Abstract:Paleoclimatology -- the study of past climate -- is relevant beyond climate science itself, such as in archaeology and anthropology for understanding past human dispersal. Information about the Earth's paleoclimate comes from simulations of physical and biogeochemical processes and from proxy records found in naturally occurring archives. Climate-field reconstructions (CFRs) combine these data into a statistical spatial or spatiotemporal model. To date, there exists no consensus spatiotemporal paleoclimate model that is continuous in space and time, produces predictions with uncertainty, and can include data from various sources. A Gaussian process (GP) model would have these desired properties; however, GPs scale unfavorably with data of the magnitude typical for building CFRs. We propose to build on recent advances in sparse spatiotemporal GPs that reduce the computational burden by combining variational methods based on inducing variables with the state-space formulation of GPs. We successfully employ such a doubly sparse GP to construct a probabilistic model of European paleoclimate from the Last Glacial Maximum (LGM) to the mid-Holocene (MH) that synthesizes paleoclimate simulations and fossilized pollen proxy data.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to construct a spatiotemporally continuous paleoclimate model with uncertainty that can synthesize paleoclimate simulations and fossil pollen proxy data. Specifically, the research aims to address the following issues: 1. **Lack of a Consensus Model**: Currently, there is no globally - spatiotemporally continuous, probabilistic paleoclimate model that can combine most of the available proxy data and simulation data and can be queried at any location and time. 2. **Computational Complexity**: Traditional Gaussian Process (GP) models have high computational complexity when dealing with large - scale data and are difficult to apply to actual paleoclimate reconstructions. To solve these problems, the authors propose a method based on doubly sparse Gaussian processes. This method reduces the computational burden by introducing inducing variables and Markov state - space representations. This method enables the model to perform spatiotemporal modeling of the annual average temperature in Europe from the Last Glacial Maximum (LGM; approximately 21,000 years ago) to the mid - Holocene (MH; approximately 6,000 years ago). ### Main Contributions - **Application of Sparse Gaussian Processes**: Using sparse variational Gaussian processes combined with Markov structures significantly reduces the computational complexity, making it suitable for large - scale paleoclimate datasets. - **Integration of Multiple Data Sources**: Combining paleoclimate simulations and fossil pollen proxy data to construct a spatiotemporally continuous probability model. - **Uncertainty Estimation**: The model not only provides predicted values but also comes with calibrated uncertainty estimates, enhancing the reliability of the results. ### Model Validation The authors verified the effectiveness of the model through the "leave - one - time - slice - out" method and showed the prediction errors and confidence intervals of the model in different time periods. The experimental results show that the model performs well in terms of prediction accuracy and uncertainty estimation. ### Future Prospects The authors plan to further improve the model, including: - Using measurement data from the early 1900s for spatial interpolation instead of relying on weighted - interpolation simulation data. - Introducing non - zero - parameter prior mean functions to capture the remaining large - scale spatiotemporal trends. - Using independent likelihood functions for different data sources (such as proxy data and simulation data). - Considering the dating uncertainty of proxy data (such as pollen). In summary, this research lays the foundation for constructing a global paleoclimate consensus model and provides an effective framework that can be applied in fields such as archaeology and paleoecology.

Spatiotemporal modeling of European paleoclimate using doubly sparse Gaussian processes

Active data based Gaussian process models for nonlinear spatiotemporal systems

Towards variance-conserving reconstructions of climate indices with Gaussian process regression in an embedding space

Nonstationary Spatial Modeling of Massive Global Satellite Data

Mixed-Stationary Gaussian Process for Flexible Non-Stationary Modeling of Spatial Outcomes

Gaussian Process Regression for Arctic Coastal Erosion Forecasting

ProSpar-GP: scalable Gaussian process modeling with massive non-stationary datasets

Warped multifidelity Gaussian processes for data fusion of skewed environmental data

Towards Efficient Modeling and Inference in Multi-Dimensional Gaussian Process State-Space Models

Modelling Spatial Compositional Data: Reconstructions of past land cover and uncertainties

Physics-Informed Variational State-Space Gaussian Processes

Highly Scalable Bayesian Geostatistical Modeling via Meshed Gaussian Processes on Partitioned Domains

Modeling massive highly-multivariate nonstationary spatial data with the basis graphical lasso

Spatio-temporal Functional Regression on Paleo-ecological Data

Spatio-Temporal Reconstructions of Global CO2-Fluxes using Gaussian Markov Random Fields

Light curve completion and forecasting using fast and scalable Gaussian processes (MuyGPs)

On Bayesian Modelling of the Uncertainties in Palaeoclimate Reconstruction

Environmental Sensor Placement with Convolutional Gaussian Neural Processes

Infinite-Horizon Gaussian Processes

Parallel cross-validation: A scalable fitting method for Gaussian process models

Predicting paleoclimate from compositional data using multivariate Gaussian process inverse prediction