Cloud Region Research Articles

Virtual image-based cloud removal for Landsat images

ABSTRACT The inevitable thick cloud contamination in Landsat images has severely limited the usability and applications of these images. Developing cloud removal algorithms has been a hot research topic in recent years. Many previous algorithms used one or multiple cloud-free image(s) in the same area acquired on other date(s) as reference image(s) to reconstruct missing pixel values. However, it remains challenging to determine the optimal reference image(s). In addition, abrupt land cover change can substantially degrade the reconstruction accuracies. To address these issues, we present a new cloud removal algorithm called Virtual Image-based Cloud Removal (VICR). For each cloud region, VICR reconstructs the missing surface reflectance by three steps: virtual image within cloud region construction based on time-series reference images, similar pixel selection using the newly proposed temporally weighted spectral distance (TWSD), and residual image estimation. By establishing two buffer zones around the cloud region, VICR allows automatic selection of the optimal set of time-series reference images. The effectiveness of VICR was validated at four testing sites with different landscapes (i.e. urban, croplands, and wetlands) and land change patterns (i.e. phenological change, abrupt change caused by flooding and tidal inundation), and the performances were compared with mNSPI (modified neighborhood similar pixel interpolator), WLR (weighted linear regression) and ARRC (AutoRegression to Remove Clouds). Experimental results showed that VICR outperformed the other algorithms and achieved higher Correlation Coefficients and lower Root Mean Square Errors in surface reflectance estimation at the four sites. The improvement is particularly noticeable at the sites with abrupt land change. By considering the difference in the contributions from the reference images, TWSD can select more reliable similar pixels to improve the prediction of abrupt change in surface reflectance. Moreover, VICR is more robust to different cloud sizes and to changing reference images. VICR is also computationally much faster than ARRC. The framework for time-series image cloud removal by VICR has great potential to be applied for large datasets processing.

A comprehensive assessment of endogenous bubbles properties in fluidized bed reactors via X-ray imaging

Properties of endogenous bubbles released during the devolatilization of a single biomass particle under inert conditions have been investigated by means of advanced X-ray imaging techniques. Distribution of void fraction showed that endogenous bubbles structure resembles that of classic bubbles observed in fluidized bed reactors, constituted by a cloud, wake and a central void region. A value of about 0.25 for the wake fraction has been obtained from experiments, which is in agreement with literature data for Geldart B particles. Volume of cloud region as a function of relative bubble velocity was generally well-described by the theoretical models of Davidson and Murray, showing effective recirculation of volatile matter around the bubble. Moreover, lack of mixing between bubbles and emulsion phase, as predicted by the Davidson's theory for classic bubbles, confirmed the bypass phenomenon observed for endogenous bubbles in previous studies. Owing to the non-invasive nature of the X-ray technique employed, it was possible to estimate the main features of endogenous bubbles with high accuracy. Knowledge provided in this work can be easily implemented to improve modelling of fluidized bed reactors applied to advanced thermochemical conversions, such as gasification and pyrolysis, of biomass and waste materials. • Devolatilization of biomass in fluidized beds generates endogenous bubbles • Advanced X-ray imaging provided comprehensive assessment of endogenous bubbles • Bubbles void fraction distribution is obtained by means of the Beer-Lambert law • Endogenous bubbles consist of a wake, cloud, and central gas-rich region • Endogenous bubbles strongly resemble traditional bubbles in bubbling regime

The VISCACHA survey – VI. Dimensional study of the structure of 82 star clusters in the Magellanic Clouds

ABSTRACT We present a study of the internal structure of 82 star clusters located at the outer regions of the Large Magellanic Cloud and the Small Magellanic Cloud using data of the VISCACHA Survey. Through the construction of the minimum spanning tree, which analyses the relative position of stars within a given cluster, it was possible to characterize the internal structure and explore the fractal or subclustered distribution for each cluster. We computed the parameters $\overline{m}$ (which is the average length of the connected segments normalized by the area), $\overline{s}$ (which is the mean points separation in units of cluster radius), and Q (the ratio of these components). These parameters are useful to distinguish between radial, homogeneous, and substructured distributions of stars. The dependence of these parameters with the different characteristics of the clusters, such as their ages and spatial distribution, was also studied. We found that most of the studied clusters present a homogeneous stellar distribution or a distribution with a radial concentration. Our results are consistent with the models, suggesting that more dynamically evolved clusters seem to have larger Q values, confirming previous results from numerical simulations. There also seems to be a correlation between the internal structure of the clusters and their galactocentric distances, in the sense that for both galaxies, the more distant clusters have larger Q values. We also paid particular attention to the effects of contamination by non-member field stars and its consequences finding that field star decontamination is crucial for these kinds of studies.

Marketing Strategy to Improve Brand Awareness: Case Study of a Cloud Service Provider in Indonesia

The World Health Organization classified Covid-19 a global pandemic, its effects have rippled across every aspect of society, including the economy. Firms are shifting their objectives and concentrating on their information technology capability to deal with the unpredictability that this pandemic has brought onto the economy. Cloud and colocation will be required to support company operations with digital infrastructure. Many of the world’s largest cloud service providers are thinking about creating cloud regions within its boundaries as a result of its tremendous development potential. Cloudgear as one of local Indonesian cloud service provider wants to take this opportunity to grow their business. Cloudgear has a comprehensive service portfolio consisting of Infrastructure as a Service (IaaS), Software as a Service (SaaS), Professional as a Software (PaaS), professional services, and multi-cloud services. However, based on the initial interview with ten cloud engineers and specialists in different companies, most of them are not aware that Cloudgear is a local Indonesian cloud service provider. Therefore, the next challenge is how to improve brand awareness of Cloudgear by formulating the right marketing strategy. Because it will strengthen the brand and develop Cloudgear competitive advantage for the company’s future growth. This research is objected to analyse Cloudgear internal, external conditions, analyse the factors that influence the brand awareness, and propose the most suitable marketing strategy in building Cloudgear brand awareness. Brand awareness is affected by advertising, social media, word of mouth and publicity. In this research, advertising, social media, word of mouth and publicity are being analysed whether it has significant positive influence to cloud service providers brand awareness or not. The hypothesis are that each of these factors have positive influence to cloud service brand awareness by performing customers analysis. Questionnaire was distributed to two hundred samples. Then, the data were processed by SPSS using multiple linear regression method. Several tests were conducted consisted of validity test, reliability test, normality test, and linearity test. After the data were valid, reliable, normal, and linear, analysis continued by doing multiple linear regression. The results are there is significant effect of each advertising, social media, word of mouth and publicity to brand awareness of cloud service provider. Determination test is resulted these variables have 50.1% influence. When F Test was conducted, the result was these variables if tested simultaneously or simultaneously have an effect on cloud service brand awareness. Therefore, the proposed marketing strategy for Cloudgear is related to improvement in advertising, social media, word of mouth and publicity in order to increase their brand awareness as a local Indonesian cloud service provider.

Venus Life Finder Habitability Mission: Motivation, Science Objectives, and Instrumentation

For over half a century, scientists have contemplated the potential existence of life within the clouds of Venus. Unknown chemistry leaves open the possibility that certain regions of the Venusian atmosphere are habitable. In situ atmospheric measurements with a suite of modern instruments can determine whether the cloud decks possess the characteristics needed to support life as we know it. The key habitability factors are cloud particle droplet acidity and cloud-layer water content. We envision an instrument suite to measure not only the acidity and water content of the droplets (and their variability) but additionally to confirm the presence of metals and other non-volatile elements required for life’s metabolism, verify the existence of organic material, and search for biosignature gases as signs of life. We present an astrobiology-focused mission, science goals, and instruments that can be used on both a large atmospheric probe with a parachute lasting about one hour in the cloud layers (40 to 60 km) or a fixed-altitude balloon operating at about 52 km above the surface. The latter relies on four deployable mini probes to measure habitability conditions in the lower cloud region. The mission doubles as a preparation for sample return by determining whether a subset of cloud particles is non-liquid as well as characterizing the heterogeneity of the cloud particles, thereby informing sample collection and storage methods for a return journey to Earth.

Computational simulation of the bombardment of molecular clump by realistic cosmic ray field employing GEANT4 code

ABSTRACT Here, we present calculations on the energy delivered (and heating) by realistic cosmic rays (CRs) field at a typical molecular clump. The current model describes, with unprecedented spatial resolution, the energy delivery by a realistic CR field in molecular clumps. The calculations were performed employing the Geant4 code (considering full cascade physical processes and hadron physics) considering the cosmic ray field taken from the Voyager spacecraft measurements in the interstellar medium. The results showed that the total energy deposition rate, considering light particles (protons, electrons and alphas), medium-mass ions and heavy-ions, ranges from 400 MeV/g/s in the outer layer (at 105 AU) to roughly 100 MeV/g/s in the inner layer of the model (below 0.1 AU). The main energy deposition rate is due to the incoming protons. Incoming alphas represent 15–20 per cent of the energy deposition. In the deep core of the cloud, the fraction of energy delivered by medium-mass ions, electrons, and heavy ions are 5 per cent, 2.5 per cent, and 1 per cent, respectively. The heating induced by cosmic rays seems to affect mostly the regions above ∼500 AU. Considering a balanced heat model with warm dust grains (T∼16–18 K), we observe a small bump in temperature at 2000–5000 AU. We suggest this temperature enhancement by CRs might have some affect on the molecular formation or cometary formation in pristine Oort cloud region inside the Solar System.

Dust extinction map of the Galactic plane based on the VVV survey data

ABSTRACT Dust extinction is one of the most reliable tracers of the gas distribution in the Milky Way. The near-infrared (NIR) Vista Variables in the Vía Láctea (VVV) survey enables extinction mapping based on stellar photometry over a large area in the Galactic plane. We devise a novel extinction mapping approach, XPNICER, by bringing together VVV photometric catalogues, stellar parameter data from StarHorse catalogues, and previously published X percentile and PNICER extinction mapping techniques. We apply the approach to the VVV survey area, resulting in an extinction map that covers the Galactic disc between 295° ≲ l ≲ 350° and −2° ≲ b ≲ 2°, and the Galactic bulge between −10° ≲ b ≲ 5°. The map has 30 arcsec spatial resolution and it traces extinctions typically up to AV ∼ 10–20 mag and maximally up to AV ∼ 30 mag. We compare our map to previous dust-based maps, concluding that it provides a high-fidelity extinction-based map, especially in its ability to recover both the diffuse dust component of the Galaxy and moderately extincted giant molecular cloud regions. The map is especially useful as independent, extinction-based data on the Galactic dust distribution and applicable for a wide range of studies from individual molecular clouds to the studies of the Galactic stellar populations.

The homogeneity of chemical abundances in H ii regions of the Magellanic Clouds

ABSTRACT We use very deep spectra obtained with the Ultraviolet-Visual Echelle Spectrograph at the Very Large Telescope to derive physical conditions and chemical abundances of four H ii regions of the Large Magellanic Cloud (LMC) and four H ii regions of the Small Magellanic Cloud (SMC). The observations cover the spectral range 3100–10400 Å with a spectral resolution of Δλ ≥ λ/11600, and we measure 95–225 emission lines in each object. We derive ionic and total abundances of O, N, S, Ne, Ar, Cl, and Fe using collisionally excited lines. We find average values of 12 + log (O/H) = 8.37 in the LMC and 8.01 in the SMC, with standard deviations of σ = 0.03 and 0.02 dex, respectively. The S/O, Ne/O, Ar/O, and Cl/O abundance ratios are very similar in both clouds with σ = 0.02–0.03 dex, which indicates that the chemical elements are well mixed in the interstellar medium of each galaxy. The LMC is enhanced in N/O by ∼0.20 dex with respect to the SMC, and the dispersions in N/O, σ = 0.05 dex in each cloud, are larger than those found for the other elements. The derived standard deviations would be much larger for all the abundance ratios, up to 0.20 dex for N/O, if previous spectra of these objects were used to perform the analysis. Finally, we find a wide range of iron depletions in both clouds with more than 90 per cent of the iron atoms deposited on to dust grains in most objects.

Discovery of non-metastable ammonia masers in Sagittarius B2

We report the discovery of widespread maser emission in non-metastable inversion transitions of NH3 toward various parts of the Sagittarius B2 molecular cloud and star-forming region complex. We detect masers in the J, K = (6, 3), (7,4), (8,5), (9,6), and (10,7) transitions toward Sgr B2(M) and Sgr B2(N), an NH3 (6,3) maser in Sgr B2(NS), and NH3 (7,4), (9,6), and (10,7) masers in Sgr B2(S). With the high angular resolution data of the Karl G. Jansky Very Large Array (JVLA) in the A-configuration, we identify 18 maser spots. Nine maser spots arise from Sgr B2(N), one from Sgr B2(NS), five from Sgr B2(M), and three in Sgr B2(S). Compared to our Effelsberg single-dish data, the JVLA data indicate no missing flux. The detected maser spots are not resolved by our JVLA observations. Lower limits to the brightness temperature are > 3000 K and reach up to several 105 K, manifesting the lines’ maser nature. In view of the masers’ velocity differences with respect to adjacent hot molecular cores and/or UCH II regions, it is argued that all the measured ammonia maser lines may be associated with shocks caused either by outflows or by the expansion of UCH II regions. Overall, Sgr B2 is unique in that it allows us to measure many NH3 masers simultaneously, which may be essential in order to elucidate their thus far poorly understood origin and excitation.

Disentangling emission from star-forming regions in the Magellanic Clouds: Linking [O iii]λ88 µm and 24 µm

Context. The [O iii]λ88 µm line is observed in many galaxies including our neighboring Magellanic Clouds and is a well-known tracer of H ii regions, while the 24 µm continuum emission has often been used to trace warm dust in the ionized phases of galaxies. The association of both the [O iii]λ88 µm line and 24 µm in galaxies to star formation motivates this study to determine their observational relation. Aims. This study explores the link between the [O iii]λ88 µm and 24 µm continuum in star-forming regions in the Magellanic Clouds. We also explore the local conditions driving the relation between those tracers. Methods. We compared observations with 1D Cloudy models consisting of an H ii region plus a photodissociation region (PDR) component, varying the stellar age, the initial density (at the illuminated edge of the cloud), and the ionization parameter. We introduced a new parameter, cPDR, to quantify the proportion of emission arising from PDRs and that with an origin in H ii regions along each line of sight. We used the ratio ([C ii]+[O i])/[O iii] as a proxy for the ratio of PDR versus H ii region emission, and compared it to the [O iii]/24 µm ratio. The use of [O iii]/24 µm and [O iii]/70 µm together allowed us to constrain the models most efficiently. Results. We find a correlation over at least 3 orders of magnitude in [O iii]λ88 µm and 24 µm continuum. This correlation is seen for spatially resolved maps of the Magellanic Cloud regions as well as unresolved galaxy-wide low metallicity galaxies of the Dwarf Galaxy Survey. We also find that most of the regions have low proportions of PDRs along the lines of sight (<12%), while a limited area of some of the mapped regions can reach 30–50%. For most lines of sight within the star-forming regions we have studied in the Magellanic Clouds, H ii regions are the dominant phase. Conclusions. We propose the use of the correlation between the [O iii]λ88 µm and 24 µm continuum as a new predictive tool to estimate, for example, the [O iii]λ88 µm when the 24 µm continuum is available or inversely. This can be especially useful to prepare for Atacama Large Milimeter Array (ALMA) observations of [O iii]λ88 µm in high-z galaxies. The simple and novel method we developed may also provides a way to disentangle different phases along the line of sight, when other 3D information is not available.

DcTr: Noise-robust point cloud completion by dual-channel transformer with cross-attention

Current point cloud completion research mainly utilizes the global shape representation and local features to recover the missing regions of 3D shape for the partial point cloud. However, these methods suffer from inefficient utilization of local features and unstructured points prediction in local patches, hardly resulting in a well-arranged structure for points. To tackle these problems, we propose to employ Dual-channel Transformer and Cross-attention (CA) for point cloud completion (DcTr). The DcTr is apt at using local features and preserving a well-structured generation process. Specifically, the dual-channel transformer leverages point-wise attention and channel-wise attention to summarize the deconvolution patterns used in the previous Dual-channel Transformer Point Deconvolution (DCTPD) stage to produce the deconvolution in the current DCTPD stage. Meanwhile, we employ cross-attention to convey the geometric information from the local regions of incomplete point clouds for the generation of complete ones at different resolutions. In this way, we can generate the locally compact and structured point cloud by capturing the structure characteristic of 3D shape in local patches. Our experimental results indicate that DcTr outperforms the state-of-the-art point cloud completion methods under several benchmarks and is robust to various kinds of noise.

FaaScinating Resilience for Serverless Function Choreographies in Federated Clouds

Cloud applications often benefit from deployment on serverless technology Function-as-a-Service (FaaS), which may instantly spawn numerous functions and charges users for the period when serverless functions are running. Maximum benefit is achieved when functions are orchestrated in a workflow or function choreographies (FCs). However, many provider limitations specific for FaaS, such as maximum concurrency or duration often increase the failure rate, which can severely hamper the execution of entire FCs. Current support for resilience is often limited to function retries or try-catch, which are applicable within the same cloud region only. To overcome these limitations, we introduce rAFCL, a middleware platform that maintains reliability of complex FCs in federated clouds. In order to support resilient FC execution under rAFCL, our model creates an alternative strategy for each function based on the required availability specified by the user. Alternative strategies are not restricted to the same cloud region, but may contain alternative functions across five providers, invoked concurrently in a single alternative plan or executed subsequently in multiple alternative plans. With this approach, rAFCL offers flexibility in terms of cost-performance trade-off. We evaluated rAFCL by running three real-life applications across three cloud providers. Experimental results demonstrated that rAFCL outperforms the resilience of AWS Step Functions, increasing the success rate of entire FC by 53.45%, while invoking only 3.94% more functions with zero wasted function invocations. rAFCL significantly improves availability of entire FCs to almost 1 and survives even after massive failures of alternative functions.

Relating Vertical Velocity and Cloud/Precipitation Properties: A Numerical Cloud Ensemble Modeling Study of Tropical Convection

AbstractFundamental relationships exist between cloud and precipitation development and their dynamic processes. Latent heat released by cloud/precipitation formation affects cloud vertical motions, which in turn affect convective cloud development. Here, a cloud‐resolving model is used to relate cloud properties and latent heating with cloud drafts using 15‐day simulations for an oceanic and continental environment. The results show condensation, deposition, and freezing occur mainly in moderate (3–5 m s−1) to strong (>10 m s−1) updrafts, evaporation and sublimation mainly in weak (1–2 m s−1) to moderate downdrafts, and melting in moderate updrafts and downdrafts. Active updrafts cover only a small percentage of the model domain but contribute significantly to the latent heat release and are associated with large proportions of the hydrometeors. Active updrafts with vertical velocities exceeding 1 and 2 m s−1 account for more than 75% and 50%, respectively, of the condensation, deposition, and freezing in both the oceanic and continental cases. However, active downdrafts with vertical velocity magnitudes exceeding |1 m s−1| account for less than 40% and 25%, respectively, of the evaporation and sublimation. More evaporation and sublimation than condensation and deposition occur in the inactive cloud regions. Sensitivity tests are also conducted to assess the impact of model grid spacing (1,000 m vs. 250 m) and microphysical schemes (3 ice classes vs. 4 ice classes) on latent heat release and hydrometeor amount. The results show that model resolution had more impact than the microphysics on the simulated cloud properties in both cases.

Optical Remote Sensing Image Cloud Detection with Self-Attention and Spatial Pyramid Pooling Fusion

Cloud detection is a key step in optical remote sensing image processing, and the cloud-free image is of great significance for land use classification, change detection, and long time-series landcover monitoring. Traditional cloud detection methods based on spectral and texture features have acquired certain effects in complex scenarios, such as cloud–snow mixing, but there is still a large room for improvement in terms of generation ability. In recent years, cloud detection with deep-learning methods has significantly improved the accuracy in complex regions such as high-brightness feature mixing areas. However, the existing deep learning-based cloud detection methods still have certain limitations. For instance, a few omission alarms and commission alarms still exist in cloud edge regions. At present, the cloud detection methods based on deep learning are gradually converted from a pure convolutional structure to a global feature extraction perspective, such as attention modules, but the computational burden is also increased, which is difficult to meet for the rapidly developing time-sensitive tasks, such as onboard real-time cloud detection in optical remote sensing imagery. To address the above problems, this manuscript proposes a high-precision cloud detection network fusing a self-attention module and spatial pyramidal pooling. Firstly, we use the DenseNet network as the backbone, then the deep semantic features are extracted by combining a global self-attention module and spatial pyramid pooling module. Secondly, to solve the problem of unbalanced training samples, we design a weighted cross-entropy loss function to optimize it. Finally, cloud detection accuracy is assessed. With the quantitative comparison experiments on different images, such as Landsat8, Landsat9, GF-2, and Beijing-2, the results indicate that, compared with the feature-based methods, the deep learning network can effectively distinguish in the cloud–snow confusion-prone region using only visible three-channel images, which significantly reduces the number of required image bands. Compared with other deep learning methods, the accuracy at the edge of the cloud region is higher and the overall computational efficiency is relatively optimal.

Distributed YSOs in the Perseus Molecular Cloud from the Gaia and LAMOST Surveys

Identifying the young optically visible population in a star-forming region is essential for fully understanding the star formation event. In this paper, we identify 211 candidate members of the Perseus molecular cloud based on Gaia astrometry. We use LAMOST spectra to confirm that 51 of these candidates are new members, bringing the total census of known members to 856. The newly confirmed members are less extincted than previously known members. Two new stellar aggregates are identified in our updated census. With the updated member list, we obtain a statistically significant distance gradient of 4.84 pc deg−1 from west to east. Distances and extinction corrected color–magnitude diagrams indicate that NGC 1333 is significantly younger than IC 348 and the remaining cloud regions. The disk fraction in NGC 1333 is higher than elsewhere, consistent with its youngest age. The star formation scenario in the Perseus molecular cloud is investigated and the bulk motion of the distributed population is consistent with the cloud being swept away by the Per-Tau Shell.

Cloud removal for optical remote sensing imagery using the SPA-CycleGAN network

Optical remote sensing images play an important role in many fields due to their natural visual representations. However, clouds obstruct a large proportion of recorded spectral signals and greatly affect the availability of high-quality optical images. In addition, the progress of cloud removal methods has long been restricted by the deficiency in theoretical basis and low computing power. With the advent of deep learning technology, this problem can be well addressed. We propose a network based on style transfer theory to remove thin and thick clouds in optical remote sensing images. An input data concatenation procedure with synthetic aperture radar (SAR) images is used to provide auxiliary information for the recovery of missing data caused by clouds. Moreover, the spatial attention mechanism adept at generating saliency information of image objects is also used to locate the clouds and corresponding shadows to make the network concentrate on the cloud region reconstruction task. Furthermore, to make the spatial attention mechanism self-adapted to the practical characteristics of clouded images, a loss function including a cloud mask is designed to assist the training step of the network. The proposed network is trained with a global open-source dataset and a region-specified dataset. The results of ablation experiments and comparison with baseline methods demonstrate that a high accuracy is guaranteed by the proposed model, even under high cloud coverage conditions [mean absolute error (MAE) = 0.0883, root-mean-square error (RMSE) = 0.1775, peak signal-to-noise ratio (PSNR) = 23.8, and structural similarity (SSIM) = 0.742). Using optical images independently to remove clouds results in poor results (MAE = 0.1156, RMSE = 0.2825, PSNR = 21.6, and SSIM = 0.596), which need to be addressed with data concatenation, such as SAR context information. Spatial attention mechanism plays an important role in improving the accuracy of cloud removal (MAE = 0.0251, RMSE = 0.0363, PSNR = 27.1, and SSIM = 0.876). For the missing values of optical images caused by cloud coverage, the Spatial-attention-CycleGAN proposed provides a potential solution for information recovery.

Introducing Improved Transformer to Land Cover Classification Using Multispectral LiDAR Point Clouds

The use of Transformer-based networks has been proposed for the processing of general point clouds. However, there has been little research related to multispectral LiDAR point clouds that contain both spatial coordinate information and multi-wavelength intensity information. In this paper, we propose networks for multispectral LiDAR point cloud point-by-point classification based on an improved Transformer. Specifically, considering the sparseness of different regions of multispectral LiDAR point clouds, we add a bias to the Transformer to improve its ability to capture local information and construct an easy-to-implement multispectral LiDAR point cloud Transformer (MPT) classification network. The MPT network achieves 78.49% mIoU, 94.55% OA, 84.46% F1, and 0.92 Kappa on the multispectral LiDAR point cloud testing dataset. To further extract the topological relationships between points, we present a standardization set abstraction (SSA) module, which includes the global point information while considering the relationships among the local points. Based on the SSA module, we propose an advanced version called MPT+ for the point-by-point classification of multispectral LiDAR point clouds. The MPT+ network achieves 82.94% mIoU, 95.62% OA, 88.42% F1, and 0.94 Kappa on the same testing dataset. Compared with seven point-based deep learning algorithms, our proposed MPT+ achieves state-of-the-art results for several evaluation metrics.

Identification of synoptic patterns for extreme rainfall events associated with landfalling typhoons in China during 1960–2020

Extreme rainfall associated with landfalling typhoon (ERLTC) can cause severe disasters and economic impacts throughout China. Improving the accuracy of ERLTC forecasts is therefore crucial in disaster prevention and mitigation. The top 26 ERLTC events in China during 1960–2020 are investigated based on multi-source datasets. These ERLTC events are categorized into five main types according to the geographical location of the extreme precipitation and its position relative to the tropical cyclone (TC) center, namely: the typhoon inner-core rainfall in Taiwan (TWIC), typhoon inverted trough rainfall in Taiwan (TWIT), weak typhoon rainfall in Hainan (HNWK), strong typhoon rainfall in Zhejiang (ZJST) and inland typhoon remnant rainfall (ILRM). All the ERLTC events occurred in the weakening stage of TC after reaching its lifetime maximum intensity in convective cloud (TBB ≤ −32 °C) regions over complex local terrain. The translational speeds of 20 TCs (76.9% of the total) were smaller than the climatological average (20.6 km h−1) during the extreme precipitation events. The differences are as follows: the TWIC and TWIT types are featured with different season, track and water vapor channel although both occurred in Taiwan. The other three types are distinguished by spinning track and strong convective cloud for HNWK type, strong TC intensity and binary TC interactions for ZJST type; and stagnation and strong westerly trough activity for ILRM type, respectively. These results are expected to provide useful clues for an in-depth understanding of ERLTC events over China.

Macro- and Micro-physical Characteristics of Different Parts of Mixed Convective-stratiform Clouds and Differences in Their Responses to Seeding

This study investigates the cloud macro- and micro-physical characteristics in the convective and stratiform regions and their different responses to the seeding for mixed convective-stratiform clouds that occurred in Shandong province on 21 May 2018, based on the observations from the aircraft, the Suomi National Polar-Orbiting Partnership (NPP) satellite, and the high-resolution Himawari-8 (H8) satellite. The aircraft observations show that convection was deeper and radar echoes were significantly enhanced with higher tops in response to seeding in the convective region. This is linked with the conversion of supercooled liquid droplets to ice crystals with released latent heat, resulting in strengthened updrafts, enhanced radar echoes, higher cloud tops, and more and larger precipitation particles. In contrast, in the stratiform cloud region, after the Silver Iodide (AgI) seeding, the radar echoes become significantly weaker at heights close to the seeding layer, with the echo tops lowered by 1.4–1.7 km. In addition, a hollow structure appears at the height of 6.2–7.8 km with a depth of about 1.6 km and a diameter of about 5.5 km, and features such as icing seeding tracks appear. These suggest that the transformation between droplets and ice particles was accelerated by the seeding in the stratiform part. The NPP and H8 satellites also show that convective activity was stronger in the convective region after seeding; while in the stratiform region, a cloud seeding track with a width of 1–3 km appears 10 km downstream of the seeding layer 15 minutes after the AgI seeding, which moves along the wind direction as width increases.

Analysis of the Characteristics and Evolution Mechanisms of a Bow-Shaped Squall Line in East China Observed with Dual-Polarization Doppler Radars

To gain a deeper understanding of the formation and evolutionary mechanisms of a bow-shaped squall line (BSL) that occurred in East China on 10 May 2021, observations from S-band dual-polarization radars, a disdrometer and other instruments are used to investigate the characteristics and evolution of the kinematic, microphysical and radar echo structure within the squall line during its formative and mature stages. The results are as follows. The updraft induced by upper-level divergence and vertical thermal instability induced by the cold source at the middle and top of the troposphere provided environmental conditions suitable for the formation and strengthening of a squall line. The characteristics of the vertical vorticity at the leading edge of the squall line provided a good indication of its echo structure and evolutionary trend. The mechanism behind a new echo phenomenon—double high-differential reflectivity (ZDR) bands—observed in plan position indicator scans produced by the dual-polarization radar is investigated from the kinematic and microphysical structural perspectives. The evolutionary characteristics of the microphysical structure of the bulk of the squall line and its trailing stratiform cloud region are analyzed based on the quasi-vertical profiles retrieved from the S-band dual-polarization radar in Quzhou. Moreover, a conceptual model describing this type of BSL with a trailing region of stratiform rain in the warm sector is developed to provide technical support for the monitoring and early warning of BSLs.