Prediction of 9-1-1 call volumes for emergency event detection

A multi-dimensional linear predictor of 9-1-1 (emergency) call volumes was built and used to automatically detect emergency events. This is illustrated by analyzing the emergency calls generated in two emergency events in the San Francisco Bay Area. The predictor can help emergency service providers recognize the occurrence of anomalously large numbers of 9-1-1 calls and subsequently map the spatiotemporal extent of widescale emergency events such as earthquakes and fires, complementing the effort of individual 9-1-1 emergency call takers.