arXiv Science⌕ Search

arXiv subjects

Julien Emile-Geay

Publications and source records attributed to Julien Emile-Geay.

4 recordsLinked to original sources

Coupled multiscale paleoclimate reconstruction with four-dimensional variational data assimilation

Paleoclimate archives extend climate knowledge beyond the instrumental era, registering different seasons, variables, time averages, and memory lengths. A longstanding problem is to integrate these heterogeneous sources of information within a unified methodology. Here we present a new data-assimilation framework, Last Millennium Reanalysis 4D-Var (LMR4D-Var), which reconstructs climate trajectories from these heterogeneous datasets while balancing errors in the model, observations, and initial conditions. We compare results using LMR4D-Var to assimilate proxies from PAGES2k, Temp12k, and borehole temperature profiles without treating them as instantaneous equivalents. Instrumental validation demonstrates that LMR4D-Var achieves the highest skill in reconstructing global mean temperature among existing reconstructions. Borehole assimilation preserves skill against withheld annually resolved records, increases agreement between reconstructed 300--2000-m ocean heat content (OHC) and independent estimates, and yields a cooler reconstructed Little Ice Age ocean. Modern 130-year OHC trends in the 0--300-m and 300--2000-m layers exceed their respective reconstructed pre-1870 Common Era trend distributions. Results for Temp12k demonstrate assimilation of decadal-to-millennial records and the potential for Holocene and deeper-time applications with suitable emulators.

physics.ao-ph↗

Imputing Missing Near-Surface Air Temperature using Graph Attention Networks with Distance-Aware Edge Features

We introduce a graph attention network for reconstructing near-surface air temperature fields from incomplete observations. The model represents each grid cell as a node connected to its $k$ nearest neighbors, with attention weights that depend on node features, distance, and land-ocean transitions. This graph representation avoids the polar distortions associated with treating the globe as a flat latitude-longitude image, as most deep learning reconstruction methods do. This is particularly relevant for historical temperature records, which extend back to the mid-nineteenth century but remain incomplete, with the largest gaps early in the record and at high latitudes. We train our model on CMIP6 CESM2 historical simulations masked with the time-varying HadCRUT5 observational coverage from 1850 to 2014. We then compare our approach with ordinary kriging, GraphEM, and the LaMa inpainting network on a held-out realization. The graph model achieves the lowest spatial RMSE (0.997 K), 10.1% lower than LaMa and approximately 31% lower than kriging and GraphEM. The largest gains occur in the Arctic and Antarctic, while LaMa remains slightly more accurate in the densely observed tropics. The graph model's error also falls faster as coverage increases, so its advantage over LaMa grows by more than a third from the 1850s to the 2000s. An ablation study shows that the model's skill comes mainly from the geometric graph construction rather than from auxiliary physical covariates.

cs.LG↗

Efficient Reconstructions of Common Era Climate via Integrated Nested Laplace Approximations

Paleoclimate reconstruction on the Common Era (1-2000AD) provide critical context for recent warming trends. This work leverages integrated nested Laplace approximations (INLA) to conduct inference under a Bayesian hierarchical model using data from three sources: a state-of-the-art prox database (PAGES 2k), surface temperature observations (HadCRUT4), and latest estimates of external forcings. INLA's computational efficiency allows to explore several model formulations (with or without forcings, explicitly modeling internal variability or not), as well as five data reduction techniques. Two different validation exercises find a small impact of data reduction choices, but a large impact for model choice, with best results for the two models that incorporate external forcings. These models confirm that man-made greenhouse gas emissions are the largest contributor to temperature variability over the Common Era, followed by volcanic forcing. Solar effects are indistinguishable from zero. INLA provide an efficient way to estimate the posterior mean, comparable with the much costlier Monte Carlo Markov Chain procedure, but with wider uncertainty bounds. We recommend using it for exploration of model designs, but full MCMC solutions should be used for proper uncertainty quantification.

stat.AP↗

Statistical paleoclimate reconstructions via Markov random fields

Understanding centennial scale climate variability requires data sets that are accurate, long, continuous and of broad spatial coverage. Since instrumental measurements are generally only available after 1850, temperature fields must be reconstructed using paleoclimate archives, known as proxies. Various climate field reconstructions (CFR) methods have been proposed to relate past temperature to such proxy networks. In this work, we propose a new CFR method, called GraphEM, based on Gaussian Markov random fields embedded within an EM algorithm. Gaussian Markov random fields provide a natural and flexible framework for modeling high-dimensional spatial fields. At the same time, they provide the parameter reduction necessary for obtaining precise and well-conditioned estimates of the covariance structure, even in the sample-starved setting common in paleoclimate applications. In this paper, we propose and compare the performance of different methods to estimate the graphical structure of climate fields, and demonstrate how the GraphEM algorithm can be used to reconstruct past climate variations. The performance of GraphEM is compared to the widely used CFR method RegEM with regularization via truncated total least squares, using synthetic data. Our results show that GraphEM can yield significant improvements, with uniform gains over space, and far better risk properties. We demonstrate that the spatial structure of temperature fields can be well estimated by graphs where each neighbor is only connected to a few geographically close neighbors, and that the increase in performance is directly related to recovering the underlying sparsity in the covariance of the spatial field. Our work demonstrates how significant improvements can be made in climate reconstruction methods by better modeling the covariance structure of the climate field.

stat.AP↗