This study focuses on daytime-only detection of pyroCb using GOES-15 imager data for 41 large wildfires observed during the 2013 fire season in North America. We use a simple thresholding technique to isolate potential pyroCb using multiple assumptions: 1) pyroCb occur close to significant fire events; 2) pyroCb are optically thick, 3) pyroCb reach high altitudes, and 4) entrained smoke reduces cloud top particle size thereby increasing albedo in the shortwave infrared. Based on these assumptions, the algorithm first identifies clouds near fires that are cold in the longwave (high altitude and optically thick). Microphysical changes are identified by a shortwave brightness temperature that is significantly higher than in longwave. For the 41 fires studied, the resulting algorithm proved capable of detecting pyroCb as observed by the scientific community and was also able to detect a significant number of pyroCb that were not previously identified. This algorithm can be extended to next-generation geostationary sensors such as the Advanced Baseline Imager (GOES-R) and the Advanced Himawari Imager (Himawari-8), which will likely provide even greater ability to identify and detect stratospheric smoke injections.