Transport of the smoke plume from Chiado’s fire in Lisbon (Portugal) sensed by atmospheric electric field measurements

The Chiado’s fire that affected the city centre of Lisbon (Portugal) occurred on 25th August 1988 and had a significant human and environmental impact. This fire was considered the most significant hazard to have occurred in Lisbon city centre after the major earthquake of 1755. A clear signature of this fire is found in the atmospheric electric field data recorded at Portela meteorological station about 8 km NE from the site where the fire started at Chiado. Measurements were made using a Benndorf electrograph with a probe at 1 m height. The atmospheric electric field reached 510 V/m when the wind direction was coming from SW to NE, favourable to the transport of the smoke plume from Chiado to Portela. Such observations agree with predictions using Hysplit air mass trajectory modelling and have been used to estimate the smoke concentration to be ~0.4 mg/m3. It is demonstrated that atmospheric electric field measurements were therefore extremely sensitive to Chiado’s fire. This result is of particular current interest in using networks of atmospheric electric field sensors to complement existing optical and meteorological observations for fire monitoring.

Mobile sink based fault diagnosis scheme for wireless sensor networks

Network diagnosis in Wireless Sensor Networks (WSNs) is a difficult task due to their improvisational nature, invisibility of internal running status, and particularly since the network structure can frequently change due to link failure. To solve this problem, we propose a Mobile Sink (MS) based distributed fault diagnosis algorithm for WSNs. An MS, or mobile fault detector is usually a mobile robot or vehicle equipped with a wireless transceiver that performs the task of a mobile base station while also diagnosing the hardware and software status of deployed network sensors. Our MS mobile fault detector moves through the network area polling each static sensor node to diagnose the hardware and software status of nearby sensor nodes using only single hop communication. Therefore, the fault detection accuracy and functionality of the network is significantly increased. In order to maintain an excellent Quality of Service (QoS), we employ an optimal fault diagnosis tour planning algorithm. In addition to saving energy and time, the tour planning algorithm excludes faulty sensor nodes from the next diagnosis tour. We demonstrate the effectiveness of the proposed algorithms through simulation and real life experimental results.