Since the industrial revolution, anthropogenic activities have resulted in unprecedented carbon emissions that exceed the carbon sink capacity of terrestrial and marine ecosystems, with CO₂ concentrations increasing by approximately 30% over the last few decades to 418 μL/L by 2022. To meet the needs to effectively implement carbon emission management, lidar detection based on the differential absorption principle has been proposed using the existing technology. With an average global cloud coverage of up to 60%, there are many cloud echo signals in addition to ground and ocean echo signals when laser penetrates the atmosphere to the ground, and effective use of cloud echo signals can improve data utilization and contribute to the analysis of carbon sources and sinks. For complex cloud echo signals, an outlier screening method based on the absolute deviation from the median is proposed to extract signals, which can separate multi-layer cloud echo signals and the signals in which the cloud and ground echo signals co-exist. In addition, this paper analyzes the detection capability of cloud signals, studies effective data processing methods for cloud echo signals to calculate the CO₂ column concentration on the cloud, and compares the results with the data obtained with in-situ instrument.

This paper uses data from an airborne flight experiment conducted by the integrated path differential absorption (IPDA) lidar system in Qinhuangdao in March 2019. Firstly, the types of signals that may be received by the IPDA lidar are analyzed, and an outlier screening method based on the absolute deviation from the median is proposed to extract the echo signals. A correction method for the target altitudes is proposed, and the extraction results are compared with those extracted using the traditional minimum method. Secondly, the online and offline monitor signals and echo signals of the clouds are analyzed, and data suitable for CO₂ column concentration inversion are selected. Then, the relative reflectance of the clouds is calculated using the offline monitor signals and echo signals. The relationship between the relative reflectance and cloud density is investigated by combining the cloud density distribution with the cloud images taken by the airborne camera. Finally, the CO₂ column concentration on the clouds is obtained by using the differential optical thickness and the integral weighting function corrected for Doppler shift, and the results are compared with the trends of single-point CO₂ concentrations measured by in-situ instruments. Meanwhile, the relationship between cloud altitude and CO₂ column concentration on the clouds is also investigated.

According to the signal extraction results, the amount of valid cloud signal extracted using the outlier screening method based on the absolute deviation from the median is 1.9 times greater than that extracted using the minimum value method (Fig.10). The relative reflectance of the clouds is obtained by the offline monitor signals energy and the echo signals energy. Based on the experiment, the relative reflectance of the clouds over the mountainous area is 0.1897, that over the residential area is 0.1418, and that over the sea is 0.1656 (Fig.12). The number of signal points in the time-altitude grid area is counted to get the cloud density distribution, and combined with the photographs taken by the airborne camera (Figs.13 and 14), the cloud tops are found to be around 3400 m for thin clouds and 3800 m for thick clouds on the day of the experiment. We use differential optical depth and the integral weighting function corrected for Doppler shift to calculate the concentrations of CO₂ on clouds (Fig.15). On the day of the experiment, the average CO₂ column concentration over the cloud is 415.98 μL/L, the average CO₂ column concentration over the residential area is 416.96 μL/L, and the average CO₂ column concentration over the sea area is 413.92 μL/L, with an overall average value of 416.23 μL/L. The trend is consistent with those of the single-point CO₂ concentrations measured by the in-situ instrument. The altitude variation of CO₂ column concentrations over clouds (Fig.16) is calculated to show the distribution of CO₂ column concentrations over clouds with altitude in different regions.

In this paper, cloud echo signals from an airborne IPDA lidar are investigated to obtain the CO₂ column concentration. An outlier screening method based on the absolute deviation from the median to extract cloud echo signals is proposed, which can improve the effective utilization of the data measured by the airborne platforms. The stable cloud echo signals from the level flight phase of the airborne experiment are selected for analysis. The offline wavelength echo signals are used to calculate the relative reflectance of the clouds, and the different spatial distributions of the clouds and their reflectance variations in the experiment are analyzed. The CO₂ column concentration data on the clouds are analyzed and compared with the data from the onboard in-situ carbon dioxide instrument equipped on the same aircraft. The CO₂ concentration trends measured by the two methods are in good agreement, with an average deviation of 2.8 μL/L. The research conducted in this paper provides an important reference for processing the cloud echo signals of the spaceborne lidar to calculate the CO₂ concentration.