Subregional Model Research Articles

Coupled-Region Visual Tracking Formulation Based on a Discriminative Correlation Filter Bank

The visual tracking algorithm based on discriminative correlation filter (DCF) has shown excellent performance in recent years, especially as the higher tracking speed meets the real-time requirement of object tracking. However, when the target is partially occluded, the traditional single discriminative correlation filter will not be able to effectively learn information reliability, resulting in tracker drift and even failure. To address this issue, this paper proposes a novel tracking-by-detection framework, which uses multiple discriminative correlation filters called discriminative correlation filter bank (DCFB), corresponding to different target sub-regions and global region patches to combine and optimize the final correlation output in the frequency domain. In tracking, the sub-region patches are zero-padded to the same size as the global target region, which can effectively avoid noise aliasing during correlation operation, thereby improving the robustness of the discriminative correlation filter. Considering that the sub-region target motion model is constrained by the global target region, adding the global region appearance model to our framework will completely preserve the intrinsic structure of the target, thus effectively utilizing the discriminative information of the visible sub-region to mitigate tracker drift when partial occlusion occurs. In addition, an adaptive scale estimation scheme is incorporated into our algorithm to make the tracker more robust against potential challenging attributes. The experimental results from the OTB-2015 and VOT-2015 datasets demonstrate that our method performs favorably compared with several state-of-the-art trackers.

Predicting increasing high severity area burned for three forested regions in the western United States using extreme value theory

More than 70 years of fire suppression by federal land management agencies has interrupted fire regimes in much of the western United States. The result of missed fire cycles is a buildup of both surface and canopy fuels in many forest ecosystems, increasing the risk of severe fire. The frequency and size of fires has increased in recent decades, as has the area burned with high severity in some ecosystems. A number of studies have examined controls on high severity fire occurrence, but none have yet determined what controls the extent of high severity fire. We developed statistical models predicting high severity area burned for the western United States and three sub-regions—the Northern Rocky Mountains, Sierra Nevada Mountains, and Southwest. A simple model with maximum temperature the month of fire, annual normalized moisture deficit and location explains area burned in high severity fire in our west-wide model, with the exception of years with especially large areas burned with high severity fire: 1988, 2002. With respect to mitigation or management of high severity fire, understanding what drives extreme fire years is critical. For the sub-regional models, topography, spring temperature and snowpack condition, and vegetation condition class variables improved our prediction of high severity burned area in extreme fire years. Fire year climate is critical to predicting area burned in high severity fire, especially in extreme fire years. These models can be used for scenario analyses and impact assessments to aid management in mitigating negative impacts of high severity fire.

Designing regional-level stakeholder engagement processes: striving for good governance while meeting the challenges of scale

ABSTRACTStakeholder engagement processes have sought to ensure that state government meets public trust and good governance obligations to citizens. As the expectations of stakeholders and state agencies change, and management focuses on landscape-level interventions, a change in the level at which agencies engage the public is needed. This involves tradeoffs, as different levels call for different engagement design and implementation considerations. To understand how these differences affect decision making, we examine a regional engagement model for deer management in New York that was piloted to replace a sub-regional model. We identify concerns with the old model, objectives for the redesigned model, and explain the logistical and good governance considerations that informed its design. We share our evaluation of the model's process and outcomes, including implications for program design and scale. Overall, despite the pilot model's attention to design components aimed at addressing potential barriers to regional engagement as well as limitations of the previous engagement model, the pilot did not meet many of its objectives, especially those related to representation, resulting in some of the same concerns associated with the model it was intended to enhance and replace. Implications of this for regional-level engagement efforts are discussed.

Scenarios of land-use change in a deforestation corridor in the Brazilian Amazon: combining two scales of analysis

Local, regional, and global processes affect deforestation and land-use changes in the Brazilian Amazon. Characteristics are: direct conversions from forest to pasture; regional processes of indirect land-use change, described by the conversion of pastures to cropland, which increases the demand for pastures elsewhere; and teleconnections, fueled by the global demands for soybeans as animal fodder. We modeled land-use changes for two scenarios Trend and Sustainable Development for a hot spot of land-use change along the BR-163 highway in Mato Grosso and Pará, Brazil. We investigated the differences between a coupled modeling approach, which incorporates indirect land-use change processes, and a noncoupled land-use model. We coupled the regional-scale LandSHIFT model, defined for Mato Grosso and Pará, with a subregional model, alucR, covering a selected corridor along the BR-163. The results indicated distinct land-use scenario outcomes from the coupled modeling approach and the subregional model quantification. We found the highest deforestation estimates returned from the subregional quantification of the Trend scenario. This originated from the strong local dynamics of past deforestation and land-use changes. Land-use changes exceeded the demands estimated at regional scale. We observed the lowest deforestation estimates at the subregional quantification of the Sustainable Development story line. We highlight that model coupling increased the representation of scenario outcomes at fine resolution while providing consistency across scales. However, distinct local dynamics were explicitly captured at subregional scale. The scenario result pinpoints the importance of policies to aim at the cattle ranching sector, to increase land tenure registration and enforcement of environmental laws.

Improvements for the Western North Atlantic, Caribbean and Gulf of Mexico ADCIRC Tidal Database (EC2015)

This research details the development and validation of an updated constituent tidal database for the Western North Atlantic, Caribbean and Gulf of Mexico (WNAT) region, referred to as the EC2015 database. Regional databases, such as EC2015, provide much higher resolution than global databases allowing users to more accurately define the tidal forcing on smaller sub-region domains. The database last underwent major updates in 2001 and was developed using the two-dimensional, depth-integrated form of the coastal hydrodynamic model, ADvanced CIRCulation (ADCIRC), which solves the shallow-water equations in the generalized wave continuity equation form. Six main areas of improvement are examined: (1) placement of the open ocean boundary; (2) higher coastal resolution using Vertical Datum (VDatum) models; (3) updated bathymetry from global databases; (4) updated boundary forcing compared using two global tidal databases; (5) updated bottom friction formulations; and (6) improved model physics by incorporating the advective terms in ADCIRC. The skill of the improved database is compared to that of its predecessor and is calculated using harmonic data from the National Oceanic and Atmospheric Administration Center for Operational Oceanographic Products and Services (NOAA CO-OPS) stations and historic International Hydrographic Organization (IHO) data. Overall, the EC2015 database significantly reduces errors realized in the EC2001 database and improves the quality of coastal tidal constituents available for smaller sub-regional models in the Western North Atlantic, Caribbean and Gulf of Mexico (WNAT) region.

Forecast of drifter trajectories using a Rapid Environmental Assessment based on CTD observations

A high resolution submesoscale resolving ocean model was implemented in a limited area north of Island of Elba where a maritime exercise, named Serious Game 1 (SG1), took place on May 2014 in the framework of the project MEDESS-4MS (Mediterranean Decision Support System for Marine Safety). During the exercise, CTD data have been collected responding to the necessity of a Rapid Environmental Assessment, i.e. to a rapid evaluation of the marine conditions able to provide sensible information for initialisation of modelling tools, in the scenario of possible maritime accidents. The aim of this paper is to evaluate the impact of such mesoscale-resolving CTD observations on short-term forecasts of the surface currents, within the framework of possible oil-spill related emergencies. For this reason, modelling outputs were compared with Lagrangian observations at sea: the high resolution modelled currents, together with the ones of the coarser sub-regional model WMED, are used to force the MEDSLIK-II oil-spill model to simulate drifter trajectories. Both ocean models have been assessed by comparing the prognostic scalar and vector fields as an independent CTD data set and with real drifter trajectories acquired during SG1. The diagnosed and prognosed circulation reveals that the area was characterised by water masses of Atlantic origin influenced by small mesoscale cyclonic and anti-cyclonic eddies, which govern the spatial and temporal evolution of the drifter trajectories and of the water masses distribution. The assimilation of CTD data into the initial conditions of the high resolution model highly improves the accuracy of the short-term forecast in terms of location and structure of the thermocline and positively influence the ability of the model in reproducing the observed paths of the surface drifters.

Improving Short Utterance Speaker Recognition by Modeling Speech Unit Classes

Short utterance speaker recognition (SUSR) is highly challenging due to the limited enrollment and/or test data. We argue that the difficulty can be largely attributed to the mismatched prior distributions of the speech data used to train the universal background model (UBM) and those for enrollment and test. This paper presents a novel solution that distributes speech signals into a multitude of acoustic subregions that are defined by speech units, and models speakers within the subregions. To avoid data sparsity, a data-driven approach is proposed to cluster speech units into speech unit classes, based on which robust subregion models can be constructed. Further more, we propose a model synthesis approach based on maximum likelihood linear regression (MLLR) to deal with no-data speech unit classes. The experiments were conducted on a publicly available database SUD12. The results demonstrated that on a text-independent speaker recognition task where the test utterances are no longer than 2 seconds and mostly shorter than 0.5 seconds, the proposed sub-region modeling offered a 21.51% relative reduction in equal error rate (EER), compared with the standard GMM-UBM baseline. In addition, with the model synthesis approach, the performance can be greatly improved in scenarios where no enrollment data are available for some speech unit classes.

Translational neuroscience of basolateral amygdala lesions: Studies of Urbach-Wiethe disease.

Urbach-Wiethe disease (UWD) is an extremely rare autosomal recessive disorder characterized by mutations in the extracellular matrix protein 1 gene on chromosome 1. Typical clinical manifestations include voice hoarseness in early infancy and neuropsychiatric, laryngeal, and dermatological pathologies later in life. Neuroimaging studies have revealed a pattern of brain calcification often but not exclusively leading to selective bilateral amygdala damage. A large body of work on amygdala lesions in rodents exists, generally employing a subregion model focused on the basolateral amygdala (BLA) and the central-medial amygdala. However, human work usually considers the amygdala as a unified structure, not only complicating the translation of animal findings to humans but also providing a unique opportunity for further research. To compare data from rodent models with human cases and to complement existing data from Europe and North America, a series of investigations was undertaken on UWD subjects with selective BLA damage in the Namaqualand region, South Africa. This review presents key findings from this work, including fear processing, social-economic behavior, and emotional conflict processing. Our findings are broadly consistent with and support rodent models of selective BLA lesions and show that the BLA is integral to processing sensory stimuli and exhibits inhibitory regulation of responses to unconditioned innate fear stimuli. Furthermore, our findings suggest that the human BLA mediates calculative-instrumental economic behaviors and may compromise working memory via competition for attentional resources between the BLA salience detection system and the dorsolateral prefrontal cortex working memory system.

Housebuilding, demographic change and affordability as outcomes of local planning decisions: Exploring interactions using a sub-regional model of housing markets in England

The recent history of planning policy in England leaves the treatment of housing supply in a confused position. Much depends on the analysis of evidence on the current and future need and demand for housing for groups of inter-related localities. Arguably the most appropriate spatial scale of analysis of market adjustment processes is the functional urban sub-region, and this article discusses the development and application of an economic market model at this level, applied to c. 100 functional housing market areas across England. This goes beyond most previous analyses of UK housing markets, which have lacked explicit treatment of land-use planning and the supply process, while also recognising the economic and supply influences on demographic change. Based mainly on short-medium panel datasets and some micro surveys, component models for migration, household formation, prices, rents, and new construction are estimated and combined with simpler labour market and demographic accounting to build a simulation model which can explore the potential impacts of economic, demographic and planning policy scenarios. Through a range of examples applied to groupings of sub-regions, and two local case studies, this paper focuses on the way housebuilding, demographics and markets adjust to local decisions and the implications of this ‘outcome-oriented’ approach for planning. The results confirm a wider literature in showing that planning restriction raises housing costs, but underlines that the effects are quite gradual, particularly given the rather partial response of new housebuilding itself to planning decisions. They also show the interdependence between nearby sub-regions, and the likely adverse effects of leaving decisions to the local level, as well as the somewhat mixed impact of different levels of economic growth.

Linking land use/land cover with climatic and geomorphologic factors in regional mean annual streamflow models with geospatial regression approach

Estimates of annual streamflow in connection with key natural and anthropogenic factors are necessary and important for different purposes, such as water resource planning and management, sediment and nutrient loading in streams and rivers, hydropower, and navigation. This study is an attempt to use the spatial statistical regression approach to develop regression models for mean annual streamflow at regional scale while adequately dealing with the common spatial dependency issue in input and output variables used in regression models. The proposed modeling approach is illustrated with a case study of the Upper Mississippi River Basin. The R-squared and the Nash–Sutcliffe model efficiency coefficient of the regional model were 0.993 and 0.985, respectively, while those of the sub-regional model were 0.995 and 0.990, respectively. Methodologically, the proposed model provided an effective way to utilize an extensive spatial dataset of various climatic, geomorphologic, and land cover variables for a large region like the Upper Mississippi River Basin to assess and compare the impact of various factors on mean annual streamflow at regional scale. Furthermore, the model was able to handle spatial dependency in data.

Bioenergy Supply and Environmental Impacts on Cropland: Insights from Multi‐market Forecasts in a Great Lakes Subregional Bioeconomic Model

AbstractUsing subregional models of crop production choices in central Wisconsin and southwest Michigan, we predict biomass production, land use, and environmental impacts with details that are unavailable from national scale models. When biomass prices are raised exogenously, we find that the subregional models overestimate the supply, the land use, and the beneficial environmental aspects of perennial biomass crops. Multi‐market price feedbacks tied to realistic policy parameters predict high threshold absolute prices for biomass to enter production, resulting in intensified production of biomass from annual grain crops with damaging environmental impacts. Multi‐market feedbacks also predict regional specialization in energy biomass production in areas with lower yields of food crops. Policies promoting biofuels will not necessarily generate environmental benefits in the absence of environmental regulations.

Effects of environmental and anthropogenic drivers on Amur tiger distribution in northeastern China

AbstractWe examined environmental and anthropogenic factors drive range loss in large mammals, using presence data of Amur tigers opportunistically collected between 2000 and 2012, and anthropogenic and environmental variables to model the distribution of the Amur tiger in northeastern China. Our results suggested that population distribution models of different subregions showed different habitat factors determining tiger population distribution patterns. Where farmland cover was over 50 km2 per pixel (196 km2), distance was within 15 km to the railway in Changbaishan and road density (length per pixel) increased in Wandashan, the relative probability of Amur tiger occurrence exhibited monotonic avoidance responses; however, where distance was within 150 km of the Sino‐Russia border, the occurrence probability of Amur tiger was relatively high. We analyzed the avoidance or preference responses of Amur tiger distribution to elevation, snow depth and Viewshed. Furthermore, different subregional models detected a variety of spatial autocorrelation distances due to different population clustering patterns. We found that spatial models significantly improved model fits for non‐spatial models and made more robust habitat suitability predications than that of non‐spatial models. Consequently, these findings provide useful guidance for habitat conservation and management.

A relocatable ocean model in support of environmental emergencies

During the Costa Concordia emergency case, regional, subregional, and relocatable ocean models have been used together with the oil spill model, MEDSLIK-II, to provide ocean currents forecasts, possible oil spill scenarios, and drifters trajectories simulations. The models results together with the evaluation of their performances are presented in this paper. In particular, we focused this work on the implementation of the Interactive Relocatable Nested Ocean Model (IRENOM), based on the Harvard Ocean Prediction System (HOPS), for the Costa Concordia emergency and on its validation using drifters released in the area of the accident. It is shown that thanks to the capability of improving easily and quickly its configuration, the IRENOM results are of greater accuracy than the results achieved using regional or subregional model products. The model topography, and to the initialization procedures, and the horizontal resolution are the key model settings to be configured. Furthermore, the IRENOM currents and the MEDSLIK-II simulated trajectories showed to be sensitive to the spatial resolution of the meteorological fields used, providing higher prediction skills with higher resolution wind forcing.

Bypass flow computations on the LOFA transient in a VHTR

Bypass flow in the prismatic gas-cooled very high temperature reactor (VHTR) is not intentionally designed to occur, but is present in the gaps between graphite blocks. Previous studies of the bypass flow in the core indicated that the cooling provided by flow in the bypass gaps had a significant effect on temperature and flow distributions for normal operating conditions. However, the flow and heat transports in the core are changed significantly after a Loss of Flow Accident (LOFA). This study aims to study the effect and role of the bypass flow after a LOFA in terms of the temperature and flow distributions and for the heat transport out of the core by natural convection of the coolant for a 1/12 symmetric section of the active core which is composed of images and mirror images of two sub-region models. The two sub-region models, 9 × 1/12 and 15 × 1/12 symmetric sectors of the active core, are employed as the CFD flow models using computational grid systems of 70.2 million and 117 million nodes, respectively. It is concluded that the effect of bypass flow is significant for the initial conditions and the beginning of LOFA, but the bypass flow has little effect after a long period of time in the transient computation of natural circulation.

Assessing impact of land uses on land salinization in the Yellow River Delta, China using an integrated and spatial statistical model

Intensification of agriculture and industry in salinized areas poses a risk of secondary salinization. Thus, comprehensive and spatially explicit assessments are needed to assist government in developing ecologically sound policies. Few assessments have comprehensively quantified the impacts of multiple anthropogenic activities on salinization as environmental interferences and salinity autocorrelation are largely neglected. This study tried to perform such an assessment by identifying the nature of human impacts on salinization from three aspects in the Yellow River Delta (YRD) of China. A versatile GIS-based spatial autoregression (SAR) was applied to nine selected explainable variables in six sub-region models. Sub-region model was verified as an effective tool of normalizing environmental interferences because more useful spatial information was provided compared to the whole region model. GIS-SAR model fit better and performed better in quantifying human activities, compared to the conventional ordinary least square regression (OLSR) model, as SAR can deal with spatial autocorrelation in soil salinity. Among the well-defined key determinants, oil exploitation and saline aquaculture were aggregative to salinization but only in originally highly saline sub-regions, such as coastal zone and Gleyic Solonchaks (coastal saline moisture soil) area. Two agricultural activities, crop plantation and fertilization, were mainly ameliorators in most sub-regions. The most effective salinization alleviation occurred in moderately saline sub-regions, such as floodplain and Salic Fluvisols (saline moisture soil) area, which benefitted from the development of agroforests and farm ponds. The SAR sub-region model is spatially explicit for spotting the hazardous areas and some suggestions were also provided for the policy makers.

Modelling yearly flooding extent of the Waza-Logone floodplain in northern Cameroon based on MODIS and rainfall data

The Sahelian floodplains are of high ecological and economical importance, providing water and fresh pasture in the dry season. A spatial model is presented to predict the yearly flooding extent of the Waza-Logone floodplain based on cumulative runoff in the catchment area and estimations of the soil moisture prior to the flooding. Observations of flooding extent were based on thresholding 16-day composite Moderate Resolution Imaging Spectroradiometer (MODIS) shortwave infrared (SWIR) images. The Soil Conservation Service Curve Number (SCS-CN) method was used to calculate cumulative runoff within the Logone catchment area based on rainfall estimates (RFEs) for Africa. MODIS SWIR images acquired prior to the flooding were used as indicators for soil moisture. The mean observed flooding extent of the Waza-Logone floodplain during the period 2000–2005 was 6747 km2 with a standard deviation of 1838 km2. Multiple regression analysis was performed to create a predictive model forecasting flooding extent 1.5 months in advance with a coefficient of determination (R 2) equal to 0.957. Multiple regression modelling was also performed for three subregions separately. The 1.5-month forecast model for the Waza subregion resulted in the highest accuracy (R 2 = 0.950). A floodwater distribution map was created for this subregion model, allowing determination where the flooding occurs for an estimated flood size. The average additional error caused by the mapping procedure was 138 km2, which is relatively small compared to an average flooded area of 3211 km2 for the Waza subregion. As the flooding extent in the Waza-Logone floodplain is highly correlated to the amount of natural resources available in the dry season, the model may be a valuable tool for sustainable management of these resources.

Simulated village locations in Thailand: a multi-scale model including a neural network approach

The simulation of rural land use systems, in general, and rural settlement dynamics in particular has developed with synergies of theory and methods for decades. Three current issues are: linking spatial patterns and processes, representing hierarchical relations across scales, and considering nonlinearity to address complex non-stationary settlement dynamics. We present a hierarchical simulation model to investigate complex rural settlement dynamics in Nang Rong, Thailand. This simulation uses sub-models to allocate new villages at three spatial scales. Regional and sub-regional models, which involve a nonlinear space-time autoregressive model implemented in a neural network approach, determine the number of new villages to be established. A dynamic village niche model, establishing pattern-process link, was designed to enable the allocation of villages into specific locations. Spatiotemporal variability in model performance indicates the pattern of village location changes as a settlement frontier advances from rice-growing lowlands to higher elevations. Experiments results demonstrate this simulation model can enhance our understanding of settlement development in Nang Rong and thus gain insight into complex land use systems in this area.

Multilevel Modeling of the Forest Resource Dynamics

We examine the theoretical and applications-specific issues relating to modeling the temporal and spatial dynamics of forest ecosystems, based on the principles of investigating dy- namical models. When developing the predictive dynamical models of forest resources, there is a possibility of achieving uniqueness of the solutions to equations by taking into account the initial and boundary conditions of the solution, and the conditions of the geographical environment. We present the results of a computer modeling and predictive mapping for the regional and subregional model of the forest resource dynamics.

Model‐Based Clustering in a Brook Trout Classification Study within the Eastern United States

AbstractData collected on the population status (extirpated or present) of brook trout Salvelinus fontinalis at the landscape level across the eastern United States is useful for identifying important stressors and predicting brook trout status at the watershed level. However, when dealing with data compiled over a large region, a single model may not adequately describe relationships between variables. To find models with better classification performance, we used a Monte Carlo model‐based clustering method with logistic regression models to obtain subregions with good predictive performance. To subdivide the eastern United States into subregions, we used Voronoi tessellations with randomly selected centers. The average fraction correctly classified for fit was the criterion used when searching for optimal models within clusters. Logistic regression models were chosen by stepwise selection based on five explanatory variables: Elevation, percentage of forested land, percentage of agricultural land, road density, and an environmental factor (depositional NO3 and SO4). Application of the method to the brook trout status data set resulted in six subregions and improved predictive ability by approximately 20% relative to the single regionwide model. Resulting models were also more interpretable than the single model and reflected effects at a smaller spatial scale. In contrast to results of the single model, the role of elevation in the six subregional models was consistent with expectations and indicated an increased probability of brook trout presence with increased elevation. The resulting models should be useful for identification and prioritizing sites for restoration and recovery programs.

Upscaling Discrete Fracture Characterizations to Dual-Porosity, Dual-Permeability Models for Efficient Simulation of Flow With Strong Gravitational Effects

Summary The geological complexity of fractured reservoirs requires the use of simplified models for flow simulation. This is often addressed in practice by using flow modeling procedures based on the dual-porosity, dual-permeability concept. However, in most existing approaches, there is not a systematic and quantitative link between the underlying geological model [in this case, a discrete fracture model (DFM)] and the parameters appearing in the flow model. In this work, a systematic upscaling procedure is presented to construct a dual-porosity, dual-permeability model from detailed discrete fracture characterizations. The technique, referred to as a multiple subregion (MSR) model, represents an extension of an earlier method that did not account for gravitational effects. The subregions (or subgrid) are constructed for each coarse block using the iso-pressure curves obtained from local pressure solutions of a discrete fracture model over the block. The subregions thus account for the fracture distribution and can represent accurately the matrix-matrix and matrix-fracture transfer. The matrix subregions are connected to matrices in vertically adjacent blocks (as in a dual-permeability model) to capture phase segregation caused by gravity. Two-block problems are solved to provide fracture-fracture flow effects. All connections in the coarse-scale model are characterized in terms of upscaled transmissibilities, and the resulting coarse model can be used with any connectivity-based reservoir simulator. The method is applied to simulate 2D and 3D fracture models, with viscous, gravitational, and capillary pressure effects, and is shown to provide results in close agreement with the underlying DFM. Speedups of approximately a factor of 120 are observed for a complex 3D example.