Emulation of high-resolution land surface models using sparse Gaussian processes with application to JULES

Abstract

Abstract. Land surface models are typically integrated into global climate projections, but as their spatial resolution increases the prospect of using them to aid in local policy decisions becomes more appealing. If these complex models are to be used to make local decisions, then a full quantification of uncertainty is necessary, but the computational cost of running just one full simulation at high resolution can hinder proper analysis. Statistical emulation is an increasingly common technique for developing fast approximate models in a way that maintains accuracy but also provides comprehensive uncertainty bounds for the approximation. In this work, we developed a statistical emulation framework for land surface models, enabling fast predictions at a high resolution. To do so, our emulation framework acknowledges, and makes use of, the multitude of contextual data that are often fed into land surface models (sometimes called forcing data, or driving data), such as air temperature or various soil properties. We use The Joint UK Land Environment Simulator (JULES) as a case study for this methodology, and perform initial sensitivity analysis and parameter tuning to showcase its capabilities. The JULES is perhaps one of the most complex land surface models and so our success here suggests incredible gains can be made for all types of land surface model.

Microclimate, an important part of ecology and biogeography

“…The fusion of statistical and mechanistic approaches to model microclimates shows promise for developing mechanistically informed and computationally efficient methods. The application of statistical model emulation techniques that reproduce the behaviour of more complex models using techniques routinely adopted in other areas of climate modelling could significantly reduce computational run times (Baker et al., 2022). Further implementation requires a break‐down of traditional barriers between disciplines as far apart as ecology, meteorology and computer science (Briscoe et al., 2023).…”

Comment on egusphere-2024-978

“…This ability to capture more variability than the other models, which is closer to the observed variability, can potentially improve the represented variability in CLM5PFT if the suitable variation can be modeled at the right time and location. This spatiotemporal discrepancy of simulated and observed GPP and ET variability could potentially be solved with optimized PFT parameters (Baker et al, 2022;Birch et al, 2021;Cheng et al, 2021;Dagon et al, 2020;Deng et al, 2021;Fisher et al, 2019b).…”

Constraining the carbon cycle in JULES-ES-1.0

“…To combat this explosion of dimensions and effort, we chose the pragmatic option to perturb each of the PFTs together, by a multiplication factor for each parameter. This choice was computationally convenient for a top-down, globally averaged experiment, but we see great potential for optimising the values of these parameters for individual PFTs in further work (a good example is Baker et al (2022)). Many of the "multiplication factors" varied the parameter range between a half (0.5) and double (2) the parameter standard value.…”