Fine Spatial Scales Research Articles

Fine-Scale Spatial Genetic Structure of Remnant Populations of Abies religiosa, in a Temperate Forest in Central Mexico

Genetic structure of a population can be molded by the resistance of the landscape or the distance between populations that function as barriers to gene flow. We analyzed the population genetic structure of Abies religiosa on a fine spatial scale and examined isolation models by resistance and distance. We collected vegetative tissue from populations located at the altitudinal extremes of the distribution range of the species on three slopes of La Malinche National Park (LMNP) (South, North, and East) in central Mexico. Genomic DNA was obtained using the CTAB 2X method, and eight microsatellite chloroplast loci were amplified. The genetic structure was identified based on an Analysis of Molecular Variance, a Discriminant Analysis of Principal Components with cross-validation and a spatial Principal Component Analysis using the Gabriel-type connectivity network. The isolation hypotheses were evaluated by constructing partial Mantel tests using Reciprocal Causal Modeling and Maximum Likelihood Population Effects models. A genetic structure of isolation by resistance to elevation was identified, and two genetic groups were recognized: one including populations of the South slope and the other comprising populations of the North and East slopes. We detected in Abies religiosa populations of the LMNP an isolation by resistance to elevation that determines the genetic structure, and the greatest genetic exchange between groups of populations located at higher altitudes. It is suggested to promote the connectivity between slopes through assisted migration and immediately halt land-use changes, as part of the actions to preserve genetic diversity in the LMPN. This study contributes to the knowledge of the spatial genetic structure of species at risk in the Mexican temperate forest for their conservation.

Influence of urbanisation on the phenology of evergreen coniferous and deciduous broadleaf trees in Madrid (Spain)

Plant phenology plays a crucial role in terrestrial ecosystems and is a key indicator of ecological changes. It is influenced by alterations in meteorological conditions at micro- and mesoscale in cities as a result of urbanisation and the intensification of anthropogenic activities. One example is the urban heat island (UHI), i.e. the increase in temperature in city cores compared to their surroundings. In this study, we used plant phenological data from the High-Resolution Vegetation Phenology and Productivity dataset of the Copernicus Land Monitoring Service to explore the relationship between urbanisation intensity (i.e., proportion of built-up area) and the phenology and productivity of broadleaf deciduous and evergreen coniferous trees in the city of Madrid and its surroundings, one of the largest metropolitan areas in Europe. We found the average start of season date advanced by 0.79–11.79 days in broadleaf deciduous trees in highly urbanised areas, and was delayed by 5.77–9.16 days in evergreen coniferous trees. Although differences were also observed for the end and length of the growing season, these patterns showed great interannual variability. Both broadleaf deciduous and evergreen coniferous trees were most productive in highly urbanised areas. Although previous studies have explored the response of different land-cover types to urbanisation, to our knowledge, none have looked at this relationship at a fine spatial scale in a Mediterranean city with different vegetation functional groups. The analysis of more specific classes of vegetation enables a better understanding of the phenological response of vegetation to urbanisation intensity.

A digital close range photogrammetric observation system for measuring soil surface morphology during ongoing rainfall

Soil erosion is a hydro-geomorphic process occurring over the earth’ surface. The observation of soil surface morphology during ongoing rainfall at fine spatial and temporal scales are critical for the study of soil erosion. A digital close range photogrammetric observation system based on wireless networking technology was explored and established in this study. Through wireless networking, multi-cameras were concurrently controlled, solving the problems of synchronous acquisition of underlying digital image, big data transmission and storage. Image sensors collected data in parallel based on network command, and noise on the images such as raindrops was eliminated based on the K-means clustering algorithm by serialization technology during data acquisition. After parallel preprocessing, high density 3D point clouds and digital elevation models (DEMs) were reconstructed. The evolution of soil surface topography was dynamically monitored by instantaneous image acquisition at different time intervals during ongoing rainfall. The results showed that the measurement precision of the established system could reach a sub-millimeter level with the minimum single measurement error being 0.0069 mm. The maximum error between check point coordinates generated by photogrammetry and those generated by a Total station was 5.7 mm on the horizontal axes and 9.6 mm on the vertical axis. The temporal and spatial resolutions of the observation system were 1 min and 2 mm, respectively. Compared with a runoff sediment collection method, the average relative error of the system for estimating soil loss was 5.63%, and the accuracy of single observation was up to 99.58%. The observation methods explored in our study provided a reliable way to monitor soil erosion processes, which was also helpful to analyze the role and influence mechanism of soil erosion in surface hydrological process.

Age-specific allocation of glucosinolates within plant reproductive tissues

The Optimal Defense Theory (ODT) predicts that the distribution of defenses within a plant should mirror the value and vulnerability of each tissue. Although the ODT has received much experimental support, very few studies have examined defense allocation among reproductive tissues and none assessed simultaneously how these defenses evolve with age. We quantified glucosinolates in perianths, anthers and pistils at different bud maturity stages (i.e., intermediate flower buds, old flower buds and flowers) of undamaged and mechanically damaged plants of an annual brassicaceous species. The youngest leaf was used as a reference for vegetative organs, since it is predicted to be one of the most defended. In line with ODT predictions, reproductive tissues were more defended than vegetative tissues constitutively, and within the former, pistils and anthers more defended than perianths. No change in the overall defense level was found between bud maturity stages, but a significant temporal shift was observed between pistils and anthers. Contrary to ODT predictions, mechanical damage did not induce systemic defenses in leaves but only in pistils. Our results show that defense allocation in plant reproductive tissues occurs at fine spatial and temporal scales, extending the application framework of the ODT. They also demonstrate interactions between space and time in fine-scale defense allocation.

Influences on koala habitat selection across four local government areas on the far north coast of NSW

AbstractConserving habitats crucial for threatened koala (Phascolarctos cinereus) populations requires rating habitat quality from a fine spatial scale to patches, landscapes and then regions. The koala has a specialized diet focused on the leaves of a suite of Eucalyptus species. We asked: what are the key regional influences on habitat selection by koalas in the far north coast of New South Wales? We addressed this question by investigating the multi‐scale factors, and within‐scale and cross‐scale interactions, that influence koala habitat selection and distribution across four local government areas on the far north coast of New South Wales. We assembled and analysed a large data set of tree selection, identified by the presence of scats, in a wide range of randomly selected 5 × 5 km grids across the region. This resulted in more than 9000 trees surveyed for evidence of koala use from 302 field sites, together with associated biophysical site features. The dominant factor influencing habitat use and koala occurrence was the distribution of five Eucalyptus species. Koalas were more likely to use medium‐sized trees of these species where they occurred on soils with high levels of Colwell phosphorous. We also identified new interactions among the distribution of preferred tree species and soil phosphorous, and their distribution with the amount of suitable habitat in the surrounding landscape. Our study confirmed that non‐preferred species of eucalypts and non‐eucalypts are extensively used by koalas and form important components of koala habitat. This finding lends support to restoring a mosaic of koala‐preferred tree species and other species recognized for their value as shelter. Our study has provided the ecological foundation for developing a novel regional‐scale approach to the conservation of koalas, with adaptability to other wildlife species.

Understanding the spatiotemporal distribution of snow refugia in the rain-snow transition zone of north-central Idaho

Knowledge of snow cover distribution and disappearance dates over a wide range of scales is imperative for understanding hydrological dynamics and for habitat management of wildlife species that rely on snow cover. Identification of snow refugia, or places with relatively late snow disappearance dates (SDDs) compared to surrounding areas, is especially important as climate change alters snow cover timing and duration. The purpose of this study was to increase understanding of snow refugia in complex terrain spanning the rain-snow transition zone at fine spatial and temporal scales. To accomplish this objective, we used remote cameras to provide relatively high temporal and spatial resolution measurements on snowpack conditions. We built linear models to relate SDDs at the monitoring sites to topoclimatic and canopy cover metrics. One model to quantify SDDs included elevation, aspect, and an interaction between canopy cover and cold-air pooling potential. High-elevation, north-facing sites in cold-air pools (CAPs) had the latest SDDs, but isolated lower-elevation points also exhibited relatively late potential SDDs. Importantly, canopy cover had a much stronger effect on SDDs in CAPs than in non-CAPs, indicating that best practices in forest management for snow refugia could vary across microtopography. A second model that included in situ hydroclimate observations (December–February (DJF) temperature and March 1 snow depth) indicated that March 1 snow depth had little impact on SDD at the coldest winter temperatures, and that DJF temperatures had a stronger effect on SDD at lower snow depths, implying that the relative importance of snowfall and temperature could vary across hydroclimatic contexts in their impact on snow refugia. This new understanding of factors influencing snow refugia can guide forest management actions to increase snow retention and inform management of snow-dependent wildlife species in complex terrain.

A 21-year dataset (2000–2020) of gap-free global daily surface soil moisture at 1-km grid resolution

Global soil moisture estimates from current satellite missions are suffering from inherent discontinuous observations and coarse spatial resolution, which limit applications especially at the fine spatial scale. This study developed a dataset of global gap-free surface soil moisture (SSM) at daily 1-km resolution from 2000 to 2020. This is achieved based on the European Space Agency - Climate Change Initiative (ESA-CCI) SSM combined product at 0.25° resolution. Firstly, an operational gap-filling method was developed to fill the missing data in the ESA-CCI SSM product using SSM of the ERA5 reanalysis dataset. Random Forest algorithm was then adopted to disaggregate the coarse-resolution SSM to 1-km, with the help of International Soil Moisture Network in-situ observations and other optical remote sensing datasets. The generated 1-km SSM product had good accuracy, with a high correlation coefficent (0.89) and a low unbiased Root Mean Square Error (0.045 m3/m3) by cross-validation. To the best of our knowledge, this is currently the only long-term global gap-free 1-km soil moisture dataset by far.

Population genomics and conservation management of the threatened black-footed tree-rat (Mesembriomys gouldii) in northern Australia

Genomic diversity is a fundamental component of Earth’s total biodiversity, and requires explicit consideration in efforts to conserve biodiversity. To conserve genomic diversity, it is necessary to measure its spatial distribution, and quantify the contribution that any intraspecific evolutionary lineages make to overall genomic diversity. Here, we describe the range-wide population genomic structure of a threatened Australian rodent, the black-footed tree-rat (Mesembriomys gouldii), aiming to provide insight into the timing and extent of population declines across a large region with a dearth of long-term monitoring data. By estimating recent trajectories in effective population sizes at four localities, we confirm widespread population decline across the species’ range, but find that the population in the peri-urban area of the Darwin region has been more stable. Based on current sampling, the Melville Island population made the greatest contribution to overall allelic richness of the species, and the prioritisation analysis suggested that conservation of the Darwin and Cobourg Peninsula populations would be the most cost-effective scenario to retain more than 90% of all alleles. Our results broadly confirm current sub-specific taxonomy, and provide crucial data on the spatial distribution of genomic diversity to help prioritise limited conservation resources. Along with additional sampling and genomic analysis from the far eastern and western edges of the black-footed tree-rat distribution, we suggest a range of conservation and research priorities that could help improve black-footed tree-rat population trajectories at large and fine spatial scales, including the retention and expansion of structurally complex habitat patches.

Spatiotemporal Trends in Self-Reported Mask-Wearing Behavior in the United States: Analysis of a Large Cross-sectional Survey.

Face mask wearing has been identified as an effective strategy to prevent the transmission of SARS-CoV-2, yet mask mandates were never imposed nationally in the United States. This decision resulted in a patchwork of local policies and varying compliance, potentially generating heterogeneities in the local trajectories of COVID-19 in the United States. Although numerous studies have investigated the patterns and predictors of masking behavior nationally, most suffer from survey biases and none have been able to characterize mask wearing at fine spatial scales across the United States through different phases of the pandemic. Urgently needed is a debiased spatiotemporal characterization of mask-wearing behavior in the United States. This information is critical to further assess the effectiveness of masking, evaluate the drivers of transmission at different time points during the pandemic, and guide future public health decisions through, for example, forecasting disease surges. We analyzed spatiotemporal masking patterns in over 8 million behavioral survey responses from across the United States, starting in September 2020 through May 2021. We adjusted for sample size and representation using binomial regression models and survey raking, respectively, to produce county-level monthly estimates of masking behavior. We additionally debiased self-reported masking estimates using bias measures derived by comparing vaccination data from the same survey to official records at the county level. Lastly, we evaluated whether individuals' perceptions of their social environment can serve as a less biased form of behavioral surveillance than self-reported data. We found that county-level masking behavior was spatially heterogeneous along an urban-rural gradient, with mask wearing peaking in winter 2021 and declining sharply through May 2021. Our results identified regions where targeted public health efforts could have been most effective and suggest that individuals' frequency of mask wearing may be influenced by national guidance and disease prevalence. We validated our bias correction approach by comparing debiased self-reported mask-wearing estimates with community-reported estimates, after addressing issues of a small sample size and representation. Self-reported behavior estimates were especially prone to social desirability and nonresponse biases, and our findings demonstrated that these biases can be reduced if individuals are asked to report on community rather than self behaviors. Our work highlights the importance of characterizing public health behaviors at fine spatiotemporal scales to capture heterogeneities that may drive outbreak trajectories. Our findings also emphasize the need for a standardized approach to incorporating behavioral big data into public health response efforts. Even large surveys are prone to bias; thus, we advocate for a social sensing approach to behavioral surveillance to enable more accurate estimates of health behaviors. Finally, we invite the public health and behavioral research communities to use our publicly available estimates to consider how bias-corrected behavioral estimates may improve our understanding of protective behaviors during crises and their impact on disease dynamics.

Habitat severity characteristics structure soil communities at regional and local spatial scales along the Antarctica Peninsula

AbstractAntarctic soils provide an excellent setting to test biogeographical patterns across spatial and environmental scales given their relatively simple communities and the dominance of physical factors that create strong environmental gradients. Additional urgency is given by the fact that their unique terrestrial communities are the subject of conservation efforts in a rapidly changing environment. We investigated relationships of soil community assembly and alpha and beta diversity with climatic and environmental parameters across regional and local scales in Maritime Antarctica. We sampled from a regional gradient of sites that differ in habitat severity, ranging from relatively favourable to harsher physicochemical conditions. At the regional scale, bacterial community characteristics and microarthropod abundance varied along this severity gradient, but most measures of fungal communities did not. Microarthropod and microbial communities differed in which soil and climate parameters were most influential, and the specific parameters that influenced each taxon differed across broad and fine spatial scales. This suggests that conservation efforts will need to focus on a large variety of habitat characteristics to successfully encompass diversity across taxa. Because beta diversity was the result of species turnover, conservation efforts also cannot focus on only the most biodiverse sites to effectively preserve all aspects of biodiversity.

Ground‐based infrared thermometry reveals seasonal evapotranspiration patterns in semiarid rangelands

AbstractDetailed assessment of small‐scale heterogeneity in local surface water balance is essential to accurate estimation of evapotranspiration in semiarid climates. However, meteorological approaches are often impractical to implement in sites with sparse and diverse vegetation composition, especially with seasonally variable leaf canopy features. Ground‐based infrared thermometry (TIR) provides spatially and temporally continuous resolution of surface skin temperature that can be directly related to the land surface energy balance. We made repeated measurements with a portable TIR camera to capture seasonal replicates for patch scale heat images for four sagebrush communities. The heat images near peak foliage and near the end of the growing season were compared by computation of surface energy fluxes from TIR sensing to surface energy balance (SEB) and Bowen ratio (BR). Estimates of sensible (H) and latent heat flux (LE) were evaluated with eddy covariance measurements to disaggregate the expression of seasonal phenology of sagebrush species across wetness and elevation. Estimations showed reasonable agreement with ground‐based LE observations for most cases (r2 = 0.59–0.76 for SEB and 0.22–0.72 for BR; root mean squared error = 73.4–106.4 W m−2 for SEB and 109.9–204.0 W m−2 for BR). Predictability declined as the fraction of senescent foliage increased in dry conditions. The field trials suggest the methods have the potential for monitoring land surface energy fluxes and plant health at a very fine spatial scale. The ability to partition heat fluxes from various plant communities over a range of moisture availability will provide valuable information associated with the consumptive water use and phenological processes in the semiarid West.

On the Morphodynamics of a Wide Class of Large‐Scale Meandering Rivers: Insights Gained by Coupling LES With Sediment‐Dynamics

AbstractIn meandering rivers, interactions between flow, sediment transport, and bed topography affect diverse processes, including bedform development and channel migration. Predicting how these interactions affect the spatial patterns and magnitudes of bed deformation in meandering rivers is essential for various river engineering and geoscience problems. Computational fluid dynamics simulations can predict river morphodynamics at fine temporal and spatial scales but have traditionally been challenged by the large scale of natural rivers. We conducted coupled large‐eddy simulation and bed morphodynamics simulations to create a unique database of hydro‐morphodynamic data sets for 42 meandering rivers with a variety of planform shapes and large‐scale geometrical features that mimic natural meanders. For each simulated river, the database includes (a) bed morphology, (b) three‐dimensional mean velocity field, and (c) bed shear stress distribution under bankfull flow conditions. The calculated morphodynamics results at dynamic equilibrium revealed the formation of scour and deposition patterns near the outer and inner banks, respectively, while the location of point bars and scour regions around the apexes of the meander bends is found to vary as a function of the radius of curvature of the bends to the width ratio. A new mechanism is proposed that explains this seemingly paradoxical finding. The high‐fidelity simulation results generated in this work provide researchers and scientists with a rich numerical database for morphodynamics and bed shear stress distributions in large‐scale meandering rivers to enable systematic investigation of the underlying phenomena and support a range of river engineering applications.

Evaluation of the near-surface wind field over the Adriatic region: local wind characteristics in the convection-permitting model ensemble

We present the first evaluation of the wind field from the ensemble of kilometer-scale simulations from the CORDEX-Flagship Pilot Study on convection, with focus on the Adriatic region. Kilometer-scale climate models, also known as convection-permitting models (CPMs), produce a good representation of small-scale topographic features and consequently a more detailed depiction of dynamical and thermal circulations. These enable a reliable view of climate characteristics of the wind field, especially in coastal regions and over complex terrain, such as the Adriatic region. We investigate the (potential) added value introduced by CPMs compared to classical “cumulus-parametrized” regional climate models (RCMs), reanalysis and station observations. For this purpose, wind components at 10 m level are used at 3-hourly frequency. All simulations cover a 10-year period, extending from 2000 to 2009. In terms of the standard statistical parameters such as correlation coefficient and temporal standard deviation, CPMs are very dependent on their parent RCM performance. However, the orographic forcing emphasizes the potential added value and CPMs contain some fine spatial scale variability (i.e., stronger extremes by 25% and more accurate wind direction) that is absent in coarser RCMs and reanalysis. The potential added value is higher in the cold season compared to the warm season due to the proportion of severe wind events. CPMs reproduce well the typical wind regimes along the Adriatic coast, namely Bora and Sirocco. The benefit of using CPMs is especially pronounced in simulating Bora maximum wind speeds in northern Adriatic and Sirocco frequencies in southern Adriatic. Based on our overall analysis, we conclude that CPMs provide added value compared to coarser models, especially in the complex coastal terrain.

Mapping Fire Severity in Southwest China Using the Combination of Sentinel 2 and GF Series Satellite Images.

Fire severity mapping can capture heterogeneous fire severity patterns over large spatial extents. Although numerous remote sensing approaches have been established, regional-scale fire severity mapping at fine spatial scales (<5 m) from high-resolution satellite images is challenging. The fire severity of a vast forest fire that occurred in Southwest China was mapped at 2 m spatial resolution by random forest models using Sentinel 2 and GF series remote sensing images. This study demonstrated that using the combination of Sentinel 2 and GF series satellite images showed some improvement (from 85% to 91%) in global classification accuracy compared to using only Sentinel 2 images. The classification accuracy of unburnt, moderate, and high severity classes was significantly higher (>85%) than the accuracy of low severity classes in both cases. Adding high-resolution GF series images to the training dataset reduced the probability of low severity being under-predicted and improved the accuracy of the low severity class from 54.55% to 72.73%. RdNBR was the most important feature, and the red edge bands of Sentinel 2 images had relatively high importance. Additional studies are needed to explore the sensitivity of different spatial scales satellite images for mapping fire severity at fine spatial scales across various ecosystems.

Health effects of exposure to particulate matter and physical activity by individual sustainable commuting modes

Promotion of sustainable commuting modes (e.g., walking, cycling and electric mobility) is considered as a promising solution to achieve sustainability. Active travel may also increase physical activity, which is beneficial for health. But it increases health burdens by directly exposing commuters to air pollution. There is a lack of generalizable quantitative approaches to comprehensively evaluate the health outcomes of sustainable plans during the design period. In this research, we aim to examine the long-term combined health impacts of commuting by walking, cycling, and e-moped in Cangzhou, China, a populous industrial city with high air pollution levels, to inform decision-making prior to the implementation of sustainable strategies. A taxi-based mobile remote sensing system was implemented to create a link-level fine particulate matter (PM2.5) concentration map to overcome the insufficiency of sparse fixing monitoring network, which allows researchers to determine joint relative risks of all-cause mortality at a finer spatial scale. Results show that the benefits always outweighed the risks for active commuting at the case study site, even in high polluted days. E-moped commuters experienced a negative burden. In our case, encouraging these commuters to travel during off-peak periods to reduce exposure might not be effective and could even lead to greater PM2.5 exposure than during peak periods (up to 0.37 μg/km). The proposed method can be applied to different locations, providing decision makers with the information needed to target interventions to achieve sustainability.

Climatic, land-use and socio-economic factors can predict malaria dynamics at fine spatial scales relevant to local health actors: Evidence from rural Madagascar.

While much progress has been achieved over the last decades, malaria surveillance and control remain a challenge in countries with limited health care access and resources. High-resolution predictions of malaria incidence using routine surveillance data could represent a powerful tool to health practitioners by targeting malaria control activities where and when they are most needed. Here, we investigate the predictors of spatio-temporal malaria dynamics in rural Madagascar, estimated from facility-based passive surveillance data. Specifically, this study integrates climate, land-use, and representative household survey data to explain and predict malaria dynamics at a high spatial resolution (i.e., by Fokontany, a cluster of villages) relevant to health care practitioners. Combining generalized linear mixed models (GLMM) and path analyses, we found that socio-economic, land use and climatic variables are all important predictors of monthly malaria incidence at fine spatial scales, via both direct and indirect effects. In addition, out-of-sample predictions from our model were able to identify 58% of the Fokontany in the top quintile for malaria incidence and account for 77% of the variation in the Fokontany incidence rank. These results suggest that it is possible to build a predictive framework using environmental and social predictors that can be complementary to standard surveillance systems and help inform control strategies by field actors at local scales.

Statistical Downscaling of Seasonal Forecasts of Sea Level Anomalies for U.S. Coasts

AbstractIncreasing coastal inundation risk in a warming climate will require accurate and reliable seasonal forecasts of sea level anomalies at fine spatial scales. In this study, we explore statistical downscaling of monthly hindcasts from six current seasonal prediction systems to provide a high‐resolution prediction of sea level anomalies along the North American coast, including at several tide gauge stations. This involves applying a seasonally invariant downscaling operator, constructing by linearly regressing high‐resolution (1/12°) ocean reanalysis data against its coarse‐grained (1°) counterpart, to each hindcast ensemble member for the period 1982–2011. The resulting high‐resolution coastal hindcasts have significantly more deterministic skill than the original hindcasts interpolated onto the high‐resolution grid. Most of this improvement occurs during summer and fall, without impacting the seasonality of skill noted in previous studies. Analysis of the downscaling operator reveals that it boosts skill by amplifying the most predictable patterns while damping the less predictable patterns.

Quantifying rarity of intraspecific diversity at multiple spatial scales by combining fine-grain citizen-based data across national boundaries

Quantifying rarity at the intraspecific level is an important task for conservation biologists because rare phenotypes have a higher extinction risk than common ones. However, quantitative assessments of rarity are lacking at this level, preventing to adequately protect rare and endangered variants that are part of the pool of common species. Our aim was to take benefit of high-resolution volunteer-based occurrence data to assess quantitatively rarity patterns of facultative paedomorphosis, in the Alpine newt (Ichthyosaura alpestris). This polyphenism is composed of two morphs, a rare adult phenotype retaining gills and a common phenotype metamorphosing. We included multiple spatial scales (i.e., resolution) and combined 15,613 occurrence records from the atlas databases of the seven countries in the European Alps to understand how spatial scale may affect rarity metrics. Our analyses revealed that the rarer phenotype was 292 times rarer than the common phenotype, occupying a very small and fragmented area of occupancy. Yet, rarity depended on the spatial resolution of the data. Rarity estimates were up to 1300 times lower at the coarser than at the finer spatial scale. Both the rarity of the rare phenotype and the presence of threats (i.e., fish introductions) make paedomorphs critically endangered whereas the common phenotype was widespread at all spatial scales. Altogether, these results show how rarity metrics can differ between spatial scales for rare and common organisms, suggesting that fine-grain data should be used to assess intraspecific rarity. They also show that combining datasets from distribution atlases is efficient to estimate rarity.

High-frequency time series comparison of Sentinel-1 and Sentinel-2 satellites for mapping open and vegetated water across the United States (2017–2021)

Frequent observations of surface water at fine spatial scales will provide critical data to support the management of aquatic habitat, flood risk and water quality. Sentinel-1 and Sentinel-2 satellites can provide such observations, but algorithms are still needed that perform well across diverse climate and vegetation conditions. We developed surface inundation algorithms for Sentinel-1 and Sentinel-2, respectively, at 12 sites across the conterminous United States (CONUS), covering a total of >536,000 km2 and representing diverse hydrologic and vegetation landscapes. Each scene in the 5-year (2017–2021) time series was classified into open water, vegetated water, and non-water at 20 m resolution using variables from Sentinel-1 and Sentinel-2, as well as variables derived from topographic and weather datasets. The Sentinel-1 algorithm was developed distinct from the Sentinel-2 model to explore if and where the two time series could potentially be integrated into a single high-frequency time series. Within each model, open water and vegetated water (vegetated palustrine, lacustrine, and riverine wetlands) classes were mapped. The models were validated using imagery from WorldView and PlanetScope. Classification accuracy for open water was high across the 5-year period, with an omission and commission error of only 3.1% and 0.9% for the Sentinel-1 algorithm and 3.1% and 0.5% for the Sentinel-2 algorithm, respectively. Vegetated water accuracy was lower, as expected given that the class represents mixed pixels. The Sentinel-2 algorithm showed higher accuracy (10.7% omission and 7.9% commission error) relative to the Sentinel-1 algorithm (28.4% omission and 16.0% commission error). Patterns over time in the proportion of area mapped as open or vegetated water by the Sentinel-1 and Sentinel-2 algorithms were charted and correlated for a subset of all 12 sites. Our results showed that the Sentinel-1 and Sentinel-2 algorithm open water time series can be integrated at all 12 sites to improve the temporal resolution, but sensor-specific differences, such as sensitivity to vegetation structure versus pixel color, complicate the data integration for mixed-pixel, vegetated water. The methods developed here provide inundation at 5-day (Sentinel-2 algorithm) and 12-day (Sentinel-1 algorithm) time steps to improve our understanding of the short- and long-term response of surface water to climate and land use drivers in different ecoregions.

Unsupervised learning reveals interpretable latent representations for translucency perception.

Humans constantly assess the appearance of materials to plan actions, such as stepping on icy roads without slipping. Visual inference of materials is important but challenging because a given material can appear dramatically different in various scenes. This problem especially stands out for translucent materials, whose appearance strongly depends on lighting, geometry, and viewpoint. Despite this, humans can still distinguish between different materials, and it remains unsolved how to systematically discover visual features pertinent to material inference from natural images. Here, we develop an unsupervised style-based image generation model to identify perceptually relevant dimensions for translucent material appearances from photographs. We find our model, with its layer-wise latent representation, can synthesize images of diverse and realistic materials. Importantly, without supervision, human-understandable scene attributes, including the object's shape, material, and body color, spontaneously emerge in the model's layer-wise latent space in a scale-specific manner. By embedding an image into the learned latent space, we can manipulate specific layers' latent code to modify the appearance of the object in the image. Specifically, we find that manipulation on the early-layers (coarse spatial scale) transforms the object's shape, while manipulation on the later-layers (fine spatial scale) modifies its body color. The middle-layers of the latent space selectively encode translucency features and manipulation of such layers coherently modifies the translucency appearance, without changing the object's shape or body color. Moreover, we find the middle-layers of the latent space can successfully predict human translucency ratings, suggesting that translucent impressions are established in mid-to-low spatial scale features. This layer-wise latent representation allows us to systematically discover perceptually relevant image features for human translucency perception. Together, our findings reveal that learning the scale-specific statistical structure of natural images might be crucial for humans to efficiently represent material properties across contexts.