Search

Precipitation images play an important role in meteorological forecasting and flood forecasting, but how to characterize precipitation images and conduct rainfall similarity analysis is challenging and meaningful work. This paper proposes a rainfall similarity research method based on deep learning by using precipitation images. The algorithm first extracts regional precipitation, precipitation distribution, and precipitation center of the precipitation images and defines the similarity measures, respectively. Additionally, an ensemble weighting method of Normalized Discounted Cumulative Gain-Improved Particle Swarm Optimization (NDCG-IPSO) is proposed to weigh and fuse the three extracted features as the similarity measure of the precipitation image. During the experiment on similarity search for daily precipitation images in the Jialing River basin, the NDCG@10 of the search results reached 0.964, surpassing other methods. This indicates that the method proposed in this paper can better characterize the spatiotemporal characteristics of the precipitation image, thereby discovering similar rainfall processes and providing new ideas for hydrological forecasting.

Precipitation images play an important role in meteorological forecasting and flood forecasting, but how to characterize precipitation images and conduct rainfall similarity analysis is challenging and meaningful work. This paper proposes a rainfall similarity research method based on deep learning by using precipitation images. The algorithm first extracts regional precipitation, precipitation distribution, and precipitation center of the precipitation images and defines the similarity measures, respectively. Additionally, an ensemble weighting method of Normalized Discounted Cumulative Gain-Improved Particle Swarm Optimization (NDCG-IPSO) is proposed to weigh and fuse the three extracted features as the similarity measure of the precipitation image. During the experiment on similarity search for daily precipitation images in the Jialing River basin, the NDCG@10 of the search results reached 0.964, surpassing other methods. This indicates that the method proposed in this paper can better characterize the spatiotemporal characteristics of the precipitation image, thereby discovering similar rainfall processes and providing new ideas for hydrological forecasting. [ABSTRACT FROM AUTHOR]

Parameters of the normalized gamma particle size distribution (PSD) have been retrieved from the Precipitation Image Package (PIP) snowfall observations collected during the International Collaborative Experiment–PyeongChang Olympic and Paralympic winter games (ICE-POP 2018). Two of the gamma PSD parameters, the mass-weighted particle diameter Dmass and the normalized intercept parameter NW, have median values of 1.15–1.31 mm and 2.84–3.04 log(mm−1 m−3), respectively. This range arises from the choice of the relationship between the maximum versus equivalent diameter, Dmx–Deq, and the relationship between the Reynolds and Best numbers, Re–X. Normalization of snow water equivalent rate (SWER) and ice water content W by NW reduces the range in NW, resulting in well-fitted power-law relationships between SWER/NW and Dmass and between W/NW and Dmass. The bulk descriptors of snowfall are calculated from PIP observations and from the gamma PSD with values of the shape parameter μ ranging from −2 to 10. NASA's Global Precipitation Measurement (GPM) mission, which adopted the normalized gamma PSD, assumes μ = 2 and 3 in its two separate algorithms. The mean fractional bias (MFB) of the snowfall parameters changes with μ, where the functional dependence on μ depends on the specific snowfall parameter of interest. The MFB of the total concentration was underestimated by 0.23–0.34 when μ = 2 and by 0.29–0.40 when μ = 3, whereas the MFB of SWER had a much narrower range (from −0.03 to 0.04) for the same μ values. [ABSTRACT FROM AUTHOR]

Providing reliable and accurate high-resolution meteorological data is very significant to guide the rapid response to extreme weather conditions. However, due to the constraints of computing power and simulation time, the spatial resolution of existing all global climate models is low, and it is unable to provide meteorological data with more precise resolution at local scale. In this research, a deep learning downscaling model called two-stage multi-scale feature extraction network (TSMFN) is proposed. By combining ERA5 reanalysis data and terrain data, spatial downscaling of global precipitation measurement mission precipitation data is carried out. Specifically, in the first stage of the network, several multi-scale residual Inception blocks are used to extract multi-scale features of low-resolution precipitation data; several residual-based residual multi-scale cross blocks are used to fully excavate multi-scale features after the fusion of multiple data. In the second stage of the TSMFN, the output feature map after the fusion of the previous stage is fused with the high-resolution monthly average precipitation data, and several progressive multistage Swin Transformer blocks are constructed to overcome the problems that the reconstructed image of a general convolutional neural network is smooth and cannot reflect the real spatial distribution of precipitation. Finally, a hybrid loss function combining L 1 loss function and focal frequency loss function is proposed to alleviate the ill-posedness of the downscaling. By comparing the proposed algorithm with some advanced deep learning downscaling algorithms, the results of the experiment show that the TSMFN model is significantly better in terms of generated image quality and multiple evaluation indexes, and the model has stronger generalization ability. [ABSTRACT FROM AUTHOR]

Abstract Radar observation variables reflect the precipitation amount of strong convective precipitation processes, which accurate forecast is an important difficulty in weather forecasting. Current forecasting methods are mostly based on radar echo extrapolation, which has the insufficiency of input information and the ineffectiveness of model architecture. This paper presents a Bidirectional Long Short-Term Memory forecasting method for strong convective precipitation based on the attention mechanism and residual neural network (ResNet-Attention-BiLSTM). First, this paper uses ResNet to effectively extract the key information of extreme weather and solves the problem of regression to the mean of the prediction model by learning the residuals of the radar observation data. Second, this paper uses the attention mechanism to adaptively weight the fusion of the features to enhance the extraction of the important features of the precipitation image data. On this basis, this paper presents a novel spatio-temporal reasoning method for radar observations and establishes a precipitation forecasting model, which captures the past and future time-order relationship of the sequence data. Finally, this paper conducts experiments based on the real collected data of a strong convective precipitation process and compares its performance with the existing models, the mean absolute percentage error of this model was reduced by 15.94% (1 km), 18.72% (3 km), and 14.91% (7 km), and the coefficient of determination ( $$R^{2}$$ R 2 ) was increased by 10.89% (1 km), 9.61% (3 km), and 9.29% (7 km), which proves the state of the art and effectiveness of this forecasting model.

This paper presents the first method for synthesizing seasonal transition of terrain textures for an input heightfield. Our method reproduces a seamless transition of terrain textures according to the seasons by learning measured data on the earth using a convolutional neural network. We attribute the main seasonal texture transition to vegetation and snow, and control the texture synthesis not only with the input heightfield but also with the annual temperature and precipitation based on Köppen's climate classification as well as insolation at the location. We found that month-by-month synthesis yields incoherent transitions, while a naïve conditioning with explicit temporal information (e.g., month) degrades generalizability due to the north–south hemisphere difference. To address these issues, we introduce a simple solution—periodic conditioning on the annual data without explicit temporal information. Our experiments reveal that our method can synthesize plausible seasonal transitions of terrain textures. We also demonstrate large-scale texture synthesis by tiling the texture output. [ABSTRACT FROM AUTHOR]

Tetanus is a life-threatening bacterial disease in humans and many animal species caused by the neurotoxin tetanospasmin produced by Clostridium tetani. Antitoxins obtained from horses and humans are primarily used to treat this disease. However, there are several clinical side effects and disadvantages associated with the use of these antitoxins. Current techniques for diagnosing tetanus use monoclonal antibodies produced in mice. These antibodies have several advantages, such as their homogeneity and specificity. In contrast, a notable feature of polyclonal antibodies, especially egg yolk antibodies, i.e. immunoglobulin-Y (IgY), extracted from poultry, is that they can be generated in greater quantities than mammalian antibodies (IgG). In this study, 22-week-old chickens were immunized with C. tetani toxoid and adjuvant (Freund's complete and incomplete adjuvants) via injection into the chest muscle. The immunization process was completed by administering two booster injections at 4-week intervals. Total antibody titers were observed to reach their highest level in the serum of blood samples taken 14 days after the last immunization. IgY antibodies were isolated noninvasively from the eggs of immunized and nonimmunized chickens using the polyethylene glycol (6000) extraction protocol. Immunological analyses confirmed that the purified IgY antibodies were produced specifically for the C. tetani toxoid. The specific tetanus antibodies obtained in this study may be valuable therapeutic tools as alternatives to current treatments for tetanus in humans and domestic animals. [ABSTRACT FROM AUTHOR]

The Jambato Harlequin toad (Atelopus ignescens), a formerly abundant species in the Andes of Ecuador, faced a dramatic population decline in the 1980s, with its last recorded sighting in 1988. The species was considered Extinct by the IUCN until 2016, when a fortuitous discovery of one Jambato by a local boy reignited hope. In this study, we present findings from an investigation conducted in the Angamarca parish, focusing on distribution, abundance, habitat preferences, ecology, disease susceptibility, and dietary habits of the species. In one year we identified 71 individuals at different stages of development in various habitats, with a significant presence in agricultural mosaic areas and locations near water sources used for crop irrigation, demonstrating the persistence of the species in a complex landscape, with considerable human intervention. The dietary analysis based on fecal samples indicated a diverse prey selection, primarily comprising arthropods such as Acari, Coleoptera, and ants. Amphibian declines have been associated with diseases and climate change; notably, our study confirmed the presence of the pathogen Batrachochytrium dendrobatidis (Bd), but, surprisingly, none of the infected Jambatos displayed visible signs of illness. When analyzing climatic patterns, we found that there are climatic differences between historical localities and Angamarca; the temporal analysis also exposes a generalized warming trend. Finally, in collaboration with the local community, we developed a series of management recommendations for terrestrial and aquatic environments occupied by the Jambato. [ABSTRACT FROM AUTHOR]

In a Lagrangian frame of reference, the accuracy of rainfall systems predicted by nowcasting algorithms can be improved by incorporating the growth and decay of the rainfall. The scale dependence of predictability of growth and decay of continental-scale precipitating systems is studied with the help of the U.S. national radar composites. The growth and decay of precipitating systems is estimated in a time interval t by correcting the precipitation image for advection and rotation at time t + τ with respect to the precipitation image at time t and then subtracting the former from the latter. Results show that the two-dimensional correlation of growth and decay has an elliptical structure, indicating that growth and decay is nonisotropic. The probability density function of precipitation intensities and of growth and decay follows a Gaussian distribution. The scale-dependence analysis of growth and decay patterns indicates that the growth and decay of rainfall may be predictable up to about 2 h for scales larger than 250 km. [ABSTRACT FROM AUTHOR]

CO2 flooding is an economically feasible and preferred carbon capture, storage, and utilization technology. Asphaltene deposition is a common problem in the process of CO2 injection because it may cause reservoir damage. The mechanism of asphaltene precipitation damage to the formation remains elusive. Experiments were conducted to reveal the pore-scale formation damage mechanism in ultra-low permeability reservoirs caused by asphaltene precipitation during CO2 flooding. Initially, the precipitation onset point for asphaltene within the crude oil-CO2 system was determined using a high-pressure tank equipped with visual capabilities. Subsequently, CO2 flooding experiments were conducted on ultra-low permeability cores under miscible and immiscible conditions, with the support of nuclear magnetic resonance (NMR) to quantitatively evaluate the impact of asphaltene precipitation on ultra-low permeability reservoirs. The results indicate that within the pressure range from the asphaltene precipitation onset point to the minimum miscibility pressure (MMP). The level of asphaltene precipitation rises as CO2 injection pressure increases. In the miscible flooding stage, asphaltene precipitation can still occur, but to a lesser extent. Notably, asphaltene deposition predominantly occurs in larger pores; above the MMP, the permeability decreases significantly as asphalt particles agglomerate, resulting in notable pore-throat blockages. While asphaltene deposition has a minimal impact on porosity, the bridging effect of asphaltene particles reduces permeability. [ABSTRACT FROM AUTHOR]

The synergistic effects of boron (B) and rare earth (RE) elements on the microstructure and stress rupture properties were investigated in a Ni-based superalloy. The stress rupture lifetime at 650 °C/873 MPa significantly increased with the addition of B as a single element. Furthermore, the stress rupture lifetime reached its peak (303 h), with a certain amount of B and RE added together in test alloys. Although the grain size and morphology of the γ′ phase varied a little with the change in B and RE addition, they were not considered to be the main reasons for stress rupture performance. The enhancement in stress rupture lifetime was mostly attributed to the segregation of the B and RE elements, which increased the binding force of the grain boundary and improved its strength and plasticity. In addition, the enrichment of B and RE inhabited the precipitation of carbides along grain boundaries. Furthermore, nano-scale RE precipitates containing sulfur (S) and phosphorus (P) were observed to be distributed along the grain boundaries. The purification of grain boundaries by B and RE elements was favorable to further improve the stress rupture properties. [ABSTRACT FROM AUTHOR]

This paper argues for the absenting presence of an embodied theoretical architecture of materialist harmonious dialectic in Hubert Damisch's engagement with Chinese art and architecture around the cloud. This connects with recentmedia studies discourse where environment both conditions and is being with an alternative architecture that is different from a centralized elevated arche. At the same time, the essay positively explores the decolonizing structural potential fordisplacingEurocentrismandidealisminthemaking of theory and art and architecture history by opening up to global collaborative knowledge formation and reconnecting theory with practice. [ABSTRACT FROM AUTHOR]

Identifying the defects process in concurrent recrystallization and precipitation during aging a supersaturated solid solution is essential for understanding their interaction mechanisms and for manipulating the microstructure, but was rarely done at the atomic scale. Herein, the concurrent recrystallization and precipitation in the supersaturated hexagonal Sm(Co, Fe, Cu, Zr) 7.5 alloys were studied through detailed transmission electron microscopy investigations, where the recrystallization, growth of recrystallized subgrains (cells) and precipitates stem from the gradual formation and dissociation of defects, including basal stacking faults (SFs), vacancies and excess interstitial atoms. The diffusion-controlled glides of < a >-type partial dislocations associated with the SFs not only transform the matrix from the mixture of hexagonal Sm(Co, Fe, Cu, Zr) 7 (1:7H) and Sm 2 (Co, Fe, Cu, Zr) 17 (2:17H) to Sm-depleted rhombohedral Sm 2 (Co, Fe, Cu, Zr) 17 (2:17R) cells but also provide continuous diffusion channels to reduce the point defects to form the Sm-enriched hexagonal Sm(Co, Fe, Cu, Zr) 5 (1:5H) cell boundary precipitates and Zr-enriched rhombohedral (Sm, Zr)(Co, Fe, Cu) 3 (1:3R) platelets. It indicates a diffusion-controlled displacive phase transformation mechanism, characterized by the composition-dependent 2:17R' intermediate phase due to incomplete basal slip and incomplete solute partitioning. The growth velocities of both recrystallized cells and precipitates are closely related to the defects density, being faster due to the high defects density at early stage, and being slower due to the reduced defects density at later stage. A basal slip model is proposed to explain the formation and dissociation of defects along with the stacking period change and the simultaneous formation of continuous atomic diffusion channels. These new findings may yield a deep understanding of the interaction between recrystallization and precipitation. Image, graphical abstract [ABSTRACT FROM AUTHOR]

In hypothetical accidental conditions, zirconium-based nuclear fuel claddings can absorb high hydrogen contents (up to several thousand wppm) and be exposed to high temperatures (β Zr phase temperature range) before being cooled. This paper thoroughly investigates the microstructural and microchemical evolutions that take place in such conditions. Two zirconium-based alloys and unalloyed zirconium were pre-charged with hydrogen at various contents up to 3300 wppm and heat-treated at 1000 or 850 °C. Neutron diffraction analyses were performed in situ upon slow cooling from 700 °C and at room temperature. In the materials containing 3300 wppm of hydrogen, β Zr progressively transforms into α Zr during slow cooling then extensively transforms into α Zr and δ ZrH2-x hydrides precipitate via a eutectoid reaction. Thermodynamic predictions at equilibrium are in good agreement with the experimental results. However, depending on the cooling scenario and the average hydrogen content, the precipitation of γ ZrH hydrides, potentially metastable, is evidenced below 350 °C and a significant amount of hydrogen can remain in solid solution in α Zr. These metallurgical evolutions and the evolution of the different phase lattice parameters are strongly influenced by the partitioning of oxygen and hydrogen (revealed by electron probe and elastic recoil detection microanalyses) that occurs during the β Zr to α Zr transformation and hydride precipitation. Image, graphical abstract [ABSTRACT FROM AUTHOR]

The thermodynamic properties of α-Al and other phases (GP zones, θ'', θ' and θ) in the Al-rich part of the Al-Cu system have been obtained by means of the cluster expansion formalism in combination with statistical mechanics. This information was used to build the Al-rich part of the Al-Cu phase-diagram taking into account vibrational entropic contributions for θ', as those of the other phases were negligible. The simulation predictions of the phase boundaries between α-Al and either θ'', θ' or θ phases as a function of temperature are in good agreement with experimental data and extend the phase boundaries to a wider temperature range. The DFT calculations reveal the presence of a number of metastable Guinier-Preston-zone type configurations that may coexist with α-Al and θ'' at low temperatures. They also demonstrate that θ' is the stable phase below 550K but it is replaced by θ above this temperature due to the vibrational entropic contribution to the Gibbs energy of θ'. This work shows how the combination of cluster expansion and statistical mechanics can be used to expand our knowledge of the phase diagram of metallic alloys and to provide Gibbs free energies of different phases that can be used as input in mesoscale simulations of precipitation. Image, graphical abstract [ABSTRACT FROM AUTHOR]

• Contents of As, Cd, Co, Cu, Ni, U and rare earth elements are high in mine waters. • Plants in mining area are enriched in As, Cr, Cu, Ni, Pb, U and REE. • Moss samples show the highest concentrations of determined elements. • Medium REE anomalies (Sm through Dy) are found in plants and mine waters. • The δ34S values in plants imprint stable S isotope ratios of sulfates in AMD waters. The abundances of trace elements, a low pH of water and soil in areas impacted by the acid mine drainage (AMD) may cause an excessive uptake of potentially toxic elements and nutritional imbalances in plants. Metal-tolerant, native plants are used for revegetation of degraded mining areas. We established levels of selected trace elements and stable sulfur isotopes in the above-ground plant biomass collected in a mining area in south-central Poland. In 2016, 20 samples of the most common species were collected from sites with a different influence of acid mine drainage and analyzed for trace elements by the inductively coupled plasma mass spectrometry technique. On the basis of the results obtained in 2016, the most contaminated site was selected for a more detailed study, in which sulfur contents and stable sulfur isotope ratios were determined together with trace elements in 17 samples. The results confirmed that the plants native to the AMD area efficiently accumulated trace elements, especially As and rare earth elements. Mosses showed the highest content of trace elements, but exhibited the lowest concentrations of sulfur accompanied by the highest δ34S values. It has been shown for the first time that stable sulfur isotope composition of AMD plants in south-central Poland is significantly depleted in the 34S isotope showing an average δ34S value of –10.5‰ in comparison with positive δ34S values in local vegetation growing outside the AMD area and in local precipitation. Image, graphical abstract [ABSTRACT FROM AUTHOR]

An initially single phase high entropy alloy (HEA) Al 0.3 CoCrFeNi was irradiated by 3 MeV Au ions to a fluence of 6 × 1015 cm−2 (∼31 dpa at damage peak) at four different temperatures ranging from 250 °C to 650 °C. Transmission electron microscopy (TEM) and Atom probe tomography (APT) were employed to study the evolution of structural damage and phase stability with irradiation temperature. Al 0.3 CoCrFeNi exhibited a similar evolution of irradiation-induced defects with temperature as compared with conventional FCC alloys. At 250 °C and 350 °C, most of the visible irradiation-induced defects were faulted 1/3〈111〉 dislocation loops. As the irradiation temperature increased to 500 °C, perfect 1/2〈110〉 dislocation loops were observed along with the faulted loops. At the highest irradiation temperature 650 °C, only dislocation lines and networks could be observed. Regarding phase stability, the 3 MeV Au irradiation was observed to suppress the precipitation of (Ni, Al)-enriched nano clusters and the L12 ordered structure at irradiation temperatures 250 °C to 500 °C whereas precipitation of the B2 ordered structure was accelerated at 650 °C. This resulted in qualitatively opposite precipitation behavior between the ion irradiated damage region and unirradiated region at 500 °C and 650 °C. The opposite phase stability of the ion-irradiated damage region and unirradiated region at different temperatures is attributed to the competing effects of ballistic dissolution vs irradiation enhanced diffusion on precipitation. Image, graphical abstract [ABSTRACT FROM AUTHOR]

Hourly precipitation forecasting is considered a spatiotemporal sequence forecasting problem that plays an increasingly important role in early warning of rainfall-induced floods and secondary disasters. Current popular precipitation forecasting methods usually ignore the spatial autocorrelation features, resulting in limited spatial information representations and extractions. In this work, a deformable convolution long short-term memory model considering spatial autocorrelation (DConvLSTM-SAC) is proposed for short-term precipitation forecasting. The proposed model extracts irregularly distributed precipitation information by deformable convolution with better generalization ability and calculates the spatial autocorrelation of inputs by calculating the local Moran index of precipitation images to fully utilize the spatial distribution information of features. In the experiment, the consecutive 3-h precipitation image sequences are used as the input data for the model to predict the precipitation images for the next 3 h. Four commonly used heuristic forecasting methods are adopted as baseline models, including decision tree regression (DTR), random forest regression (RF), recurrent neural network (RNN) and convolutional long short-term memory (ConvLSTM) methods. The proposed DConvLSTM-SAC model exhibits higher predictive accuracy than the four baseline models during the experiments and effectively improves the spatial patterns, with the most significant improvement in the first-hour lead time. The average R2 values for the first-hour prediction are 0.842, 0.838, 0.834, 0.850 and 0.892 for the DTR, RF, RNN, ConvLSTM and DConvLSTM-SAC models, respectively. As a result, the average R2 increased by 4.96%, and the average RMSE (MAE) decreased by 15.21% (9.01%) for the proposed DConvLSTM-SAC model relative to the ConvLSTM method during the first-hour prediction. • Deformable convolution is better suited to extract precipitation information. • Spatial autocorrelation fully utilizes the aggregation information of precipitation. • The DConvLSTM-SAC model outperforms DTR, RF, RNN and ConvLSTM predictions. [ABSTRACT FROM AUTHOR]

The optical flow technique has advantages in motion tracking and has long been employed in precipitation nowcasting to track the motion of precipitation fields using ground radar datasets. However, the performance and forecast time scale of models based on optical flow are limited. Here, we present the results of the application of the deep learning method to optical flow estimation to extend its forecast time scale and enhance the performance of nowcasting. It is shown that a deep learning model can better capture both multispatial and multitemporal motions of precipitation events compared with traditional optical flow estimation methods. The model comprises two components: 1) a regression process based on multiple optical flow algorithms, which more accurately captures multispatial features compared with a single optical flow algorithm; and 2) a U-Net-based network that trains multitemporal features of precipitation movement. We evaluated the model performance with cases of precipitation in South Korea. In particular, the regression process minimizes errors by combining multiple optical flow algorithms with a gradient descent method and outperforms other models using only a single optical flow algorithm up to a 3-h lead time. Additionally, the U-Net plays a crucial role in capturing nonlinear motion that cannot be captured by a simple advection model through traditional optical flow estimation. Consequently, we suggest that the proposed optical flow estimation method with deep learning could play a significant role in improving the performance of current operational nowcasting models, which are based on traditional optical flow methods. Significance Statement: The purpose of this study is to improve the accuracy of short-term rainfall prediction based on optical flow methods that have been employed for operational precipitation nowcasting. By utilizing open-source libraries, such as OpenCV, and commonly applied machine learning techniques, such as multiple linear regression and U-Net networks, we propose an accessible model for enhancing prediction accuracy. We expect that the improvement in prediction accuracy will significantly improve the practical application of operational precipitation nowcasting. [ABSTRACT FROM AUTHOR]

Chemical reduction of the Li+@C 60 cation by decamethylferrocene was carried out to obtain neutral Li+@C 60 •– (simply denoted as Li@C 60). The method is scalable and does not demand long reaction times unlike electrolytic reduction routes. Powder X-ray diffraction and Raman and EPR spectroscopic measurements of the Li@C 60 solid sample are consistent with the presence mainly of (Li@C 60) 2 dimers together with remaining Li+@C 60 •– monomer species due to lack of crystallization time in the course of formation and precipitation. Image 1 [ABSTRACT FROM AUTHOR]

In this study, wet oxidation of sewage sludge and struvite-precipitation processes were integrated to safely manage sludge and to recover organic acids and phosphorus. To understand this process integration, the effects of temperature, treatment time, and sludge concentration during wet oxidation, and pH and magnesium dosage during struvite precipitation on total suspended solids (TSS), volatile suspended solids (VSS), organic acids production, and phosphorus recovery were investigated. The results showed that TSS and VSS destruction, organic acids production, and phosphorus recovery from sludge into the liquid phase are more sensitive to temperature and time than sludge concentration during wet oxidation. These are novel findings of this work. Furthermore, pH during struvite precipitation had a strong effect on phosphorus recovery. Wet-oxidised sludge was mainly converted to acetic acid, and its concentration increased with increasing temperature and time. These results may be used for optimised operation and revenue generation from wastewater treatment processes when integrated with wet oxidation and struvite precipitation. Image 107636 • Wet oxidation (WO) of sewage sludge and struvite-precipitation processes were integrated. • Sludge was safely managed and valuables (e.g. phosphorous) were recovered. • Outputs e.g. TSS, VSS, organics production, and phosphorus recovery were investigated. • WO processing parameters had a stronger effect on outputs than struvite-precipitation. [ABSTRACT FROM AUTHOR]

The development of long range order in nickel-chromium alloys is of great technological interest but the kinetics and mechanisms of the transformation are poorly understood. The present research utilizes a combined computational and experimental approach to elucidate the mechanism by which phosphorus accelerates the ordering rate of stoichiometric Ni 2 Cr in Ni-Cr alloys. A series of Ni-33%Cr-x%P samples (in atomic percent) were fabricated with phosphorus concentrations, x = <0.005–0.1 at.% and aged between 373 and 470 °C for times up to 3000 h. The first-principles modeling considers fcc Ni with dilute P as a reasonable approximation for the complex Ni-Cr-P alloy system. Calculation results show a pronounced enhancement of vacancy transport by vacancy-solute pair diffusion via consecutive exchange and rotation jumps of vacancies associated with the phosphorus atom. The energy barriers of these two migration paths are at least 0.35 eV lower than that of vacancy-atom exchange in pure Ni solvent. The analytical diffusion model predicts enhanced solvent diffusion by 2 orders of magnitude for 0.1 at.% P at 400–500 °C. The model prediction is in good agreement with the evolution of micro-hardness. We characterize the micro-hardness result by a kinetic ordering model, showing a significant decrease of the activation energy of ordering transformation. These results help gauge the risk of industrial alloys developing long range order which increases strength but degrades ductility and toughness. Specifically, minor alloying additions that bind with excess vacancies and lower the vacancy migration barrier can greatly accelerate hardening via Ni 2 Cr precipitation. Image 1 [ABSTRACT FROM AUTHOR]

Abstract Tritium is a radioisotope of hydrogen and a component of the water molecule. It is a marker for reservoirs such as the stratosphere, troposphere, and oceans involved in the hydrological cycle. Tritium monitoring is an essential research tool in hydro-climate, dating for water and recharge groundwater. The Isotope Hydrology Laboratory has collected monthly precipitation samples in Hanoi for tritium concentration analysis. This paper reports the tritium concentrations in precipitation in the city from 2011 to 2016. The results show that monthly tritium concentration reached a maximum of 7.07 Tritium Units (TU) in August 2011. The mean annual tritium concentration stabilized from 2.03 to 3.36 TU. It suggests that tritium in monitoring station precipitation is predominantly natural. The seasonal variation trend of 3H in precipitation at the Hanoi station is similar to those monitored at the Hong Kong station. The correlation of tritium and rainfall was also estimated. Graphical abstract Statistical distribution and seasonal trends of tritium concentration in Hanoi's precipitation. Image 1 Highlights • The monthly concentrations of tritium in rainwater from 2011 to 2016 in Hanoi are described by a Gaussian distribution. • Tritium in precipitation at the monitoring station is predominantly natural. • The seasonality of tritium in rainwater mainly reflects the different transport patterns of moisture under the monsoon regime. • Correlations between tritium concentrations and monthly precipitation amount are estimated. [ABSTRACT FROM AUTHOR]

Skilful and localised daily weather forecasts for upcoming seasons are desired by climate-sensitive sectors. Various General circulation models routinely provide such long lead time ensemble forecasts, also known as seasonal climate forecasts (SCF), but require downscaling techniques to enhance their skills from historical observations. Traditional downscaling techniques, like quantile mapping (QM), learn empirical relationships from pre-engineered predictors. Deep-learning-based downscaling techniques automatically generate and select predictors but almost all of them focus on simplified situations where low-resolution images match well with high-resolution ones, which is not the case in ensemble forecasts. To downscale ensemble rainfall forecasts, we take a two-step procedure. We first choose a suitable deep learning model, very deep super-resolution (VDSR), from several outstanding candidates, based on an ensemble forecast skill metric, continuous ranked probability score (CRPS). Secondly, via incorporating other climate variables as extra input, we develop and finalise a very deep statistical downscaling (VDSD) model based on CRPS. Both VDSR and VDSD are tested on downscaling 60 km rainfall forecasts from the Australian Community Climate and Earth-System Simulator Seasonal model version 1 (ACCESS-S1) to 12 km with lead times up to 217 days. Leave-one-year-out testing results illustrate that VDSD has normally higher forecast accuracy and skill, measured by mean absolute error and CRPS respectively, than VDSR and QM. VDSD substantially improves ACCESS-S1 raw forecasts but does not always outperform climatology, a benchmark for SCFs. Many more research efforts are required on downscaling and climate modelling for skilful SCFs. [ABSTRACT FROM AUTHOR]

Remote sensing has emerged as a new technique for collecting farmland data due to its rapid advancement, rising popularity, and application in social production practice. In order to understand and manage farmland resources in China, it is essential to account for and monitor high-standard farmland and its usage. Therefore, this work used satellite remote sensing empowered with various abilities for monitoring high-standard farmland by employing GF-2 high-resolution satellite images to identify targets and objects in Hebei and Guangdong provinces. Farmland occupation and utilization were analyzed by detecting destructions, underutilization, and overutilization, and converting farmland for other economic activities registered on a special field sheet for quantification. A statistical summary was compiled for the two provinces, and the results reveal that high-standard farmland irregularities were detected in both Hebei and Guangdong provinces. However, in Hebei province, this was due to domestic purposes, such as building home shelters and domestic factories. On a contract, the result shows that in Guangdong province, farmland was being converted for economic purposes on an industrial scale, such as high residential apartment blocks and new industrial zones, and environmental destruction. Furthermore, the results reveal that there is still a steady and continuous decline in arable land due to accelerated industrialization and population pressure, especially in the Guangdong provinces, which is a threat to national food security. The high interpretation accuracy demonstrates that high-resolution remote sensing is an effective farmland monitoring tool that can be used to advance policy formulation. [ABSTRACT FROM AUTHOR]

For the purpose of improving the scientific nature, reliability, and accuracy of flood forecasting, it is an effective and practical way to construct a flood forecasting scheme and carry out real-time forecasting with consideration of different rain patterns. The technique for rain pattern classification is of great significance in the above-mentioned technical roadmap. With the rapid development of artificial intelligence technologies such as machine learning, it is possible and necessary to apply these new methods to assist rain classification applications. In this research, multiple machine learning methods were adopted to study the time-history distribution characteristics and conduct rain pattern classification from observed rainfall time series data. Firstly, the hourly rainfall data between 2003 and 2021 of 37 rain gauge stations in the Pi River Basin were collected to classify rain patterns based on the universally acknowledged dynamic time warping (DTW) algorithm, and the classifications were treated as the benchmark result. After that, four other machine learning methods, including the Decision Tree (DT), Long- and Short-Term Memory (LSTM) neural network, Light Gradient Boosting Machine (LightGBM), and Support Vector Machine (SVM), were specifically selected to establish classification models and the model performances were compared. By adjusting the sampling size, the influence of different sizes on the classification was analyzed. Intercomparison results indicated that LightGBM achieved the highest accuracy and the fastest training speed, the accuracy and F1 score were 98.95% and 98.58%, respectively, and the loss function and accuracy converged quickly after only 20 iterations. LSTM and SVM have satisfactory accuracy but relatively low training efficiency, and DT has fast classification speed but relatively low accuracy. With the increase in the sampling size, classification results became stable and more accurate. Besides the higher accuracy, the training efficiency of the four methods was also improved. [ABSTRACT FROM AUTHOR]

In the semiarid Bulal transboundary catchment of southern Ethiopia, groundwater is the only reliable drought-resilient water source. The central and southern parts of the catchment are dominantly overlain by the transboundary aquifers of the Bulal basalts, while the basement rocks outcrop in the eastern part. This study uses an integrated geographic information system (GIS), remote sensing (RS), and analytical hierarchal process (AHP) to identify and delineate the groundwater potential zones of the semiarid Bulal catchment within the Ethiopian territory. Based on their relative importance to groundwater occurrence and movement, ten input parameters were chosen. According to Saaty's AHP approach, the input themes and each of their distinct features were given normalized weights. A composite groundwater potential zone index (GWPZI) map was generated by integrating all the input layers employing the GIS-overlay analysis technique. The map was validated using the yield of wells from the catchment. The GWPZI map depicts four groundwater potential zones: high (representing 27% of the total area), moderate (20%), low (28%), and very low (25%). The geological feature has the greatest influence on the distribution of groundwater potential. Areas with high potential are mainly overlain by the Bulal basaltic flow, while low groundwater potential zones are in the regolith over the basement rocks. Unlike conventional methods, our novel approach is effective in identifying relatively shallow GWPZs throughout the catchment, and it can be applied in similar semiarid regions. The GWPZI map serves as a quick guide for effectively planning, managing, and developing the catchment's groundwater resources. [ABSTRACT FROM AUTHOR]

Our world is in a state of critical transition demanding new, creative, ecosystemically fit and sustainable responses to complex challenges. We need both new types of knowledge and new modes of knowledge production. Interdisciplinarity and Transdisciplinarity have the potential to support more congruently complex forms of knowledge (differentiated, integrated, recursive, emergent, ecosystemically fit). Their success is dependent on a deeper understanding of their own organizational complexity. In this paper, I highlight key knowledge gaps and research questions for the development of a richer knowledge base to guide the intentional management and facilitation of Interdisciplinarity and Transdisciplinary Relations toward creative and abductive outcomes. I defend the investigation of creativity and abduction, as hallmarks of the complexity of Interdisciplinarity and Transdisciplinarity, from a process, relational and complexity-informed perspective, mobilizing contributions from Psychology. I discuss Psychology's modes of engagement with Interdisciplinarity and Transdisciplinarity in addressing complex challenges. In this context, I introduce the notion of "dissolution" as an Interdisciplinary and Transdisciplinary relational process supporting the theoretical, methodological and pragmatic enrichment or transformation and increased complexity of different disciplines, bodies of knowledge or modes of knowing. Finally, I propose a new domain for research and practice: a ("Dissolved") Psychology of Interdisciplinary and Transdisciplinary Relations. [ABSTRACT FROM AUTHOR]

Numerical weather prediction (NWP) models are atmospheric simulations that imitate the dynamics of the atmosphere and provide high-quality forecasts. One of the most significant limitations of NWP is the elevated amount of computational resources required for its functioning, which limits the spatial and temporal resolution of the outputs. Traditional meteorological techniques to increase the resolution are uniquely based on information from a limited group of interest variables. In this study, we offer an alternative approach to the task where we generate precipitation maps based on the complete set of variables of the NWP to generate high-resolution and short-time precipitation predictions. To achieve this, five different deep learning models were trained and evaluated: a baseline, U-Net, two deconvolution networks and one conditional generative model (Conditional Generative Adversarial Network; CGAN). A total of 20 independent random initializations were performed for each of the models. The predictions were evaluated using skill scores based on mean absolute error (MAE) and linear error in probability space (LEPS), equitable threat score (ETS), critical success index (CSI) and frequency bias after applying several thresholds. The models showed a significant improvement in predicting precipitation, showing the benefits of including the complete information from the NWP. The algorithms doubled the resolution of the predictions and corrected an over-forecast bias from the input information. However, some new models presented new types of bias: U-Net tended to mid-range precipitation events, and the deconvolution models favored low rain events and generated some spatial smoothing. The CGAN offered the highest-quality precipitation forecast, generating realistic outputs and indicating possible future research paths. [ABSTRACT FROM AUTHOR]

Global Satellite Mapping of Precipitation (GSMaP), Climate Hazards Group InfraRed Preconception with Station data (CHIRPS), Tropical Rain Measurement Mission Multisatellite Precipitation Analysis (TRMM 3B42 V7) and Rainfall Estimation from Soil Moisture Observations (SM2RAIN) are satellite precipitation products with high applicability, but their applicability in hydrological research in arid mountainous areas is not clear. Based on precipitation and runoff data, this study evaluated the applicability of each product to hydrological research in a typical mountainous basin (the Qaraqash River basin) in an arid region by using two methods: a statistical index and a hydrological model (Soil and Water Assessment Tool, SWAT). Simulation results were evaluated by Nash efficiency coefficient (NS), relative error (PBIAS) and determination coefficient (R2). The results show that: (1) The spatial distributions of precipitation estimated by these four products in the Qaraqash River basin are significantly different, and the multi-year average annual precipitation of GSMaP is 97.11 mm, which is the closest to the weather station interpolation results. (2) On the annual and monthly scales, GSMaP has the highest correlation (R ≥ 0.82) with the observed precipitation and the smallest relative error (BIAS < 6%). On the seasonal scale, the inversion accuracy of GSMaP in spring, summer and autumn is significantly higher than other products. In winter, all four sets of products perform poorly in estimating the actual precipitation. (3) Monthly runoff simulations based on SM2RAIN and GSMaP show good fitting (R2 > 0.6). In daily runoff simulation, GSMaP has the greatest ability to reproduce runoff changes. The study provides a reference for the optimization of precipitation image data and hydrological simulation in data-scarce areas. [ABSTRACT FROM AUTHOR]

Quantitative precipitation estimation (QPE) plays an important role in meteorology and hydrology. Currently, multichannel Doppler radar image is used for QPE based on some traditional methods like the Z − R relationship, which struggles to capture the complicated non-linear spatial relationship. Encouraged by the great success of using Deep Learning (DL) segmentation networks in medical science and remoting sensing, a UNet-based network named Reweighted Regression Encoder–Decoder Net (RRED-Net) is proposed for QPE in this paper, which can learn more complex non-linear information from the training data. Firstly, wavelet transform (WT) is introduced to alleviate the noise in radar images. Secondly, a wider receptive field is obtained by taking advantage of attention mechanisms. Moreover, a new Regression Focal Loss is proposed to handle the imbalance problem caused by the extreme long-tailed distribution in precipitation. Finally, an efficient feature selection strategy is designed to avoid exhaustion experiments. Extensive experiments on 465 real processes data demonstrate that the superiority of our proposed RRED-Net not only in the threat score (TS) in the severe precipitation (from 17.6% to 39.6%, ≥20 mm/h) but also the root mean square error (RMSE) comparing to the traditional Z-R relationship-based method (from 2.93 mm/h to 2.58 mm/h, ≥20 mm/h), baseline models and other DL segmentation models. [ABSTRACT FROM AUTHOR]

Near-real-time precipitation retrieval plays an important role in the study of the evolutionary process of precipitation and the prevention of disasters caused by heavy precipitation. Compared with ground-based precipitation observations, the infrared precipitation estimations from geostationary satellites have great advantages in terms of geographical coverage and temporal resolution. However, precipitation retrieved from multispectral infrared data still faces challenges in terms of accuracy, especially in extreme cases. In this paper, we propose a new paradigm for satellite multispectral infrared data retrieval of precipitation and construct a new model called PrecipGradeNet. This model uses FY-4A L1 FDI data as the input, IMERG precipitation data as the training target, and improves the precipitation retrieval accuracy by grading the precipitation intensity through Res-UNet, a semantic segmentation network. To evaluate the precipitation retrieval of the model, we compare the retrieval results with the FY-4A L2 QPE operational product to the IMERG precipitation. IMERG is considered as the ground truth. We evaluate the precipitation retrieval from the precipitation fall area identification, the precipitation intensity interval discrimination, and the precipitation quantification. Experimental results show that PrecipGradeNet has better overall performance compared with the FY-4A QPE product in precipitation fall area identification with POD increased by 48% and CSI and HSS improved by 21% and 14%. PrecipGradeNet also has better performance in light precipitation with POD increased by 114% and CSI and HSS improved by 64% and 52%, and better overall precipitation quantification, with RMSE and CC improved by 16% and 15%. In addition, PrecipGradeNet avoids the overall bias in the low and extreme high precipitation cases. Therefore, the new paradigm proposed in this paper has the potential to improve the retrieval accuracy of satellite precipitation estimation products. This study suggests that the application of semantic segmentation methods may provide a new path to correct the intensity bias of the satellite-based precipitation products. [ABSTRACT FROM AUTHOR]

Copyright of Advances in Atmospheric Sciences is the property of Springer Nature and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

A new method that automatically determines the modality of an observed particle size distribution (PSD) and the representation of each mode as a gamma function was used to characterize data obtained during the High Altitude Ice Crystals and High Ice Water Content (HAIC-HIWC) project based out of Cayenne, French Guiana, in 2015. PSDs measured by a 2D stereo probe and a precipitation imaging probe for particles with maximum dimension (Dmax) > 55 μm were used to show how the gamma parameters varied with environmental conditions, including temperature (T) and convective properties such as cloud type, mesoscale convective system (MCS) age, distance away from the nearest convective peak, and underlying surface characteristics. Four kinds of modality PSDs were observed: unimodal PSDs and three types of multimodal PSDs (Bimodal1 with breakpoints 100 ± 20 μm between modes, Bimodal2 with breakpoints 1000 ± 300 μm, and Trimodal PSDs with two breakpoints). The T and ice water content (IWC) are the most important factors influencing the modality of PSDs, with the frequency of multimodal PSDs increasing with increasing T and IWC. An ellipsoid of equally plausible solutions in (No–λ–μ) phase space is defined for each mode of the observed PSDs for different environmental conditions. The percentage overlap between ellipsoids was used to quantify the differences between overlapping ellipsoids for varying conditions. The volumes of the ellipsoid decrease with increasing IWC for most cases, and (No–λ–μ) vary with environmental conditions related to distribution of IWC. HIWC regions are dominated by small irregular ice crystals and columns. The parameters (No–λ–μ) in each mode exhibit mutual dependence. [ABSTRACT FROM AUTHOR]

Rainfall forecasting plays a key role in mitigating environmental risks in urban areas, which are subject to increasing hydrogeological risk due to transformations in the urban landscape. We present a new technique for probabilistic precipitation nowcasting by generating an ensemble of equiprobable forecasts, which is especially useful for weather radars with limited spatial range, and that can be used operationally on devices with low computational capacity. The ensemble members are obtained by a novel stochastic noise generation process, consistent with the spatial scales not resolved by the prediction model, which allows continuous downscaling of the output of a deep generative neural network. Through an in-depth analysis of the results, for precipitation accumulated in the first hour, measured by all the most robust skill indicators, extended to an entire year of data at 5-min time resolution, we demonstrate that the proposed procedure is able to provide calibrated, reliable, and sharp ensemble rainfall forecasts with a quality comparable or superior to the state-of-the-art classical alternative optical flow technique. The ensemble generation procedure we propose is sufficiently general to be applied in principle to other deterministic architectures as well, thus enabling their generalization in probabilistic terms. [ABSTRACT FROM AUTHOR]

Based on Landsat MSS/TM/OLI (1975-2018), MODIS NDVI (2000-2015), and DMSP/OLS (1992-2013) multitemporal remote sensing data, as well as annual average precipitation and temperature data from meteorological stations (1980-2018) and SRTM DEM data, this study extracted information on climate change, human activity, and eco-environmental factors in the Zijin Mountain (ZJM) area of Shanxi Province, and then analyzed the spatiotemporal changes in ecological structure, landscape structure, and climate response. The results revealed that in the ZJM area some sustainable development indicators associated with human activity, ecology, climate, and landscape geography have changed. Specifically, over the 44-year period from 1975 to 2018, the urban and rural construction land area increased by 4.29 times; in the 22-year period from 1992 to 2013, the regional light index has increased by 26.9 times; in the past 32 years, that is, from 1987 to 2018, the average vegetation index of the entire basin increased by about 54.7%, the vegetation cover exhibited a gradual positive trend, and the land use types in the ZJM area were mainly converted from bare land to grassland or forest. Compared with 1987, the area of bare land decreased 32.5% by 2018, whereas the forest and grassland areas increased by 85.7% and 40.7%, respectively. In addition, over the 44-year period from 1975 to 2018, the road density increased by 8.57 times. The area affected by human activities, comprised roads, urban and rural construction land, and their buffer areas, accounted for 64.87% of the total study area. As a result, the fragmentation index of landscape patches has increased by 6.81 times since 1987. We conclude that in the ZJM area, the ecosystem type has transformed from an indigenous mountainous grassland to mountainous arid agriculture, and from an isolated habitat to a fragmented habitat. [ABSTRACT FROM AUTHOR]

Precipitation forecasting is an important mission in weather science. In recent years, data-driven precipitation forecasting techniques could complement numerical prediction, such as precipitation nowcasting, monthly precipitation projection and extreme precipitation event identification. In data-driven precipitation forecasting, the predictive uncertainty arises mainly from data and model uncertainties. Current deep learning forecasting methods could model the parametric uncertainty by random sampling from the parameters. However, the data uncertainty is usually ignored in the forecasting process and the derivation of predictive uncertainty is incomplete. In this study, the input data uncertainty, target data uncertainty and model uncertainty are jointly modeled in a deep learning precipitation forecasting framework to estimate the predictive uncertainty. Specifically, the data uncertainty is estimated a priori and the input uncertainty is propagated forward through model weights according to the law of error propagation. The model uncertainty is considered by sampling from the parameters and is coupled with input and target data uncertainties in the objective function during the training process. Finally, the predictive uncertainty is produced by propagating the input uncertainty in the testing process. The experimental results indicate that the proposed joint uncertainty modeling framework for precipitation forecasting exhibits better forecasting accuracy (improving RMSE by 1 %–2 % and R2 by 1 %–7 % on average) relative to several existing methods, and could reduce the predictive uncertainty by ∼28 % relative to the approach of Loquercio et al. (2020). The incorporation of data uncertainty in the objective function changes the distributions of model weights of the forecasting model and the proposed method can slightly smooth the model weights, leading to the reduction of predictive uncertainty relative to the method of Loquercio et al. (2020). The predictive accuracy is improved in the proposed method by incorporating the target data uncertainty and reducing the forecasting error of extreme precipitation. The developed joint uncertainty modeling method can be regarded as a general uncertainty modeling approach to estimate predictive uncertainty from data and model in forecasting applications. [ABSTRACT FROM AUTHOR]

In the Southeastern U. S. there are Concentrated Animal Feeding Operations (CAFOs) that emit a variety of gases, including SO2. Sulfur is emitted as reduced sulfur compounds and can react in the atmosphere to produce SO2. It is expected that the concentration and spread of SO2 emissions from these sources would differ between wet and dry periods. In this research, SO2 emissions from locations representing CAFOs and its dispersion over the southeastern U.S. were simulated through sensitivity experiments using the Weather Research and Forecasting-Chemistry (WRF- Chem) model. Simulations were performed for dry periods and precipitation events that occurred over western Kentucky between July 7 and 13 July 2012. The study found that spatial coverage of SO2 dispersion originating from these locations was reduced during precipitation events and expanded during dry periods. The average concentration of SO2 over the study area was also higher during the breaks between precipitation events than during precipitation. The highest concentrations of SO2 exceeding 1,000 pptv remained within close range of the emission locations for the majority of the simulations, except for when local surface wind speeds were high. Most emissions from the locations remained limited to the surface and lower levels of the atmosphere (850 mb). [ABSTRACT FROM AUTHOR]

Numerical Weather Prediction models (NWP) are atmospheric simulations that imitate the dynamics of the atmosphere and provide high-quality forecasts. One of the most significant limitations of NWP is the elevated amount of computational resources required for its functioning, which limits the spatial and temporal resolution of the outputs. Traditional meteorological techniques to increase the resolution are based uniquely on information from a limited group of interest variables. In this study, we offer an alternative approach to the task where we generate precipitation maps based on the complete set of variables of the NWP to generate high-resolution and short-time precipitation predictions. To achieve this, five different deep learning models were trained and evaluated: baseline, U-Net, two deconvolutional networks, and one conditional generative model (CGAN). A total of 20 independent random initializations were performed for each of the models. The predictions were evaluated using MAE and LEPS-based skill scores, ETS, CSI, and frequency bias after applying several thresholds. The models showed a significant improvement in predicting precipitation showing the benefits of including the complete information from the NWP. The algorithms increased the resolution of the predictions and corrected an over-forecast bias from the input information. However, some new models presented new types of bias: U-Net tended to mid-range precipitation events, and the deconvolutional models favored low rain events and generated some spatial smoothing. The CGAN offered the highest quality precipitation forecast generating realistic outputs and indicating possible future research paths. [ABSTRACT FROM AUTHOR]

Copyright of American Journal of Botany is the property of Wiley-Blackwell and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The primary objective of precipitation forecasting is to accurately predict short-term precipitation at a high resolution within a specific area, which is a significant and intricate challenge. Traditional models often struggle to capture the multidimensional characteristics of precipitation clouds in both time and space, leading to imprecise predictions due to their expansion, dissipation, and deformation. Recognizing this limitation, we introduce ConvLSTM-UNet, which leverages spatiotemporal feature extraction from meteorological images. ConvLSTM-UNet is an efficient convolutional neural network (CNN) based on the classical UNet architecture, equipped with ConvLSTM and improved deep separable convolutions. We evaluate our approach on the generic time series dataset Moving MNIST and the regional precipitation dataset of the Netherlands. The experimental results show that the proposed method has better spatiotemporal prediction skills than other tested models, and the mean squared error is reduced by more than 7.2%. In addition, the visualization results of the precipitation forecast show that the approach has a better ability to capture heavy precipitation, and the texture details of the precipitation forecast are closer to the ground truth. [ABSTRACT FROM AUTHOR]