the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 16 Apr 2026
Using artificial intelligence to monitor live fuel moisture content across France, based on a high-resolution land surface analysis
Abstract. Live Fuel Moisture Content (LFMC) is one of the most critical variables for understanding fire dynamics, particularly within forest environments. This study intends to develop a daily LFMC indicator to support operational fire danger management services in France. The product is based on in situ observations from the French National Forest Office and is generated using a lightweight expressive neural network model. The network has been designed to generalise well over time and space. It can be integrated directly into land surface models to enable real-time monitoring of vegetation’s hydric status. The modelling framework combines outputs from a physically based land surface model and satellite-derived leaf area index (LAI) observations, providing high-resolution, spatially consistent estimates of land surface over France. To evaluate the model’s generalisation capacity, we implemented complementary cross-validation strategies to test interannual robustness, spatial transferability, and to simulate an operational deployment scenario. Additionally, we performed a robustness analysis to quantify the sensitivity of predictions to training variability. The results demonstrate a strong ability to estimate the range and dynamics of LFMC across most of France. They also identify regions where additional in situ sampling or improved representation could reduce epistemic uncertainty and enhance the reliability of the model.
- Preprint
(1950 KB) - Metadata XML
-
Supplement
(1145 KB) - BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2026-1247', Krishna Rao, 20 Jul 2026
-
RC2: 'Comment on egusphere-2026-1247', Anonymous Referee #2, 25 Oct 2026
This study aims to develop a daily Live Fuel Moisture Content (LFMC) indicator to support operational fire-danger management services in France. The study investigates the extent to which daily LFMC estimates, derived through an emulator of a Météo-France land-surface model, to provide information for fire-management applications. The topic is relevant and potentially useful for operational wildfire-risk assessment. However, several issues should be addressed before the publication listed below.
This study aims to develop a daily Live Fuel Moisture Content (LFMC) indicator to support operational fire-danger management services in France. Specifically, the study investigates the extent to which daily LFMC estimates, derived using an emulator of a Météo-France land-surface model, can provide useful information for fire-management applications. The topic is relevant and potentially valuable for operational wildfire-risk assessment. However, several issues listed below should be addressed before published.Major Comments
- Table 2 lists ten predictors selected for the model, but it is not clear how vegetation type or species composition is represented. In this study, LFMC measurements from multiple plant species at the same observation site appear to be averaged to represent shrub-layer LFMC. However, without information on the relative abundance, fractional coverage, or biomass contribution of each species, it is difficult to determine whether a simple arithmetic mean adequately represents the actual moisture condition of the shrub fuel layer. This issue may be particularly important when species at the same site exhibit substantially different seasonal or drought-response characteristics. In such cases, an unweighted species average may not accurately represent the LFMC of the actual fuel mixture at the site. The authors should clarify how vegetation type and species composition are represented in the model and discuss this as an potential source of uncertainty.
- The manuscript repeatedly refers to modeled LFMC values as “predictions.” However, it is not clear whether these values are truly predictions in the forecasting sense. Prediction generally implies estimation of a future state using information available prior to that time. In the present study, the LFMC appears instead to be diagnosed or estimated from land-surface model variables and other environmental predictors. If future meteorological forcing is not used, the term “LFMC estimation” would be more appropriate than “LFMC prediction.”
- Line 332: the manuscript states “The drop in NSE for the FC in 2020 is due to the first substantial expansion of observation sites beyond the Mediterranean basin, which occurred in 2020.” If the spatial expansion of the observational network has an important impact on model evaluation, this change should also be reflected in Figure 1.
- Section 3.2: Section 3.2 compares LFMC maps for individual dates, including the date corresponding to the lowest LFMC in 2022, with the same calendar date in another year. However, comparison of two individual days may not be sufficient to demonstrate a robust climatological or interannual spatial gradient. A single-day LFMC field can be strongly influenced by short-term meteorological variability. If the authors intend to demonstrate persistent spatial differences or anomalous LFMC conditions in 2022 relative to 2023, monthly or weekly mean fields would provide a more robust comparison. In addition, overlaying observed wildfire locations or burned areas on these maps could provide useful context regarding the potential relevance of the LFMC product to fire-danger applications. This point is further discussed in Comment 6 below.
- Section 3.3: The manuscript attributes the relatively large uncertainty primarily to the limited representation of temperate deciduous broadleaf forest within the observational dataset. While this explanation is plausible, it may not be the only contributing factor. Other sources of uncertainty could include complex terrain, spatial heterogeneity in soil moisture, uncertainties in meteorological forcing, differences in vegetation species composition, limitations in the land-surface model’s representation of plant water stress, and etc. Without an independent analysis that separates or evaluates these effects, it is difficult to attribute the poor model performance mainly to the underrepresentation of a specific forest type. The authors should discuss other plausible sources of uncertainty.
- Section 4.2: LFMC is an important component of wildfire danger because it influences fire behavior. However, LFMC alone does not determine whether a wildfire will occur. Ignition is a separate and essential component by natural or anthropogenic sources. Even under extremely low-LFMC conditions, a wildfire will not occur without an ignition source. Therefore, LFMC should be interpreted as an indicator of the potential for successful ignition and subsequent fire spread, rather than as a direct predictor of fire occurrence at a specific location and time. For this reason, correspondence between simulated low LFMC and a single observed fire event is insufficient to demonstrate consistency with real fire occurrence. A more appropriate evaluation would include a larger sample of fire cases and assess whether observed fire occurrence, fire spread, or burned area systematically corresponds to low-LFMC conditions identified by the model.
Minor Comments
- Table 1: The sum of the training and validation samples does not appear to match the reported total number of observations. Please clarify the reason.
- Line 101: “LMFCdry from LMFCfresh” should be corrected to: “LFMCdry from LFMCfresh”
Citation: https://doi.org/10.5194/egusphere-2026-1247-RC2
Data sets
Supplementary data: Using artificial intelligence to monitor live fuel moisture content across France, based on a high-resolution land surface analysis Yann Baehr, Pierre Vanderbecken, Bertrand Bonan, Catherine Robert, Mathieu Regimbeau, François Pimont, Kevyn Raynal, Xiangzhuo Liu, Remi Savazzi, Moncef Garouani, Josiane Mothe, Nemesio Rodriguez-Fernandez, Lionel Jarlan, and Jean-Christophe Calvet https://zenodo.org/records/18850161
Model code and software
Supplementary data: Using artificial intelligence to monitor live fuel moisture content across France, based on a high-resolution land surface analysis Yann Baehr, Pierre Vanderbecken, Bertrand Bonan, Catherine Robert, Mathieu Regimbeau, François Pimont, Kevyn Raynal, Xiangzhuo Liu, Remi Savazzi, Moncef Garouani, Josiane Mothe, Nemesio Rodriguez-Fernandez, Lionel Jarlan, and Jean-Christophe Calvet https://zenodo.org/records/18850161
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 437 | 142 | 51 | 630 | 77 | 52 | 42 |
- HTML: 437
- PDF: 142
- XML: 51
- Total: 630
- Supplement: 77
- BibTeX: 52
- EndNote: 42
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Understanding fire activity in Mediterranean ecosystems is paramount, especially under a changing climate that has driven recent increases in fire activity and associated hazards. Live Fuel Moisture Content (LFMC) is a critical indicator of fire danger, yet it remains poorly understood in spatial detail due to limited observational data and a lack of continuous, wall to wall maps. In this context, this paper is a very welcome contribution to the literature.
I view this manuscript as a "3 in 1" contribution: the authors not only produce daily LFMC estimates across France and adjacent regions, but they also provide actionable insights for selecting future sampling sites and evaluate the relationship between estimated LFMC and regional fire outbreaks at the canton level. The manuscript is well written, and the discussions regarding the physical mechanisms linking leaf area index modulated by soil water capacity are insightful. I commend the authors for extending their work beyond map generation to actively inform future sampling strategies.
However, I have several major questions and suggestions regarding methodological details that currently hinder a full evaluation of the findings. Addressing these points, I think, will significantly strengthen the clarity and impact of the paper.
1. Cross Species LFMC Averaging
While I recognize the constraints imposed by available data on species distribution, it is unclear why LFMC was averaged across species, especially given the authors' own observation that "the species are too varied" (L. 90). Extensive literature demonstrates significant inter species variation in LFMC responses under identical climate conditions (e.g., Martin StPaul et al., 2018; Pimont et al., 2019; Yebra et al., 2024).
2. Physical Mechanism Linking LAI/WFC to LFMC
Table 2 shows that the top feature linked to LAI is LAI/WFC (LAI normalized by soil water field capacity), with Section 4.6 providing details of a plausible mechanism where leaf growth under limited soil capacity leads to competing forces on plant water status, thus lowering live fuel moisture. While this mechanism is plausible for certain conditions, does it hold universally across different plant functional types?
I am not familiar with the shrub types in France, so feel free to point me (and the manuscript's readers) to existing literature if such variability does not exist in France.
3. Derivation of the "Interest Score"
The "Interest Score" concept presented in Figure 10 is creative and practical for future sampling strategy design. However, the methodology behind it is insufficiently detailed.
4. Linking LFMC to Ignitions (Figure 9)
Figure 9 presents an intriguing relationship between simulated LFMC and ignition timing, but several critical details and controls are missing:
As I write this, I acknowledge that this depth of analysis for an experiment that is not the central focus of the paper might be considered too much. But I think half-results are more dangerous that no results. Since there is already plenty of other contributions in the paper, I leave it to the authors to decide whether they include full details of the analysis and make the experiment more robust or leave it out from the paper and follow up with a future study.
Minor Comments
Krishna Rao
References