The reliability of single precision computations in the simulation of deep soil heat diffusion in a land surface model
Bibliographic record
Abstract
Climate models need discretized numerical algorithms and finite precision arithmetic to solve their differential equations. Most efforts to date have focused on reducing truncation errors due to discretization effects, whereas rounding errors due to the use of floating-point arithmetic have received little attention. However, there are increasing concerns about more frequent occurrences of rounding errors in larger parallel computing platforms (due to the conflicting needs of stability and accuracy vs. performance), and while this has not been the norm in climate and forecast models using double precision, this could change with some models that are now compiled with single precision, which raises questions about the validity of using such low precision in climate applications. For example, processes occurring over large time scales such as permafrost thawing are potentially more vulnerable to this issue. In this study we analyze the theoretical and experimental effects of using single and double precision on simulated deep soil temperature from the Canadian LAnd Surface Scheme (CLASS), a state-of-the-art land surface model. We found that reliable single precision temperatures are limited to depths of less than about 20–25 m while double precision shows no loss of accuracy to depths of at least several hundred meters. We also found that, for a given precision level, model accuracy deteriorates when using smaller time steps, further reducing the usefulness of single precision. There is thus a clear danger of using single precision in some climate model applications, in particular any scientifically meaningful study of deep soil permafrost must at least use double precision. In addition, climate modelling teams might well benefit from paying more attention to numerical precision and roundoff issues to offset the potentially more frequent numerical anomalies in future large-scale parallel climate applications.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".