Drone-Based IoT Monitoring of Urban CO₂ Levels in Makassar: Spatio-Temporal Analysis Across Varying Heights

Urban air quality degradation from rising CO₂ is acute in rapidly developing tropical cities such as Makassar, Indonesia. We deploy a drone-based Internet of Things (IoT) platform for real-time CO₂ monitoring, integrating low-cost sensors (NDIR, MQ135, MG811) on a DJI Phantom 4 with cloud streaming to Firebase. Measurements were collected at five sites, namely Jl. AP. Pettarani, Jl. Ahmad Yani, Jl. Sultan Hasanuddin, Jl. Nusantara, and KIMA at 08:00, 12:00, and 16:00 in September 2024 while vertically profiling 1–20 m with three repeat flights per site and time. Descriptive statistics and one-way ANOVA with Tukey HSD assessed spatio-temporal differences; Pearson correlation quantified cross-sensor agreement. Results show marked spatial and diurnal variability: Jl. AP. Pettarani exhibits the highest mean concentration (442.5 ppm), likely due to flyover-induced trapping, whereas Jl. Ahmad Yani records the lowest (390.0 ppm). Vertical profiles reveal mid-altitude peaks in street-canyon and industrial settings, and dilution with height in greener areas, indicating ventilation contrasts. Preprocessing removed outliers and applied temperature-humidity corrections to low-cost sensors. Differences across locations and times are statistically significant (p < 0.05), and cross-sensor correlations are strong (r ≈ 0.88–0.96) after correction. Compared with fixed ground stations, the system provides fine-scale three-dimensional coverage and real-time visualization useful for field decisions. Limitations include payload-constrained endurance and intermittent data loss in obstructed areas. Findings support targeted interventions, improving canyon ventilation around flyovers and expanding urban greenery relevant to Makassar and similar tropical cities.