“…Accordingly, predicting where we can restrict fire growth is decisive, and current decision support systems aim to integrate estimates of wildfire containment probability (Wei, Thompson, Scott, O'Connor, & Dunn, 2019). The common strategies to produce such probabilities rely on binary statistical models (O'Connor, , modeling algorithms (Eghbal, Hugh, & Aaron, 2018), or expert-informed indices (Rodríguez y Silva, Molina Martínez, & González-Cabán, 2014). These models leverage historical fire perimeters and landscape features such as ground accessibility, presence of unburnable barriers, vegetation impedance, local fire danger conditions, or wildfire behavior (Holsinger, Parks, & Miller, 2016;Narayanaraj & Wimberly, 2011;O'Connor et al, 2017;Price, Borah, & Maier, 2013;Rodríguez y Silva et al, 2014).…”