A nonparametric bootstrap method for spatial data

Castillo-Páez, Sergio; Fernández-Casal, Rubén; García-Soidán, Pilar

doi:10.1016/j.csda.2019.01.017

Cited by 10 publications

(5 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…For further geostatistical analyses, the influential points detected by the estimation of the local Moran's I test were removed from the database. For each soil property, semivariograms were calculated using local polynomial kernel smoothing of linearly binned semivariances, using the np.svar function of the npsp R package (Castillo‐Páez et al, 2019). A variogram is defined as the variance of the difference between field values measured at two locations (Shekhar et al, 2017).…”

Section: Methodsmentioning

confidence: 99%

Small‐scale spatial variability of Technosol properties in a chronosequence of reclamation of dredged‐sediment landfills

et al. 2023

View full text Add to dashboard Cite

Active reclamation is often necessary to ensure a transformation of mining waste into Technosols-"soils dominated or strongly influenced by human-made material"and restore its utility and environmental value. The objective of this study is to assess the spatial variation of the physicochemical properties relevant for reclamation and Technosol formation on dredged-sediment landfills left by alluvial gold mining in a chronosequence of reclamation (0, 4, 8, and 12 years). We hypothesize a higher spatial dependency of most soil properties with increasing time of Technosol formation and an overall homogenization of the soil resulting from pedogenetic processes. Our results show early signs of Technosol pedogenesis with strong physicochemical changes after only few years of formation. We observed that older Technosols (>4 years) are more acidic, have less nutrient and organic matter content, higher exchangeable cations, and less signs of compaction than nonrevegetated landfills.The content of organic matter, phosphorus, and exchangeable cations show the highest spatial variability in Technosols of all ages. In older Technosols, most soil properties showed less spatial variability than in younger Technosols. A multivariate geostatistical approach allowed the delineation of zones with distinctive physicochemical properties within areas of the chronosequence. The results show that reclamation and Technosol formation lead to spatially dependent fragmentation processes reflected in more and smaller clustered zones in Technosols after 12 years of formation. From the perspective of reclamation management, understanding the spatial variability of highly heterogeneous Technosols where substantial changes can be observed within small distances can support the development of site-specific reclamation strategies suitable to the characteristics of each field as well as the determination of its potential uses.

show abstract

Section: Methodsmentioning

confidence: 99%

Small‐scale spatial variability of Technosol properties in a chronosequence of reclamation of dredged‐sediment landfills

et al. 2023

View full text Add to dashboard Cite

show abstract

“…For example, if the model's mean function were non-constant, say a parametric function of the spatial coordinates and/or covariates, then this could be readily accommodated by adjusting how the residuals above are calculated. Different variations on the bootstrap procedure itself are also available (e.g., Castillo-Páez et al 2019) whose use in the GPrC algorithm will be explored in subsequent work.…”

Section: Spatial Datamentioning

confidence: 99%

Calibrating generalized predictive distributions

Martin

2021

Preprint

View full text Add to dashboard Cite

In prediction problems, it is common to model the data-generating process and then use a model-based procedure, such as a Bayesian predictive distribution, to quantify uncertainty about the next observation. However, if the posited model is misspecified, then its predictions may not be calibrated-that is, the predictive distribution's quantiles may not be nominal frequentist prediction upper limits, even asymptotically. Rather than abandoning the comfort of a model-based formulation for a more complicated non-model-based approach, here we propose a strategy in which the data itself helps determine if the assumed model-based solution should be adjusted to account for model misspecification. This is achieved through a generalized Bayes formulation where a learning rate parameter is tuned, via the proposed generalized predictive calibration (GPrC) algorithm, to make the predictive distribution calibrated, even under model misspecification. Extensive numerical experiments are presented, under a variety of settings, demonstrating the proposed GPrC algorithm's validity, efficiency, and robustness.

show abstract

“…The CNPB procedure is a modification of the previous NPB approach, considering a procedure to correct the resulting bias in the nonparametric estimator of the variogram. In the geostatistical framework, more accurate results have been obtained using this technique (Castillo-Páez et al, 2019). Specifically, the following adjustments are performed in the previous generic bootstrap algorithm.…”

Section: Corrected Nonparametric Residual Bootstrap (Cnpb)mentioning

confidence: 99%

“…However, no matter the method used to remove the trend, either parametric or nonparametric, the direct use of residuals gives rise to biased variogram estimates, especially at large lags (see Cressie, 1993, Section 3.4.3). To solve this problem, the CNPB approach is a modification of the NPB method, but including a bias-corrected algorithm for the dependence estimation (see Fernández-Casal and Francisco-Fernández, 2014;Castillo-Páez et al, 2019).…”

Section: Introductionmentioning

confidence: 99%

A computational validation for nonparametric assessment of spatial trends

Meilán-Vila¹,

Fernández-Casal²

2020

Preprint

Self Cite

View full text Add to dashboard Cite

The analysis of continuously spatially varying processes usually considers two sources of variation, namely, the large-scale variation collected by the trend of the process, and the small-scale variation. Parametric trend models on latitude and longitude are easy to fit and to interpret. However, the use of simple parametric models for characterizing spatially varying processes may lead to misspecification problems if the model is not appropriate. Recently, Meilán-Vila et al. ( 2019) proposed a goodness-offit test based on an L 2 -distance for assessing a parametric trend model with correlated errors, under random design, comparing a parametric and a nonparametric trend estimators. The present work aims to provide a detailed computational analysis of the behavior of this approach using different bootstrap algorithms for calibration, under a fixed-design geostatistical framework. Asymptotic results for the test are provided and an extensive simulation study, considering complexities that usually arise in geostatistics, is carried out to illustrate the performance of the proposal.

show abstract

A nonparametric bootstrap method for spatial data

Cited by 10 publications

References 29 publications

Small‐scale spatial variability of Technosol properties in a chronosequence of reclamation of dredged‐sediment landfills

Small‐scale spatial variability of Technosol properties in a chronosequence of reclamation of dredged‐sediment landfills

Calibrating generalized predictive distributions

A computational validation for nonparametric assessment of spatial trends

Contact Info

Product

Resources

About