Hybrid AI framework in urban stress detection for smart nature-based interventions towards sustainable urban development

Introduction Rapid urbanization coupled with climate change has escalated multidimensional urban stress, particularly in expanding cities characterized by high population density, unmanaged land expansion, infrastructure imbalance, and ecological vulnerability. Conventional urban stress assessment approaches struggle to integrate heterogeneous spatial datasets, model complex spatial dependencies, and generate actionable intervention priorities. This study proposes a novel hybrid AI framework for urban stress detection to support smart nature-based interventions using Global Human Settlement Layer (GHSL) datasets to address these limitations. Methods Multi-temporal GHSL-derived indicators including built-up density, settlement morphology, and population distribution were spatially aggregated and normalized into standardized stress matrices. The pipeline processed these datasets using the independent pattern component extractor (IPCE) to reduce dimensionality while isolating independent spatial stress components. Next, the Spatial Stress Graph Optimizer (SSGO) modeled and optimized spatial stress interactions across topological node networks. The optimized representations were evaluated through an ANFIS-ALTO Prophet Network (AAPN) to generate uncertainty-aware urban intervention priority scores, which were subsequently ranked by the Ripper Sort Prioritization Mechanism (RSPM) according to urgency. Results Experimental evaluation using spatially blocked validation demonstrates that the proposed framework achieves high predictive accuracy (R 2 = 0.93), reduces classification noise across complex urban–rural fringes, and successfully identifies critical intervention micro-zones. Discussion The findings indicate that combining GHSL spatial datasets with a graph-optimized hybrid neural architecture resolves the analytical fragmentation inherent in traditional indicator-based approaches. By providing objective, uncertainty-aware rankings, this framework offers a scalable decision-support mechanism to guide city planners in deploying targeted, climate-resilient Nature-Based Interventions.

Authors

Institutions

Publication Details

Journal
Frontiers in Sustainability
Published
2026-11-20
DOI
https://doi.org/10.3389/frsus.2026.1894183
Primary Topic
Environmental Quality and Pollution
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Hybrid AI framework in urban stress detection for smart nature-based interventions towards sustainable urban development

Seema Rawat, Rajiv Iyer, Garima Shukla, Dipak Raskar et al.
5 links
Frontiers in Sustainability
Environmental Quality and Pollution
article

Hybrid AI framework in urban stress detection for smart nature-based interventions towards sustainable urban development

Seema Rawat, Rajiv Iyer, Garima Shukla, Dipak Raskar, Deepak Kumar, A. Saranya
article en

Abstract

Introduction Rapid urbanization coupled with climate change has escalated multidimensional urban stress, particularly in expanding cities characterized by high population density, unmanaged land expansion, infrastructure imbalance, and ecological vulnerability. Conventional urban stress assessment approaches struggle to integrate heterogeneous spatial datasets, model complex spatial dependencies, and generate actionable intervention priorities. This study proposes a novel hybrid AI framework for urban stress detection to support smart nature-based interventions using Global Human Settlement Layer (GHSL) datasets to address these limitations. Methods Multi-temporal GHSL-derived indicators including built-up density, settlement morphology, and population distribution were spatially aggregated and normalized into standardized stress matrices. The pipeline processed these datasets using the independent pattern component extractor (IPCE) to reduce dimensionality while isolating independent spatial stress components. Next, the Spatial Stress Graph Optimizer (SSGO) modeled and optimized spatial stress interactions across topological node networks. The optimized representations were evaluated through an ANFIS-ALTO Prophet Network (AAPN) to generate uncertainty-aware urban intervention priority scores, which were subsequently ranked by the Ripper Sort Prioritization Mechanism (RSPM) according to urgency. Results Experimental evaluation using spatially blocked validation demonstrates that the proposed framework achieves high predictive accuracy (R 2 = 0.93), reduces classification noise across complex urban–rural fringes, and successfully identifies critical intervention micro-zones. Discussion The findings indicate that combining GHSL spatial datasets with a graph-optimized hybrid neural architecture resolves the analytical fragmentation inherent in traditional indicator-based approaches. By providing objective, uncertainty-aware rankings, this framework offers a scalable decision-support mechanism to guide city planners in deploying targeted, climate-resilient Nature-Based Interventions.

Frontiers in SustainabilityVol. 7
Texas Tech University (US), Nalanda University (IN), Astana IT University (KZ), Amity University Mumbai (IN)
Sustainable cities and communities
Environmental Quality and Pollution
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon