Uncertainty Map Research Articles

965: Individualised dose mapping uncertainty estimation for head and neck cancer re-irradiation.

965: Individualised dose mapping uncertainty estimation for head and neck cancer re-irradiation.

Multivariate Geostatistics for Mapping of Transmissivity and Uncertainty in Karst Aquifers

Due to their complex morphology, karst terrains are particularly more fragile and vulnerable to environmental damage compared to most natural systems. Their hydraulic properties, such as their transmissivity (T) and spatial variability, can be relevant for understanding groundwater flow and, consequently, for the sustainable management of water resources. The application of geostatistical methods allows for spatial interpolation and mapping based on observations combined with uncertainty quantification. Direct measurements of T are typically scarce, while those of the specific capacity (Sc) are more frequent. We established a linear and spatial relationship between the logarithms of T and Sc measured in 174 wells in a semi-arid karst region in northeastern Brazil. These relationships were used to construct a cross-variogram, whose Linear Model of Coregionalization proved valid. The values and the cross-variogram of logT and logSc were used to generate interpolations over 2554 values of logSc, which did not spatially coincide with logT. We used ordinary co-kriging (CO-OK) and conditional sequential Gaussian co-simulation (CO-SGS) to generate the interpolations. The cross-variogram of logT and logSc, when considering 174 wells, was isotropic with an exponential structure, a nugget effect of approximately 20% of the sill, and a range of 5 km. Cross-validation indicated an optimal number of 10 neighboring wells used in CO-OK, and we used 500 stochastic realizations in CO-SGS, which were then used to generate maps of logT estimates, deviations derived from the interpolations, and probabilistic scenarios. The resulting transmissivity maps are relevant for the design of groundwater management strategies, including stochastic approaches where the transmissivity realizations can be used to parameterize multiple executions of numerical flow models.

Applications and challenges of digital soil mapping in Africa

The mapping of soils in Africa is at least a century old. We currently have access to various maps depicting mapping units locally and for the continent. In the past two decades, there has been a growing interest in alternatives for generating soil maps through digital soil mapping (DSM) techniques. There are, however, numerous challenges pertaining to the implementation of DSM in Africa, such as the unavailability of appropriate covariates, age and positional error in the measurements, low sampling density, and spatial clustering of the soil data used to fit and validate the models. This review aims to investigate the current state of DSM in Africa, identify challenges specific to implementing DSM in Africa and the ways it has been solved in the literature. We found that nearly half of African countries had an existing digital soil map covering either a local or national area, and that most studies were performed at a local extent. Soil carbon was the most common property under study, whereas soil hydraulic variables were seldom reported. Nearly all studies performed mapping for the topsoil up to 30 cm and calculated validation statistics using existing datasets but without collecting a post-mapping probability sample. Few studies (i.e., 11%) reported an estimate of map uncertainty. Half of the studies had in mind a downstream application (e.g., soil fertility assessment) in the map generation. We further correlated the area of study and sampling density and found a strong negative relationship. About 30% of the studies relied on legacy soil datasets and had a lack of sufficient spatial coverage of their area of study. From this review, we highlight some research opportunities and suggest improvements in the current methodologies. Future research should focus on capacity building in DSM, new data collection, and legacy data rescue. New initiatives, that should be initiated and led from within the continent, could support the long-term monitoring of soils and updating of soil information systems while ensuring their contextualised usability. This pairs with better delivery of existing DSM studies to stakeholders and the generation of a value-added proposition to governmental institutions.

Post-wildfire boreal forest vegetation cover change mapping via information fusion for secondary disaster risk assessments

Post-wildfire vegetation cover damage and loss can escalate the risks of secondary disasters such as flood, landslide, and water contamination, particularly in a major wildfire affected region where human settlements are situated. In assessments of the secondary disaster risks, the post-wildfire vegetation cover change plays a key role in influencing the distribution and intensity of the risks. In this work, a processing framework for mapping post-wildfire vegetation cover changes through information fusion has been generated and tested using Landsat8 and WorldView imagery data. The test site was the boreal forest region surrounding Fort McMurray, Alberta, Canada, affected by a massive wildfire in May 2016. The derived map results indicate that the fusion process in the framework is effective for generation of post-wildfire vegetation cover change information. The use of WorldView data revealed more variation details in distribution of the vegetation cover burn damages than use of Landsat data. Moreover, the uncertainty in vegetation burn severity using Landsat-based Differenced Normalized Burn Ratio (dNBR) index exists in the areas with low dNBR reading values due to the sub-pixel effect. The forest burn severity measured with dNBR index can be underestimated due to the quick herbaceous cover recovery after wildfire. These uncertainties in the post-wildfire vegetation cover mapping should be taken into consideration when the derived information is being used for risk assessments.

Assessing the potential of multi-source remote sensing data for cropland soil organic matter mapping in hilly and mountainous areas

Cropland soil organic matter (SOM) is recognized as a significant carbon reservoir in terrestrial ecosystems. Digital mapping of SOM in croplands is essential for comprehending the global carbon cycle. Accurately mapping cropland SOM using multi-source remote sensing data has been effectively incorporated into prediction models across various scales. However, the impact of multi-source remote sensing data on cropland SOM mapping outcomes in hilly and mountainous regions remains insufficiently understood. In this study, Jiangyou City, located in Sichuan Province, China, was chosen as a representative example of hilly and mountainous regions. Fifteen distinct feature combinations were devised using three remote sensing variables (Sentinel-1, Sentinel-2, and Landsat-8) along with DEM data. Feature selection was conducted using the Boruta algorithm. Subsequently, the RF, SVR, Cubist, and INLA-SPDE models were adopted to create spatially detailed distribution maps of cropland SOM for the region. Additionally, an uncertainty analysis was performed on the cropland SOM mapping results. The results indicate the following: (1) The INLA-SPDE model, which integrates both data information and spatial structure, achieves the highest accuracy and the less uncertainty in cropland SOM mapping, with an R2 of 0.647 and an RMSE of 4.227 g/kg. (2) Optical imagery is more important than SAR images, but their combination enhances model accuracy. Specifically, Sentinel-2 data has a significant impact cropland SOM prediction in hilly and mountainous areas, followed by Landsat-8 data. (3) The predicted spatial distribution patterns of cropland SOM by the four models show consistency, indicating lower SOM content in the southwest and higher SOM content in the central and northeast regions. This study provides valuable references for future large-scale and high-spatial cropland SOM prediction, highlighting the importance of spatial resolution for precise SOM prediction accuracy in hilly and mountainous regions.

Improved Robustness for Deep Learning-based Segmentation of Multi-Center Myocardial Perfusion MRI Datasets Using Data Adaptive Uncertainty-guided Space-time Analysis

Fully automatic analysis of myocardial perfusion MRI datasets enables rapid and objective reporting of stress/rest studies in patients with suspected ischemic heart disease. Developing deep learning techniques that can analyze multi-center datasets despite limited training data and variations in software (pulse sequence) and hardware (scanner vendor) is an ongoing challenge. Datasets from 3 medical centers acquired at 3T (n = 150 subjects; 21,150 first-pass images) were included: an internal dataset (inD; n = 95) and two external datasets (exDs; n = 55) used for evaluating the robustness of the trained deep neural network (DNN) models against differences in pulse sequence (exD-1) and scanner vendor (exD-2). A subset of inD (n = 85) was used for training/validation of a pool of DNNs for segmentation, all using the same spatiotemporal U-Net architecture and hyperparameters but with different parameter initializations. We employed a space-time sliding-patch analysis approach that automatically yields a pixel-wise "uncertainty map" as a byproduct of the segmentation process. In our approach, dubbed Data Adaptive Uncertainty-Guided Space-time (DAUGS) analysis, a given test case is segmented by all members of the DNN pool and the resulting uncertainty maps are leveraged to automatically select the "best" one among the pool of solutions. For comparison, we also trained a DNN using the established approach with the same settings (hyperparameters, data augmentation, etc.). The proposed DAUGS analysis approach performed similarly to the established approach on the internal dataset (Dice score for the testing subset of inD: 0.896 ± 0.050 vs. 0.890 ± 0.049; p = n.s.) whereas it significantly outperformed on the external datasets (Dice for exD-1: 0.885 ± 0.040 vs. 0.849 ± 0.065, p < 0.005; Dice for exD-2: 0.811 ± 0.070 vs. 0.728 ± 0.149, p < 0.005). Moreover, the number of image series with "failed" segmentation (defined as having myocardial contours that include bloodpool or are noncontiguous in ≥1 segment) was significantly lower for the proposed vs. the established approach (4.3% vs. 17.1%, p < 0.0005). The proposed DAUGS analysis approach has the potential to improve the robustness of deep learning methods for segmentation of multi-center stress perfusion datasets with variations in the choice of pulse sequence, site location or scanner vendor.

Target-based georeferencing of terrestrial radar images using TLS point clouds and multi-modal corner reflectors in geomonitoring applications

Terrestrial Radar Interferometry (TRI) is widely adopted in geomonitoring applications due to its capability to precisely observe surface displacements along the line of sight, among other key characteristics. As its deployment grows, TRI is also increasingly used to monitor smaller and more dispersed geological phenomena, where the challenge is their precise localization in 3d space if the pose of the radar interferometer is not known beforehand. To tackle this challenge, we introduce a semi-automatic target-based georeferencing method for precisely aligning TRI data with 3d point clouds obtained using long-range Terrestrial Laser Scanning (TLS). To facilitate this, we developed a multi-modal corner reflector (mmCR) that serves as a common reference point recognizable by both technologies, and we accompanied it with a semi-automatic data-processing pipeline, including the algorithms for precise center estimation. Experimental validation demonstrated that the corner reflector can be localized within the TLS data with a precision of 3–5 cm and within the TRI data with 1–2 dm. The targets were deployed in a realistic geomonitoring scenario to evaluate the implemented workflow and the achievable quality of georeferencing. The post-georeferencing mapping uncertainty was found to be on a decimeter level, matching the state-of-the-art results using dedicated targets and achieving more than an order of magnitude lower uncertainty than the existing data-driven approaches. In contrast to the existing target-based approaches, our results were achieved without laborious visual data inspection and manual target detection and on significantly larger distances, surpassing 2 km. The use of the developed mmCR and its associated data-processing pipeline extends beyond precise georeferencing of TRI imagery to TLS point clouds, allowing for alternatively georeferencing using total stations, mapping quality evaluation as well as on-site testing and calibrating TRI systems within the application environment.

Spatial distribution of soil rare earth elements in Sicily (Italy)

Rare earth elements (REEs) are becoming increasingly interesting as indicators of soil processes, and modelling their spatial distribution has become a fundamental requirement for this purpose. This study aims to model and quantify the spatial distribution of soil REEs taking into account their compositional nature in their mapping, and to assess the associated spatial uncertainty. In particular, the cerium anomaly (Ce⁎) and the ratio of the sum (Σ) of light to heavy REEs (ΣLREEs/ΣHREEs) were calculated and analysed. The study was carried out in Sicily (Italy), one of the largest (25,832 km2) and most populous island in the Mediterranean Sea. Soil REE data were obtained from the soil database of Sicily region. Soil REEs were transformed into a vector of isometric log-ratio (ilr) coordinates and analysed using a geostatistical approach for predicting their values at unsampled locations and generating 100 REEs realizations using turning bands simulation. Each REE element and Ce⁎ were mapped as well as the ΣLREEs/ΣHREEs ratio. The joint variability of the investigated REEs and the ΣLREEs/ΣHREEs ratio provided insights into their abundance and distribution patterns. From the simulated realizations, standard deviation maps were calculated for the Ce⁎ and the ΣLREEs/ΣHREEs ratio, providing a measure of their spatial uncertainty. Although with some limitations, the study established a first baseline of the Ce⁎ and the ΣLREEs/ΣHREEs ratio. The maps of spatial uncertainty may be a useful tool for planning future soil sampling and optimise the choice of new sampling locations.

Quantitative evaluation of uncertainty and interpretability in machine learning-based landslide susceptibility mapping through feature selection and explainable AI

Landslide susceptibility mapping (LSM) is essential for determining risk regions and guiding mitigation strategies. Machine learning (ML) techniques have been broadly utilized, but the uncertainty and interpretability of these models have not been well-studied. This study conducted a comparative analysis and uncertainty assessment of five ML algorithms—Random Forest (RF), Light Gradient-Boosting Machine (LGB), Extreme Gradient Boosting (XGB), K-Nearest Neighbor (KNN), and Support Vector Machine (SVM)—for LSM in Inje area, South Korea. We optimized these models using Bayesian optimization, a method that refines model performance through probabilistic model-based tuning of hyperparameters. The performance of these algorithms was evaluated using accuracy, Kappa score, and F1 score, with accuracy in detecting landslide-prone locations ranging from 0.916 to 0.947. Among them, the tree-based models (RF, LGB, XGB) showed competitive performance and outperformed the other models. Prediction uncertainty was quantified using bootstrapping and Monte Carlo simulation methods, with the latter providing a more consistent estimate across models. Further, the interpretability of ML predictions was analyzed through sensitivity analysis and SHAP values. We also expanded our investigation to include both the inclusion and exclusion of predictors, providing insights into each significant variable through a comprehensive sensitivity analysis. This paper provides insights into the predictive uncertainty and interpretability of ML algorithms for LSM, contributing to future research in South Korea and beyond.

Estimating the uncertainties of satellite derived soil moisture at global scale

This study attempts to derive the uncertainty of the soil moisture estimation from passive microwave satellite mission at global scale. To do so, the approach is based on the sensitivity of the Soil Moisture and Ocean Salinity (SMOS) soil moisture retrieval quality to the land surface characteristics within its footprint (presence of forest, topography, open water bodies, sand, clay, bulk density and soil organic carbon content). First, we performed a global assessment of SMOS using in situ measurements from the International Soil Moisture Network (ISMN) as reference, with more than 1900 ISMN stations and 10 years of SMOS data. This assessment shows that the ubRMSD scores vary greatly between locations (with a mean of 0.074 m3m−3 and an interquartile range of 0.030 m3m−3). Second, the scores are analyzed for different surface conditions within the satellite footprint. The best agreement between the ground measurement and SMOS time series are obtained for low forest cover, low topographic complexity, and marginal presence of open water bodies within the SMOS footprint. Soil parameters also have an impact, with better scores for sandier soils with a high bulk-density and low soil organic carbon content. Finally, we propose to extrapolate the obtained relationships, using a multiple linear regression, in order to derive a global map of SMOS uncertainties based on surface conditions. This map of predicted uncertainties show a diverse range of ubRMSD values across the globe (with a mean of 0.076 m3m−3 and an interquartile range of 0.031 m3m−3) depending on the surface characteristics. At the ISMN site location, the predicted ubRMSD shows similar results than the comparison between SMOS and the in situ measurements. The map of predicted SMOS ubRMSD represents an upper bound estimate of the SMOS uncertainty, as it includes the uncertainties of the in situ sensor measurements and the scale mismatch. Further investigations will focus on the different components of this uncertainty budget to obtain a better assessment of the absolute uncertainties of SMOS soil moisture retrievals across the globe.

Using spatial aggregation of soil multifunctionality maps to support uncertainty‐aware planning decisions

AbstractTo ensure soil preservation, it is essential to incorporate the soil's ability to provide ecosystem services into the spatial planning process. For well‐informed planning decisions, stakeholders need spatially explicit information on the state of the soils and the functions they fulfil, with sufficient spatial resolution and quantified uncertainty. It has been shown that Digital Soil Mapping (DSM) products can provide such information. However, in some cases, fine spatial resolution coupled with high levels of uncertainty may lead stakeholders to overlook the inherent uncertainties in the information. Spatial aggregation of DSM products opens up a promising avenue for obtaining maps that are more tailored to the users' scales of decision making while facilitating uncertainty communication. In this perspective, we propose a new spatial aggregation approach relying on spatially constrained agglomerative clustering (AC). The spatial aggregation approach is applied to a 25‐m‐resolution soil potential multifunctionality index (SPMI) map developed for the coastal plain of the Occitanie Region. This DSM product was increasingly aggregated to obtain SPMI maps of different resolutions displaying two distinct areal metrics: proportions of area above a given threshold of SPMI, and mean SPMI. Each map was evaluated through a set of indicators selected for their potential impact on user decision making: mean spatial resolution, overall predicted uncertainty, quantity of information and mean within‐unit variability. The maps were compared with respect to these indicators to other maps obtained with alternative aggregation methods employed in DSM literature (maps aggregated according to some administrative units and QuadMaps). We show that all the tested aggregation methods produced a substantial decrease of the map uncertainty with moderate loss of spatial resolution. However, only AC preserved the fine spatial pattern of the initial DSM product while enabling fine tuning of the uncertainty displayed to end‐users. We show that AC can simplify the identification of extensive regions characterized by low uncertainty without losing information regarding soil multifunctionality, thereby facilitating and enhancing the efficiency of planning decisions.

Georectifying drone image data over water surfaces without fixed ground control: Methodology, uncertainty assessment and application over an estuarine environment

Light-weight consumer-grade drones have the potential to provide geospatial image data to study a broad range of oceanic processes. However, rigorously tested methodologies to effectively and accurately geolocate and rectify these image data over mobile and dynamic water surfaces, where temporally fixed points of reference are unlikely to exist, are limited. We present a simple to use automated workflow for georectifying individual aerial images using position and orientation data from the drone's on-board sensor (i.e. direct-georectification). The presented methodology includes correcting for camera lens distortion and viewing angle and exploits standard mathematics and camera data processing techniques. The method is used to georectify image datasets from test flights with different combinations of altitude and camera angle. Using a test site over land, directly-georectified images, as well as the same images georectified using standard photogrammetry software, are evaluated using a network of known ground control points. The novel methodology performs well with the camera at nadir (both 10 m and 25 m above ground level) and exhibits a mean spatial accuracy of ±1 m. The same accuracy is achieved when the camera angle is 30° at 10 m above ground level but decreases to ±2.9 m at 30° and 25 m. The accuracy changes because the uncertainties are a function of the altitude and angle of the camera versus the ground. Drone in-flight positioning errors can reduce the accuracy further to ±5 m with the camera at 30° and 25 m. An ensemble approach is used to map the uncertainties within the camera field-of-view to show how they change with viewing distance and drone position and orientation. The complete approach is demonstrated over an estuarine environment that includes the shoreline and open water, producing results consistent with the land-based field-tests of accuracy. Overall, the workflow presented here provides a low cost and agile solution for direct-georectification of drone-captured image data over water surfaces. This approach could be used for collecting and processing image data from drones or ship-mounted cameras to provide observations of ocean colour, sea-ice, ocean glitter, sea surface roughness, white-cap coverage, coastal water quality, and river plumes. The Python scripts for the complete image georectification workflow, including uncertainty map generation, are available from https://github.com/JamieLab/SArONG.

Elucidating Uncertainty in Heat Vulnerability Mapping: Perspectives on Impact Variables and Modeling Approaches.

Heat vulnerability maps are vital for identifying at-risk areas and guiding interventions, yet their relationship with health outcomes is underexplored. This study investigates the uncertainty in heat vulnerability maps generated using health outcomes and various statistical models. We constructed vulnerability maps for 167 municipalities in Korea, focusing on the mild and severe health impacts of heat waves on morbidity and mortality. The outcomes included incidence rates of heat-related outpatient visits (morbidity) and attributable mortality rates (mortality) among individuals aged 65 years and older. To construct these maps, we utilized 11 socioeconomic variables related to population, climate, and economic factors. Both linear and nonlinear statistical models were employed to assign these socioeconomic variables to heat vulnerability. We observed variations in the crucial socioeconomic variables affecting morbidity and mortality in the vulnerability maps. Notably, nonlinear models depicted the spatial patterns of health outcomes more accurately than linear models, considering the relationship between health outcomes and socioeconomic variables. Our findings emphasize the differences in the spatial distribution of heat vulnerability based on health outcomes and the choice of statistical models. These insights underscore the importance of selecting appropriate models to enhance the reliability of heat vulnerability maps and their relevance for policy-making.

Risk-Targeted Seismic Design Maps with Aleatory and Epistemic Uncertainty in Tehran and Surrounding Areas

ABSTRACT This study aims to introduce a risk-targeted probabilistic model that facilitates seismic design hazard mapping. Here, we investigate how to incorporate hazard regional variations and different structural fragility typologies so as to achieve a uniform target risk across the map. This paper takes Tehran, the capital city, and its surrounding area as a case study and explores the highly digitized hazard curves conditioned on the time-dependent occurrence of an earthquake at a 50-year exposure time. We introduce a new set of parameterized quantile functions for Risk-targeted design Intensity Measure (IMR) as the seismic design hazard. A group of IMR maps is subsequently prepared based on a 1% collapse probability during the next 50-year lifetime. Furthermore, the quantile function corresponding to each fragility type is able to show the aleatory uncertainty at each site on the region map. The uncertainty maps illustrate relatively less inherent variability than what exists in the hazard curves themselves. Then, we tackle the source of uncertainty arising from percentile hazard curves into IMR, known as epistemic uncertainty, by deriving a new closed-form expression, which allows for a sampling-free estimation of the epistemic uncertainty. The risk-targeted seismic design hazard with its accompanying uncertainties can lead to a promising seismic design hazard with potential applications for the insurance industry, urban planners, and risk-aware building owners.

High-resolution mapping of tree species and associated uncertainty by combining aerial remote sensing data and convolutional neural networks ensemble

Mapping tree species diversity is essential for monitoring and managing forest ecosystems. Automating ecoforestry mapping using remote sensing images remains an important challenge due to the tremendous variability in forest covers and the conditions under which images used for classification are acquired. Deep learning algorithms have been increasingly used over the last few years to analyse remote sensing data for mapping tree species. However, most of these studies focus only on a small number of species or a limited area and avoid providing a spatially explicit representation of the uncertainty related to the predictions, rendering them unsuitable for operational use.In this study, we used an ensemble of convolutional neural networks to map forest species and land cover types across a 10,000 km2 area in Quebec, Canada, spanning mixed and boreal forests. We built a georeferenced label database to train and test nine models, which resulted from the combination of three training datasets and three-commonly used convolutional neural network architectures (VGG16, ResNet50v2, Densenet121). These models were trained on multiband aerial photographs and a canopy height model derived from airborne lidar point clouds and used to map the diversity and distribution of tree species and land cover types. The level of agreement among models was used to generate uncertainty maps. The performance of the super-ensemble using 1311 independent forest inventory plots and assessed the extent to which uncertainty maps could serve as a spatially explicit indicator of model performance.The super-ensemble achieved 90% global accuracy. Our results indicated that the performance of the super-ensemble surpassed that of all individual architectures while also showing a positive effect of the canopy height model on performance. The comparison of the super-ensemble map with the proportion of basal area measured in forest inventory plots confirmed the reliability of the model over the study area. Our results also indicated that mapping the inter-model agreement provides a reliable spatially explicit estimate of model performance. The robustness and reliability of the proposed approach support its use in an operational context while also providing a conceptual framework to evaluate the reliability of uncertainty maps.

Digital soil mapping of soil burn severity

AbstractFire alters soil hydrologic properties leading to increased risk of catastrophic debris flows and post‐fire flooding. As a result, US federal agencies map soil burn severity (SBS) via direct soil observation and adjustment of rasters of burned area reflectance. We developed a unique application of digital soil mapping (DSM) to map SBS in the Creek Fire which burned 154,000 ha in the Sierra Nevada. We utilized 169 ground‐based observations of SBS in combination with raster proxies of soil forming factors, pre‐fire fuel conditions, and fire effects to vegetation to build a digital soil mapping model of soil burn severity (DSMSBS) using a random forest algorithm and compared the DSMSBS map to the established SBS map. The DSMSBS model had a cross‐validation accuracy of 48%. The established technique had 46% agreement between field observations and pixels. However, since the established technique is manual, it could not be compared to the DSMSBS model via cross‐validation. We produced SBS class uncertainty maps, which showed high prediction probabilities around field observations, and low probabilities away from field observations. SBS prediction probabilities could aid post‐fire assessment teams with sample prioritization. We report 107 km2 more area classified as high and moderate SBS compared to the established technique. We conclude that blending soil forming factors based mapping and vegetation burn severity mapping can improve SBS mapping. This represents a shift in SBS mapping away from validating remotely sensed reflectance imagery and toward a quantitative soil landscape model, which incorporates both fire and soils information to directly predict SBS.

Benchmarking the performance and uncertainty of machine learning models in estimating scour depth at sluice outlets

ABSTRACT This study investigates the performance of six machine learning (ML) models – Random Forest (RF), Adaptive Boosting (ADA), CatBoost (CAT), Support Vector Machine (SVM), Lasso Regression (LAS), and Artificial Neural Network (ANN) – against traditional empirical formulas for estimating maximum scour depth after sluice gates. Our findings indicate that ML models generally outperform empirical formulas, with correlation coefficients (CORR) ranging from 0.882 to 0.944 for ML models compared with 0.835–0.847 for empirical methods. Notably, ANN exhibited the highest performance, followed closely by CAT, with a CORR of 0.936. RF, ADA, and SVM performed competitive metrics around 0.928. Variable importance assessments highlighted the dimensionless densimetric Froude number (Fd) as significantly influential, particularly in RF, CAT, and LAS models. Furthermore, SHAP value analysis provided insights into each predictor's impact on model outputs. Uncertainty assessment through Monte Carlo (MC) and Bootstrap (BS) methods, with 1,000 iterations, indicated ML's capability to produce reliable uncertainty maps. ANN leads in performance with higher mean values and lower standard deviations, followed by CAT. MC results trend towards optimistic predictions compared with BS, as reflected in median values and interquartile ranges. This analysis underscores the efficacy of ML models in providing precise and reliable scour depth predictions.

UGEE-Net: Uncertainty-guided and edge-enhanced network for image splicing localization

Image splicing, a prevalent method for image tampering, has significantly undermined image authenticity. Existing methods for Image Splicing Localization (ISL) struggle with challenges like limited accuracy and subpar performance when dealing with imperceptible tampering and multiple tampered regions. We introduce an Uncertainty-Guided and Edge-Enhanced Network (UGEE-Net) for ISL to tackle these issues. UGEE-Net consists of two core tasks: uncertainty guidance and edge enhancement. We employ Bayesian learning to model uncertainty maps of tampered regions, directing the model's focus to challenging pixels. Simultaneously, we employ a frequency domain-auxiliary edge enhancement strategy to imbue localization features with global contour information and fine-grained local details. These mechanisms work in parallel, synergistically boosting performance. Additionally, we introduce a cross-level fusion and propagation mechanism that effectively utilizes contextual information for cross-layer feature integration and leverages channel-level correlations for cross-layer feature propagation, gradually enhancing the localization feature's details. Experiment results affirm UGEE-Net's superiority in terms of detection accuracy, robustness, and generalization capabilities. Furthermore, to meet the growing demand for high-quality datasets in image forensics, we present the HTSI12K dataset, which includes 12,000 spliced images with imperceptible tampering traces and diverse categories, rendering it suitable for real-world auxiliary model training.

Uncovering mangrove range limits using very high resolution satellite imagery to detect fine‐scale mangrove and saltmarsh habitats in dynamic coastal ecotones

AbstractMangroves are important ecosystems for coastal biodiversity, resilience and carbon dynamics that are being threatened globally by human pressures and the impacts of climate change. Yet, at several geographic range limits in tropical–temperate transition zones, mangrove ecosystems are expanding poleward in response to changing macroclimatic drivers. Mangroves near range limits often grow to smaller statures and form dynamic, patchy distributions with other coastal habitats, which are difficult to map using moderate‐resolution (30‐m) satellite imagery. As a result, many of these mangrove areas are missing in global distribution maps. To better map small, scrub mangroves, we tested Landsat (30‐m) and Sentinel (10‐m) against very high resolution (VHR) Planet (3‐m) and WorldView (1.8‐m) imagery and assessed the accuracy of machine learning classification approaches in discerning current (2022) mangrove and saltmarsh from other coastal habitats in a rapidly changing ecotone along the east coast of Florida, USA. Our aim is to (1) quantify the mappable differences in landscape composition and complexity, class dominance and spatial properties of mangrove and saltmarsh patches due to image resolution; and (2) to resolve mapping uncertainties in the region. We found that the ability of Landsat to map mangrove distributions at the leading range edge was hampered by the size and extent of mangrove stands being too small for detection (50% accuracy). WorldView was the most successful in discerning mangroves from other wetland habitats (84% accuracy), closely followed by Planet (82%) and Sentinel (81%). With WorldView, we detected 800 ha of mangroves within the Florida range‐limit study area, 35% more mangroves than were detected with Planet, 114% more than Sentinel and 537% more than Landsat. Higher‐resolution imagery helped reveal additional variability in landscape metrics quantifying diversity, spatial configuration and connectedness among mangrove and saltmarsh habitats at the landscape, class and patch scales. Overall, VHR satellite imagery improved our ability to map mangroves at range limits and can help supplement moderate‐resolution global distributions and outdated regional maps.

Mapping surface sediment characteristics in enclosed shallow‐marine environments using spatially balanced designs and the random forest algorithm

AbstractMapping the sedimentary character of the seafloor in large water‐filled basins is fundamental for understanding landform dynamics to inform research, management, intervention and conservation actions. Seabed mapping methods have undergone considerable development in the last two decades, including the uptake of machine learning approaches for sediment size prediction and classification. However, predictions of surficial sediment characteristics are often hindered by the availability of ground truthing data, their arrangement in space and the modelling approach chosen. Spatially informed sampling designs provide an opportunity to significantly improve the accuracy and uncertainty of predicted sediment distributions. In this study, we apply a machine learning algorithm to predict sediment distributions across Port Phillip Bay, a large (1930 km2) structurally controlled estuary on the southeast coast of Victoria, Australia. Surface sediment samples (n = 252) were collected using a spatially balanced design, ensuring that sampling effort was spread evenly within the embayment with increased sampling intensity placed in more heterogeneous areas. Surficial textural metrics were modelled using the random forest algorithm with bathymetric and hydrodynamic predictor variables. Models highlighted trends in sediment grain size, sorting and composition consistent with predicted wave‐ and current‐induced sediment mobilisation. Model predictions were accurate (normalised‐root mean squared error [NRMSE]: 0.14–0.16); however, standard error was not homogeneous across the study area. Uncertainty maps highlighted areas where additional sampling effort may be needed, including areas where transitional bathymetry impacted surficial sediment character and areas of anthropogenic modifications to the seabed. This study shows the benefits of undertaking spatially informed sample design, block cross‐validation during model fitting and quantifying spatial uncertainty in predictive maps to accurately quantify the fundamental boundary conditions of sediment size. The results of this study are intended to inform local coastal management, including beach renourishment activities. However, approaches outlined are applicable to any study where the seafloor grain size is a fundamental variable in understanding landscape change.