Integrated Google Traffic and Field Speed Dataset for Flood-Prone Urban Areas: Loja, Ecuador (2025)

Research Hypothesis
This comprehensive dataset was developed to test whether integrating Google Traffic data with field measurements can improve urban traffic modeling accuracy in medium-sized cities with incomplete digital coverage.

Dataset Composition
The dataset comprises two complementary components: (1) Field Speed Measurements: 1,501 georeferenced vehicle speed observations collected via Survey123 mobile application between July 7-18, 2025 (07:00-24:00), and (2) Google Traffic Observations: typical traffic condition records collected across a complete weekly cycle using Google Traffic's color-coded system. Combined, these datasets provide comprehensive spatial and temporal coverage of traffic patterns in Loja, Ecuador's flood-vulnerable areas.

Data Collection Methodology
Field data collection involved GPS-enabled smartphones recording vehicle speeds, coordinates, and timestamps while operators traveled through flood-prone streets in taxis and personal vehicles. Google Traffic data acquisition used systematic manual observation of color-coded traffic states (Green, Yellow, Orange, Red, None) at predetermined coordinates across all days and hourly intervals. Both datasets employed WGS84 coordinate system and covered the same geographic study area defined by flood vulnerability and incomplete Google Traffic coverage.

Data Quality
Field measurements underwent quality control, removing six invalid records. The remaining validated speeds, ranging from 0 to 70 km/h. Google Traffic observations were successfully assigned to specific road networks. An integration analysis revealed that adjustment factors for Google Traffic data systematically either undershoot or overshoot the ground speed data. These adjustments varied by road type and time period, highlighting the need for local calibration of data from digital platforms.

Notable Findings
The dataset reveals significant coverage gaps in Google Traffic for medium-sized cities. A speed distribution analysis shows that most field measurements fall between 10 and 40 km/h, which is consistent with typical urban traffic patterns. Temporal analysis identified periods of peak data density.

Applications and Interpretation
This hybrid dataset allows for the development of calibrated traffic models for emergency planning, particularly for flood scenario analysis. Researchers can calculate platform-specific bias corrections, develop integrated routing algorithms, and validate commercial traffic platforms in similar urban contexts. The comprehensive temporal coverage supports the analysis of congestion patterns, while the spatial precision enables the integration of infrastructure mapping. The data structure is designed to be easily replicated in other medium-sized cities that face similar digital coverage limitations and climate-related mobility challenges.