Calculate the height of the target from the sea level based on the peak position of the signal; Step 6. Use statistical features and height thresholds to filter the matched signals and obtain valid cloud echo signals; Step 7. Calculate the differential absorption optical depth (DAOD) based on the voltage integration values of the cloud's online and offline echo signals; Step 8: Calculate the integral weight function (IWF) using temperature, humidity, and pressure meteorological data combined with Doppler shift correction; Step 9: Calculate the carbon dioxide column concentration (XCO2) above the cloud based on the differential absorption optical depth obtained in step 7 and the integral weight function obtained in step 8. Perform multi-point averaging of the carbon dioxide column concentration based on different resolution requirements. In step 4, the outlier screening method based on median absolute deviation is used to screen the data set to be screened. That is, the points in the signal point samples that differ from the median by more than three times the converted median absolute deviation are selected as echo signal points. The median absolute deviation (MAD) is defined as: MAD=median(|X i -meadian(X i )|)i=1,...,n Among them, median(X) represents the median of the data, X i represents the data point; the specific method is to filter out the points in the signal point sample whose difference from the median is more than three times the converted median absolute deviation, which is the echo signal point, that is, satisfying: X i >MAD+3·s X i <MAD-3·σ When MAD is used as a consistent estimator of the standard deviation σ, σ = k·MAD, where k is the proportional factor constant. For normally distributed data, k≈1.4826; In step 5, the height H of the target from the sea level is calculated based on the peak position of the signal, and the formula is as follows: H=h-D·cosγ p ·cosγ R D=0.5×(d on +d off ) Where h is the flight height of the payload, γ p is the flight pitch angle, γ R is the flight roll angle, D is the distance from the payload to the target, d on and d off are the distances from the load to the target center of mass calculated using the Online pulse and the Offline pulse, respectively.

2. The method for determining the carbon dioxide column concentration based on the laser radar cloud echo signal according to claim 1, characterized in that: The consistency processing of the integration range described in step 4 means that the peak positions of the corresponding online and offline echo signals and the number of signal points taken before and after the peak are consistent.

3. The method for determining the carbon dioxide column concentration based on the laser radar cloud echo signal according to claim 1, characterized in that: The height threshold in step 6 is set according to the difference between the target height H from sea level and the elevation at the corresponding longitude and latitude in the digital elevation model (DEM).

4. The method for determining the carbon dioxide column concentration based on the laser radar cloud echo signal according to claim 1, characterized in that: Step 7: Calculate the differential absorption optical depth (DAOD) based on the voltage integration values of the cloud's online and offline echo signals. The formula is as follows: Among them, (P on ) i and (P off ) i are the power of the i-th point on the echo signal selected online and offline after the integration range consistency processing, (P 0 on ) i and (P 0 off ) i are the powers of the i-th point on the monitoring signal selected online and offline after integration range consistency processing.

5. The method for determining the carbon dioxide column concentration based on the laser radar cloud echo signal according to claim 4, characterized in that: Step 8 uses the temperature, humidity, and pressure meteorological data combined with Doppler shift correction to calculate the integral weight function (IWF). The formula is as follows: Where NA is Avogadro's constant, P is atmospheric pressure, T is atmospheric temperature, Δν is the frequency shift caused by the Doppler effect, σ is the absorption cross section of carbon dioxide molecules at r, and R i is the ideal gas constant, is the H2O dry air mixture ratio.

6. The method for determining the carbon dioxide column concentration based on the laser radar cloud echo signal according to claim 5, characterized in that: In step 9, the columnar concentration of carbon dioxide above the cloud (XCO2) is calculated using the following formula: 。 7. The method for determining the carbon dioxide column concentration based on the laser radar cloud echo signal according to claim 1, characterized in that: The payload flight information includes flight attitude angle, flight altitude, and flight speed.

Images