Ubiquitous geo-sensing for context-aware analysis: exploring relationships between environmental and human dynamics

Area of Science:

• Urban dynamics and spatio-temporal analysis
• Human behavior studies
• Environmental science and meteorology

Background:

• Ubiquitous geo-sensing facilitates context-aware analysis of interconnected phenomena.
• Holistic understanding of spatio-temporal processes through context-aware analysis is underrepresented in scientific literature.
• Investigating the interplay between collective human behavior and weather offers novel insights into urban systems.

Purpose of the Study:

• To analyze collective human behavior in the context of weather.
• To explore complex relationships between spatio-temporal phenomena (human behavior and weather).
• To provide novel insights into the dynamics of urban systems.

Main Methods:

• Utilized aggregated mobile phone data as a proxy for collective human behavior.
• Integrated weather data from climate stations in Udine, Italy.
• Developed a hybrid approach combining spatio-temporal statistical methods, including explanatory factor analysis and Maximal Information Coefficient (MIC).

Main Results:

• Explanatory factor analysis effectively differentiates between normal and adverse weather conditions.
• The Maximal Information Coefficient (MIC) quantified the relationship between adverse weather and variations in mobile network traffic.
• Generated three spatially referenced maps illustrating weather-human relationships and collective dynamics.

Conclusions:

• The study demonstrates a quantifiable link between weather conditions and collective human behavior patterns.
• Spatially explicit analysis reveals localized impacts of weather on human mobility.
• The findings highlight the potential of geo-sensing for understanding urban system dynamics and identify areas for future research.