Fragile greening: High-frequency turnover limits urban cooling benefits in a dryland metropolis

Global arid cities are increasingly promoting urban greening to alleviate the effect of urban heat island; however, the resilience and functional performance of these artificial ecosystems remain uncertain. Using Urumqi, a major metropolis in Central Asia, as a case study, this article examines the long-term development trajectory (1990–2024) and the underlying processes of its “artificial oasis”. A multi-method framework was employed, integrating Bayesian Estimator of Abrupt change, Seasonal change, and Trend (BEAST), Continuous Change Detection and Classification (CCDC), statistical diagnostics of long-term trends and persistence, and bivariate NDVI–SUHI coupling analysis to uncover latent ecological vulnerabilities masked by decades of policy-driven greening under conditions of water scarcity and climatic stress. The results indicate that urban greening intensified markedly following a major structural regime shift around 2013. Nevertheless, this apparent recovery is dominated by anti-persistent dynamics, frequency vegetation turnover, and widespread future degradation trajectories, indicating weak ecological memory and a strong dependence on sustained external inputs. Consequently, increases in vegetation greenness have not translated proportionally into urban heat mitigation. While large, intensively managed parks maintain relatively stable cooling effects, agricultural and peri-urban green spaces exhibit pronounced interannual variability and intermittent functional failure. Joint NDVI–SUHI trajectories further reveal that greening and cooling are not linearly coupled; instead, thermal regulation is primarily governed by vegetation stability, structural maturity, and spatial configuration rather than greenness alone. Overall, the key challenge for urban greening in arid cities lies not in continued spatial expansion, but in fostering ecological memory through stable, spatially coherent, and structurally mature urban green space systems. By explicitly linking long-term vegetation dynamics with thermal responses, this framework provides a resilience-oriented lens for evaluating urban greening and offers actionable insights for climate adaptation and sustainable urban planning in dryland regions.