Spatial structure in urban water consumption: A Bayesian space-time analysis of Bogotá (2019–2024)
- 1Departamento de Estadística, Facultad de Ciencias, Universidad Nacional de Colombia - Sede Bogotá, Bogotá D.C., Colombia
- 2Universidad Nacional de Colombia – Sede de La Paz, Escuela de Pregrados, La Paz, Colombia
Abstract
In this study, we apply fully Bayesian spatio-temporal models to locality-level water consumption in Bogotá, Colombia, from 2019 through 2024. An additive model identifies spatial differences after adjusting for common temporal dynamics and city-wide climatic and socioeconomic time series. Because these covariates are shared across all localities, they provide temporal adjustment but cannot explain between-locality spatial variation. A formal sensitivity analysis adding a structured space–time interaction substantially improves model fit (WAIC: 604.0 versus 2502.7; DIC: 333.9 versus 2500.5), showing that locality deviations are not adequately represented as constant throughout the study period. Student-t likelihoods indicate heavy residual tails (posterior mean degrees of freedom: 3.02 in the additive model and 2.60 in the interaction model), but the Gaussian interaction model retains substantially better information criteria than its Student-t counterpart. The principal spatial main effects remain highly correlated across likelihoods (r = 0.975). The results therefore support spatially heterogeneous temporal dynamics and show that the main spatial ordering is robust to heavy-tailed errors. Author summary: Water consumption in cities is influenced by processes that operate over both time and space. In this study, we analyze monthly water consumption across localities in Bogotá between 2019 and 2024 using statistical models that account for relationships between neighboring areas and temporal dependence. The precipitation, energy, and population variables used in the model are city-wide time series; they help describe temporal variation but cannot explain why localities differ from one another. We also test whether each locality maintains the same deviation from the city-wide trajectory over the entire period. A model allowing spatial patterns to change over time fits substantially better, including during the COVID-19 lockdown and reopening period. The results therefore show that territorial differences are dynamic rather than constant and highlight the need for locality-level temporal information in urban water-demand analysis.
Related
- Normalised, accessible, and addictive: Socioecological drivers of gutka consumption among urban cloth shopkeepers in Karachi, Pakistan
- Fruit and vegetable consumption among people living with HIV at a large Urban HIV Clinic in Central Uganda: A cross-sectional study
- Digital financial literacy and household consumption: Total expenditure versus consumption structure in China
- A bayesian inference and neural network model for predicting physical fitness consumption in table tennis athletes
- Spatial and Temporal Evaluation of Violent and Total Crime Following Community-Led Urban Greening
BibTeX
@article{cruzreyes2026spatial,
title = {{Spatial structure in urban water consumption: A Bayesian space-time analysis of Bogotá (2019–2024)}},
author = {Danna Lesley Cruz-Reyes and Daniel Leonardo Ramírez Orozco},
journal = {PLOS Water},
year = {2026},
month = sep,
doi = {10.1371/journal.pwat.0000512},
url = {https://journals.plos.org/water/article?id=10.1371/journal.pwat.0000512},
note = {CC BY 4.0},
}
RIS
TY - JOUR
TI - Spatial structure in urban water consumption: A Bayesian space-time analysis of Bogotá (2019–2024)
AU - Danna Lesley Cruz-Reyes
AU - Daniel Leonardo Ramírez Orozco
PY - 2026
DA - 2026/09/16
JO - PLOS Water
DO - 10.1371/journal.pwat.0000512
UR - https://journals.plos.org/water/article?id=10.1371/journal.pwat.0000512
AB - In this study, we apply fully Bayesian spatio-temporal models to locality-level water consumption in Bogotá, Colombia, from 2019 through 2024. An additive model identifies spatial differences after adjusting for common temporal dynamics and city-wide climatic and socioeconomic time series. Because these covariates are shared across all localities, they provide temporal adjustment but cannot explain between-locality spatial variation. A formal sensitivity analysis adding a structured space–time interaction substantially improves model fit (WAIC: 604.0 versus 2502.7; DIC: 333.9 versus 2500.5), showing that locality deviations are not adequately represented as constant throughout the study period. Student-t likelihoods indicate heavy residual tails (posterior mean degrees of freedom: 3.02 in the additive model and 2.60 in the interaction model), but the Gaussian interaction model retains substantially better information criteria than its Student-t counterpart. The principal spatial main effects remain highly correlated across likelihoods (r = 0.975). The results therefore support spatially heterogeneous temporal dynamics and show that the main spatial ordering is robust to heavy-tailed errors. Author summary: Water consumption in cities is influenced by processes that operate over both time and space. In this study, we analyze monthly water consumption across localities in Bogotá between 2019 and 2024 using statistical models that account for relationships between neighboring areas and temporal dependence. The precipitation, energy, and population variables used in the model are city-wide time series; they help describe temporal variation but cannot explain why localities differ from one another. We also test whether each locality maintains the same deviation from the city-wide trajectory over the entire period. A model allowing spatial patterns to change over time fits substantially better, including during the COVID-19 lockdown and reopening period. The results therefore show that territorial differences are dynamic rather than constant and highlight the need for locality-level temporal information in urban water-demand analysis.
SN - 2767-3219
ER -