Scholarlink

<p>Urban expansion in subtropical metropolitan regions intensifies land use/land cover (LULC) transformations and alters land surface temperature (LST) dynamics, directly affecting urban climate resilience and environmental sustainability. This study evaluates historical LULC and LST changes (1995–2017) and predicts future scenarios (2028–2039) in Porto Alegre, southern Brazil, using Landsat imagery integrated with a Cellular Automata–Artificial Neural Network (CA-ANN) model implemented in the MOLUSCE platform. LULC maps were generated using supervised Maximum Likelihood Classification. Model validation achieved an overall accuracy of 77.6% with a Kappa coefficient of 0.66, indicating moderate agreement. LST validation resulted in an overall accuracy of 74.67% and a Kappa coefficient of 0.64. The results reveal a continuous expansion of built-up areas accompanied by an increase in higher LST classes, highlighting the spatial coupling between urban growth and surface thermal intensification. Beyond reproducing historical dynamics, the CA-ANN model demonstrated satisfactory predictive performance in capturing spatial transition patterns based on past urbanization trends. The proposed framework advances existing LULC–LST modeling approaches by integrating categorical thermal dynamics within a CA-ANN simulation environment, enabling the simultaneous assessment of land transition processes and thermal pattern evolution. Unlike many previous studies that analyze LULC and LST separately or focus only on land transition modeling, this integrated approach explicitly links thermal dynamics with urban expansion processes. By applying this framework to a humid subtropical metropolitan region in South America, this study provides spatially explicit projections of both land cover change and associated thermal dynamics, supporting climate-sensitive urban planning and long-term mitigation strategies.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Assessment and Prediction of LULC and LST Changes Using Remote Sensing and CA-ANN Algorithm: a Study from Porto Alegre, Brazil

  • 期刊论文

  • 发表日期：

    2026年4月29日

查看全文
  • Eduardo André Kaiser,
  • Silvia Beatriz Alves Rolim,
  • Atílio Efrain Bica Grondona,
  • Bijeesh Kozhikkodan Veettil

摘要

Urban expansion in subtropical metropolitan regions intensifies land use/land cover (LULC) transformations and alters land surface temperature (LST) dynamics, directly affecting urban climate resilience and environmental sustainability. This study evaluates historical LULC and LST changes (1995–2017) and predicts future scenarios (2028–2039) in Porto Alegre, southern Brazil, using Landsat imagery integrated with a Cellular Automata–Artificial Neural Network (CA-ANN) model implemented in the MOLUSCE platform. LULC maps were generated using supervised Maximum Likelihood Classification. Model validation achieved an overall accuracy of 77.6% with a Kappa coefficient of 0.66, indicating moderate agreement. LST validation resulted in an overall accuracy of 74.67% and a Kappa coefficient of 0.64. The results reveal a continuous expansion of built-up areas accompanied by an increase in higher LST classes, highlighting the spatial coupling between urban growth and surface thermal intensification. Beyond reproducing historical dynamics, the CA-ANN model demonstrated satisfactory predictive performance in capturing spatial transition patterns based on past urbanization trends. The proposed framework advances existing LULC–LST modeling approaches by integrating categorical thermal dynamics within a CA-ANN simulation environment, enabling the simultaneous assessment of land transition processes and thermal pattern evolution. Unlike many previous studies that analyze LULC and LST separately or focus only on land transition modeling, this integrated approach explicitly links thermal dynamics with urban expansion processes. By applying this framework to a humid subtropical metropolitan region in South America, this study provides spatially explicit projections of both land cover change and associated thermal dynamics, supporting climate-sensitive urban planning and long-term mitigation strategies.