Data-driven urban planning: Real population estimation using water consumption records and tourism data

Accurately estimating de facto population in dynamic urban environments is essential for predictive planning, service provision, and Smart City resilience, particularly in areas with strong seasonal fluctuations. Census-based approaches lack temporal resolution and overlook transient populations, while alternative methods often rely on restricted data, high computational demands, or indirect proxies. This study presents a scalable computational framework for estimating de facto urban population at monthly resolution by integrating high-resolution smart-meter water consumption records with census data, climatic normalization, and tourism indicators. The approach integrates IoT-based remote-reading records, supply-contract databases, census information, and official accommodation registers within an end-to-end ETL pipeline to infer monthly population variations in Alicante (Spain) between 2022 and 2024. A thermal-correction model isolates temperature-driven changes in water demand, improving estimation accuracy. K-means clustering with Dynamic Time Warping identifies neighborhood-level seasonal patterns, distinguishing stable residential areas, high-tourism zones, and districts with marked summer vacancy. Results reveal a robust methodology for population estimation, rigorously verified through out-of-sample validation for 2024, yielding a Mean Absolute Percentage Error (MAPE) of 6.34% and an R 2 = 0.9874. Between January and July 2024, Alicante experienced an estimated 8.78% population increase (+32,749 people). Coastal tourist neighborhoods showed summer rises above 50%, whereas central medium-to-high-income areas exhibited declines up to −22.2% due to seasonal displacement to second residences. Consolidated residential areas remained stable year-round, with deviations below 5%. The methodology offers a low-cost, high-resolution, and privacy-preserving tool for strategic decision-making in Smart Cities and is fully replicable in urban contexts equipped with domestic smart-metering infrastructure to capture resident and floating population dynamics.