Uncertainty Quantification and Monte Carlo Simulations to Enhance SDG 11.3.1 Monitoring

Abstract. As urbanization accelerates globally, efficient land use is essential for sustainable development. Sustainable Development Goal (SDG) 11 includes Target 11.3, which promotes sustainable urbanization and is assessed through Indicator 11.3.1, measuring land use efficiency (LUE) via the ratio of Land Consumption Rate (LCR) to Population Growth Rate (PGR), known as LCRPGR. While this metric is valuable for guiding urban planning, it could be affected by variability in Earth Observation (EO) data products, especially differences in built-up area definitions and classification errors that lead to data quality issues. This paper presents a comprehensive approach to improving SDG 11.3.1 monitoring by (1) quantifying the impact of EO data variability on LCR and LCRPGR estimates, highlighting the importance of standardized built-up area definitions; (2) adapting a bias-adjustment methodology to correct for over- and underestimations in EO-derived built-up area estimates, thus enhancing accuracy; and (3) incorporating Monte Carlo simulations to quantify uncertainties in LCR and LCRPGR due to classification errors. The findings indicate that definitions and adjustments significantly influence the SDG 11.3.1 metrics. Monte Carlo simulations provide essential insights into the confidence intervals of LCR and LCRPGR values, revealing the degree of uncertainty tied to EO data accuracy. This study supports more reliable urban planning and policy formulation by ensuring LCR and LCRPGR values reflect actual urban dynamics in a better way, enabling robust, equitable comparisons across cities, countries, and SDG regions.