Cloud Features Research Articles

Polarization Characteristics of Massive HVI Debris Clouds Using an Improved Monte Carlo Ray Tracing Method for Remote Sensing Applications

As a signature phenomenon of massive hypervelocity impacts (HVIs) in space, debris clouds provide critical optical information for satellite remote sensing and the assessment of large-scale impacts. However, studies of the optical scattering properties of debris clouds remain limited, and existing vector radiative transfer (VRT) methods struggle to accurately simulate the optical characteristics of these complex scatterers. To address this gap, this paper presents an improved Monte Carlo VRT program (PGS–MC) for multicomponent polydisperse scatterers to precisely evaluate the radiation and polarization characteristics of complex scatterers. Based on the Monte Carlo ray tracing (MCRT) method, our program introduces a particle grouping strategy (PGS) to further emphasize the importance of accounting for optical property discrepancies between different materials and particle sizes, thus significantly improving the fidelity of VRT simulations. Moreover, our program, developed using the compute unified device architecture (CUDA), can be run parallelly on graphics processing units (GPUs), which effectively reduces the computational time. The validation results indicated that the developed PGS–MC program can accurately and efficiently simulate the polarization of complex 3D scatterers. A further investigation showed that the polarization characteristics of debris clouds are highly sensitive to parameters such as the angle between the incident and detection directions, number density, particle size distribution, debris material, and wavelength. In addition, the polarization imaging of debris clouds offers distinct advantages over intensity imaging. This study offers guidance for analyzing the VRT properties of massive HVI debris clouds. Additionally, it provides a practical tool and concrete ideas for modeling the polarization characteristics of various complex scatterers, such as aircraft contrails and clouds, etc.

Evolution Characteristics of Convective Clouds With Relatively Small Scales Over South China

AbstractAs a crucial element in the Earth's system, development of convective clouds is still insufficiently understood and simulated in both weather and climate models, particularly for small‐scale regional convective clouds. In this study, a series of convective clouds cases with relatively small scales are selected over south China, and the evolution characteristics of those convective clouds are investigated using high spatiotemporal resolution geostationary satellite data. Statistical results show that the shorter the life cycle or the smaller the area, the higher the proportion of convective clouds. Notably, approximately 79.23% of convective clouds have a life cycle of less than 3 hr, and 63.81% have an area of less than 500 km2 for selected cases. In addition, there are significant differences in the cloud characteristics and meteorological parameters during various convective cloud stages and durations. Nevertheless, the relative proportions of convective clouds at three identified stages remain relatively constant with almost no dependence on duration of convective clouds, which are 33.50%, 23.92%, and 42.58% for the developing, mature, and dissipation stages, respectively. In addition, we find that the cloud‐top cooling rate during the developing stage also affects the characteristics of the later stage of convective clouds. Quantitatively, the average cloud area and duration changed by 157.03 km2 and 0.17 hr when the cloud‐top cooling rate varies by 15 K/h.

SuPrNet: Super Proxy for 4D occupancy forecasting

Spacetime (4D) occupancy forecasting explicitly models the future development of self-driving vehicles (SDVs) themselves and their surrounding environment, which is highly beneficial for their downstream tasks. Currently, using point cloud forecasting as a proxy to self-supervisedly optimize the 4D occupancy forecasting framework has significant application prospects due to its independence from labeled data. However, existing methods have the following shortcomings: (1) Insufficient mining of spatiotemporal information in the scene. (2) Neglecting the extraction of regional deep features. (3) Inadequate learning ability for spatiotemporal predictive learning. Based on this, we propose SuPrNet, constructing a ”Super Proxy” to address the aforementioned issues. Specifically, SuPrNet first embeds the spatial and temporal information of the current frame into point-wise features through a 4D hybrid representation combined with multi-planar and grid features (4DPG). Then, a Hierarchical Encoder–Decoder Module (HED) with sparsity consideration is used to structure the point cloud and extract deep features. Thanks to the structured feature map representation, SuPrNet proceeds to use a spatiotemporal predictive learning module (STL) to implicitly predict future feature maps. Finally, we decode these feature maps to a specific size as the forecasting of future occupancy. The 4DPG, HED, and STL in the model are designed to address the aforementioned issues, and experiments on Nuscene and KITTI-Odometry show that our method achieves the best 4D occupancy forecasting and point cloud forecasting accuracy, surpassing previous state-of-the-art models by approximately 30%. Additionally, we have expanded and validated the application prospects of SuPrNet through various downstream tasks. Our project is available at https://github.com/AlanLiangC/4DPCF.

High‐Frequency Isotope Compositions Reveal Different Cloud‐Top and Vertical Stratiform Rainfall Structures in the Inland Tropics of Brazil

AbstractUnderstanding the key drivers controlling rainfall stable isotope variations in inland tropical regions remains a global challenge. We present novel high‐frequency isotope data (5–30 min intervals) to disentangle the evolution of six stratiform rainfall events (N = 112) during the passage of convective systems in inland Brazil (September 2019–June 2020). These systems produced stratiform rainfall of variable cloud features. Depleted stratiform events (δ18Oinitial ≤ −4.2‰ and δ18Omean ≤ −6.1‰) were characterized by cooler cloud‐top temperatures (≤−38°C), larger areas (≥48 km2), higher liquid‐ice ratios (≥3.1), and higher melting layer heights (≥3.8 km), compared to enriched stratiform events (δ18Oinitial ≥ −3.8‰ and δ18Omean ≥ −5.1‰). Cloud vertical structure variability was reflected in a wide range of δ18O temporal patterns and abrupt shifts in d‐excess. Our findings provide a new perspective to the ongoing debate about isotopic variability and the partitioning of rainfall types across the tropics.

In-situ aircraft observations of aerosol and cloud microphysical characteristics of mixed-phase clouds over the North China Plain

Aerosol-cloud interactions play a vital role in climate change. This study leverages observations from the King-350 aircraft over the North China Plain on November 29, 2019, to examine aerosol and cloud microphysical characteristics of mixed-phase clouds. Through detailed vertical and spectral distributions, we investigate aerosol, cloud droplet, and ice crystal distributions in seeded clouds (SC) and natural precipitating clouds (NPC) within the same cloud system. From the vertical profile, SC and NPC have similar vertical distributions of aerosol and cloud droplets, with over 95 % of aerosols concentrated below 1600 m near the ground. Cloud droplets are more evenly distributed within the two clouds, cloud droplet number concentrations (Nc) in SC were three orders of magnitude higher than in NPC. Ice water content (IWC) and ice crystal number concentration (Ni) show distinct layer preferences—accumulating predominantly in SC's top layer and NPC's middle layer. From spectral distribution, a smaller proportion of cloud droplets (40–50 μm in diameter, the same hereafter) in SC compared to NPC. Rimed ice crystals and globular graupel (1325–1550 μm in diameter) were in SC, while plate and irregular ice crystals (300–450 μm) were in NPC with an order of magnitude higher than in SC. These microphysical differences highlight the complexity of cloud seeding efficacy, which varies based on cloud conditions and microphysical properties. In the first seeding, Ni increased by 1–2 orders of magnitude (125–300 μm) in the high Nc (Nc > 1.11 × 105 L−1) region. Seeding in low Nc (Nc < 1.11 × 105 L−1) regions was hard to be effective, especially in low Nc and low liquid water content (LWC) (LWC < 0.122 g/m3) regions. In the second seeding, ice crystals (125–250 μm) produced by the first seeding enhance the seeding efficiency. The responded regions were more sensitive to subsequent seeding, resulting in stronger reactions or longer duration.

Lidar cloud and aerosol layer detection method based on point cloud filtering

A point cloud filtering method is presented for atmospheric layer detection from lidar data. The method involves rising edge event recognition based on a wavelet transform function. Density-based clustering was then utilized to separate the real boundary from the original noisy point clouds based on continuous distribution characteristics of cloud and aerosol layer. Tests were carried out to verify the performance of our algorithm with synthetic lidar signals with noise. The layer base detection error within ± 5 bins was achieved for signals with SNRs higher than 3. Even for SNRs higher than 1, high consistency was still observed between retrieved results with our method and a visual analysis. These results indicate that our algorithm is suitable for unsupervised detection with large time-series datasets, such as CALIOP.

The Impact of Cloud Computing on System Lifecycle Methodologies: an Innovative Approach to Scalability Flexibility

This research analyzes the impact of cloud computing on traditional system lifecycle methodologies and offers an innovative approach that leverages the advantages of cloud computing. Qualitative research was conducted with a literature study on conventional system lifecycle methodologies and cloud computing. The results show that cloud computing enables increased scalability and flexibility through a service-based approach, automation, and elastic management of resources. This research provides recommendations for implementing a hybrid lifecycle methodology that leverages the advantages of both cloud and traditional systems.This paper discusses the impact of cloud computing on system lifecycle methodologies. Cloud computing offers an innovative approach to system scalability and flexibility by allowing computing and storage resources to scale dynamically according to demand. Traditional system lifecycle methodologies are becoming less relevant for rapidly changing cloud environments, so they need to be adapted to respond to the scalability and flexibility capabilities offered by cloud computing.This paper evaluates the impact of the shift towards cloud computing on the methodologies used in the system development lifecycle. Cloud computing offers the ability to scale elastically according to demand, which opens up opportunities to adopt a more adaptable and flexible approach to system development. This paper discusses how traditional system lifecycle methodologies can be adapted to accommodate key cloud computing characteristics such as scalability, flexibility, and virtualization of computing resources. This paper reviews the implications of the move towards a cloud computing paradigm on the methodologies used in developing and maintaining software systems. The shift towards cloud-based environments offers the advantages of scalability and flexibility in dynamically managing system resources. However, these key characteristics of cloud computing pose challenges to traditional system lifecycle methodologies designed for on-premise environments. This paper discusses how system lifecycle methodologies can be adapted to better support system development in cloud computing environments.

Research on Student's T-Distribution Point Cloud Registration Algorithm Based on Local Features.

LiDAR offers a wide range of uses in autonomous driving, remote sensing, urban planning, and other areas. The laser 3D point cloud acquired by LiDAR typically encounters issues during registration, including laser speckle noise, Gaussian noise, data loss, and data disorder. This work suggests a novel Student's t-distribution point cloud registration algorithm based on the local features of point clouds to address these issues. The approach uses Student's t-distribution mixture model (SMM) to generate the probability distribution of point cloud registration, which can accurately describe the data distribution, in order to tackle the problem of the missing laser 3D point cloud data and data disorder. Owing to the disparity in the point cloud registration task, a full-rank covariance matrix is built based on the local features of the point cloud during the objective function design process. The combined penalty of point-to-point and point-to-plane distance is then added to the objective function adaptively. Simultaneously, by analyzing the imaging characteristics of LiDAR, according to the influence of the laser waveform and detector on the LiDAR imaging, the composite weight coefficient is added to improve the pertinence of the algorithm. Based on the public dataset and the laser 3D point cloud dataset acquired in the laboratory, the experimental findings demonstrate that the proposed algorithm has high practicability and dependability and outperforms the five comparison algorithms in terms of accuracy and robustness.

Application of the Qira'ah Mubah Method in Content on the Mubadalah.Id Website (In April 2023)

This research examines the relevance of Qira'ah Mubadalah's values and Kartini's thoughts in the content presented by Mubadalah.id during April 2023. The focus of the research is to understand the extent to which understanding and appreciation of Kartini's struggle is reflected in the content, as well as the influence and relevance of kartinian values in discussions and articles published on the platform. This relevance can be related to the values and perspective of Qira'ah Mub is one that has 3 indicators, namely a perspective (minzhar) that humanizes humans. Second, how to read (qira'ah) reference texts by placing men and women as subjects and complete humans. Third, the way of grouping or in Arabic terms (qa'idah) experiences, distractions, or cooperation between men and women is reflected in the word cloud and word frequency tables. This research is a type of qualitative research using text data in the media, samples and techniques for sampling news/documents (text) in the form of content on the Mubadalah.id website totaling 20 articles in the April 2023 time period. Data Sources and Data Collection Methods used are primary data collected from the website. Data is retrieved by downloading articles and saving them using pdf. And in data processing, data is processed using Nvivo software by utilizing the word cloud feature. The research results show that the content of Mubadalah.id predominantly discusses the issues of women, Kartini, figures, migrant workers, and the like. The word "woman" appeared 216 times, indicating the significance of the issue of women in the content. Therefore, this research provides insight into the potential of preaching gender equality through social media and confirms the relevance of Kartini's values in a contemporary context.

A feasibility study of dosimetry for breast cancer radiotherapy based on body surface changes.

The requirement for precise and effective delivery of the actual dose to the patient grows along with the complexity of breast cancer radiotherapy. Dosimetry during treatment has become a crucial component of guaranteeing the efficacy and security. To propose a dosimetry method during breast cancer radiotherapy based on body surface changes. A total of 29 left breast cancer radiotherapy cases were retroactively retrieved from an earlier database for analysis. Non-rigid image registration and dose recalculation of the planning computed tomography (CT) referring to the Cone-beam computed tomography were performed to obtain dose changes. The study used 3D point cloud feature extraction to characterize body surface changes. Based on the correlation proof, a mapping model is developed between body surface changes and dose changes using neural network framework. The MSE metrics, the Euclidean distances of feature points and the 3D gamma pass rate metric were used to assess the prediction accuracy. A strong correlation exist between body surface changes and dose changes (first canonical correlation coefficient=0.950). For the dose deformation field and dose amplitude difference in the test set, the MSE of the predicted and actual values were 0.136 pixels and 0.229 cGy, respectively. After deforming the planning dose into a deformed one, the feature points' Euclidean distance between it and the recalculated dose changes from 9.267±1.879mm to 0.456±0.374mm. The 3D gamma pass rate of 90% or higher for the 2mm/2% criteria were achieved by 80.8% of all cases, with a minimum pass rate of 75.9% and a maximum pass rate of 99.6%. Pass rate for the 3mm/2% criteria ranged from 87.8% to 99.8%, with 92.3% of the cases having a pass rate of 90% or higher. This study provides a dosimetry method that is non-invasive, real-time, and requires no additional dose for breast cancer radiotherapy.

Machine learning potentials for global multi-timescale diffuse irradiance estimation: Synthesizing ground observations, time-series, and environmental features

Separating diffuse horizontal irradiance (DHI) from ground-based global horizontal irradiance observations is critical owing to lack of direct DHI observations. Existing models are often site-specific and time-bound, thereby limiting their universal applicability. To address this, this study thoroughly explores machine learning (ML) for constructing global separation models. We develop 36 models using three ML algorithms—light gradient boosting machine (LightGBM), generalized additive model (GAM), and geographically weighted artificial neural network (GWANN)—alongside three data splitting strategies for minutely, 10-minutely, hourly, and daily timescales. LightGBM and GAM outperform GWANN in accuracy, with LightGBM excelling in interpretability, efficiency, and handling missing values, while GWANN exhibits superior stability. Compared with completely random splitting, model accuracy decreases by 2.26 % and 2.16 % for station- and date-based splitting, respectively. Prediction accuracy varies across timescales, with minutely models typically considered the best. Key predictors are Kt, solar zenith angle, and altitude, complemented by the significant influence of time-series Kt, aerosol, and cloud features. The influence of various factors on model accuracy differs and is scale-dependent. Among them, time-series Kt variability factors notably enhance ML-based predictions, especially at minutely scales. The LightGBM model outperforms classic models in overall and site-specific accuracy and adaptability, but its computational efficiency declines, especially with time-series variability factors. The findings highlight that ML-based separation models necessitate careful selection of algorithm, data splitting strategy, and input variables, fully considering study region, data situation, and temporal and spatial scales. This research offers a potential solution for constructing global separation models and insights for model improvement.

Irradiance separation model parameter estimation from historical cloud cover statistical properties

Irradiance separation models allow the decomposition of Global Horizontal Irradiance into Diffuse Horizontal and Direct Normal Irradiances. These models need fitting to the irradiance characteristics of the location of interest using locally measured ground data. For locations that only measure Global Horizontal Irradiance, current state of the art establishes the use of parameters obtained for another location that measures the three components, with similar climate characteristics. Nevertheless, this results in a lack of localized character for estimates and requires fitting model parameters for all possible climates, which can be infeasible given data availability. This work presents a novel approach based on the hypothesis that the separation model's parameters are a function of the statistical properties of satellite-derived cloud cover estimates. The proposed methodology was evaluated in 23 sites covering all main Köppen-Geiger climatic types and different cloud coverage properties using the Boland-Ridley-Lauret diffuse fraction model. The model performs similarly as locally adjusting the model, with Root Mean Square Errors of 0.087–0.15 diffuse fraction units versus 0.077–0.127 for locally optimized parameters, and offers adequate performance across climates and cloud characteristics. These results encourage future research by generalizing parameter estimation for other diffuse fraction models. The main applications for this research are the estimation of irradiance components where no local data is available for model fitting and the enhancement or complementarity of satellite estimates of surface irradiance. Furthermore, it allows the estimation of missing irradiance components due to equipment failure in locations with insufficient data for a representative, locally adapted model.

CarvingNet: Point cloud completion by stepwise refining multi-resolution features

In the field of 3D vision, 3D point cloud completion is a crucial task in many practical applications. Current methods use Transformer's Encoder-Decoder framework to predict the missing part of the point cloud features at low resolution, which does not fully utilize the feature information at multiple resolutions and can result in the loss of the object's geometric details. In this paper, we present a novel point cloud completion method, CarvingNet, which, to the best of our knowledge, is the first to apply the U-Net architecture to the point cloud completion task by operating directly on unordered point cloud features at multiple resolutions. Firstly, we gradually expand the receptive field and use cross-attention to purify the features of the missing part of the point cloud at each resolution and to generate the contour features of the complete point cloud at the last obtained resolution. Then, we gradually reduce the receptive field and use cross-attention to refine the features of the complete point cloud at each resolution and to generate the features of the complete point cloud with rich details at the last obtained resolution. To obtain point cloud features at different resolutions, we specifically design the up-sampling module and down-sampling module for disordered point cloud features. Furthermore, we improve the FoldingNet network to make it more suitable for generating high-quality dense point clouds. The experimental results demonstrate that our proposed CarvingNet achieves the performance of existing state-of-the-art methods on the ShapeNet-55, ShapeNet-34, and KITTI benchmarks.

Correction of Thin Cirrus Absorption Effects in Landsat 8 Thermal Infrared Sensor Images Using the Operational Land Imager Cirrus Band on the Same Satellite Platform.

Data from the Operational Land Imager (OLI) and the Thermal Infrared Sensor (TIRS) instruments onboard the Landsat 8 and Landsat 9 satellite platforms are subject to contamination by cloud cover, with cirrus contributions being the most difficult to detect and mask. To help address this issue, a cirrus detection channel (Band 9) centered within the 1.375-μm water vapor absorption region was implemented on OLI, with a spatial resolution of 30 m. However, this band has not yet been fully utilized in the Collection 2 Landsat 8/9 Level 2 surface temperature data products that are publicly released by U.S. Geological Survey (USGS). The temperature products are generated with a single-channel algorithm. During the surface temperature retrievals, the effects of absorption of infrared radiation originating from the warmer earth's surfaces by ice clouds, typically located in the upper portion of the troposphere and re-emitting at much lower temperatures (approximately 220 K), are not taken into consideration. Through an analysis of sample Level 1 TOA and Level 2 surface data products, we have found that thin cirrus cloud features present in the Level 1 1.375-μm band images are directly propagated down to the Level 2 surface data products. The surface temperature errors resulting from thin cirrus contamination can be 10 K or larger. Previously, we reported an empirical and effective technique for removing thin cirrus scattering effects in OLI images, making use of the correlations between the 1.375-μm band image and images of any other OLI bands located in the 0.4-2.5 μm solar spectral region. In this article, we describe a variation of this technique that can be applied to the thermal bands, using the correlations between the Level 1 1.375-μm band image and the 11-μm BT image for the effective removal of thin cirrus absorption effects. Our results from three data sets acquired over spatially uniform water surfaces and over non-uniform land/water boundary areas suggest that if the cirrus-removed TOA 11-μm band BT images are used for the retrieval of the Level 2 surface temperature (ST) data products, the errors resulting from thin cirrus contaminations in the products can be reduced to about 1 K for spatially diffused cirrus scenes.

Physical and optical properties of cirrus and subvisible cirrus clouds over Arabian sea and Bay of Bengal region

Cirrus clouds play a crucial role in regulating the Earth’s radiation budget, and this paper aims to contribute to a deeper understanding of their optical and geometrical properties over the Arabian Sea (AS) and Bay of Bengal (BoB) regions using CALIPSO data. Comprehensive statistics are derived, encompassing mean values of cirrus cloud top, base altitude, geometrical thickness, cloud optical depth, and temperature. Over the AS region, the mean values are 15.10 ± 1.50 km, 12.63 ± 1.75 km, 2.52 ± 1.37 km, 0.4 ± 0.58, and -62.30 ± 10.6 (°C), respectively. For the BoB region, the corresponding values are 15.43 ± 1.51 km, 12.72 ± 1.74 km, 2.71 ± 1.46 km, 0.49 ± 0.67, and -64.35 ± 10.7 (°C). A larger spread in optical depth and a higher frequency of occurrence for both cirrus and subvisible cirrus (SVC) clouds were observed over BoB compared to AS. Additionally, the study delves into SVC cloud characteristics, emphasizing their thinness and higher base altitudes compared to cirrus clouds. This comprehensive investigation contributes valuable insights into the distinctive properties of cirrus and SVC clouds in these regions, enhancing our knowledge of atmospheric processes and their implications for climate modelling and predictions.

Complementary pseudo multimodal feature for point cloud anomaly detection

Point cloud anomaly detection is steadily emerging as a promising research area. Recognizing the importance of feature descriptiveness in this task, this study introduces the Complementary Pseudo Multimodal Feature (CPMF), which combines local geometrical information extracted by 3D handcrafted descriptors with global semantic information extracted from 2D pre-trained neural networks. Specifically, to leverage 2D pre-trained neural networks for point-wise feature extraction, this study projects original point clouds into multi-view images. These images are then fed into a pre-trained 2D neural network for informative 2D modality feature extraction. Following the 2D–3D correspondence, the multi-view 2D modality features are projected back to 3D space and aggregated to obtain point-wise 2D modality features. Finally, the point-wise 3D and 2D modality features are fused to derive the CPMF for point cloud anomaly detection. Extensive experiments conducted on MVTec 3D and Real3D datasets demonstrate the complementary capacity between 2D and 3D modality features and the effectiveness of CPMF. Notably, CPMF achieves a significantly higher object-level AUROC of 95.15% compared to other methods on the MVTec 3D benchmark. Code is available at https://github.com/caoyunkang/CPMF.

A fast point cloud registration method based on spatial relations and features

Abstract Point cloud registration plays a crucial role in mobile robot localization, map building and three-dimensional (3D) model reconstruction. However, it remains challenged by issues such as compromised accuracy and sluggish efficiency, posing significant obstacles in achieving precise and timely alignments. Therefore, we propose a lightweight and fast point cloud registration method. Firstly, we mesh the 3D point cloud, compared with the traditional gridded point cloud method, it achieves initial point cloud registration by preserving the curvature characteristics of the internal point cloud, and utilizing the spatial relationship between grid cells and the quantitative relationship between the internal point cloud. Moreover, we adopt an iterative nearest point based on KD-Tree to realize the fine registration. So, our method does not necessitate intricate feature analysis and data training, and is resilient to similar transformations, non-uniform densities and noise. Finally, we conduct point cloud registration experiments using multiple publicly available point cloud datasets and compare them with several point cloud registration methods. The results demonstrate it is able to accomplish the point cloud registration quickly and exhibit high accuracy. More importantly, it maintains its efficacy and robustness even in the presence of noisy and defective point clouds.

Memory-Augmented 3D Point Cloud Semantic Segmentation Network for Intelligent Mining Shovels.

The semantic segmentation of the 3D operating environment represents the key to intelligent mining shovels' autonomous digging and loading operation. However, the complexity of the operating environment of intelligent mining shovels presents challenges, including the variety of scene targets and the uneven number of samples. This results in low accuracy of 3D semantic segmentation and reduces the autonomous operation accuracy of the intelligent mine shovels. To solve these issues, this paper proposes a 3D point cloud semantic segmentation network based on memory enhancement and lightweight attention mechanisms. This model addresses the challenges of an uneven number of sampled scene targets, insufficient extraction of key features to reduce the semantic segmentation accuracy, and an insufficient lightweight level of the model to reduce deployment capability. Firstly, we investigate the memory enhancement learning mechanism, establishing a memory module for key semantic features of the targets. Furthermore, we address the issue of forgetting non-dominant target point cloud features caused by the unbalanced number of samples and enhance the semantic segmentation accuracy. Subsequently, the channel attention mechanism is studied. An attention module based on the statistical characteristics of the channel is established. The adequacy of the expression of the key features is improved by adjusting the weights of the features. This is done in order to improve the accuracy of semantic segmentation further. Finally, the lightweight mechanism is studied by adopting the deep separable convolution instead of conventional convolution to reduce the number of model parameters. Experiments demonstrate that the proposed method can improve the accuracy of semantic segmentation in the 3D scene and reduce the model's complexity. Semantic segmentation accuracy is improved by 7.15% on average compared with the experimental control methods, which contributes to the improvement of autonomous operation accuracy and safety of intelligent mining shovels.

Automated Phenotypic Trait Extraction for Rice Plant Using Terrestrial Laser Scanning Data.

To quickly obtain rice plant phenotypic traits, this study put forward the computational process of six rice phenotype features (e.g., crown diameter, perimeter of stem, plant height, surface area, volume, and projected leaf area) using terrestrial laser scanning (TLS) data, and proposed the extraction method for the tiller number of rice plants. Specifically, for the first time, we designed and developed an automated phenotype extraction tool for rice plants with a three-layer architecture based on the PyQt5 framework and Open3D library. The results show that the linear coefficients of determination (R2) between the measured values and the extracted values marked a better reliability among the selected four verification features. The root mean square error (RMSE) of crown diameter, perimeter of stem, and plant height is stable at the centimeter level, and that of the tiller number is as low as 1.63. The relative root mean squared error (RRMSE) of crown diameter, plant height, and tiller number stays within 10%, and that of perimeter of stem is 18.29%. In addition, the user-friendly automatic extraction tool can efficiently extract the phenotypic features of rice plant, and provide a convenient tool for quickly gaining phenotypic trait features of rice plant point clouds. However, the comparison and verification of phenotype feature extraction results supported by more rice plant sample data, as well as the improvement of accuracy algorithms, remain as the focus of our future research. The study can offer a reference for crop phenotype extraction using 3D point clouds.

Physically Explainable Deep Learning for Convective Initiation Nowcasting Using GOES-16 Satellite Observations

Abstract Convective initiation (CI) nowcasting remains a challenging problem for both numerical weather prediction models and existing nowcasting algorithms. In this study, an object-based probabilistic deep learning model is developed to predict CI based on multichannel infrared GOES-16 satellite observations. The data come from patches surrounding potential CI events identified in Multi-Radar Multi-Sensor Doppler weather radar products over the Great Plains region from June and July 2020 and June 2021. An objective radar-based approach is used to identify these events. The deep learning model significantly outperforms the classical logistic model at lead times up to 1 h, especially on the false alarm ratio. Through case studies, the deep learning model exhibits dependence on the characteristics of clouds and moisture at multiple altitudes. Model explanation further reveals that the contribution of features to model predictions is significantly dependent on the baseline, a reference point against which the prediction is compared. Under a moist baseline, moisture gradients in the lower and middle troposphere contribute most to correct CI forecasts. In contrast, under a clear-sky baseline, correct CI forecasts are dominated by cloud-top features, including cloud-top glaciation, height, and cloud coverage. Our study demonstrates the advantage of using different baselines in further understanding model behavior and gaining scientific insights.