Aerial retardant drops are widely used in wildfire suppression, yet their effectiveness in slowing fire spread remains difficult to quantify at scale. This study evaluates their impact on wildfire rate of spread (ROS) using a framework that combines observed and counterfactual (synthetic) drop locations from 62 Oregon wildfires. Synthetic drops were generated to simulate a no-suppression baseline, enabling comparison of ROS with and without suppression. We trained two random forest classifiers: one with real and synthetic drops (full model) and one with only synthetic drops. Both incorporated environmental and topographic features to predict whether spread slowed following a drop. While the full model performed well, the real-synthetic indicator had low feature importance, offering limited causal evidence that aerial suppression consistently reduced spread. The synthetic-only model produced similar performance, suggesting that drops with observed ROS reductions often coincided with favorable environmental and topographic conditions and may have occurred independent of suppression. These findings highlight the challenges of evaluating suppression at scale and emphasize the need for finer data, detailed operational records, and advanced modeling to better assess the role of aerial fire retardant drops in future wildfire management activities.
Local raster texture is an important attribute that can be used for sample design and raster-based classification, regression, and clustering. However, determining the appropriate kernel weights to quantify texture can be challenging and can vary depending on the remote sensing or computer vision task. We present an automated approach and companion Jupyter notebook that uses principal component analysis (PCA) and convolution kernels to quantify orthogonal intensity and texture metrics. Our approach linearly projects raster data across covariance axes, learns kernel cell weights, and provides a straightforward mechanism to retain as much information as possible within the raster dataset while removing redundancy. Using this PCA convolution approach, we reduced the dimensionality of a Landsat 8 scene and corresponding five by five analysis window from 175 to 10 independent intensity and texture components that maintained more than 95% of the variation within the raster data. Applying our PCA convolution methodology, we created a 10-band raster dataset that can be used to inform sample design or as orthogonal predictors in classification, regression, and clustering models.