This is a recap of the Envibayes Workshop on Complex Environmental Data.

The purpose of this recap is to consolidate notes, with a focus on ideas I think might have potential for personal research pursuits.

The recap is ordered by talk, though some notes are only tangentially related to the talk under which they appear.

Veronica Berrocal Stationarity or non-stationarity: that is the question

  • Mark Risser’s work, especially on non-stationary GP kernals for massive datasets

  • four broad methods of nonstationary GP’s

    1. covariates in covariance matrix
    2. covariate-informed local partitioning
    3. identification of local stationarity regions
    4. new class of prior (eg. CUSP-MRA prior)
  • joint Bayesian data fusion - eg. Nonstationary spatiotemporal Bayesian data fusion for pollutants in the near-road environment ( Gilana 2019)

  • regression-based Bayesian data fusion - eg. Model evaluation and spatial interpolation by Bayesian combination of observations with outputs from numerical models ( Fuentes 2005 )

  • covariate-driven segmentation - eg. Nonstationary spatial prediction of soil organic carbon: Implications for stock assessment decision making ( Risser 2019)

  • fast approximation methods for fully Bayesian ensembles of different spatial prediction methods

Andee Kaplan Improving Bayesian inference for streaming data

  • Making Recursive Bayesian Inference Accessible ( Hooten 2021)
  • Generative Filtering for Recursive Bayesian Inference with Streaming Data ( Taylor 2023
  • Sequential Markov Chain Monte Carlo ( Yang 2013

Maryclare Griffin Log-Gaussian Cox Process Modeling of Large Spatial Lightning Data using Spectral and Laplace Approximations

  • Vecchia-Laplace approximations of generalized Gaussian processes for big non-Gaussian spatial data ( Zilber 2019)

  • Circular embeddings are highly efficient

Student Papers

  • A dynamic spatial filtering approach to mitigate underestimation bias in field calibrated low-cost sensor air-pollution data ( Heffernan) -pm2.5 downscaler inverse regression
  • Dynamic Population Models with Temporal Preferential Sampling to Infer Phenology ( Schwob 2022
  • New spatial models for integrating standardized detection-nondetection and opportunistic presence-only data: application to estimating risk factors associated to powerline-induced death of birds Sicacha-Parada 2023

Amy Herring Low Rank Longitudinal Factor Regression

  • connections to brain imaging?

Henry Scharf Predicting fine-scale taxonomic variation in landscape vegetation using large satellite imagery data sets

Maggie Johnson Tracking plant stress from space: Improving estimates of evapotranspiration through spatiotemporal data fusion

  • Multisensor Fusion of Remotely Sensed Vegetation Indices Using Space-Time Dynamic Linear Models ( Johnson 2021)

Yawen Guan A Bayesian Hierarchical Approach for Modeling Tree Cover Change

Cory Zigler Bayesian Causal Inference with Uncertain Physical Process Interference

  • Solving the SPDE - A Mechanistic Model of Annual Sulfate Concentrations in the United States ( Wikle 2022)

Amy Braverman Statistical Challenges for the Next Generation of NASA’s Earth Observing Satellites

  • Post hoc Uncertainty Quantification for Remote Sensing Observing Systems Braverman 2021
  • how to preserve local information on a global scale?

Robert Gramacy Contour Location for Reliability in Airfoil Simulation Experiments using Deep Gaussian Processes

  • precursor: Bayesian inference for non-stationary spatial covariance structure via spatial deformations Schmidt 2003
  • layers of Gaussians: Elliptical slice sampling Murray 2010
  • Vechia approx: Vecchia-Approximated Deep Gaussian Processes for Computer Experiments Sauer 2022

Abhirup Datta Combining machine learning with Gaussian processes for geospatial data

  • machine learning for nonlinear geospatial analysis

    1. residual kriggin (Olsen 2020)
    2. adding spatial covariates (Wang 2019)
    3. Embed ML in spatial GLMM (this is best)
  • Random Forests for Spatially Dependent Data Saha 2021

  • IDEA: uncertainty quantification for graphical neural nets, get CI’s for mean, stoch gradient MCMC

Matt Koslovsky A Bayesian model for measurement error in multinomial data

  • zero inflated dirichlet model
  • data augmentation: normalized random measures
  • Bayesian Inference for Logistic Models Using Pólya–Gamma Latent Variables ( Polson 2013
  • A Bayesian zero-inflated Dirichlet-multinomial regression model for multivariate compositional count data ( Koslovsky 2023)

Alexandra Schmidt Temporal misalignment in geostatistical

  • spatiotemporal methods in env. epi

Ephraim Hanks A Mixture of OU-processes Framework for Jointly Modeling Animal Movement and Species Distribution Data

Toryn Schafer Inverse reinforcement learning for animal behavior in the environment

  • RL <–> ABM <–> LMDP: Bayesian inverse reinforcement learning for collective animal movement Schafer


  • Calibration identifiability issue - Bayesian calibration of computer models ( Kennedy 2002)
  • Dunson