Dominant Gradient Research Articles

Response of understory vegetation to long-term occupancy by introduced Pinus species in temperate deciduous hardwood forests

In many temperate regions of the world, conifer species have been planted to stabilize soils and promote site recovery on former hardwood sites that were cleared for agriculture. In many areas of eastern North America, these conifer plantings consisted of introduced (non-native) Pinus species planted on abandoned agricultural land once dominated by mesophytic hardwood species. These plantings constitute a shift in overstory composition away from native hardwood species with nutrient-rich litter that decomposes more quickly towards Pinus species with recalcitrant litter that may alter soil chemistry and nutrient availability. To examine how edaphic conditions associated with long-term introduced Pinus species occupancy are related to forest regeneration and herbaceous-layer diversity and composition, we sampled a total of 97 plots in planted Pinus echinata and Pinus strobus stands and naturally regenerated hardwood stands growing on two ecological landtype phases (ELTPs) of southern Indiana, USA forests, Fagus-Acer saccharum/Arisaema Mesic Ridges, and Acer saccharinum/Boehmeria Bottomlands. We collected vegetation data and analyzed soil samples to examine herbaceous-layer species distribution across gradients using non-metric multidimensional scaling (NMS) ordination. Two-way ANOVA was used to examine differences in individual species and species functional groups across stand types (P. echinata, P. strobus, and native hardwoods) and ELTPs. Our results show that differences in soil chemistry resulting from Pinus spp. occupancy were associated with differences in the composition and distribution of herbaceous-layer species in ordination space. Species across stand types and ELTPs were distributed across dominant gradients related to litter depth, cation exchange capacity, cation content, and soil aluminum concentration. Hardwood sites had significantly greater herbaceous-layer cover (139.6 ± 8.0 %) than P. echinata (48.5 ± 6.2 %) or P. strobus sites (81.7 ± 7.7 %), as well as greater herbaceous-layer species richness and diversity (mean species richness was 46.9 ± 2.1 on hardwood stands vs. 33.3 ± 1.6 and 36.4 ± 2.0 on P. echinata and P. strobus stands, respectively). Pinus echinata stands contained a greater density of woody regeneration, including Quercus spp. (201 ± 48 saplings ha−1) and Fagus grandifolia stems (415 ± 82 saplings ha−1), both of which occurred in greater density than Acer saccharum (163 ± 93 saplings ha−1) and A. rubrum (70 ± 64 saplings ha−1). Our results suggest that pine occupancy has created divergent successional trajectories in comparison to hardwood stands. These differing trajectories may offer both challenges and opportunities for restoration efforts. For example, the greater abundance of Quercus reproduction under P. echinata on ridges, combined with less productive soils, may allow Quercus stems to be promoted into the canopy with less competition from mesophytic competitors.

GEAR: Learning graph neural network explainer via adjusting gradients

Graph neural network (GNN) explainers aim to elucidate the prediction behavior of GNNs, facilitating their wide adoption in high-stakes tasks. Current approaches typically define multiple objective functions to construct comprehensive and accurate explainers. Nevertheless, optimizing GNN explainers with multiple objective functions is challenging because of conflicts between the gradients among these objectives, which may result in suboptimal solutions. To eliminate potential conflicts and enhance explainer optimization, we introduce GEAR, which is a novel framework that adjusts the gradients to optimize the GNN explainer. Specifically, we attempt to define comprehensive objectives from multiple perspectives that are crucial for optimizing GNN explainers, including fidelity, sparsity, counterfactual reasoning, and connectivity. Subsequently, we purposefully determine the dominant gradient and angle threshold of the conflicting gradients from a geometric perspective. More importantly, we propose a simple yet effective gradient adjuster to refine the gradients during explainer optimization. Experiments on synthetic and real-world datasets demonstrate the effectiveness of the proposed GEAR. In addition, state-of-the-art explainers with the incorporated gradient adjuster outperform their counterparts without the proposed gradient adjuster. The data and code are available at https://github.com/ymzhang0103/GEAR.

Accurate weather forecasting with dominant gradient boosting using machine learning

Accurate weather forecasting with dominant gradient boosting using machine learning

A common ericoid shrub modulates the diversity and structure of fungal communities across an arbuscular to ectomycorrhizal tree dominance gradient.

Differences between arbuscular (AM) and ectomycorrhizal (EcM) trees strongly influence forest ecosystem processes, in part through their impact on saprotrophic fungal communities. Ericoid mycorrhizal (ErM) shrubs likely also impact saprotrophic communities given that they can shape nutrient cycling by slowing decomposition rates and intensifying nitrogen limitation. We investigated the depth distributions of saprotrophic and EcM fungal communities in paired subplots with and without a common understory ErM shrub, mountain laurel (Kalmia latifolia L.), across an AM to EcM tree dominance gradient in a temperate forest by analyzing soils from the organic, upper mineral (0-10cm), and lower mineral (cumulative depth of 30cm) horizons. The presence of K. latifolia was strongly associated with the taxonomic and functional composition of saprotrophic and EcM communities. Saprotrophic richness was consistently lower in the Oa horizon when this ErM shrub species was present. However, in AM tree-dominated plots, the presence of the ErM shrub was associated with a higher relative abundance of saprotrophs. Given that EcM trees suppress both the diversity and relative abundance of saprotrophic communities, our results suggest that separate consideration of ErM shrubs and EcM trees may be necessary when assessing the impacts of plant mycorrhizal associations on belowground communities.

Sourcing aliphatic and polycyclic aromatic hydrocarbons (PAHs) in Jinga shrimp (Metapenaeus affinis) muscle tissues and surface sediments (study case: Northwest Persian Gulf).

This study addresses the sources of aliphatic hydrocarbons (AHs) and polycyclic aromatic hydrocarbons (PAHs) in the surface sediments of the northwestern Persian Gulf and the muscle tissues of Jinga shrimp (Metapenaeus affinis), a commercially important aquatic species. In November 2018, 28 Jinga shrimp samples were systematically collected from four key fishing areas in Behrgan and Khormusi: Imam Khomeini Port (S1), Mahshahr Port (S2), Sejafi (S3), and Behrgan Wharf (S4). Additionally, sediment samples were collected from these locations, and AHs and PAHs concentrations were analyzed using gas chromatography-mass spectrometry (GC-MS). The average aliphatic concentration in Jinga shrimp was 4800.32 (μg g-1 DW), exceeding the sediment samples' 2496.69 (μg g-1 DW) estimate. Hydrocarbon component analysis revealed EPA priority list (PAH-16) and measured PAHs (PAH-29) concentrations in Jinga shrimp ranging from 1095.8 to 2698.3 (ng g-1 DW) and in sediments from 653.6 to 1019.5 (ng g-1 DW). Elevated AHs and PAHs in Jinga shrimp, compared to sediments, suggest a petrogenic source, notably at station S4 near Behrgansar and Nowruz oil fields. Low molecular weight (LMW) compounds dominated in both shrimp and sediment PAHs. Aliphatic composition profiles in shrimps closely mirrored sediment profiles, illustrating an even-to-odd carbon dominance gradient. Diagnostic ratio examinations of hydrocarbons indicated pervasive petroleum derivatives in the environment. This study establishes a direct correlation between hydrocarbon concentrations in shrimp and sediment samples and the corresponding aliphatic groups, PAH-16, and PAH-29. The findings underscore the potential of Jinga shrimp as a reliable indicator of hydrocarbon pollution in the northwestern Persian Gulf.

Global marine gravity gradient tensor inverted from altimetry-derived deflections of the vertical: CUGB2023GRAD

Abstract. Geodetic applications of altimetry have largely been inversions of gravity anomaly. Previous studies of Earth's gravity gradient tensor mostly presented only the vertical gravity gradient (VGG). However, there are six unique signals that constitute the gravity gradient tensor. Gravity gradients are signals suitable for detecting short-wavelength topographic and tectonic features. They are derived from double differentiation of the disturbing potential and hence are susceptible to noise amplification which was exacerbated by low across-track resolution of altimetry data in the past. However, current generation of altimetry observations have improved spatial resolutions, with some better than 5 km. Therefore, this study takes advantage of current high-resolution altimetry datasets to present CUGB2023GRAD, a global (latitudinal limits of ±80°) 1 arcmin model of Earth's gravity gradient tensor over the oceans using deflections of the vertical as inputs in the wavenumber domain. The results are first assessed via Laplace's equation, whereby the resultant residual gradient is virtually zero everywhere. Further analysis at local regions in the Arctic and south Indian oceans showed that Txy, Txz and Tyz are the most dominant gravity gradients for bathymetric studies. This proves that bathymetric signatures in the non-diagonal tensor components are worth exploiting. Bathymetric coherence analysis of Tzz over the Tonga Trench showed strong correlation with multibeam shipboard depths. This study proves that current generation of altimetry geodetic missions can effectively resolve Earth's gravity gradient tensor. The CUGB2023GRAD model data can be freely accessed at https://doi.org/10.5281/zenodo.10511125 (Annan et al., 2024).

Nanoparticle Deep-Subwavelength Dynamics Empowered by Optical Meron-Antimeron Topology.

Optical meron is a type of nonplanar topological texture mainly observed in surface plasmon polaritons and highly symmetric points of photonic crystals in the reciprocal space. Here, we report Poynting-vector merons formed at the real space of a photonic crystal for a Γ-point illumination. Optical merons can be utilized for subwavelength-resolution manipulation of nanoparticles, resembling a topological Hall effect on electrons via magnetic merons. In particular, staggered merons and antimerons impose strong radiation pressure on large gold nanoparticles (AuNPs), while focused hot spots in antimerons generate dominant optical gradient forces on small AuNPs. Synergistically, differently sized AuNPs in a still environment can be trapped or orbit in opposite directions, mimicking a coupled galaxy system. They can also be separated with a 10 nm precision when applying a flow velocity of >1 mm/s. Our study unravels a novel way to exploit topological textures for optical manipulation with deep-subwavelength precision and switchable topology in a lossless environment.

Belowground competition and niche partitioning between the Babassu palm and Urochloa grass in eastern Amazonian pastures

AbstractMuch of deforested Amazonia is dominated by pastures planted with exotic Urochloa (syn Brachiaria) grass, containing interspersed, ruderal Babassu palm (Attalea speciosa). This two‐species system offers the opportunity to investigate how two very competitive species interact. We investigated three monospecific Urochloa pastures containing dispersed Babassu palm clusters in the eastern periphery of Amazonia. We mapped 0‐ to 50‐cm Babassu and Urochloa root profiles at three distances representing contrasting relative dominance: “within clusters” (0.4–0.6 m from adult palm, Babassu‐dominated), “close” (2.5–4.0 m), and “far” (>8.0–10.0 m, Urochloa‐dominated). Total Babassu root share count was 76.0% within Babassu clusters, as opposed to Urochloa root share count (80.0%) far from clusters, confirming our experimental dominance gradient. Babassu and Urochloa root diameter composition differed strongly, whereas 44.6% of all Babassu roots were mid‐diameter or coarse; 83.0% of all Urochloa roots were fine graminoid. Vertical root profiles likewise differed: Babassu coarse and mid‐diameter roots concentrated within clusters at 20–50 cm depth, well protected against disturbance, though Babassu mid‐diameter roots did extend up to 10 m, indicating active foraging at long distances. By contrast, Urochloa mid‐diameter and coarse roots concentrated in the topsoil (70.4% in top 20 cm). Notably, 45.0% of all Urochloa fine roots but only 29.3% of deeper rooting Babassu concentrated in the top 0–10 cm. Both Urochloa mid‐diameter (R = 0.49) and fine roots (R = 0.40) correlated positively with Urochloa aboveground biomass. Fine‐root fine‐scale variability was high, with high‐density hotspots of both species in the topsoil. Babassu vertical root profiles were unaffected by the degree of interspecific competitive confrontation, whereas Urochloa fine roots were significantly shallower within than outside of Babassu clusters. Though differing Babassu and Urochloa vertical root profiles do support the notion of niche expansion caused by Babassu clusters within the vast Urochloa pastures throughout Amazonia, we also find a strong horizontal and vertical niche overlap with no indication of fine‐scale root segregation, indicating direct competition between both species, possibly in a dynamic state of “mutual invasability.” In practical terms, our findings call for the diversification of Amazonian pastures beyond this two‐species system in order to increase productivity and ecological efficiency.

A Crystallographic Basis for Critically Evaluating the Mechanisms for Secondary Grain Formation During Directional Solidification

AbstractA crystallography-based method is presented for the critical appraisal of possible mechanisms that trigger the formation of secondary grains during directional solidification. The method permits an analysis of a large population of defects, while avoiding the pitfalls of the metallographic sectioning approach that is affected by dendrite stereology. Here, the nickel-base superalloy CMSX-4, an alloy commonly used for single crystal turbine blade applications, is studied. All secondary grains originate exclusively at the external surface and when the off-axial primary $$\langle 0\,0\,1\rangle$$ ⟨ 0 0 1 ⟩ crystal orientations are measured, are evident at both the converging and diverging dispositions of the single crystal primary dendrites without a noticeable bias. Almost all of the secondary grains have low misorientations, with an average misorientation between 5 to 15 deg. No systematic deviation between the individual $$\langle 0\,0\,1\rangle$$ ⟨ 0 0 1 ⟩ orientations of the secondary grain and the single crystal is observed. A significant twist contribution about an axis within ~ 30 deg from one of the secondary arms occurs when primary arms converge on the external surface, but both twist and tilt prevail for the diverging case. Both nucleation and buoyancy driven thermo-solutal convection can be eliminated as potential mechanisms. Thermo-mechanical deformation is deduced to be the most likely mechanism; deformation must originate in the vicinity of the primary dendrite tips. It is proposed that dendrite deflection arises primarily from the resistance encountered by the primary tips with the external surface during axial contraction in the presence of a dominant vertical thermal gradient.

Emergent cerebral venous stenting: A valid treatment option for fulminant idiopathic intracranial hypertension

BackgroundFulminant idiopathic intracranial hypertension (FIIH) is characterized by rapid, severe, progressive vision loss and often treated surgically. Cerebral transverse venous stenting (CTVS) is efficacious in IIH patients, but emergent CTVS in FIIH is rarely reported. We present our experience with emergent CTVS in patients with FIIH. MethodsSince 01/2019, an institutional protocol allowed emergent CTVS in FIIH patients with bilateral transverse sinus stenosis and gradient pressure > 15 on digital subtraction angiography (DSA). We retrospectively analyzed a prospective registry of all IIH patients with details of neurological and neuro-ophthalmological assessments before and after treatment, and subjective assessments of headache and tinnitus were made pre-and post-procedure. Results259 IIH patients, including 49 who underwent CTVS, were registered. Among them, five female patients met inclusion criteria for FIIH and underwent emergent CTVS. FIIH patients were younger (18.8 ± 1.64 vs 27.7 ± 4.85, p < 0.01), mean BMI was lower (30.8 ± 10.57 vs 34.6 ± 4.3, p < 0.01), and lumbar puncture opening pressure higher (454 ± vs 361 ± 99.4, p < 0.01) than that of IIH patients. They presented with acute visual loss, severe headache, papilledema, significant bilateral transverse sinus stenosis on CT-venography, and mean dominant side gradient pressure of 26.4 ± 6.2 on DSA. CTVS was performed without significant complications, resulting in remarkable improvement in headache, optical coherence tomography, and visual fields within 1 week. At 1-year follow-up (four patients) and 6-month follow-up (1 patient), there was complete resolution of papilledema and headache, and marked improvement in visual acuity. ConclusionsIn these patients, emergent-CTVS was a safe and effective treatment option for FIIH. Further evaluation is warranted.

Directional droplet/bubble transportation on slippery photothermal responsive laser-induced graphene/polyimide membrane

Droplet and bubble manipulation on functional surfaces is a critical issue in numerous applications such as microreactor, microfluidic, biological, and gas transport. Despite recent efforts in tuning the physicochemical properties of responsive surfaces to impart functionalities in response to different external stimuli, it remains challenges to manipulate both the droplet and the bubble simultaneously on responsive surfaces. Here, we report an intelligent slippery laser-induced graphene/polyimide (SLIG/PI) membrane with excellent photothermal responsive property and superior oil absorption capacity. Through local near-infrared radiation (NIR) irradiation, we can simultaneously manipulate both the droplet and the bubble transportation on the SLIG/PI membrane in a controllable fashion. Owing to a dominant wettability gradient force, the antigravity droplet transportation and the antibuoyancy bubble movement can be simultaneously achieved on the inclined SLIG/PI surface. Moreover, both the droplet and the bubble can be steered to move on desirable traces by constructing the SLIG/PI membrane with programmable pathways. We envision that this work might provide important insights into the construction of intelligent devices for controllable droplet/bubble motion.

Linkages between Plant Community Composition and Soil Microbial Diversity in Masson Pine Forests.

Plant species identity influences soil microbial communities directly by host specificity and root exudates, and indirectly by changing soil properties. As a native pioneer species common in early successional communities, Masson pine (Pinus massoniana) forests are widely distributed in subtropical China, and play a key role in improving ecosystem productivity. However, how pine forest composition, especially the dominance of plant functional groups, affects soil microbial diversity remains unclear. Here, we investigated linkages among woody plant composition, soil physicochemical properties, and microbial diversity in forests along a dominance gradient of Masson pine. Soil bacterial and fungal communities were mainly explained by woody plant community composition rather than by woody species alpha diversity, with the dominance of tree (without including shrub) species and ectomycorrhizal woody plant species accounting for more of the variation among microbial communities than pine dominance alone. Structural equation modeling revealed that bacterial diversity was associated with woody plant compositional variation via altered soil physicochemical properties, whereas fungal diversity was directly driven by woody plant composition. Bacterial functional groups involved in carbohydrate and amino acid metabolism were negatively correlated with the availability of soil nitrogen and phosphorus, whereas saprotrophic and pathogenic fungal groups showed negative correlations with the dominance of tree species. These findings indicate strong linkages between woody plant composition than soil microbial diversity; meanwhile, the high proportion of unexplained variability indicates great necessity of further definitive demonstration for better understanding of forest-microbe interactions and associated ecosystem processes.

Photochemically Induced Marangoni Patterning of Polymer Bilayers

Surface-tension gradients created along a polymer film by patterned photochemical reactions are a powerful tool for creating surface topography. Here, we use mathematical modeling to explore a strategy for patterning photochemically inactive polymers by coupling a light-sensitive and light-insensitive polymer to form a polymer bilayer. The light-sensitive polymer forms the top layer, and the most dominant surface-tension gradients are introduced at the interface between this layer and air. Lubrication theory is used to derive nonlinear partial differential equations describing the heights of each layer, and linear analysis and nonlinear simulations are performed to characterize interface dynamics. Patterns form at both the polymer-air and polymer-polymer interfaces at early thermal annealing times as a result of Marangoni stresses but decay on prolonged thermal annealing as a result of the dissipative mechanisms of capillary leveling and photoproduct diffusion, thus setting a limit to the maximum individual layer deformation. Simulations also show that the bottom-layer features can remain "trapped", i.e., exhibit no significant decay, even while the top layer topography has dissipated. We study the effects of two key parameters, the initial thickness ratio and the viscosity ratio of the two polymers, on the maximum deformation attained in the bottom layer and the time taken to attain this deformation. We identify regions of parameter space where the maximum bottom-layer deformation is enhanced and the attainment time is reduced. Overall, our study provides guidelines for designing processes to pattern photochemically inactive polymers and create interfacial topography in polymer bilayers.

Functional hierarchy of the angular gyrus and its underlying genetic architecture.

The angular gyrus (AG), given its rich connectivity and its location where multisensory information converges, is a functionally and anatomically heterogeneous structure. Using the state-of-the-art functional gradient approach and transcription-neuroimaging association analysis, we sought to determine whether there is an overarching hierarchical organization of the AG and if so, how it is modulated by the underlying genetic architecture. Resting-state functional MRI data of 793 healthy subjects were obtained from discovery and validation datasets. Functional gradients of the AG were calculated based on the voxel-wise AG-to-cerebrum functional connectivity patterns. Combined with the Allen Human Brain Atlas, we examined the spatial correlations between the AG functional gradient and gene expression. The dominant gradient topography showed a dorsoanterior-ventroposterior hierarchical organization of the AG, which was related to its intrinsic geometry. Concurrently, AG functional subdivisions corresponding to canonical functional networks (behavioral domains) were distributed along the dominant gradient in a hierarchical manner, that is, from the default mode network (abstract cognition) at one extreme to the visual and sensorimotor networks (perception and action) at the other extreme. Remarkably, we established a link between the AG dominant gradient and gene expression, with two gene sets strongly contributing to this link but diverging on their functional annotation and specific expression. Our findings represent a significant conceptual advance in AG functional organization, and may introduce novel approaches and testable questions to the investigation of AG function and anatomy in health and disease.

Modeling the potential distribution of floristic assemblages of high Andean wetlands dominated by Juncaceae and Cyperaceae in the Argentine Puna

Aims: The aim of this work was to model the distribution of suitable environmental conditions of vegas with specific floristic characteristics. Vegas are high Andean wetlands that represent the main sequestered carbon stocks, biodiversity hotspots, and water regulating systems in the region. In these wetlands, plant communities are the main biological factor that determines functional processes, and plant species assemblages are associated with different ecogeographic features. Study area: Argentine Central Andean Puna ecoregion. Methods: For two different floristic assemblages of vegas, we develop ecological niche models of n-dimensional minimum volume ellipsoids through NicheToolBox, then obtain potential distribution maps. One floristic assemblage was dominated by the cushion-structured plant Oxychloe andina (Juncaceae) and the other by plants of the Cyperaceae family. Results: Elevation and precipitation were the main environmental factors determining the distribution of the two floristic assemblages. Juncaceae dominated vegas tend to be located in high, humid, and cold places, while Cyperaceae vegas are found at a lower elevation, with less humidity, and higher temperatures. According to the dominant climatic gradient in the region, potential distribution maps show that vegas of Juncaceae are commonly found towards the Northeast of the Puna while Cyperaceae vegas are more frequent at lower elevations to the South of the region. Conclusions: This study represents the first approach to niche modeling based on plant communities in vegas of the Argentine Puna, providing knowledge on the environmental factors that limit their distribution. This information could serve as a planning tool in a region exposed to growing perturbations such as mining and climate change. Taxonomic reference: Zuloaga et al. (2019). Abbreviations: AUC = Area Under the ROC Curve; NDVI = Normalized Difference Vegetation Index; ROC = Receiver Operating Characteristic.

GradCa: Generalizing to unseen domains via gradient calibration

In domain generalization (DG) problem, we hope to train a robust model from multiple source domains and generalize it to unseen domains. However, the trained model can not perform well in the unseen domain since the domain-invariant representation is hard to learn across multiple source domains. In this work, we propose a novel gradient optimization strategy named gradient calibration (GradCa) which optimizes the dominant gradients and conflicting gradients without learning extra parameters. The dominant gradients are suppressed for addressing to be inclined to the optimization directions of specific domains. Then two strategies (average and sign-mask) are designed for further alleviating the conflicting gradients. These types of gradients are effectively calibrated for improving the domain-invariant representation learning. Extension experiments on four benchmark datasets have shown the competitive results and the effectiveness of improving the model generalization with GradCa.

Query-Efficient Adversarial Attack With Low Perturbation Against End-to-End Speech Recognition Systems

With the widespread use of automated speech recognition (ASR) systems in modern consumer devices, attack against ASR systems have become an attractive topic in recent years. Although related white-box attack methods have achieved remarkable success in fooling neural networks, they rely heavily on obtaining full access to the details of the target models. Due to the lack of prior knowledge of the victim model and the inefficiency in utilizing query results, most of the existing black-box attack methods for ASR systems are query-intensive. In this paper, we propose a new black-box attack called the Monte Carlo gradient sign attack (MGSA) to generate adversarial audio samples with substantially fewer queries. It updates an original sample based on the elements obtained by a Monte Carlo tree search. We attribute its high query efficiency to the effective utilization of the dominant gradient phenomenon, which refers to the fact that only a few elements of each origin sample have significant effect on the output of ASR systems. Extensive experiments are performed to evaluate the efficiency of MGSA and the stealthiness of the generated adversarial examples on the DeepSpeech system. The experimental results show that MGSA achieves 98% and 99% attack success rates on the LibriSpeech and Mozilla Common Voice datasets, respectively. Compared with the state-of-the-art methods, the average number of queries is reduced by 27% and the signal-to-noise ratio is increased by 31%.

Refined revealing the chain path of multiple ecosystem services under diverse environmental factor gradients

A fragile karst ecosystem presents ecosystem services affected by the environment and interactions between ecosystem services. However, the ecosystem in the karst area is characterized by low environmental capacity and low resistance to force disturbance. Current research mainly focuses on the trade-off/synergy of ecosystem services in the karst area yet it lacks inductive analysis of the multiple chain path among ecosystem services. This paper quantitatively identified dominant factors influencing spatial differentiation of surface runoff, soil moisture, sediment yield, and net primary productivity (NPP) and determined the chain path. The chain paths of surface runoff–soil moisture–NPP and NPP–surface runoff–sediment yield were analyzed. The results showed that land use and soil type were the dominant factors, and chain effects of ecosystem services were diverse under the various dominant factor gradients. The mediation effects of paddy soil (97.21 %) and mountain meadow soil (55.56 %) were high, and surface runoff had a greater impact on NPP by affecting soil moisture. Among the diverse land use types, the mediation effect of surface runoff on NPP affecting sediment yield varied greatly (from 5 to 100 %). In addition, its variation trend was consistent with that of the soil moisture as the mediation variable. The mediation effect of surface runoff on construction land was the highest (99.43 %). This study provides the scientific basis for selecting more effective water and soil conservation measures by analyzing the chain relationships of multiple ecosystem services under different environmental factor gradients.

Comparing the three-dimensional morphologies of urban buildings along the urban-rural gradients of 91 cities in China

Urban buildings and their three-dimensional (3D) morphologies play important roles in influencing urban climates and the diffusion of air pollutants. Previous studies on the 3D morphologies of urban buildings were mainly conducted within cities due to limited data, and we still lack an understanding of their features among cities, especially along urban-rural gradients. Here, we analyzed the urban-rural gradients of the 3D morphologies of buildings for 91 cities in China by using two groups of morphologic indicators and the concentric buffer method. Then, we used a hierarchical clustering method based on dynamic time warping to classify the gradients. Our results showed that >75 % of grids have a mean height of 9 m to 16 m, and their building coverage ranged from 20 % to 60 %. The gradients of five morphology indicators could be clustered to 9 patterns, and the dominant urban-rural gradient having the average height and building coverage decreasing over distance. This study provided a methodological framework for the quantitative analysis of the urban-rural gradient of the 3D morphologies of urban buildings, and can provide information on mitigate city-scale climate issues.

Enhanced Elman spike neural network based sentiment analysis of online product recommendation

The major task of Natural Language Processing (NLP) is Sentiments Analysis (SA) or Opinion Mining (OM). It captures the opinion, trust, and feelings of every user’s about the related products to find whether the attitude of users is positive, negative, or neutral. By this, companies can make necessary changes in the product for customer satisfaction. The majority of the existing approaches on sentiment analysis have inaccurate and takes more time. Therefore, in this manuscript, an Enhanced Elman Spike Neural Network based Sentiment Analysis of Online Product recommendation is proposed (EESNN-SA-OPR). Here, filtering Collaborative (FC) and product to product (P–P) similarity are used as new recommendation systems. The aim of Collaborative Filtering is “to predict the best shops and product–product​ similarity is to predict the best products”. Initially, the datas are taken from Amazon reviews database. Then the input data is pre-processed. The trilateral smoothing filtering (TRSF) is used to remove the content which is no more needed and texts related filtering. After that, the features like manufacturing price (MRP), Manufacturing date (MFD), discounts, offers, ratings in qualities are extracted by using Dominant Gradient Local Ternary Pattern descriptor (DGLTPD) technique. At last, enhanced Elman spike neural network classifies the product recommendation as excellent, good, very good, bad and very bad. The proposed method is executed in MATLAB and its performance is analyzed under some performance metrics, like mean absolute error (MAE), mean squared error (MSE), mean absolute percentage error (MAPE), accuracy, F-Score, recall and precision. The proposed EESNN-SA-OPR method provides 23.14%, 15.96%, 31.54% higher accuracy and 12.33%, 21.31%, 41.09% lower mean absolute error compared with the existing techniques, like Sentiment analysis of online product reviews utilizing DLMNN and future prediction of online product utilizing IANFIS (DLMNN-SA-OPR), a machine learning-based sentiment analysis of online product reviews with new term weighting along feature selection method (LSIBA-ENN-SA-OPR) and Sentiment Analysis on the Reviews of Online Product utilizing Optimized RNN-LSTM along Support Vector Machine (RNN-LSTM-SA-OPR) respectively.