Integration of MODIS NDVI data into Tuscany meteorological fire risk bulletin

The Regional Government of Tuscany (Central Italy), together with the LaMMA Consortium and the National Research Council – Institute for BioEconomy (CNR IBE), have implemented an operational workflow and a Spatial Decision Support System (SDSS) which provide a daily wildfire danger rating for each municipality in the region. The system is driven by the Canadian Forest Fire Weather Index (FWI), the components of which are calculated from in-situ observations and numerical weather predictions to quantify the predisposition of forest fuels to ignition and spread. During the peak risk season (approximately from June to September), a dedicated Bulletin is produced daily, containing FWI metrics, fire-weather indices and additional atmospheric parameters over the administrative areas (municipalities) of Tuscany Region. The AIB on-duty operator uses this information to provide operational guidance on anticipated fires, their behaviors, and regional hotspots requiring heightened readiness. Because NDVI reflects vegetation stress and fuel condition, this work explores the integration of vegetation status information from MODIS NDVI images into the bulletin. Fire statistics for Tuscany (number of fires) over 25 years (2000–2024) are therefore analyzed together with MOD13Q1 NDVI imagery (16‑day, 250 m) and negative relationships with fire activity are defined, mostly complementary to meteorological danger indices. A preliminary composite FWI+NDVI indicator was tested and compared to the conventional FWI. The results obtained show an improvement in the model capability to predict the number of fires. On average, the determination coefficient between the conventional and modified indices increases from 0.38 to 0.51, suggesting that NDVI can provide additional predictive information. These findings open the way to operationally include an NDVI-based layer/index in the Tuscany fire risk bulletin