A 32-year species-specific live fuel moisture content dataset for southern California chaparral
Data files
Jan 12, 2026 version files 2.43 GB
-
lfmc_observations.csv
780.99 KB
-
README.md
3.37 KB
-
sba_lfmc_1987-2019.nc
2.43 GB
Abstract
Live fuel moisture content (LFMC) strongly affects the behavior of wildland fire, resulting in its incorporation into wildfire spread models and danger ratings. In this dataset, over ten thousand LFMC observations were combined with predictor variables from Landsat imagery and the Weather Research and Forecasting model to train species-specific random forest models that predict the LFMC of four fuel types—chamise, old growth chamise, black sage, and bigpod ceanothus. These models are then utilized to create a historical, 32-year long, LFMC dataset in southern California chaparral. Additionally, the high spatial and temporal sampling frequency of chamise allowed for quantile mapping bias correction to be applied. The final chamise output, which is the most robust, has a mean absolute error of 9.68% and an R2 value of 0.76. The LFMC dataset successfully captures the variability in the annual cycle, the spatial heterogeneity, and the interspecies differences, which makes it applicable for better understanding varying fire season characteristics and landscape level flammability.
Dataset DOI: 10.5061/dryad.rjdfn2zkw
Description of the data and file structure
The lfmc_observations.csv file contains all of the live fuel moisture content (LFMC) observations that were used to train species-specific, LFMC predictive random forest models. The studied species include new and old growth of chamise (Adenostoma fasciculatum), new growth of black sage (Salvia mellifera), and new growth of bigpod ceanothus (Ceanothus megacarpus).
The sba_lfmc_1987-2019.nc file contains the LFMC predictions created with the four random forest models. The predictions cover a spatial domain that stretches from San Luis Obispo, CA, to Los Angeles, CA, with 1km, semi-monthly (1st and 15th) LFMC outputs from December 1987 through June 2019. The netCDF file variables include each fuel type, as well as uncertainty calculations for each fuel type. The chamise predictions also underwent quantile mapping bias correction. The chamise outputs without the bias correction are also included.
Files and variables
File: lfmc_observations.csv
Description: LFMC observations used for training random forest models.
Variables
- site: Observation sampling site
- date: Date sample collected
- latitude: Degrees north
- longitude: Degrees east
- fuel: Fuel type
- percent: Live fuel moisture content percent (water content / dry content * 100)
File: sba_lfmc_1987-2019.nc
Description: LFMC predictions of four fuel types: new growth chamise, old growth chamise, new growth black sage, new growth bigpod ceanothus, and their corresponding uncertainty
- chamise: LFMC predictions of new growth chamise with quantile mapping bias correction applied (percent)
- chamise_uncertainty: uncertainty calculated from the variance of the decision trees (percent)
- chamise_old_growth: LFMC predictions of old growth chamise (percent)
- chamise_old_growth_uncertainty: uncertainty calculated from the variance of the decision trees (percent)
- sage_black: LFMC predictions of new growth black sage (percent)
- sage_black_uncertainty: uncertainty calculated from the variance of the decision trees (percent)
- ceanothus_bigpod: LFMC predictions of new growth bigpod ceanothus (percent)
- ceanothus_bigpod_uncertainty: uncertainty calculated from the variance of the decision trees (percent)
- chamise_no_bias_correction: LFMC predictions of new growth chamise without quantile mapping bias correction applied (percent)
*NaN values correspond to grid cells where model predictor data was not available.
Code/software
No software is needed to view the data. The codebase used to create the data is found here:
https://github.com/kcvarga7/sba_lfm_1987-2019
Access information
Data was derived from the following sources:
- National Fuel Moisture Database (https://github.com/wmjolly/pyNFMD)
- Santa Barbara County Fire Department (https://sbcfire.com/wildfire-predictive-services/)
- Landsat (NASA)
- Santa Barbara Area WRF Climatology (https://clivac.eri.ucsb.edu/clivac/SBCWRFD/index.html)
LFMC observations were acquired as the model predictands. Then, predictors were calculated, including long-term lag variables, such as 90-day precipitation, 90-day mean temperature, and 150-day mean insolation, as well as short-term lag or instantaneous predictors, such as day length, 7-day mean soil moisture, and near-infrared reflectance of vegetation (NIRv). Species-specific random forest models were then trained and tested with two techniques: 5-fold cross validation using all observations and observation site specific cross validation, where each site was individually withheld as the test data. After testing, models were trained on all predictors/predictands and used to create the species-specific LFMC datasets. Lastly, quantile mapping bias correction was applied to the modeled chamise LFMC dataset, which was possible due to the high spatial frequency of observation sites.