Geospatial Statistics Research Articles

Application of multivariate binary logistic regression grouped outlier statistics and geospatial logistic model to identify villages having unusual health-seeking habits for childhood malaria in Malawi

BackgroundEarly diagnosis and prompt treatment of malaria in young children are crucial for preventing the serious stages of the disease. If delayed treatment-seeking habits are observed in certain areas, targeted campaigns and interventions can be implemented to improve the situation.MethodsThis study applied multivariate binary logistic regression model diagnostics and geospatial logistic model to identify traditional authorities in Malawi where caregivers have unusual health-seeking behaviour for childhood malaria. The data from the 2021 Malawi Malaria Indicator Survey were analysed using R software version 4.3.0 for regressions and STATA version 17 for data cleaning.ResultsBoth models showed significant variability in treatment-seeking habits of caregivers between villages. The mixed-effects logit model residual identified Vuso Jere, Kampingo Sibande, Ngabu, and Dzoole as outliers in the model. Despite characteristics that promote late reporting of malaria at clinics, most mothers in these traditional authorities sought treatment within twenty-four hours of the onset of malaria symptoms in their children. On the other hand, the geospatial logit model showed that late seeking of malaria treatment was prevalent in most areas of the country, except a few traditional authorities such as Mwakaboko, Mwenemisuku, Mwabulambya, Mmbelwa, Mwadzama, Zulu, Amidu, Kasisi, and Mabuka.ConclusionsThese findings suggest that using a combination of multivariate regression model residuals and geospatial statistics can help in identifying communities with distinct treatment-seeking patterns for childhood malaria within a population. Health policymakers could benefit from consulting traditional authorities who demonstrated early reporting for care in this study. This could help in understanding the best practices followed by mothers in those areas which can be replicated in regions where seeking care is delayed.

Unveiling Site Visibility: A Study of Farming Communities in the Magaliesberg Region, South Africa

This paper explores the historical and contemporary significance of visibility in human interactions with their environments, particularly in the context of archaeology and the application of geographic information systems (GIS) for visibility analysis. The study highlights the role of visibility analysis in investigating not only the physical visibility of features in landscapes but also the cultural significance associated with seeing or not seeing them. It draws from the ‘visibility relates’ principle, which argues that individuals tend to establish connections with visible entities. The focus is on comparing nineteenth-century urban settlements (Kaditshwene, Molokwane, and Marothodi) in the Magaliesberg region of South Africa, particularly examining the strategic positioning of kraals within these Sotho-Tswana farming communities. These settlements are some of the more popular Late Farming Communities (AD 1300–1840) in South Africa; hence, they have archaeological background and are among the few, if not the only ones, that have LiDAR data coverage. The findings reveal distinctions in visibility at both settlement and household scalar levels, with Kaditshwene standing out as different from Marothodi and Molokwane. This suggests that kraals were strategically located to be more or less visible based on specific settlement circumstances, such as attracting people from other communities and concerns about cattle theft. This study contributes to GIS approaches to archaeological sites and landscapes in Africa and calls for more extensive use of geospatial statistics in African archaeology.

TURKISH REFUGEES IN BRAZIL AFTER THE 2016 TURKEY COUP ATTEMPT

Population geography is a subdiscipline of Geography devoted to the study of one of the aspects of geographic space, which is population flows. The flow of refugees is a topic of interest. In 2016, Turkey underwent a coup attempt which resulted in the deaths of hundreds of people and increased internal political tension. This study aimed to verify whether Turkey’s internal circumstances impacted the number of Turkish refugee requests in Brazil and if so, to describe these refugees’ general profile and their spatial distribution, without risking their safety. We used data provided by the International Migration Observatory, which is linked to the Brazilian Ministry of Justice. The analyses were performed using descriptive and geospatial statistics and showed a considerable increase in the number of Turkish refugees in Brazil from 2016, as well as their concentration in the richest state of the federation. We believe these results provide relevant information to improve the Brazilian system of recognition and reception of refugees.

Examining rehabilitation access disparities: an integrated analysis of electronic health record data and population characteristics through bivariate choropleth mapping.

Despite efforts to view electronic health records (EHR) data through an equity lens, crucial contextual information regarding patients' social environments remains limited. Integrating EHR data and Geographic Information Systems (GIS) technology can give deeper insights into the relationships between patients' social environments, health outcomes, and geographic factors. This study aims to identify regions with the fastest and slowest access to outpatient physical therapy services using bivariate choropleth maps to provide contextual insights that may contribute to health disparity in access. This was a retrospective cohort study of patients' access timelines for the first visit to outpatient physical therapy services (n = 10,363). The three timelines evaluated were (1) referral-to-scheduled appointment time, (2) scheduled appointment to first visit time, and (3) referral to first visit time. Hot and coldspot analyses (CI 95%) determined the fastest and slowest access times with patient-level characteristics and bivariate choropleth maps that were developed to visualize associations between access patterns and disadvantaged areas using Area Deprivation Index scores. Data were collected between January 1, 2016 and January 1, 2020. EHR data were geocoded via GIS technology to calculate geospatial statistics (Gi∗ statistic from ArcGIS Pro) in an urban area. Statistically significant differences were found for all three access timelines between coldspot (i.e., fast access group) and hotspot (i.e., slow access group) comparisons (p < .05). The hotspot regions had higher deprivation scores; higher proportions of residents who were older, privately insured, female, lived further from clinics; and a higher proportion of Black patients with orthopaedic diagnoses compared to the coldspot regions. Our study identified and described local areas with higher densities of patients that experienced longer access times to outpatient physical therapy services. Integration of EHR and GIS data is a more robust method to identify health disparities in access to care. With this approach, we can better understand the intricate interplay between social, economic, and environmental factors contributing to health disparities in access to care.

Evidence from personal ornaments suggest nine distinct cultural groups between 34,000 and 24,000 years ago in Europe.

Mechanisms governing the relationship between genetic and cultural evolution are the subject of debate, data analysis and modelling efforts. Here we present a new georeferenced dataset of personal ornaments worn by European hunter-gatherers during the so-called Gravettian technocomplex (34,000-24,000 years ago), analyse it with multivariate and geospatial statistics, model the impact of distance on cultural diversity and contrast the outcome of our analyses with up-to-date palaeogenetic data. We demonstrate that Gravettian ornament variability cannot be explained solely by isolation-by-distance. Analysis of Gravettian ornaments identified nine geographically discrete cultural entities across Europe. While broadly in agreement with palaeogenetic data, our results highlight a more complex pattern, with cultural entities located in areas not yet sampled by palaeogenetics and distinctive entities in regions inhabited by populations of similar genetic ancestry. Integrating personal ornament and biological data from other Palaeolithic cultures will elucidate the complex narrative of population dynamics of Upper Palaeolithic Europe.

Using Social Media and Multi-Source Geospatial Data for Quantifying and Understanding Visitor’s Preferences in Rural Forest Scenes: A Case Study from Nanjing

Rapid urbanization has made urban forest scenes scarce resources, leading to a surge in the demand for high-quality rural forest scenes as alternative outdoor recreation spaces. Previous studies mainly applied survey methods, focusing on visitors’ feedback for different types of scenes from the perspective of visual quality evaluation. Nevertheless, the explanations of the relationships between various factors of scenes and visitors’ preferences are relatively superficial. This study sought to explore the distribution and characteristics of preferred rural forest scenes based on visitor reviews from social media, and using Geodetector, a geospatial statistics tool, to quantitatively analyzed the reasons for visitors’ preferences in terms of factors obtained from multi-source geospatial data. The findings are that (1) visitors are already satisfied with the natural environment but expect scenes that reflect the culture of tea; (2) spatial factor has a more robust interpretation of visitors’ preference, and although the Normalized Difference Vegetation Index (NDVI) and non-consumption indicators barely explain visitors’ preference solely when each of them is combined with other indicators, they can produce non-linear enhancement effects. Consequently, this study synthesizes visitors’ feedback and factors in rural forest scenes to understand visitors’ preferences, thus providing insights into human-centered planning.

Geo-visualization of Sarawak COVID-19 Publicly Available Data Employing Open-source Geospatial Software

The state government of Sarawak with the help of the Sarawak Disaster Management Committee (SDMC) has continuously made the updated information on the state COVID-19 situation and its ensuing control measures available to general public in the form of daily press statements. However, these statements are merely providing textual information on daily basis though the data are in fact rich in temporal and spatial properties. Since the onset of COVID-19 pandemic, spatiotemporal analysis becomes the key element to better understand the spread of COVID-19 in various spatial levels worldwide. Hence, there is an urgent need to convert this textual information into more valuable insights by applying geo-visualization techniques and geospatial statistics. The paper demonstrates the prospect of retrieving geospatial data from publicly available document to locate, map and analyze the spread of COVID-19 up to division level of Sarawak. Specifically, map visualization and geospatial statistical analysis are performed for the list of exposed locations, which are indeed locations visited by COVID-19 patients prior to being tested positive in Kuching division, using open-source geospatial software QGIS. It is found that these exposed locations concentrate on the build-up areas in the division and are in south-west to north-east direction of the center of Kuching in September and October 2021. Despite the number of exposed locations published is relatively small compared to the number of confirmed cases reported, both are nearly strongly correlated. The insights gained from such geospatial analysis may assist the local public health authorities to impose applicable disease control interventions at division level.

Медико-географический атлас России «Факторы риска онкологических заболеваний»

The fight against cancer remains on the agenda to preserve the health of the Russian population and improve their quality of life. One of the methods of scientific analysis of the causes of the appearance and spread of cancer has become a complex system mapping with extensive use of geospatial statistics methods. Currently, the creation of cancer atlases is widely practiced in different countries of the world, which indicates their demand as an effective tool for research of the epidemiological search. No less significant is their role in educating and mobilizing the population to improve the environmental situation and, in some cases, to obtain specific practical knowledge in solving the problems that have arisen. One of the most pressing problems of modern fundamental medicine and, in particular, oncology, the solution of which is aimed at preventing cancer, is the establishment of risk factors and links between their effects and cancer. Risk factors are behavioral, biological, genetic, environmental, social, environmental, residential and work environment potentially hazardous to human health, which increase the likelihood of developing diseases, their progression and unfavorable outcome. To study risk factors on the territory of the Russian Federation, the Faculty of Geography of Lomonosov Moscow State University with the support of the Russian Geographical Society is working on the creation of the Medico-Geographical Atlas of Russia “Cancer risk factors”. The atlas is included in a series of previously published medico-geographical atlases of Russia, prepared by the faculty with the assistance of the Russian Geographical Society [2015; 2017; 2019]. The article presents the preliminary results of the preparation of the Atlas. At present, familiarization with domestic and foreign experience in the development of cancer atlases issues has been carried out; the concept of the Atlas was developed. In accordance with it, the structure and content of the Atlas was formed. The sources for the development of the Atlas were cartographic materials available on various media; statistical data for 2000–2021; the results of public surveys on the provision of medical care in oncology; original author’s development of a number of topics; numerous publications on the identification and work with cancer risk factors. Employees of leading oncological organizations are involved in the work on the Atlas: Federal State Budgetary Institution “National Medical Research Center of Oncology named after N.N. Blokhin” of the Ministry of Health of Russia, Moscow Research Institute of Oncology named after P.A. Herzen and others.

Unintended consequences of species translocations: changes in distribution and habitat selection of mule deer following introduction of elk

Translocation of animals for conservation has a long history of successes and failures since humans began intervening with species distributions in the early part of the 20th century. Effects of translocations on other species occupying the area of introduction are rarely considered in species management plans. We hypothesized that the introduction of a large-bodied ungulate, Rocky Mountain elk (Cervus canadensis nelsoni), would cause a shift in the spatial distribution and winter habitat selection of a smaller-bodied ungulate, Rocky Mountain mule deer (Odocoileus hemionus hemionus), already occupying the study area in northeastern Nevada. We examined mule deer survey locations during an 8-year time interval from 1993 to 2001 to test hypotheses related to potential competition with or displacement of mule deer following introduction of elk in 1997. We used geospatial statistics to quantify changes in seasonal distributions of mule deer and a resource selection function (RSF) framework to model changes in selection of resources by mule deer before and after elk were translocated into our study area. Our results indicated that mule deer exhibited a shift in their core distribution by approximately 5.72 km after the introduction of elk. Mule deer changed their use of habitat by selecting shallower slopes, more north-facing aspects, and areas farther from the elk release site, where most of the elk congregated. Mule deer selected habitats with more pinyon-juniper tree cover and mixed shrublands prior to the elk translocation, indicating a potential tradeoff in thermal cover and forage quality. This research is one of the few empirical studies to describe competitive interactions between elk and mule deer on a shared winter range in North America. Implications of this research have importance for managers concerned with restoring communities to their native conditions especially where the potential for competition with non-indigenous species or other competing species with similar niche requirements exists.

Geospatial analysis of blindness within rural and urban counties.

PurposeTo determine the associations of blindness within rural and urban counties using a registry of blind persons and geospatial analytics.MethodsWe used the Oregon Commission for the Blind registry to determine the number of persons who are legally blind, as well as licensure data to determine the density of eye care providers (optometrists and ophthalmologists) within each county of the State of Oregon. We used geospatial statistics, analysis of variance, and logistic regression to determine the explanatory variables associated with blindness within counties.ResultsWe included 8350 individuals who are legally blind within the state of Oregon in the calendar year 2015. The mean observed prevalence of registered blindness was 0.21% and ranged almost 9-fold from 0.04% to 0.58% among counties (p < .001). In univariate models, higher blindness was associated with increasing median age (p = .027), minority race (p < .001), decreased median household income (p < .001), increased poverty within a county (p < .001), and higher density of ophthalmologists (p = .003). Density of optometrists was not associated with prevalence of blindness (p = .89). The final multivariable model showed higher blindness to be associated with lower median household income, higher proportion of black race, and lower proportion of Hispanic race (p < .001 for all).ConclusionGeospatial analytics identified counties with higher and lower than expected proportions of blindness even when adjusted for sociodemographic factors. Clinicians and researchers may use the methods and results of this study to better understand the distribution of individuals with blindness and the associated factors to help design public health interventions.

Geospatial insights into the controls of microbialite formation at Laguna Negra, Argentina.

Microbialites provide a record of the interaction of microorganisms with their environment constituting a record of microbial life and environments through geologic time. Our capacity to interpret this record is limited by an incomplete understanding of the microbial, geochemical, and physical processes that influence microbialite formation and morphogenesis. The modern system Laguna Negra in Catamarca Province, Argentina contains microbialites in a zone of carbonate precipitation associated with physico-chemical gradients and variable microbial community structure, making it an ideal location to study how these processes interact to drive microbialite formation. In this study, we investigated the geospatial relationships between carbonate morphology, geochemistry, and microbial community at the macro- (decimeter) to mega- (meter) scale by combining high-resolution imagery with field observations. We mapped the distribution of carbonate morphologies and allochtonously-derived volcaniclasts and correlated these with sedimentary matrices and geochemical parameters. Our work shows that the macroscale distribution of different carbonate morphologies spatially correlates with microbial mat distributions-a result consistent with previous microscale observations. Specifically, microbialitic carbonate morphologies more commonly occur associated with microbial mats while abiotically derived carbonate morphologies were less commonly associated with microbial mats. Spatial variability in the size and abundance of mineralized structures was also observed, however, the processes controlling this variability remains unclear and likely represent a combination of microbial, geochemical, and physical processes. Likewise, the processes controlling the spatial distribution of microbial mats at Laguna Negra are also unresolved. Our results suggest that in addition to the physical drivers observed in other modern environments, variability in the spatial distribution of microbialites and other carbonate morphologies at the macro- to megascale can be controlled by microbial processes. Overall, this study provides insight into the interpretation of microbialite occurrence and distributions in the geologic record and highlights the utility of geospatial statistics to probe the controls of microbialite formation in other environments.

Crop-CASMA: A web geoprocessing and map service based architecture and implementation for serving soil moisture and crop vegetation condition data over U.S. Cropland

• A data service system for the selected SMAP data and its derivative data products over CONUS is developed. • All map and geoprocessing services are implemented via the OGC standard interface. • The data service system is interoperable with web and desktop clients. • Monitoring and analyzing U.S. cropland soil moisture conditions are significantly simplified. Soil moisture is an essential parameter to understand crop conditions throughout the growing season. Collecting soil moisture data by field observation is labor-intensive, especially when attempting to obtain Conterminous United States (CONUS) geographic coverage. In addition, using soil moisture for assessing current and future crop conditions is best realized by combining soil moisture estimates with concurrent observations of crop conditions. However, until recently, this capability has not been available to the public. In this paper, we present an interoperable data service application system, the Crop Condition and Soil Moisture Analytics (Crop-CASMA) system, that facilitates the retrieval, analysis, visualization, and sharing of soil moisture data for the CONUS. This system delivers a variety of satellite remote sensing based data products that are derived from Soil Moisture Active Passive (SMAP) Level-4 data and SMAP Thermal Hydraulic disaggregation of Soil Moisture (THySM) data, as well as vegetation index data derived from Moderate Resolution Imaging Spectroradiometer (MODIS) observations. To make services interoperable and reusable, all data products are disseminated via Open Geospatial Consortium (OGC) Web Map Service (WMS) and Web Coverage Service (WCS) interface standards. Additionally, a suite of geoprocessing operations, such as geospatial statistics, time-series profile generation, PDF map production, and image composition, has been implemented in the OGC Web Processing Service (WPS) interface standard. The implementation shows the proposed web service system can significantly simplify the mapping and quantitative analysis of soil moisture and crop condition over U.S. cropland. In addition, it is interoperable with GIS software and has been successfully integrated with web-based applications.

EcoPhyloMapper: An r package for integrating geographical ranges, phylogeny and morphology

Abstract Spatial patterns of species richness, phylogenetic and morphological diversity are key to answering many questions in ecology and evolution. Across spatial scales, geographical and environmental features, as well as evolutionary history and phenotypic traits, are thought to play roles in shaping both local species communities and regional assemblages. By examining these geographical patterns, it is possible to infer how different axes of biodiversity influence one another, and how their interaction with abiotic factors has led to the spatial distribution of species assemblages—and their attributes—that we observe in the present. Although there has been interest in this area of research for some time, it has recently become more tractable to include multivariate shape data in such analyses. Shape information has the potential to provide a more direct measure of the functional morphology of species as compared to individual trait measurements and might be more relevant to understanding community composition. However, few tools currently exist to explore geographical patterns of both phylogenetic and shape diversity. We present the EcoPhyloMapper r package (epm) that aims to streamline the handling of geographical range polygons or point occurrences and integration of resulting species metacommunities with phylogenetic trees and morphological shape. Geographical maps can be generated that demonstrate spatial patterns in diversity metrics pertaining to phylogenetic similarity, multivariate shape similarity and disparity, and combinations of the two. Patterns of taxonomic, phylogenetic and shape disparity turnover can also be visualized. Biodiversity indices summarized across grid cells can easily be exported to GIS software as well as to other r packages that specialize in community assembly or geospatial statistics. This r package will facilitate the geographical exploration of multivariate shape data in concert with phylogenetic diversity, which will, in turn, support macroecological research exploring how species assemblages are structured. Furthermore, this r package should prove useful across a wide range of macroecological applications that extend beyond the study of morphology.

Spatial analysis of global Bitcoin mining

Bitcoin mining is not only the fundamental process to maintain Bitcoin network, but also the key linkage between the virtual cryptocurrency and the physical world. A variety of issues associated with it have been raised, such as network security, cryptoasset management and sustainability impacts. Investigating Bitcoin mining from a spatial perspective will provide new angles and empirical evidence with respect to extant literature. Here we explore the spatial distribution of Bitcoin mining through bottom-up tracking and geospatial statistics. We find that mining activity has been detected at more than 6000 geographical units across 139 countries and regions, which is in line with the distributed design of Bitcoin network. However, in terms of computing power, it has demonstrated a strong tendency of spatial concentration and association with energy production locations. We also discover that the spatial distribution of Bitcoin mining is dynamic, which fluctuates with diverse patterns, according to economic and regulatory changes.

Influence of geological conditions on gully distribution in the Dry–hot Valley, SW China

Geological conditions are important in the initiation and expansion of a gully. In order to explore the relationship between spatial distribution of gullies and the geological conditions, the methods of geospatial statistics and correlation analysis were used based on interpretations of high-resolution Google Earth images in the Dry–hot Valley (SW China). The results showed that the gullies had a clustering pattern in the study area (Moran’s I = 0.89, P < 0.01). The correlation between geological conditions (lithology and faults) and clustering characteristics of gullies was strong in lithology (F(6) = 139.75, p < 0.01, η2 = 0.184) and irrelevant in faults (F(2) = 0.64, p > 0.05, η2 = 0.001). The resistance to erosion of a particular lithology had a negative correlation with the degree of gully clustering. The lithologies of sandstones, mudstones and siltstones, loose quaternary sediments, of phyllite and schist were conducive to gully development; however, the conglomerate, quartzite and marble, and granite may led to low gully clustering. The results will help to understand the law of spatial distribution of gullies at the regional scale and also provide a scientific reference for their controls.

Triatomine fauna in the state of Bahia, Brazil: What changed after 40 years of the vector-control program?

ABSTRACTBackground: Neglected tropical diseases are a growing threat to global health, and endemic Chagas disease has emerged as one of the most important health problems in America. The main strategy to prevent Trypanosoma cruzi transmission is chemical control of vectors. This study presents a descriptive analysis of synanthropic triatomines before and after the implementation of a vector-control program in Bahia, Brazil. Methods: Descriptive analysis and geospatial statistics were performed on triatomine data, (1) the relative abundance and (2) proportional spatial distribution, from Bahia during two periods: (A) 1957 to 1971 and (B) 2006 to 2019. Results: We observed a decrease in the relative abundance of Panstrongylus megistus (A: n=22.032, 61.9%; B: n=1.842, 1.0%) and Triatoma infestans (A: n=1.310, 3.7%; B: n=763, 0.43%), as well as an increase in the relative abundance of T. sordida (A: n=8.314, 23.4%, B: n=146.901, 81.6%) and T. pseudomaculata (A: n=894, 2.5%, B: n=16.717, 9.3%). Conclusions: Our results indicate a clear reduction in the occurrence of P. megistus and T. infestans (last record in 2015) and an increase in the relative abundance and geographical distribution of T. sordida and T. pseudomaculata after 40 years of the vector-control program. The high frequency of other triatomine species in the municipalities of the state of Bahia and their abundance in recent years highlight the need to reinforce permanent entomological surveillance actions to prevent Chagas disease.

Geography of Disparity: Connecting COVID-19 Vulnerability and Social Determinants of Health in Colorado

The COVID-19 pandemic has drawn greater attention to social determinants of health and associated health inequities, which disproportionately affect vulnerable populations and places in the U.S. In this study, we explored geographic patterns of local-level COVID-19 vulnerability and associations with social and health determinants across Colorado. To conceptualize social and health determinants and how together they generate risk and exposure, we integrated the concepts of social vulnerability and syndemic to situate COVID-19 vulnerability within a broader hazards of place framework. Using geospatial statistics and GIS, we estimated census tract-level rates of COVID-19, which are not yet available in Colorado, and mapped areas of high and low incidence risk. We also developed composite indices that characterized social and health vulnerabilities to measure multivariate associations with COVID-19 rates. The findings revealed hotspots of persistent risk in mountain communities since the pandemic emerged in Colorado, as well as clusters of risk in the Urban Front Range’s central and southern counties, and across many parts of eastern Colorado. Vulnerability analyses indicate that COVID-19 rates were associated with mental health and chronic conditions along with social determinants that represent inequities in education, income, healthcare access, and race/ethnicity (minority percent of population), which may have disproportionately exposed some communities more than others to infection and severe health outcomes. Overall, the findings provide geographic health information about COVID-19 and vulnerability context, which may better inform local decision-making for interventions and policies that support equity of social determinants of health. Supplemental data for this article is available online at https://doi.org/10.1080/08964289.2021.2021382 .

Assessment of earthquake-induced landslide inventories and susceptibility maps using slope unit-based logistic regression and geospatial statistics

Inventories of seismically induced landslides provide essential information about the extent and severity of ground effects after an earthquake. Rigorous assessment of the completeness of a landslide inventory and the quality of a landslide susceptibility map derived from the inventory is of paramount importance for disaster management applications. Methods and materials applied while preparing inventories influence their quality, but the criteria for generating an inventory are not standardized. This study considered five landslide inventories prepared by different authors after the 2015 Gorkha earthquake, to assess their differences, understand the implications of their use in producing landslide susceptibility maps in conjunction with standard landslide predisposing factors and logistic regression. We adopted three assessment criteria: (1) an error index to identify the mutual mismatches between the inventories; (2) statistical analysis, to study the inconsistency in predisposing factors and performance of susceptibility maps; and (3) geospatial analysis, to assess differences between the inventories and the corresponding susceptibility maps. Results show that substantial discrepancies exist among the mapped landslides. Although there is no distinct variation in the significance of landslide causative factors and the performance of susceptibility maps, a hot spot analysis and cluster/outlier analysis of the maps revealed notable differences in spatial patterns. The percentages of landslide-prone hot spots and clustered areas are directly proportional to the size of the landslide inventory. The proposed geospatial approaches provide a new perspective to the investigators for the quantitative analysis of earthquake-triggered landslide inventories and susceptibility maps.

Spatially compounded surge events: an example from hurricanes Matthew and Florence

Abstract. The simultaneous rise of tropical-cyclone-induced flood waters across a large hazard management domain can stretch rescue and recovery efforts. Here we present a means to quantify the connectedness of maximum surge during a storm with geospatial statistics. Tide gauges throughout the extensive estuaries and barrier islands of North Carolina deployed and operating during hurricanes Matthew (n=82) and Florence (n=123) are used to compare the spatial compounding of surge for these two disasters. Moran's I showed the occurrence of maximum storm tide was more clustered for Matthew compared to Florence, and a semivariogram analysis produced a spatial range of similarly timed storm tide that was 4 times as large for Matthew than Florence. A more limited data set of fluvial flooding and precipitation in eastern North Carolina showed a consistent result – multivariate flood sources associated with Matthew were more concentrated in time as compared to Florence. Although Matthew and Florence were equally intense, they had very different tracks and speeds which influenced the timing of surge along the coast.

High Performance Multivariate Geospatial Statistics on Manycore Systems

Modeling and inferring spatial relationships and predicting missing values of environmental data are some of the main tasks of geospatial statisticians. These routine tasks are accomplished using multivariate geospatial models and the cokriging technique. The latter requires the evaluation of the expensive Gaussian log-likelihood function, which has impeded the adoption of multivariate geospatial models for large multivariate spatial datasets. However, this large-scale cokriging challenge provides a fertile ground for supercomputing implementations for the geospatial statistics community as it is paramount to scale computational capability to match the growth in environmental data coming from the widespread use of different data collection technologies. In this article, we develop and deploy large-scale multivariate spatial modeling and inference on parallel hardware architectures. To tackle the increasing complexity in matrix operations and the massive concurrency in parallel systems, we leverage low-rank matrix approximation techniques with task-based programming models and schedule the asynchronous computational tasks using a dynamic runtime system. The proposed framework provides both the dense and the approximated computations of the Gaussian log-likelihood function. It demonstrates accuracy robustness and performance scalability on a variety of computer systems. Using both synthetic and real datasets, the low-rank matrix approximation shows better performance compared to exact computation, while preserving the application requirements in both parameter estimation and prediction accuracy. We also propose a novel algorithm to assess the prediction accuracy after the online parameter estimation. The algorithm quantifies prediction performance and provides a benchmark for measuring the efficiency and accuracy of several approximation techniques in multivariate spatial modeling.