ICESat Research Articles

A Multi-Level Auto-Adaptive Noise-Filtering Algorithm for Land ICESat-2 Photon-Counting Data

Due to atmospheric scattering, solar radiation, and other factors, the Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) product data suffer from a substantial amount of background noise. This poses a significant challenge when attempting to directly utilize the raw data. Consequently, data denoising becomes an indispensable preprocessing step for its subsequent applications, such as the extraction of forest structure parameters and ground elevation data. While the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm is currently the most widely used method, it remains susceptible to complexities arising from terrain, low signal-to-noise ratio (SNR), and input parameter variations. This paper proposes an efficient Multi-Level Auto-Adaptive Noise Filter (MLANF) algorithm based on photon spatial density. Its purpose is to extract signal photons from ICESat-2 terrestrial data of different ground cover types. The algorithm follows a two-step process. Firstly, random noise photons are removed from the upper and lower regions of the signal photons through a coarse denoising process. Secondly, in the fine denoising step, the K-Nearest Neighbor (KNN) algorithm selects the K photons to calculate the slope along the track. The calculated slope is then used to rotate the direction of the searching neighborhood in the DBSCAN algorithm. The proposed algorithm was tested in eight datasets of four surface types: forest, grassland, desert, and urban, and the extraction results were compared with those from the ATL08 datasets and the DBSCAN algorithm. Based on the ground-truth signal photons obtained by visual inspection, the classification precision, recall, and F-score of our algorithm, as well as two other algorithms, were calculated. The MLANF could achieve a good balance between classification precision (97.48% averaged) and recall (97.96% averaged). Its F-score (97.69% averaged) was higher than that of the other two methods. This demonstrates that the MLANF algorithm successfully obtained a continuous surface profile from ICESat-2 datasets with different surface cover types, significant topographic relief, and low SNR.

Signal Photon Extraction Method for ICESat-2 Data Using Slope and Elevation Information Provided by Stereo Images

Laser altimetry data from the Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) contain a lot of noise, which necessitates the requirement for a signal photon extraction method. In this study, we propose a density clustering method, which combines slope and elevation information from optical stereo images and adaptively adjusts the neighborhood search direction in the along-track direction. The local classification density threshold was calculated adaptively according to the uneven spatial distribution of noise and signal density, and reliable surface signal points were extracted. The performance of the algorithm was validated for strong and weak beam laser altimetry data using optical stereo images with different resolutions and positioning accuracies. The results were compared qualitatively and quantitatively with those obtained using the ATL08 algorithm. The signal extraction quality was better than that of the ATL08 algorithm for steep slope and low signal-to-noise ratio (SNR) regions. The proposed method can better balance the relationship between recall and precision, and its F1-score was higher than that of the ATL08 algorithm. The method can accurately extract continuous and reliable surface signals for both strong and weak beams among different terrains and land cover types.

Background noise model of spaceborne photon-counting lidars over oceans and aerosol optical depth retrieval from ICESat-2 noise data

The ICESat-2 (Ice, Cloud, and land Elevation Satellite-2) photon-counting lidar is a revolutionary active remote sensing device and is extremely sensitive to both laser signal and solar background induced noise photons. The weak laser signals, such as aerosol and water subsurface backscattered laser photons, may be buried in noise in the daytime. A noise model for spaceborne photon-counting lidars is of vital importance in understanding and evaluating the performance of a lidar system, as well as in data processing and applications. From the reciprocity perspective, the main noise source for photon-counting lidars is solar background radiation, which in turn is the signal source of ocean color sensors. In the absence of clouds, it is possible to estimate the aerosol information using background noise data of ICESat-2 if the background noise is accurately modeled. In this study, a background noise model of photon-counting lidars over oceans is proposed that carefully considers the contributions of gas molecules, aerosol, water surface, subsurface, and detector dark noise. As the background rate from clouds are much larger than the sum of other factors, in this study, we only consider the cloudless atmosphere by using a cloud screening process. By inputting the system and environmental parameters, the theoretical predictions of noise level in six study areas have an average MAPE of ∼5% compared with ICESat-2 measured noise data. The results also indicate that without clouds, the variation of aerosols dominates the variation of ICESat-2 total background noise in open ocean areas. Then, based on this noise model, a method to estimate aerosol optical depths (AODs) over oceans is proposed solely on the ICESat-2 product including background noise rate and environmental data. Validation against MODIS (Moderate Resolution Imaging Spectroradiometer) daily AODs indicates that the average MAPE is <20% and the average correlation coefficient is 0.95 in six study areas. This method may provide a new AOD retrieval way over oceans in the daytime, when it is difficult to directly calculate AODs using lidar backscattered signal under strong background noise. In addition, the background noise of the ICESat-2 photon-counting lidar in its six channels (or pixels) has the potential to be regarded as the signal of an “ocean color camera” with only a single narrow green band. Based on this consideration, future works may be conducted to combine ocean color remote sensing with ICESat-2 photon-counting lidar.

Denoising and classification of urban ICESat-2 photon data fused with Sentinel-2 spectral images

ABSTRACTThe ICESat-2 (Ice, Cloud, and Land Elevation Satellite-2) can collect earth surface elevation data with high precision on a global scale. However, the collected photon data contains a large amount of background noise due to the influence of sunlight, cloud reflection, and other factors. For photon data of different scenes, how to effectively denoise and achieve accurate classification of photon point clouds is crucial for subsequent applications. This study proposes a random forest based method for denoising and classifying ICESat-2 photon data in urban areas by fusing spectral features from Sentinel-2 images and spatial distribution features from photon data. The experimental results show that the method can effectively identify various types of photons. Compared with the reference data, the overall accuracy of photon denoising and classification is 95.97% on average, and the average kappa coefficient is 94.18%. Further analysis demonstrates that the addition of sentinel-2 spectral information can effectively improve the classification accuracy of photon point clouds in urban areas, and the photon classification method of combining photon lidar data and optical images can be a promising solution to improve classification accuracy.

Evaluation and Comparison of ICESat-2 and GEDI Data for Terrain and Canopy Height Retrievals in Short-Stature Vegetation

Two space-borne light detection and ranging (LiDAR) missions, Global Ecosystem Dynamics Investigation (GEDI) and Ice, Cloud, and land Elevation Satellite-2 (ICESat-2), have demonstrated high capabilities in extracting terrain and canopy heights in forest environments. However, there have been limited studies evaluating their performance for terrain and canopy height retrievals in short-stature vegetation. This study utilizes airborne LiDAR data to validate and compare the accuracies of terrain and canopy height retrievals for short-stature vegetation using the latest versions of ICESat-2 (Version 5) and GEDI (Version 2). Furthermore, this study also analyzes the influence of various factors, such as vegetation type, terrain slope, canopy height, and canopy cover, on terrain and canopy height retrievals. The results indicate that ICESat-2 (bias = −0.05 m, RMSE = 0.67 m) outperforms GEDI (bias = 0.39 m, RMSE = 1.40 m) in terrain height extraction, with similar results observed for canopy height retrievals from both missions. Additionally, the findings reveal significant differences in terrain and canopy height retrieval accuracies between ICESat-2 and GEDI data under different data acquisition scenarios. Error analysis results demonstrate that terrain slope plays a pivotal role in influencing the accuracy of terrain height extraction for both missions, particularly for GEDI data, where the terrain height accuracy decreases significantly with increasing terrain slope. However, canopy height has the most substantial impact on the estimation accuracies of GEDI and ICESat-2 canopy heights. Overall, these findings confirm the strong potential of ICESat-2 data for terrain and canopy height retrievals in short-stature vegetation areas, and also provide valuable insights for future applications of space-borne LiDAR data in short-stature vegetation-dominated ecosystems.

On the capacity of ICESat-2 laser altimetry for river level retrieval: An investigation in the Ohio River basin

The Advanced Topographic Laser Altimeter System (ATLAS) onboard the Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2), with a footprint of 17 m and an along-track spacing of 0.7 m, is expected to improve our ability to observe small inland water bodies. The objective of this study is to evaluate the capacity of ICESat-2 for water level retrieval of medium (<1000 m in width) to small (<50 m) rivers. A water level retrieval algorithm for small rivers is proposed based on geolocated photon data to fill gaps in inland water surface height product (ATL13). Validation results indicate that the retrieved river levels are consistent with in situ measurements (RMSE = 0.24 m, R = 0.99, n = 69), and in excellent agreement with those from ATL13 (RMSE = 0.12 m, R = 1.00, n = 187). The capacity of ICESat-2 is evaluated at 186 virtual stations (VSs) over rivers from 6 to 950 m wide in the Ohio River basin, USA. Water levels are retrieved for small and medium rivers using the proposed method and from ATL13, respectively. Terrain correction is carried out to relocate the VS to the gauge station so that pairs of river levels agree in position, enabling the comparison. The results show that the accuracy of river level estimates is significantly improved after terrain correction, with RMSE decreasing from 0.44 to 0.17 m, based on 186 VSs. The results reveal that ICESat-2 can resolve and measure rivers as narrow as 6 m. RMSE of retrieved river levels ranges between 0.19 and 0.25 m for rivers narrower than 50 m, and from 0.08 to 0.15 m for those wider than 50 m, with averages of 0.24 and 0.12 m, respectively. Moreover, our results suggest that river width and land cover type do not impact the performance of altimetric estimates in the study area. Instead, the validation method is likely to deteriorate accuracy metrics, and the accuracy of retrievals is also dependent on fluvial morphology. This is the first study to explore the potential of ICESat-2 to retrieve river levels, which demonstrates its superior capacity to accurately measure rivers just down to a few meters wide, providing a new and exciting asset to hydrological studies. The findings of this work will also contribute to enhancing our understanding of inland water behavior in global hydrological processes and water resource management.

Modes of Antarctic tidal grounding line migration revealed by Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) laser altimetry

Abstract. Tide-forced short-term migration of the grounding line (GL) of Antarctic ice shelves can impact ice dynamics at the ice sheet margins and obscures assessments of long-term GL advance or retreat. However, the magnitude of tidally induced GL migration is poorly known, and the spatial patterns and modes of variability are not well characterised. Here we develop and apply a technique that uses Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) repeat-track laser altimetry to locate the inland limit of tidal ice shelf flexure for each sampled tide, enabling the magnitude and temporal variability of tidal GL migration to be resolved. We demonstrate its application at an ice plain north of Bungenstockrücken, in a region of the southern Ronne Ice Shelf subject to large ocean tides. We observe a 1300 km2 area of ephemeral grounding over which the GL migrates by up to 15 km between low and high tide and identify four distinct modes of migration: linear, asymmetric, threshold and hysteresis. The short-term movement of the GL dominates any long-term migration signal in this location, and the distribution of GL positions and modes contains information about spatial variability in the ice–bed interface. We discuss the impact of extreme tidal GL migration on ice shelf–ocean–subglacial systems in Antarctica and make recommendations for how GLs should be more precisely defined and documented in future by the community.

New estimates of pan-Arctic sea ice–atmosphere neutral drag coefficients from ICESat-2 elevation data

Abstract. The effect that sea ice topography has on the momentum transfer between ice and atmosphere is not fully quantified due to the vast extent of the Arctic and limitations of current measurement techniques. Here we present a method to estimate pan-Arctic momentum transfer via a parameterization that links sea ice–atmosphere form drag coefficients with surface feature height and spacing. We measure these sea ice surface feature parameters using the Ice, Cloud and land Elevation Satellite-2 (ICESat-2). Though ICESat-2 is unable to resolve as well as airborne surveys, it has a higher along-track spatial resolution than other contemporary altimeter satellites. As some narrow obstacles are effectively smoothed out by the ICESat-2 ATL07 spatial resolution, we use near-coincident high-resolution Airborne Topographic Mapper (ATM) elevation data from NASA's Operation IceBridge (OIB) mission to scale up the regional ICESat-2 drag estimates. By also incorporating drag due to open water, floe edges and sea ice skin drag, we produced a time series of average total pan-Arctic neutral atmospheric drag coefficient estimates from November 2018 to May 2022. Here we have observed its temporal evolution to be unique and not directly tied to sea ice extent. By also mapping 3-month aggregates for the years 2019, 2020 and 2021 for better regional analysis, we found the thick multiyear ice area directly north of the Canadian Archipelago and Greenland to be consistently above 2.0×10-3, while most of the multiyear ice portion of the Arctic is typically around ∼1.5×10-3.

A Comprehensive Evaluation of Machine Learning and Classical Approaches for Spaceborne Active-Passive Fusion Bathymetry of Coral Reefs

Satellite-derived bathymetry (SDB) techniques are increasingly valuable for deriving high-quality bathymetric maps of coral reefs. Investigating the performance of the related SDB algorithms in purely spaceborne active–passive fusion bathymetry contributes to formulating reliable bathymetric strategies, particularly for areas such as the Spratly Islands, where in situ observations are exceptionally scarce. In this study, we took Anda Reef as a case study and evaluated the performance of eight common SDB approaches by integrating Sentinel-2 images with Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2). The bathymetric maps were generated using two classical and six machine-learning algorithms, which were then validated with measured sonar data. The results illustrated that all models accurately estimated the depth of coral reefs in the 0–20 m range. The classical algorithms (Lyzenga and Stumpf) exhibited a mean absolute error (MAE), root mean square error (RMSE), and mean absolute percentage error (MAPE) of less than 0.990 m, 1.386 m, and 11.173%, respectively. The machine learning algorithms generally outperformed the classical algorithms in accuracy and bathymetric detail, with a coefficient of determination (R2) ranging from 0.94 to 0.96 and an RMSE ranging from 1.034 m to 1.202 m. The multilayer perceptron (MLP) achieved the highest accuracy and consistency with an RMSE of as low as 1.034 m, followed by the k-nearest neighbor (KNN) (1.070 m). Our results provide a practical reference for selecting SDB algorithms to accurately obtain shallow water bathymetry in subsequent studies.

DORSL-FIN: A Self-supervised Neural Network for Recovering Missing Bathymetry from ICESat-2

Bathymetric data, comprising elevations of submerged surfaces (e. g., seafloor or lake bed), constitute a critical need for a wide range of science and application focus areas, such as safety of marine navi- gation, benthic habitat mapping, flood inundation modeling, and coastal engineering. Over the past decade, the availability of near- shore bathymetric data has increased dramatically due to advances in satellite-derived bathymetry (SDB). One notable advance occurred with the 2018 launch of NASA's Ice, Cloud, and land Elevation Satellite 2 (ICESat-2), carrying the Advanced Topographic Laser Altimeter System (ATLAS). However, much like other Earth observing satellites, ATLAS is often hampered by obstructions, such as clouds, which block the sensor's view of the Earth's surface. In this study, we introduce the Deep Occlusion Recovery of Satellite Lidar From ICESat-2 Network (DORSL-FIN) to recover partially occluded bathymetric profiles. We show that DORSL-FIN is able to accurately recover occluded bathymetry and outperforms other methods of interpolation.

Observing the evolution of summer melt on multiyear sea ice with ICESat-2 and Sentinel-2

Abstract. We investigate sea ice conditions during the 2020 melt season, when warm air temperature anomalies in spring led to early melt onset, an extended melt season, and the second-lowest September minimum Arctic ice extent observed. We focus on the region of the most persistent ice cover and examine melt pond depth retrieved from Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) using two distinct algorithms in concert with a time series of melt pond fraction and ice concentration derived from Sentinel-2 imagery to obtain insights about the melting ice surface in three dimensions. We find the melt pond fraction derived from Sentinel-2 in the study region increased rapidly in June, with the mean melt pond fraction peaking at 16 % ± 6 % on 24 June 2020, followed by a slow decrease to 8 % ± 6 % by 3 July, and remained below 10 % for the remainder of the season through 15 September. Sea ice concentration was consistently high (&gt;95 %) at the beginning of the melt season until 4 July, and as floes disintegrated, it decreased to a minimum of 70 % on 30 July and then became more variable, ranging from 75 % to 90 % for the remainder of the melt season. Pond depth increased steadily from a median depth of 0.40 m ± 0.17 m in early June and peaked at 0.97 m ± 0.51 m on 16 July, even as melt pond fraction had already started to decrease. Our results demonstrate that by combining high-resolution passive and active remote sensing we now have the ability to track evolving melt conditions and observe changes in the sea ice cover throughout the summer season.

Research on Multilevel Filtering Algorithm Used for Denoising Strong and Weak Beams of Daytime Photon Cloud Data with High Background Noise

The Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) is equipped with the Advanced Topographic Laser Altimeter System (ATLAS), which can achieve high-precision ground detection. However, due to its low pulse energy and high sensitivity, it is also affected by noise when obtaining data. Therefore, data denoising is critical to the subsequent processing and application. In this study, a multilevel filtering algorithm is proposed to denoise the daytime photon cloud data with high background noise. Firstly, the Random Sample Consensus (RANSAC) algorithm is used to roughly denoise the daytime photon cloud data with high background noise, and a signal point cloud buffer is established to remove most of the noise points. Subsequently, the horizontal continuity parameter is calculated based on the photon cloud data following the rough denoising process. Based on this parameter, the shape and size of the search domain of the results of the subsequent fine denoising algorithm are adaptively improved. Finally, three filtering directions (a horizontal direction, an intra-group unified direction, and an adaptive direction for each photon) are proposed, and a hybrid algorithm combining the Ordering Points to Identify the Clustering Structure (OPTICS) density clustering algorithm and the Relative Neighboring Relationship K-nearest neighbors-based noise removal (RNR−KNNB) algorithm is employed to accurately denoise the photon cloud data in the three filtering directions. Furthermore, the RANSAC algorithm based on a sliding overlap window is used to remove outliers for the weak beam fine denoising photon cloud data. The results indicate that, for the strong beams, the denoising accuracy of the multilevel filtering algorithm in the three filtering directions is comparable (Rs ≥ 0.96, F ≥ 0.67), and they are all better than that of the ATL08 algorithm (Rs/Rn/p/F = 0.85/0.67/0.52/0.65). For weak beams, the denoising accuracy of the multilevel filtering algorithm in the horizontal direction and the intra-group unified direction is similar (Rs = 0.92, F = 0.69), and it is superior to the denoising results in the adaptive direction of each photon and the ATL08 algorithm (Rs/Rn/p/F = 0.94/0.84/0.51/0.65, 0.88/0.87/0.55/0.67, respectively). For strong and weak beams, the p-value and F-value of the denoising results of multilevel filtering algorithms in three different filtering directions increase with the increase of SNR value. It is demonstrated that SNR is an important factor affecting the denoising results of algorithms. The multilevel filtering algorithm proposed in the study can effectively achieve precise denoising of daytime photon cloud data with high background noise, and the three different filtering directions have different effects on the denoising results of strong and weak beam photon cloud data. This can provide technical and methodological guidance for subsequent photon cloud data filtering processing.

Spatiotemporal Variations of Glacier Mass Balance in the Tomur Peak Region Based on Multi-Source Altimetry Remote Sensing Data

Alpine glaciers are sensitive indicators of regional climate change, which can affect regional ecological stability and social development. Variations in glacier mass balance (GMB) are an important parameter in studying glacier change. In this study, data from the Ice, Cloud, and Land Elevation Satellite-1 (ICESat-1), the Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2), and CryoSat-2 (Ku-band) were combined, and high-resolution ALOS DEM was employed to denoise. After that, the polynomial fitting method was used to analyze the characteristics of glacier surface elevation (GSE) variations from 2003–2020 in the Tomur Peak Region of the Central Asian Tianshan Mountains and the regional GMB was calculated. Research results showed that: (1) From 2003–2020, the GSE of the Tomur Peak Region had an overall −8.95 ± 4.48 m variation, the average rate of which was −0.53 ± 0.26 m/yr (/yr is /year). Overall, elevations of most glaciers in the Tomur Peak Region had downward trends, with a rate of change of −0.5 to 0 m/yr. The fastest rate of elevation decline in the Koxkar Glacier Tongue was −1.5 m/yr. The elevation of some altimetric points in the Eastern Tomur Peak Region showed a rising state, with a maximum rate of variation of 1.0 m/yr. (2) From 2003–2020, the average GMB in the Tomur Peak Region was −1.51 ± 0.04 Gt/yr. In the region of elevation below 4000 m, small glaciers dominated, with a GMB of −0.61 ± 0.04 Gt/yr. With increasing elevation, the melting rate of glaciers gradually slowed down, but overall, the mass balance remained in a state of decline. (3) Climate was the main driving factor of GMB change in the study area. From 2003–2020, in the Tomur Peak Region, the average annual temperature continued to increase at a rate of 0.04 ± 0.02 °C/yr, and this was the main influencing factor for the negative GMB in the Tomur Peak Region. In the same period, the annual precipitation showed a rising trend with a linear variation rate of 0.12 ± 0.06 mm/yr, and the rising precipitation was the influencing factor for the gradually slowing change in the GMB in the study area.

Cost-efficient bathymetric mapping method based on massive active–passive remote sensing data

Accurate bathymetric mapping is critical for island reef research, coastal ecosystem monitoring, and nearshore engineering. Increasing amounts of remote sensing data have promoted the development of optical bathymetric remote sensing. However, large-scale and efficient bathymetric mapping of shallow oceanic islands is difficult due to the limited availability of high-quality optical remote sensing images. In this study, we propose a cost-efficient method for bathymetric mapping based on a quadratic polynomial ratio model (QPRM) of massive active–passive remote sensing data. Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) data represent active data and a priori data, and Sentinel-2 data from the Google Earth Engine represent passive data. The key step is to build a QPRM based on the median value of the blue-green band logarithmic ratio from massive Sentinel-2 data. The method is applied to six typical areas located in the Pacific Ocean, Indian Ocean, and South China Sea, and the results showed that the root mean square error (RMSE) and mean absolute error (MAE) of bathymetric mapping in the water depth range of 0–20 m were 0.49–0.71 m and 0.29–0.49 m, respectively. In addition, the RMSE decreased with an increase in the number of Sentinel-2 images, and the results were relatively stable when the number reached approximately 150. The QPRM results were also compared with that of the classical linear ratio model (CLRM) and multiple fitting methods to calculate the median water depth, and the findings showed that the QPRM method has a slight advantage in terms of accuracy and efficiency. The proposed method, which does not depend on in situ data, effectively calculates precise water depths and removes the effects of complex underwater optical paths and surface instabilities, such as waves, boats, and clouds. Therefore, it has great potential for bathymetric mapping of oceanic islands and reefs worldwide.

Retrieving Sub-Canopy Terrain from ICESat-2 Data Based on the RNR-DCM Filtering and Erroneous Ground Photons Correction Approach

Currently, the new space-based laser altimetry mission, Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2), is widely used to obtain terrain information. Photon cloud filtering is a crucial step toward retrieving sub-canopy terrain. However, an unsuccessful photon cloud filtering performance weakens the retrieval of sub-canopy terrain. In addition, sub-canopy terrain retrieval would not be accurate in densely forested areas due to existing sparse ground photons. This paper proposes a photon cloud filtering method and a ground photon extraction method to accurately retrieve sub-canopy terrain from ICESat-2 data. First, signal photon cloud data were derived from ICESat-2 data using the proposed photon cloud filtering method. Second, ground photons were extracted based on a specific percentile range of elevation. Third, erroneous ground photons were identified and corrected to obtain accurate sub-canopy terrain results, assuming that the terrain in the local area with accurate ground photons was continuous and therefore could be fitted appropriately through a straight line. Then, the signal photon cloud data obtained by the proposed method were compared with the reference signal photon cloud data. The results demonstrate that the overall accuracy of the signal photon identification achieved by the proposed filtering method exceeded 96.1% in the study areas. The sub-canopy terrain retrieved by the proposed sub-canopy terrain retrieval method was compared with the airborne LiDAR terrain measurements. The root-mean-squared error (RMSE) values in the two study areas were 1.28 m and 1.19 m, while the corresponding R2 (coefficient of determination) values were 0.999 and 0.999, respectively. We also identified and corrected erroneous ground photons with an RMSE lower than 2.079 m in densely forested areas. Therefore, the results of this study can be used to improve the accuracy of sub-canopy terrain retrieval, thus pioneering the application of ICESat-2 data, such as the generation of global sub-canopy terrain products.

Crossover adjustment of ICESat-2 satellite altimetry for the Arctic region

The Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) mission, operated by NASA, was launched in September 2018. The mission aims to obtain height measurements that create a global portrait of Earth’s radial dimension, gathering data to monitor changes of terrain including glaciers, sea ice, forests and more. It is important to understand the error budget of the observations, one component of which is radial orbit error. Apart from the altimetric ranging errors, radial orbit errors directly influence the accuracy of the measurement of sea surface height (SSH). These errors can be assessed by analyzing the difference of SSHs at ground track intersections, so-called crossover differences (XO differences). An effective approach is to model the orbit error by minimizing the residual XO difference by the least-squares method. XO differences are assumed to be caused by the difference of radial orbit errors between ascending and descending arcs, the sea surface variation, mispointing, and measurement errors. Since the sea surface variation in a short time interval and measurement errors can be considered as random variable, these residuals can be reduced by the method of XO adjustment. In our study, the RMS XO differences of ICESat-2 is significantly reduced from 17.2 cm to 8.7 cm in the Arctic region.

Sea ice surface type classification of ICESat-2 ATL07 data by using data-driven machine learning model: Ross Sea, Antarctic as an example

The ICESat-2 (IS2) ATL07 sea ice height product provides precise and accurate measurements of sea ice height for the polar regions. Although the original ATL07 product classifies the sea ice surface into snow-covered ice and open water (dark leads and specular leads), it has two critical issues: (1) it does not distinguish thin ice (gray ice) from thick/snow-covered ice or open water and (2) it involves uncertainties in the dark lead determination. To address these issues and obtain more accurate lead fraction and freeboard estimation, in this study, we assess a data-driven machine learning approach for sea ice surface type classification from the ATL07 product. A total of 17 IS2 tracks covering the Ross Sea in February, March, September, October, and November 2019 are used in this study. For training and testing the neural network (NN) models, we label the ATL07 data into three surface types: thick/snow-covered sea ice, thin/gray sea ice, and open water based on the coincident Sentinel-2 optical images. We use six variables from the ATL07 data: background rate, photon rate, relative surface height, width of height distribution, number of laser pulses, and difference in mean and median height, with the first three variables found to be the most significant in determining surface types. Our spatiotemporal test shows that the NN models can be applied to the ATL07 dataset with ∼99% accuracy in the Ross Sea, and the beam test shows that the difference in the accuracy for three beams is negligible. While the current ATL07 product captures only ∼5% of thin ice and ∼ 81% of open water correctly, our NN models capture ∼90% of both open water and thin ice leads successfully. When we qualitatively assess the surface classification of the NN models by using the coincident optical and radar images, our NN models do not show significant misclassification issues both in the daytime and nighttime and both in the winter and summer seasons. Compared to the IS2 ATL10 sea ice freeboard product, the ability of our NN model to separate thin ice and open water from snow-covered/thick ice is significant in freeboard estimation, especially for winter and highly packed sea ice areas. Open water and thin ice classes also improve mapping lead fraction and floe size distribution.

Fine-Resolution Forest Height Estimation by Integrating ICESat-2 and Landsat 8 OLI Data with a Spatial Downscaling Method for Aboveground Biomass Quantification

Rapid and accurate estimation of forest aboveground biomass (AGB) with fine details is crucial for effective forest monitoring and management, where forest height plays a key role in AGB quantification. In this study, we propose a random forest (RF)-based down-scaling method to map forest height and biomass at a 15-m resolution by integrating Landsat 8 OLI and Ice, Cloud and Land Elevation Satellite-2 (ICESat-2) LiDAR data. ICESat-2 photon data are used to derive canopy parameters along 15-m segments, which are considered sample plots for the extrapolation of discrete forest height. Fourteen variables associated with spectral features, textual features and vegetation index are extracted from pan-sharpened Landsat 8 images. A regression function is established between these variables and ICESat-2-derived forest height to produce a 15-m continuous forest height distribution data based on the 30-m forest height product using the RF algorithm. Finally, a wall-to-wall forest AGB at 15-m spatial resolution is achieved by using an allometric model specific to the forest type and height. The Jilin Province in northeast China is taken as the study area, and the forest AGB estimation results reveal a density of 61.15 Mg/ha with a standard deviation of 89.46 Mg/ha. The R2 between our predicted forest heights and the ICESat-2-derived heights reaches 0.93. Validation results at the county scale demonstrate reasonable correspondence between the estimated AGB and reference data, with consistently high R2 value exceeding 0.65. This downscaling method provides a promising scheme to estimate spatial forest AGB with fine details and to enhance the accuracy of AGB estimation, which may facilitate carbon stock measurement and carbon cycle studies.

High-resolution forest age mapping based on forest height maps derived from GEDI and ICESat-2 space-borne lidar data

Forest age is a key parameter for estimating forest growth and carbon uptake and for forest management. Remote sensing provides indirect but useful information for mapping forest age at large scales. However, existing regional and global forest age products were generated at low spatial resolutions (often 1000 m) and are not useful for most forest stands in China that are smaller than 1000 m. This study aims to map forest age at the 30 m resolution based on forest height maps mainly derived from the Global Ecosystem Dynamics Investigation (GEDI) and Ice, Cloud, and land Elevation Satellite-2 (ICESat-2) data, and analyze the roles of auxiliary data including temperature, precipitation, slope, and aspect in forest age mapping. Forest age is defined as the average age of dominant tree species within a pixel. Five commonly-used stand growth equations and twelve machine learning methods were tested for their suitability for mapping forest age of different tree species. We found that the Logistic equation performed the best among the tested stand growth equations and the Random Forest (RF) was the best among the tested machine learning methods. According to RF, forest height contributed predominantly to the variance in forest age mapping, while temperature, precipitation, slope, and aspect also had an overall non-negligible and variable contribution among different tree species. By integrating the climate and topographical variables, RF was applicable for forest age mapping without classifying the tree species. These results show that forest height maps derived from space-borne lidar data such as GEDI and ICESat-2 data are highly useful for mapping forest stand age, and the methodology developed in this study highlights a perspective for generating national and global forest age products at a high spatial resolution.

Spatially-explicit mapping annual oil palm heights in peninsular Malaysia combining ICESat-2 and stand age data

Oil palm is the most efficient oil-producing crop but its extension leads to increased deforestation in Southeast Asia. Oil palm height enables the quantitative estimation for carbon stock or palm oil yield. Nevertheless, there are still no accurate characterization of oil palm height providing information for the tradeoff between forest damage and carbon stock of oil palm in Southeast Asia. The new generation of spaceborne LiDAR provides large-extent canopy height samples, offering an opportunity for mapping the oil palm height at the regional scale yet with challenge to extrapolate the footprint heights to spatially coherent maps. Here, we proposed a new method by combining Ice, Cloud, and Land Elevation Satellite-2 (ICESat-2) footprint data with stand age that is closely related to tree height. We first developed a semi-automatic filtering algorithm to filter the low-quality ICESat-2 data, and used a change detection algorithm to optimize the planting year map of oil palm. Then, an empirical age-height model was derived by linking ICESat-2 footprint canopy height with oil palm age that was used to estimate spatially and temporally oil palm height for the whole Peninsular Malaysia. A validation with independent ICESat-2 footprint data suggests a high agreement for the height estimates in 2020 from the age-height model (R2 = 0.63; RMSE = 1.64 m) with a bias within ±3 m for >90% of the height estimates. Using the age-height model and planting year map of oil palm, we produced the first comprehensive wall-to-wall maps of long-term yearly oil palm height at a spatial resolution of 30 m in Peninsular Malaysia during 2001 through 2020. Our results suggest that the mean height of oil palm in all and regionally-disturbed areas have increased by 10.82 m and 9.29 m respectively during the last two decades in Peninsular Malaysia. Our results indicate that combining stand age and ICESat-2 footprint data has great potential in spatially-explicit mapping regional oil palm height that contributes to a better quantification for regional plantation carbon stock.