Spatial clustering groups similar spatial objects and/or patterns into distinct, spatially cohere... more Spatial clustering groups similar spatial objects and/or patterns into distinct, spatially coherent clusters and is an important component of spatial data analysis in the earth sciences. Spectral clustering is a popular modern clustering algorithm, which relies on graph-theoretical concepts. It outperforms traditional clustering algorithms, but generally does not account for spatial relationships. The aim of this study is to improve the spatial awareness of the spectral clustering by utilising a dissimilarity matrix based on nonparametric higher order spatial statistics. The proposed model-free technique can identify meaningful spatial clusters (i.e., meaningful geographical subregions) from multivariate spatial data of different scales without the need to define a model of co-dependence
This poster details the variography and the Ordinary Kriging estimation maps for catch rates (kg/... more This poster details the variography and the Ordinary Kriging estimation maps for catch rates (kg/h) of western king prawns (Peneous latisulcatus) in the Shark Ray Managed Prawn Fishery.
ObjectivesThe aim of the study was to investigate the spatial and temporal relationships between ... more ObjectivesThe aim of the study was to investigate the spatial and temporal relationships between the prevalence of COVID-19 symptoms in the community-level and area-level social deprivation.DesignSpatial mapping, generalised linear models, using time as a factor and spatial-lag models were used to explore the relationship between self-reported COVID-19 symptom prevalence as recorded through two smartphone symptom tracker apps and a range of socioeconomic factors using a repeated cross-sectional study design.SettingIn the community in Northern Ireland, UK. The analysis period included the earliest stages of non-pharmaceutical interventions and societal restrictions or ‘lockdown’ in 2020.ParticipantsUsers of two smartphone symptom tracker apps recording self-reported health information who recorded their location as Northern Ireland, UK.Primary outcome measuresPopulation standardised self-reported COVID-19 symptoms and correlation between population standardised self-reported COVID-19...
While remote camera surveys have the potential to improve the accuracy of recreational fishing es... more While remote camera surveys have the potential to improve the accuracy of recreational fishing estimates, missing data are common and require robust analytical techniques to impute. Time-lapse cameras are being used in Western Australia to monitor recreational boating activities, but outages have occurred. Generalized linear mixed effect models formulated in a fully conditional specification multiple imputation framework were used to reconstruct missing data, with climatic and some temporal classifications as covariates. Using a complete 12-month camera record of hourly counts of recreational powerboat retrievals, data were simulated based on ten observed camera outage patterns, with a missing proportion of between 0.06 and 0.61. Nine models were evaluated, including Poisson and negative binomial models, and their associated zero-inflated variants. The imputed values were cross-validated against actual observations using percent bias, mean absolute error, root mean square error, and...
Digital camera monitoring is increasingly being used to monitor recreational fisheries. The manua... more Digital camera monitoring is increasingly being used to monitor recreational fisheries. The manual interpretation of video imagery can be costly and time consuming. In an a posteriori analysis, we investigated trade-offs between the reading cost and accuracy measures of estimates of boat retrievals obtained at various sampling proportions for low, moderate and high traffic boat ramps in Western Australia. Simple random sampling, systematic sampling and stratified sampling designs with proportional and weighted allocation were evaluated to assess tradeoffs in terms of bias, accuracy, precision, coverage rate and cost in estimating the annual total number of powerboat retrievals in 10,000 jackknife resampling draws. The relative standard error (RSE ± standard deviations) obtained by the sampling designs for sampling proportions from 0.4 onwards were below a 20 % threshold for three of the sampling designs across the three boat ramps. Coverage rates of over 90 % were observed for the confidence intervals for the estimated annual number of powerboat retrievals, with low relative standard errors (RSE < 20 %). Interpreting 40 % of camera footage within a year provided the minimum level to obtain sufficient accuracy measures for all sampling designs considered. The stratified random sampling design with weighted allocation consistently resulted in the smallest variance for estimates of annual powerboat retrievals across the various sampled proportions. These findings have the potential to considerably reduce the cost of manual data interpretation, since operating cost increased linearly with increasing sampling proportion.
The particularities of geosystems and geoscience data must be understood before any development o... more The particularities of geosystems and geoscience data must be understood before any development or implementation of statistical learning algorithms. Without such knowledge, the predictions and inferences may not be accurate and physically consistent. Accuracy, transparency and interpretability, credibility, and physical realism are minimum criteria for statistical learning algorithms when applied to the geosciences. This study briefly reviews several characteristics of geoscience data and challenges for novel statistical learning algorithms. A novel spatial spectral clustering approach is introduced to illustrate how statistical learners can be adapted for modelling geoscience data. The spatial awareness and physical realism of the spectral clustering are improved by utilising a dissimilarity matrix based on nonparametric higher-order spatial statistics. The proposed model-free technique can identify meaningful spatial clusters (i.e. meaningful geographical subregions) from multiva...
Analysis of high-resolution data offers greater opportunity to understand the nature of data vari... more Analysis of high-resolution data offers greater opportunity to understand the nature of data variability, behaviours, trends and to detect small changes. Climate studies often require complete time series data which, in the presence of missing data, means imputation must be undertaken. Research on the imputation of high-resolution temporal climate time series data is still at an early phase. In this study, multiple approaches to the imputation of missing values were evaluated, including a structural time series model with Kalman smoothing, an autoregressive integrated moving average (ARIMA) model with Kalman smoothing and multiple linear regression. The methods were applied to complete subsets of data from 12 month time series of hourly temperature, humidity and wind speed data from four locations along the coast of Western Australia. Assuming that observations were missing at random, artificial gaps of missing observations were studied using a five-fold cross-validation methodology with the proportion of missing data set to 10%. The techniques were compared using the pooled mean absolute error, root mean square error and symmetric mean absolute percentage error. The multiple linear regression model was generally the best model based on the pooled performance indicators, followed by the ARIMA with Kalman smoothing. However, the low error values obtained from each of the approaches suggested that the models competed closely and imputed highly plausible values. To some extent, the performance of the models varied among locations. It can be concluded that the modelling approaches studied have demonstrated suitability in imputing missing data in hourly temperature, humidity and wind speed data and are therefore recommended for application in other fields where high-resolution data with missing values are common.
An accurate prediction of benefit in ore deposits with heterogeneous spatial variations requires ... more An accurate prediction of benefit in ore deposits with heterogeneous spatial variations requires the definition of geological domains that differentiate the types of mineralogy, alteration, and lithology, as well as the prediction of full mineral and geochemical compositions within each modeled domain and across boundaries between different domains. This paper proposes and compares various approaches (different combinations of log-ratio transformation, Gaussian and flow anamorphosis, and deterministic or probabilistic geological models) for geostatistical simulation of geochemical compositions in the presence of several geological domains. Different approaches are illustrated through an application to a nickel-cobalt laterite deposit located in Western Australia. Four rock types (ferruginous, smectite, saprolite, and ultramafic) are considered to define compositionally homogeneous domains. Geochemical compositions are comprised of six different components of interest (Fe, Al, Mg, Ni, Co, and Filler). The results suggest that the flow anamorphosis is a vital element for geostatistical modeling of geochemical composition due to its invariance properties and capability for reproducing complex patterns in input data, including:
Prediction of true classes of surficial and deep earth materials using multivariate spatial data ... more Prediction of true classes of surficial and deep earth materials using multivariate spatial data is a common challenge for geoscience modelers. Most geological processes leave a footprint that can be explored by geochemical data analysis. These footprints are normally complex statistical and spatial patterns buried deep in the high-dimensional compositional space. This paper proposes a spatial predictive model for classification of surficial and deep earth materials derived from the geochemical composition of surface regolith. The model is based on a combination of geostatistical simulation and machine learning approaches. A random forest predictive model is trained, and features are ranked based on their contribution to the predictive model. To generate potential and uncertainty maps, compositional data are simulated at unsampled locations via a chain of transformations (isometric log-ratio transformation followed by the flow anamorphosis) and geostatistical simulation. The simulated results are subsequently back-transformed to the original compositional space. The trained predictive model is used to estimate the probability of classes for simulated compositions. The proposed approach is illustrated through two case studies. In the first case study, the major crustal blocks of the Australian continent are predicted from the surface regolith geochemistry of the National Geochemical Survey of Australia project. The aim of the second case study is to discover the superficial deposits (peat) from the regional-scale soil geochemical data of the Tellus Project. The accuracy of the results in these two case studies confirms the usefulness of the proposed method for geological class prediction and geological process discovery.
The saline wetlands of the Northern Ponds, Lake MacLeod, contravene general classifications for s... more The saline wetlands of the Northern Ponds, Lake MacLeod, contravene general classifications for salt lakes because, despite enduring high levels of evaporation, they contain permanently inundated ponds that are continually supplied with seawater by a seepage face from the Indian Ocean. The present study investigated the physical, sediment and chemical characteristics of these ponds, using sonar, flow-rate measurements, microscopy and inductively coupled plasma–mass spectrometry to understand the role of biogenic and terrigenous inputs on the benthic habitats. The chemical composition of the water bodies did not differ significantly from the seawater feedstock, which facilitates biogenic sediment accumulation. The largest permanent water body differs from the other ponds in physical and chemical structure, in that, possibly because of its size, it is dominated by terrigenous inputs, which result in higher nutrient concentrations and non-biogenic sediments. The Northern Ponds represen...
Multi-element near-surface geochemistry from the National Geochemical Survey of Australia has bee... more Multi-element near-surface geochemistry from the National Geochemical Survey of Australia has been evaluated in the context of mapping the exposed to deeply buried major crustal blocks of the Australian continent. The major crustal blocks, interpreted from geophysical and geological data, reflect distinct tectonic domains comprised of early Archean to recent Cenozoic igneous, metamorphic and sedimentary rock assemblages. The geochemical data have been treated as compositional data to uniquely describe and characterize the geochemistry of the regolith overlying the major crustal blocks across Australia according to the following workflow: imputation of missing/censored data, log-ratio transformation, multivariate statistical analysis, multivariate geospatial (minimum/maximum autocorrelation factor) analysis, and classification. Using cross validation techniques, the uniqueness of each major crustal block has been quantified. The ability to predict the membership of a surface regolith sample to one or more of the major crustal blocks is demonstrated. The predicted crustal block assignments define spatially coherent regions that coincide with the known crustal blocks. In some areas, inaccurate predictions are due to uncertainty in the initial crustal boundary definition or from surficial processes that mask the crustal block geochemical signature. In conclusion, the geochemical composition of the Australian surface regolith generally can be used to map the underlying crustal architecture, despite secondary modifications due to physical transport and chemical weathering effects. This methodology is however less effective where extensive and thick sedimentary basins such as the Eromanga and Eucla basins overlie crustal blocks.
Abstract A multivariate spatial analysis was conducted on a suite of glacial till geochemical dat... more Abstract A multivariate spatial analysis was conducted on a suite of glacial till geochemical data collected over the Melville Peninsula, Nunavut, Canada. Previous studies demonstrated through the application of multivariate statistical techniques that the composition of the lake sediment geochemistry reflects the underlying geology in northern Canada. In this study, the application of minimum/maximum autocorrelation factor analysis (MAF) to glacial till geochemistry has extended the knowledge and description of the underlying geology through the recognition of spatially correlated factors that represent distinct lithologic features and glacial transport processes. Maps of posterior probabilities for the underlying lithologies were estimated based on the MA factors and compared with those from standard principal component analysis. The use of MAF provides a measured improvement in predictive mapping irrespective of the choice of lag separation.
It is a common challenge for the geosciences to jointly model the uncertainty in continuous and c... more It is a common challenge for the geosciences to jointly model the uncertainty in continuous and categorical regionalised variables and to reproduce observed spatial correlation and complex relationships in realisations. The demand for computational efficiency in the case of high-dimensional data and large simulation domains has led practitioners to utilise approaches based on decorrelation/recorrelation and independent simulation. Among such approaches the method of min/max autocorrelation factors (MAF) has proven to be a practical technique for industrial purposes. This study presents a hybrid model for joint simulation of high-dimensional continuous and categorical variables. Continuous variables are transformed to Gaussian random functions (GRFs) via anamorphosis functions and categorical variables are obtained by truncating one or more GRFs based on the plurigaussian model. MAF factors are then derived from all GRFs. After independent simulation of MAF factors, different realisations of continuous and categorical variables are obtained via back-transformation of MAF factors followed by back-transformation for continuous and truncation for categorical variables, respectively. The proposed algorithm is illustrated through a case study.
Spatial clustering groups similar spatial objects and/or patterns into distinct, spatially cohere... more Spatial clustering groups similar spatial objects and/or patterns into distinct, spatially coherent clusters and is an important component of spatial data analysis in the earth sciences. Spectral clustering is a popular modern clustering algorithm, which relies on graph-theoretical concepts. It outperforms traditional clustering algorithms, but generally does not account for spatial relationships. The aim of this study is to improve the spatial awareness of the spectral clustering by utilising a dissimilarity matrix based on nonparametric higher order spatial statistics. The proposed model-free technique can identify meaningful spatial clusters (i.e., meaningful geographical subregions) from multivariate spatial data of different scales without the need to define a model of co-dependence
This poster details the variography and the Ordinary Kriging estimation maps for catch rates (kg/... more This poster details the variography and the Ordinary Kriging estimation maps for catch rates (kg/h) of western king prawns (Peneous latisulcatus) in the Shark Ray Managed Prawn Fishery.
ObjectivesThe aim of the study was to investigate the spatial and temporal relationships between ... more ObjectivesThe aim of the study was to investigate the spatial and temporal relationships between the prevalence of COVID-19 symptoms in the community-level and area-level social deprivation.DesignSpatial mapping, generalised linear models, using time as a factor and spatial-lag models were used to explore the relationship between self-reported COVID-19 symptom prevalence as recorded through two smartphone symptom tracker apps and a range of socioeconomic factors using a repeated cross-sectional study design.SettingIn the community in Northern Ireland, UK. The analysis period included the earliest stages of non-pharmaceutical interventions and societal restrictions or ‘lockdown’ in 2020.ParticipantsUsers of two smartphone symptom tracker apps recording self-reported health information who recorded their location as Northern Ireland, UK.Primary outcome measuresPopulation standardised self-reported COVID-19 symptoms and correlation between population standardised self-reported COVID-19...
While remote camera surveys have the potential to improve the accuracy of recreational fishing es... more While remote camera surveys have the potential to improve the accuracy of recreational fishing estimates, missing data are common and require robust analytical techniques to impute. Time-lapse cameras are being used in Western Australia to monitor recreational boating activities, but outages have occurred. Generalized linear mixed effect models formulated in a fully conditional specification multiple imputation framework were used to reconstruct missing data, with climatic and some temporal classifications as covariates. Using a complete 12-month camera record of hourly counts of recreational powerboat retrievals, data were simulated based on ten observed camera outage patterns, with a missing proportion of between 0.06 and 0.61. Nine models were evaluated, including Poisson and negative binomial models, and their associated zero-inflated variants. The imputed values were cross-validated against actual observations using percent bias, mean absolute error, root mean square error, and...
Digital camera monitoring is increasingly being used to monitor recreational fisheries. The manua... more Digital camera monitoring is increasingly being used to monitor recreational fisheries. The manual interpretation of video imagery can be costly and time consuming. In an a posteriori analysis, we investigated trade-offs between the reading cost and accuracy measures of estimates of boat retrievals obtained at various sampling proportions for low, moderate and high traffic boat ramps in Western Australia. Simple random sampling, systematic sampling and stratified sampling designs with proportional and weighted allocation were evaluated to assess tradeoffs in terms of bias, accuracy, precision, coverage rate and cost in estimating the annual total number of powerboat retrievals in 10,000 jackknife resampling draws. The relative standard error (RSE ± standard deviations) obtained by the sampling designs for sampling proportions from 0.4 onwards were below a 20 % threshold for three of the sampling designs across the three boat ramps. Coverage rates of over 90 % were observed for the confidence intervals for the estimated annual number of powerboat retrievals, with low relative standard errors (RSE < 20 %). Interpreting 40 % of camera footage within a year provided the minimum level to obtain sufficient accuracy measures for all sampling designs considered. The stratified random sampling design with weighted allocation consistently resulted in the smallest variance for estimates of annual powerboat retrievals across the various sampled proportions. These findings have the potential to considerably reduce the cost of manual data interpretation, since operating cost increased linearly with increasing sampling proportion.
The particularities of geosystems and geoscience data must be understood before any development o... more The particularities of geosystems and geoscience data must be understood before any development or implementation of statistical learning algorithms. Without such knowledge, the predictions and inferences may not be accurate and physically consistent. Accuracy, transparency and interpretability, credibility, and physical realism are minimum criteria for statistical learning algorithms when applied to the geosciences. This study briefly reviews several characteristics of geoscience data and challenges for novel statistical learning algorithms. A novel spatial spectral clustering approach is introduced to illustrate how statistical learners can be adapted for modelling geoscience data. The spatial awareness and physical realism of the spectral clustering are improved by utilising a dissimilarity matrix based on nonparametric higher-order spatial statistics. The proposed model-free technique can identify meaningful spatial clusters (i.e. meaningful geographical subregions) from multiva...
Analysis of high-resolution data offers greater opportunity to understand the nature of data vari... more Analysis of high-resolution data offers greater opportunity to understand the nature of data variability, behaviours, trends and to detect small changes. Climate studies often require complete time series data which, in the presence of missing data, means imputation must be undertaken. Research on the imputation of high-resolution temporal climate time series data is still at an early phase. In this study, multiple approaches to the imputation of missing values were evaluated, including a structural time series model with Kalman smoothing, an autoregressive integrated moving average (ARIMA) model with Kalman smoothing and multiple linear regression. The methods were applied to complete subsets of data from 12 month time series of hourly temperature, humidity and wind speed data from four locations along the coast of Western Australia. Assuming that observations were missing at random, artificial gaps of missing observations were studied using a five-fold cross-validation methodology with the proportion of missing data set to 10%. The techniques were compared using the pooled mean absolute error, root mean square error and symmetric mean absolute percentage error. The multiple linear regression model was generally the best model based on the pooled performance indicators, followed by the ARIMA with Kalman smoothing. However, the low error values obtained from each of the approaches suggested that the models competed closely and imputed highly plausible values. To some extent, the performance of the models varied among locations. It can be concluded that the modelling approaches studied have demonstrated suitability in imputing missing data in hourly temperature, humidity and wind speed data and are therefore recommended for application in other fields where high-resolution data with missing values are common.
An accurate prediction of benefit in ore deposits with heterogeneous spatial variations requires ... more An accurate prediction of benefit in ore deposits with heterogeneous spatial variations requires the definition of geological domains that differentiate the types of mineralogy, alteration, and lithology, as well as the prediction of full mineral and geochemical compositions within each modeled domain and across boundaries between different domains. This paper proposes and compares various approaches (different combinations of log-ratio transformation, Gaussian and flow anamorphosis, and deterministic or probabilistic geological models) for geostatistical simulation of geochemical compositions in the presence of several geological domains. Different approaches are illustrated through an application to a nickel-cobalt laterite deposit located in Western Australia. Four rock types (ferruginous, smectite, saprolite, and ultramafic) are considered to define compositionally homogeneous domains. Geochemical compositions are comprised of six different components of interest (Fe, Al, Mg, Ni, Co, and Filler). The results suggest that the flow anamorphosis is a vital element for geostatistical modeling of geochemical composition due to its invariance properties and capability for reproducing complex patterns in input data, including:
Prediction of true classes of surficial and deep earth materials using multivariate spatial data ... more Prediction of true classes of surficial and deep earth materials using multivariate spatial data is a common challenge for geoscience modelers. Most geological processes leave a footprint that can be explored by geochemical data analysis. These footprints are normally complex statistical and spatial patterns buried deep in the high-dimensional compositional space. This paper proposes a spatial predictive model for classification of surficial and deep earth materials derived from the geochemical composition of surface regolith. The model is based on a combination of geostatistical simulation and machine learning approaches. A random forest predictive model is trained, and features are ranked based on their contribution to the predictive model. To generate potential and uncertainty maps, compositional data are simulated at unsampled locations via a chain of transformations (isometric log-ratio transformation followed by the flow anamorphosis) and geostatistical simulation. The simulated results are subsequently back-transformed to the original compositional space. The trained predictive model is used to estimate the probability of classes for simulated compositions. The proposed approach is illustrated through two case studies. In the first case study, the major crustal blocks of the Australian continent are predicted from the surface regolith geochemistry of the National Geochemical Survey of Australia project. The aim of the second case study is to discover the superficial deposits (peat) from the regional-scale soil geochemical data of the Tellus Project. The accuracy of the results in these two case studies confirms the usefulness of the proposed method for geological class prediction and geological process discovery.
The saline wetlands of the Northern Ponds, Lake MacLeod, contravene general classifications for s... more The saline wetlands of the Northern Ponds, Lake MacLeod, contravene general classifications for salt lakes because, despite enduring high levels of evaporation, they contain permanently inundated ponds that are continually supplied with seawater by a seepage face from the Indian Ocean. The present study investigated the physical, sediment and chemical characteristics of these ponds, using sonar, flow-rate measurements, microscopy and inductively coupled plasma–mass spectrometry to understand the role of biogenic and terrigenous inputs on the benthic habitats. The chemical composition of the water bodies did not differ significantly from the seawater feedstock, which facilitates biogenic sediment accumulation. The largest permanent water body differs from the other ponds in physical and chemical structure, in that, possibly because of its size, it is dominated by terrigenous inputs, which result in higher nutrient concentrations and non-biogenic sediments. The Northern Ponds represen...
Multi-element near-surface geochemistry from the National Geochemical Survey of Australia has bee... more Multi-element near-surface geochemistry from the National Geochemical Survey of Australia has been evaluated in the context of mapping the exposed to deeply buried major crustal blocks of the Australian continent. The major crustal blocks, interpreted from geophysical and geological data, reflect distinct tectonic domains comprised of early Archean to recent Cenozoic igneous, metamorphic and sedimentary rock assemblages. The geochemical data have been treated as compositional data to uniquely describe and characterize the geochemistry of the regolith overlying the major crustal blocks across Australia according to the following workflow: imputation of missing/censored data, log-ratio transformation, multivariate statistical analysis, multivariate geospatial (minimum/maximum autocorrelation factor) analysis, and classification. Using cross validation techniques, the uniqueness of each major crustal block has been quantified. The ability to predict the membership of a surface regolith sample to one or more of the major crustal blocks is demonstrated. The predicted crustal block assignments define spatially coherent regions that coincide with the known crustal blocks. In some areas, inaccurate predictions are due to uncertainty in the initial crustal boundary definition or from surficial processes that mask the crustal block geochemical signature. In conclusion, the geochemical composition of the Australian surface regolith generally can be used to map the underlying crustal architecture, despite secondary modifications due to physical transport and chemical weathering effects. This methodology is however less effective where extensive and thick sedimentary basins such as the Eromanga and Eucla basins overlie crustal blocks.
Abstract A multivariate spatial analysis was conducted on a suite of glacial till geochemical dat... more Abstract A multivariate spatial analysis was conducted on a suite of glacial till geochemical data collected over the Melville Peninsula, Nunavut, Canada. Previous studies demonstrated through the application of multivariate statistical techniques that the composition of the lake sediment geochemistry reflects the underlying geology in northern Canada. In this study, the application of minimum/maximum autocorrelation factor analysis (MAF) to glacial till geochemistry has extended the knowledge and description of the underlying geology through the recognition of spatially correlated factors that represent distinct lithologic features and glacial transport processes. Maps of posterior probabilities for the underlying lithologies were estimated based on the MA factors and compared with those from standard principal component analysis. The use of MAF provides a measured improvement in predictive mapping irrespective of the choice of lag separation.
It is a common challenge for the geosciences to jointly model the uncertainty in continuous and c... more It is a common challenge for the geosciences to jointly model the uncertainty in continuous and categorical regionalised variables and to reproduce observed spatial correlation and complex relationships in realisations. The demand for computational efficiency in the case of high-dimensional data and large simulation domains has led practitioners to utilise approaches based on decorrelation/recorrelation and independent simulation. Among such approaches the method of min/max autocorrelation factors (MAF) has proven to be a practical technique for industrial purposes. This study presents a hybrid model for joint simulation of high-dimensional continuous and categorical variables. Continuous variables are transformed to Gaussian random functions (GRFs) via anamorphosis functions and categorical variables are obtained by truncating one or more GRFs based on the plurigaussian model. MAF factors are then derived from all GRFs. After independent simulation of MAF factors, different realisations of continuous and categorical variables are obtained via back-transformation of MAF factors followed by back-transformation for continuous and truncation for categorical variables, respectively. The proposed algorithm is illustrated through a case study.
Uploads
Papers by Ute Mueller