Atmos: Crowd–sourcing Estimations about Current and Future Weather Conditions

This month saw the International Joint Conference on Pervasive and Ubiquitous Computing (UbiComp) held in the unexpectedly sunny Seattle. UbiComp is the premier conference in the field, regularly attracting over 700 participants. RECALL had a strong presence there, with team members organizing a novel Workshop on “Ubiquitous Technologies for Augmenting the Human Mind (WAHM)”. In additional, RECALL team member Evangelos Niforatos presented his work on the Atmos system in the conference’s poster session .

Atmos allows collecting empirical reports about current weather conditions.
Atmos Screenshot

Atmos is a novel approach to weather estimation. It introduces the use of participatory sensing to collect in-situ weather data, both from sensors and human input. Atmos leverages a crowd-sourcing network of mobile devices to generate highly localized information about current and future weather conditions. Participatory sensing involves the utilization of mobile devices to form interactive, collaborative sensor networks that enable users to garner, analyze and share local knowledge. Under this guise, participatory sensing exhibits a unique level of spatio-temporal coverage in observing phenomena of interest in urban spaces. The key idea behind this new paradigm is the enabling of mobile users to collect and share sensed data about their natural settings in large scale, using their mobile devices.

Evangelos’ work investigates not only how people experience current weather conditions, but also how they predict future conditions. Prior work supports that during the action of forecasting, the human mind combines currently available information with similar experiences of the past to decide on a future outcome. However, people predict by representativeness as such, they select or order future outcomes by the degree at which they represent a current situation (Kahmenan, “on the psychology of prediction”). For example, we found that reported temperature feel and sky clearness collected with the Atmos mobile app revealed a maximum in the morning hours (9:00 – 10:00) and slowly degraded in the course of the day. Looking at reports about future weather conditions, those submitted during morning hours similarly foresaw increased sky clearness and temperature feel for the next few hours, while later predictions degraded their outlook over the course of the day.

Atmos enables interesting insights into how individuals perceive and remember their immediate environment, and in particular weather conditions. Can we improve individual awareness of climate conditions, both for one’s own benefit and for better crowd-sourced predictions? Atmos is an ongoing project and a new mobile app is coming out soon! Stay tuned!

Niforatos, Evangelos et al. 2014. Atmos: A Hybrid Crowdsourcing Approach to Weather Estimation. Seattle, WA, USA: ACM. Poster. (January 6, 2014).