Drones Under Weather Pressure: Analyzing Environmental Impacts on Detection Accuracy in Disaster Response Using SUR and Traditional Models

The use of unmanned aircraft systems (UAS) in disaster response within densely populated urban areas has continued to evolve. This study examines the deployment of drones in disaster management in Tokyo, with particular attention to the environmental factors of wind speed, temperature, and precipitation, using a filtered dataset of 44,569 entries obtained from Kaggle. Specifically, the study evaluates how these environmental variables affect the efficacy of drone missions in completing surveillance tasks, drawing on Tokyo’s urban disaster response simulations as well as anthropological observations of drone flight practices using an array of regression models including a binary logit regression model, negative binomial model, and multiple linear models for individual variables affecting flight and ground detection. Finally, a joint model will be performed to determine any differences between the three individual models and determine any correlation between variables. By understanding the multi-factor influences on drone performance, disaster response teams can strengthen operational strategies for unmanned aerial vehicle (UAV) deployment in relief missions. The findings of this research can contribute to improving the strategic implementation of drones during both natural and human-made disasters in highly populated areas such as Tokyo.