| 19 Nov 2025
A daily gridded dataset of the Fire Weather Index across Canada, with calculations based on the sun’s elevation
Abstract. This article summarizes the work carried out at the ESCER (Étude et Simulation du Climat à l’Échelle Régionale, UQAM) Centre to develop the ERA5-FWI-SN dataset — a new automated, gridded dataset of the Canadian Forest Fire Weather Index (FWI) System components for Canada, at a spatial resolution of approximately 31 km. The ERA5-FWI-SN dataset is derived from the hourly ERA5 reanalysis using a new method, called the solar noon method, which is based on the sun's elevation above the local horizon at noon. This method aims to improve the conventional method for calculating the FWI System components, the UTC method, based on the maximum insolation defined via time zones (or specific areas covered by one Coordinated Universal Time zone, i.e. UTC zone), which can cover a wide longitudinal region where the same time prevails. The classical method relies on the principle that the average solar time over the territory covered by the same time zone is not too far from legal time (i.e., solar noon is not too far from legal noon), which can be problematic in regions that are very extensive in longitude, such as in Canada where only six time zones have been defined over more than 95° of longitudes. The solar noon method also allows for the correction of systematic biases associated with the UTC method, particularly those arising in gridded datasets near time zone boundaries or across the east–west extent of a single time zone. The dataset spans from 1950 to the present and is updated daily through an automation process that allows the calculation of the FWI components with a six-day lag, corresponding to the latest ERA5 available reanalysis data for download, resulting in a rapid (i.e., 6-days lag) monitoring of recent wildfire danger throughout Canada. Data were compared between solar noon and UTC methods and were tested against the Canadian Wildland Fire Information System (CWFIS) FWI calculated from observation station data. The dataset is available for download at https://doi.org/10.5683/SP3/4B18XZ, and derived visualization products can be accessed across the web platform http://feux.escer.uqam.ca/.
This article will be of interest to those working with CFFDRS components. The article content supports the data set publication. The data set is useful and of high quality. The data set publication, as submitted, is of high quality. My recommendation is publication with minor revision to the manuscript. Specific comments follow:
Line 104: Perhaps add a reference to the validation of ERA5 or say something about the confidence of using the dataset. It makes sense to use this dataset for the purpose presented here, but are there modeling issues especially with hourly precipitation that could affect the study results? In other words, what is the confidence in using this dataset and perhaps why was it chosen versus other potential datasets. In the discussion section you do make a reference to extreme-value biases in ERA5.
Line 239: Do you think that there were any issues for this direct grid point comparison in complex terrain?
Line 258: 02:00AM is a local time for which time zone?
Line 375: Do you have a sense of how different these records would look without the SN methodology?
Can you say something a bit more specific regarding management implications of the SN improvement?