In this paper a distributed event localization, tracking, and classification framework (DELTA) is presented. An event is observed and tracked by dynamically established groups. Relevant sensor data is collected at dedicated nodes (group leaders) which are destined to perform all subsequent localization and group organization tasks. Based on the collected sensor data, both the event position and the signal strength(s) of the emitted signal(s) of the event are estimated. This enables DELTA to classify events based on the estimated signal emission power. Existing approaches either focus on accurate but cost-intensive collaborative signal processing (CSP) methods or on less accurate but more cost-efficient approaches mainly focusing on minimizing the communication load. DELTA bridges this gap by providing satisfying accuracy while keeping the network load at a reasonable level. The performance of the proposed framework is evaluated by simulation as well as by implementation on real hardware. In addition, problems of closed-form linear least square solutions for the localization task are discussed.