The diffuse attenuation coefficient$K_{d}$continues to play a crucial role in oceanographic research works. Recently, Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) has shown its great ability to estimate$K_{d}$using the water column decay profiles. However, the weak water column backscattered signals are vulnerable to afterpulses and solar background noise, making this way perform not well in the daytime and in nearshore areas. In this study, a method to estimate$K_{d}$is proposed which innovatively uses ICESat-2 bathymetric signal intensities. The main principle is to calculate the attenuation in water column transmission by bathymetric lidar equations. Since the seafloor signal level is much stronger than that of the water column, a significant advantage is the greater noise immunity, i.e., the ability to operate under strong background noise and afterpulses interference. The performance is validated against the moderate-resolution imaging spectroradiometer (MODIS) ocean color measurements with mean relative differences (MRDs) of <32% using both daytime and nighttime ICESat-2 data in six sea and large lake nearshore areas. Based on the new generation of spaceborne lidar data, this study explores a new path to monitor water qualities in nearshore areas. This method is applicable where seafloor photons exist in both daytime and nighttime.

Recently, an increasing number of studies have progressively explored the radiometric properties of ICE, Cloud, and land Elevation Satellite-2 (ICESat-2), enabling this satellite and its payload to provide support for not only geometric but also radiometric applications. However, due to potential changes in the optical throughput and electronics of the instrument during flight, the essential radiometric corrections are needed for quantitative radiometric applications. In this study, using ICESat-2 signal and noise data on icesheet and desert surfaces, which have a relatively stable reflectance and clear sky, a radiometric correction method is proposed to describe the changes in instrument parameters, i.e., to estimate how much the signal and noise should be exactly scaled. With instrument and environmental parameters provided by the ICESat-2 ATL04/ATL09 product, signal and noise models are used to calculate the theoretical signals and noise levels. The theoretical predictions are then compared with the actual measured signals and noise results by ICESat-2 to obtain the scale factors (or radiometric correction factors) for signal and noise, respectively. The results indicate that all three photon counting electronics (PCEs) (corresponding to three laser pairs) exhibit very close scale factor values (i.e., ~1.9), i.e., ICESat-2 receives both signal and noise nearly double the expected values. The noise scale factors$F_{\text {noise}}$over icesheet are expected to yield the most accurate scale factors because$F_{\text {noise}}$will change very little with and without layers. We also analyze the annual average radiometric drift of ICESat-2, which indicates the decreases of ~2.5% ($F_{\text {signal}}$) and ~2.2% ($F_{\text {noise}}$) from 2019 to 2022. This study demonstrates the consistency of signal and noise in radiation and also helps to understand the radiation closure and data fusion between the passive background noise and active laser signal for a spaceborne lidar.

Recently, spaceborne active lidars relying on backscattered laser signal can observe thin clouds, but the laser beam cannot penetrate clouds with large optical thickness. The new generation photon-counting lidar on Ice, Cloud, and land Elevation Satellite-2 (ICESat-2), whose noise can be treated as observations of a green band camera, provides an excellent opportunity to fuse active and passive information to retrieve cloud optical thickness (COT). Clouds significantly increase the background noise and sharply attenuate the signal returning from oceans, which makes it feasible to observe thin and thick clouds by combining active and passive information of ICESat-2. In this study, we first derive the passive background noise and active signal models over oceans for spaceborne lidars, which considers medium contributions from clouds, aerosols, air molecules, ocean surface, and subsurface. ICESat-2 measured surface signals and noise rates in open oceans of Western Pacific are used to verify the models with auxiliary environment datasets. The results indicate that the theoretical predictions have the mean absolute error (MAE) of less than 0.31 counts for signal and the mean absolute percentage error (MAPE) of less than 35% for noise. Then, based on these theoretical models, we propose a COT estimation method combining ICESat-2 active signal and passive noise data without extra auxiliary datasets, and the MAEs between ICESat-2 retrieved COTs and Himawari-8 (H8) cloud products are less than 1.2 (COTs range from 0 to exceeding 20) over open oceans. In general, the proposed method not only expands the observation range of retrieved COTs compared to methods solely relying on signal or noise data but also has great significance for assessing the availability of lidar surface signal, i.e., producing cloud mask.

The new generation photon-counting lidar on Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) can obtain the subsurface optical properties of sea waters. Recent studies highlight the applications of deriving the lidar attenuation coefficient$K_{\mathrm {lidar}}$and then substituting$K_{\mathrm {lidar}}$into the bio-optical model to obtain more information of sea waters. As commonly used bio-optical models are built for the diffuse attenuation coefficient$K_{d}$that are traditionally derived by passive ocean color sensors, whether$K_{\mathrm {lidar}}$derived from ICESat-2 can be directly used as$K_{d}$is a fundamental question. Given that$K_{d}$is an apparent optical property (AOP) in the water column rather than an inherent optical property (IOP),$K_{d}$is closely related to the zenith angle of the incident light. The zenith angle of the incident light of the sunlight is normally tens of degrees for ocean color sensors, while the maximum laser off-nadir angle is ~1.5° for the ICESat-2 lidar. To demonstrate this issue, we select hundreds of ground tracks of ICESat-2 in both open ocean and coastal sea waters and compare the derived$K_{\mathrm {lidar}}$with their corresponding Moderate Resolution Imaging Spectroradiometer (MODIS)-derived$K_{d}$. The results indicate that the corrected results of$1.2\times K_{\mathrm {lidar}}$, instead of the direct results of$K_{\mathrm {lidar}}$, are more consistent with MODIS$K_{d}$. This study is of great significance to the better fusion of active lidar data and passive optical data in ocean observations.

Over the past few decades, spaceborne passive ocean color sensors that measure the solar radiance have provided an unprecedented source of scientific knowledge on marine biology. Recently, spaceborne active lidars that measure the backscattered laser signal from the ocean subsurface provide new insights in deriving vertical profiles of ocean subsurface and obtaining shallow water bathymetry. The solar radiation is the signal source of passive ocean color sensors but acts as the primary noise source of satellite-based lidars, which may limit the extraction and application of the weak subaqueous signal in the daytime. Based on the perspective of the reciprocity, the background noise of the ICESat-2 spaceborne photon-counting lidar in six channels (or pixels) has potential to be regarded as the signal of an “ocean color camera” with a very narrow band. In this study, the remote sensing reflectance, that is the fundamental data of ocean color sensors, is theoretically linked to and accurately transferred from ICESat-2 noise data with an average Mean Absolute Percentage Error (MAPE) of less than 20% compared to thein-situmeasurements. With this method, not only the remote sensing reflectanceRrscan be retrieved from ICESat-2 under strong background noise, which enhances the capability of ICESat-2 to monitor the diurnal variation, but also numerous quantitative applications by passive remote sensing sensors may be achievable by the noise data of spaceborne photon-counting lidars in the future. In addition, a spaceborne lidar can synchronously detect active laser signal and passive solar radiation, which may bring new insights in the data fusion and verification of active and passive techniques.