Explore the vast range of titles available.

This study introduces an approach to validate sampling points for air monitoring in environments where operations generate significant airborne particulate radioactivity. A case study of repair works in a nuclear spent fuel reprocessing facility is used to synthesis and demonstrate the methodology. We use the probability distribution of air activity measurements and the correlation between two sampling points near high particulate generating operations and the ventilation ducts. The methodology developed in this paper can be applied in task-related air monitoring scenarios for validation of sampling points. This will augment internal exposure control measures during works involving physical entries in areas with high potential of escalation of air activity (e.g., during major repairs, during decommissioning of nuclear facilities etc.) and to evaluate the sufficiency of respiratory protection.

Species distribution modeling (SDM) is widely used in ecology and conservation. Currently, the most available data for SDM are species presence-only records (available through digital databases). There have been many studies comparing the performance of alternative algorithms for modeling presence-only data. Among these, a 2006 paper from Elith and colleagues has been particularly influential in the field, partly because they used several novel methods (at the time) on a global data set that included independent presence–absence records for model evaluation. Since its publication, some of the algorithms have been further developed and new ones have emerged. In this paper, we explore patterns in predictive performance across methods, by reanalyzing the same data set (225 species from six different regions) using updated modeling knowledge and practices. We apply well-established methods such as generalized additive models and MaxEnt, alongside others that have received attention more recently, including regularized regressions, point-process weighted regressions, random forests, XGBoost, support vector machines, and the ensemble modeling framework biomod. All the methods we use include background samples (a sample of environments in the landscape) for model fitting. We explore impacts of using weights on the presence and background points in model fitting. We introduce new ways of evaluating models fitted to these data, using the area under the precision-recall gain curve, and focusing on the rank of results. We find that the way models are fitted matters. The top method was an ensemble of tuned individual models. In contrast, ensembles built using the biomod framework with default parameters performed no better than single moderate performing models. Similarly, the second top performing method was a random forest parameterized to deal with many background samples (contrasted to relatively few presence records), which substantially outperformed other random forest implementations. We find that, in general, nonparametric techniques with the capability of controlling for model complexity outperformed traditional regression methods, with MaxEnt and boosted regression trees still among the top performing models. All the data and code with working examples are provided to make this study fully reproducible.

Population-based surveys commonly use point-of-care (POC) methods with capillary blood samples for estimating Hb concentrations; these estimates need to be validated by comparison with reference methods using venous blood. In a cross-sectional study in 748 participants (17–86 years, 708 women, Hb: 5·1 to 18·2 g/dl) from Hyderabad, India, we validated Hb measured from a pooled capillary blood sample by a POC autoanalyser (Horiba ABX Micros 60OT, Hb-C-AA) by comparison with venous blood Hb measured by two reference methods: POC autoanalyser (Hb-V-AA) and cyanmethemoglobin method (Hb-V-CM). These comparisons also allowed estimation of blood sample-related and equipment-related differences in the Hb estimates. We also conducted a longitudinal study in 426 participants (17–21 years) to measure differences in the Hb response to iron folate (IFA) treatment by the capillary blood POC method compared with the reference methods. In the cross-sectional study, Bland–Altman analyses showed trivial differences between source of blood (Hb-C-AA and Hb-V-AA; mean difference, limits of agreement: 0·1, −0·8 to 1·0 g/dl) and between analytical methods (Hb-V-AA and Hb-V-CM; mean difference, limits of agreement: < 0·1, −1·8 to 1·8 g/dl). Cross-sectional anaemia prevalence estimated using Hb-C-AA did not differ significantly from Hb-V-CM or Hb-V-AA. In the longitudinal study, the Hb increment in response to IFA intervention was not different when using Hb-C-AA (1·6 ± 1·7 g/dl) compared with Hb-V-AA (1·7 ± 1·7 g/dl) and Hb-V-CM (1·7 ± 1·7 g/dl). The pooled capillary blood–autoanalyzer method (Hb-C-AA) offers a practical and accurate way forward for POC screening of anaemia.

Observational data of the Earth’s weather and climate at the level of ground-based weather stations are prone to gaps due to a variety of causes. These gaps can inhibit scientific research as they impede the use of numerical models for agricultural, meteorological and climatological applications as well as introducing analytic biases. In this research, different machine learning techniques are evaluated together with traditional approaches to gap filling automated weather station data. When filling gaps for a specific data stream, data from neighbouring weather stations are used in addition to reanalysis data from the European Centre for Medium-Range Weather Forecasts atmospheric reanalyses of the global climate, ERA-5 Land. A novel gap creation method is introduced that provides 100% coverage in sampling the dataset while ensuring that the sampled data are randomly distributed. Gap filling across a range of different gap lengths and target variables are compared using a range of error functions. The variables selected for modelling are mean air temperature, dew point, mean relative humidity and leaf wetness. Our results show that models perform best on gap-filling temperature and dew point with worst performance on leaf wetness. As expected, model performance decreases with increasing gap length. Comparison between machine learning and reanalysis approaches show very promising results from a number of the machine learning models.

UAV transmission tower inspection is the use of UAV technology for regular inspection and troubleshooting of towers on transmission lines, which helps to improve the safety and reliability of transmission lines and ensures the stability of the power supply. From the traditional manual tower boarding to the current way of manually selecting target camera shooting points from 3D point clouds to plan the inspection path of the UAV, operational efficiency has drastically improved. However, indoor planning work is still labor-consuming and expensive. In this paper, a deep learning-based point cloud transmission tower segmentation (PCTTS) model combined with the corresponding target point localization algorithm is proposed for automatic segmentation of transmission tower point cloud data and automatically localizing the key inspection component as the target point for UAV inspection. First, we utilize octree sampling with unit ball normalization to simplify the data and ensure translation invariance before putting the data into the model. In the feature extraction stage, we encode the point set information and combine Euclidean distance and cosine similarity features to ensure rotational invariance. On this basis, we adopt multi-scale feature extraction, construct a local coordinate system, and introduce the offset-attention mechanism to enhance model performance further. Then, after the feature propagation module, gradual up-sampling is used to obtain the features of each point to complete the point cloud segmentation. Finally, combining the segmentation results with the target point localization algorithm completes the automatic extraction of UAV inspection target points. The method has been applied to six kinds of transmission tower point cloud data of part segmentation results and three kinds of transmission tower point cloud data of instance segmentation results. The experimental results show that the model achieves mIOU of 94.1% on the self-built part segmentation dataset and 86.9% on the self-built instance segmentation dataset, and the segmentation accuracy outperforms that of the methods for point cloud segmentation, such as PointNet++, DGCNN, Point Transformer, and PointMLP. Meanwhile, the experimental results of UAV inspection target point localization also verify the method’s effectiveness in this paper.

The upcoming Fluorescence Explorer (FLEX) mission will provide sun-induced fluorescence (SIF) products at unprecedented spatial resolution. Thus, accurate calibration and validation (cal/val) of these products are key to guarantee robust SIF estimates for the assessment and quantification of photosynthetic processes. In this study, we address one specific component of the uncertainty budget related to SIF retrieval: the spatial representativeness of in situ SIF observations compared to medium-resolution SIF products (e.g., 300 m pixel size). Here, we propose an approach to evaluate an optimal sampling strategy to characterise the spatial representativeness of in situ SIF observations based on high-spatial-resolution SIF data. This approach was applied for demonstration purposes to two agricultural areas that have been extensively characterized with a HyPlant airborne imaging spectrometer in recent years. First, we determined the spatial representativeness of an increasing number of sampling points with respect to a reference area (either monocultural crop fields or hypothetical FLEX pixels characterised by different land cover types). Then, we compared different sampling approaches to determine which strategy provided the most representative reference data for a given area. Results show that between 3 and 13.5 sampling points are needed to characterise the average SIF value of both monocultural fields and hypothetical FLEX pixels of the agricultural areas considered in this study. The number of sampling points tends to increase with the standard deviation of SIF of the reference area, as well as with the number of land cover classes in a FLEX pixel, even if the increase is not always statistically significant. This study contributes to guiding cal/val activities for the upcoming FLEX mission, providing useful insights for the selection of the validation site network and particularly for the definition of the best sampling scheme for each site.

Occupancy models are popular for estimating the probability a site is occupied by a species of interest when detection is imperfect. Occupancy models have been extended to account for interacting species and spatial dependence but cannot presently allow both factors to act simultaneously. We propose a two‐species occupancy model that accommodates both interspecific and spatial dependence. We use a point‐referenced multivariate hierarchical spatial model to account for both spatial and interspecific dependence. We model spatial random effects with predictive process models and use probit regression to improve efficiency of posterior sampling. We model occupancy probabilities of red fox (Vulpes vulpes) and coyote (Canis latrans) with camera trap data collected from six mid‐Atlantic states in the eastern United States. We fit four models comprising a fully factorial combination of spatial and interspecific dependence to two‐thirds of camera trapping sites and validated models with the remaining data. Red fox and coyotes each exhibited spatial dependence at distances >0.8 and 0.4 km, respectively, and exhibited geographic variation in interspecific dependence. Consequently, predictions from the model assuming simultaneous spatial and interspecific dependence best matched test data observations. This application highlights the utility of simultaneously accounting for spatial and interspecific dependence.

Rock type classification is crucial for evaluating mineral resources in ore deposits and for rock mechanics. Mineral deposits are formed in a variety of rock bodies and rock types. However, the rock type identification in drill core samples is often complicated by overprinting and weathering processes. An approach to classifying rock types from drill core data relies on whole-rock geochemical assays as features. There are few studies on rock type classification from a limited number of metal grades and dry bulk density as features. The novelty in our approach is the introduction of two sets of feature variables (proxies) at sampled data points, generated by geostatistical leave-one-out cross-validation and by kriging for removing short-scale spatial variation of the measured features. We applied our proposal to a dataset from a porphyry Cu–Au deposit in Mongolia. The model performances on a testing data subset indicate that, when the training dataset is not large, the performance of the classifier (a random forest) substantially improves by incorporating the proxy features as a complement to the original measured features. At each training data point, these proxy features throw light based on the underlying spatial data correlation structure, scales of variations, sampling design, and values of features observed at neighboring points, and show the benefits of combining geostatistics with machine learning.

1. The validation of field sampling techniques is a concern for applied ecologists due to the strong model assumptions implicit in all methods. Computer simulations make replication easy, but they do not give insights into how much bias occurs in real populations. Testing sampling methods on populations of known size can establish directly how well estimators perform, but such populations are very hard to find, and replicate, and they may have unusual attributes. 2. We present a field validation of distance and double-observer methods of estimating detection probabilities on aural avian point counts. Our research is relevant to conservation agencies worldwide who design thousands of avian monitoring programmes based primarily on auditory point counts. The programmes are a critical component in the management of many avian species. 3. Our validation used a simulation system which mimics birds calling in a field environment. The system allowed us to vary singing rate, species, distance, the complexity of points, and other factors. 4. Distance methods performed poorly, primarily due to large localization errors, and estimates did not improve for simplified points. 5. For the double-observer method, two pairs of observers tended to underestimate true population size, while the third pair tended to double-count birds which overestimated the population. Detection probabilities were always higher and population estimates lower when observers subjectively matched birds compared to an objective rule and showed a slight negative bias and good precision. A simplified 45-degree matching rule did not improve the performance of double-observer estimates which had a slight positive bias and much lower precision. Double-observer estimates did improve on the simplified points. 6. Synthesis and applications. We encourage ecologists working with sampling methods to develop similar methods of working with simulated populations through use of technology. Our simulated field evaluation has demonstrated the difficulty of accurately estimating population size when limited to aural detections. Problems are related to limitations in the ability of observers to localize sound, estimate distance, and accurately identify birds during a count. Other sources of error identified are the effects of observers, singing rate, singing orientation and background noise.

ContextSpecies-specific models of landscape capability (LC) can inform landscape conservation design. Landscape capability is “the ability of the landscape to provide the environment […] and the local resources […] needed for survival and reproduction […] in sufficient quantity, quality and accessibility to meet the life history requirements of individuals and local populations.” Landscape capability incorporates species’ life histories, ecologies, and distributions to model habitat for current and future landscapes and climates as a proactive strategy for conservation planning.ObjectivesWe tested the ability of a set of LC models to explain variation in point occupancy and abundance for seven bird species representative of spruce-fir, mixed conifer-hardwood, and riparian and wooded wetland macrohabitats.MethodsWe compiled point count data sets used for biological inventory, species monitoring, and field studies across the northeastern United States to create an independent validation data set. Our validation explicitly accounted for underestimation in validation data using joint distance and time removal sampling.ResultsBlackpoll warbler (Setophaga striata), wood thrush (Hylocichla mustelina), and Louisiana (Parkesia motacilla) and northern waterthrush (P. noveboracensis) models were validated as predicting variation in abundance, although this varied from not biologically meaningful (1%) to strongly meaningful (59%). We verified all seven species models [including ovenbird (Seiurus aurocapilla), blackburnian (Setophaga fusca) and cerulean warbler (Setophaga cerulea)], as all were positively related to occupancy data.ConclusionsLC models represent a useful tool for conservation planning owing to their predictive ability over a regional extent. As improved remote-sensed data become available, LC layers are updated, which will improve predictions.