What does this research mean for the field?

The proposed multi-modal multi-kernel convolutional long short-term memory (MM-MK-ConvLSTM) model improves the accuracy and efficiency of urban growth predictions by capturing multi-scale spatial patterns, demonstrating through simulations that sustainable land-use planning requires multi-dimensional policies to mitigate both slope development and urban fragmentation. Novelty: ClaimNovelty.METHODOLOGICAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

To develop a multi-scale deep learning framework that integrates city- and neighborhood-level factors for reliable scenario forecasting of urban expansion and fragmentation.

September 8, 2026GIScience & Remote SensingOpen Access

Multi-scale urban growth modeling in metropolitan area for scenario-based forecasting of urban expansion and fragmentation in Yongin, South Korea

View Full Paper

Ask AI

Bookmark

Authors

JNJuyeong Nam CLChangyeon Lee

Discussion

Member takes

Overview

Modeling study demonstrates that multi-scale deep learning accurately forecasts metropolitan urban expansion and fragmentation, highlighting trade-offs in isolated land-use zoning policies.

Key Points

To develop a multi-scale deep learning framework that integrates city- and neighborhood-level factors for reliable scenario forecasting of urban expansion and fragmentation.
Constructed a multi-modal multi-kernel convolutional long short-term memory (MM-MK-ConvLSTM) model integrating city-level contextual encoding with attention-based multi-kernel encoder–decoder layers.
Trained the network on the Seoul Metropolitan Area, benchmarked predictions against ConvLSTM and DS-ConvLSTM architectures in Yongin City, and simulated 12 urban growth policy scenarios through 2040.
The MM-MK-ConvLSTM outperformed ConvLSTM and DS-ConvLSTM baselines across all validation metrics for late 2010s predictions while achieving faster training convergence and fewer computational operations.
The architecture captured fine-grained fragmented development patterns that single-scale baseline models failed to detect.
Scenario forecasts to 2040 showed that strict slope regulations inadvertently displace fragmented sprawl onto flat farmlands, whereas pairing high-density targets with development-scale constraints minimizes both slope loss and fragmentation.

Cite This Study

Nam et al. (2026) studied this question.

synapsesocial.com/papers/6a9fd82058e84d0ff5b47565 https://doi.org/10.1080/15481603.2026.2719955

View Full Paper

Ask AI

Bookmark

September 8, 2026GIScience & Remote SensingOpen Access

Multi-scale urban growth modeling in metropolitan area for scenario-based forecasting of urban expansion and fragmentation in Yongin, South Korea

View Full Paper

Ask AI

Bookmark

Authors

JNJuyeong Nam CLChangyeon Lee

Discussion

Member takes

Overview

Modeling study demonstrates that multi-scale deep learning accurately forecasts metropolitan urban expansion and fragmentation, highlighting trade-offs in isolated land-use zoning policies.

Key Points

To develop a multi-scale deep learning framework that integrates city- and neighborhood-level factors for reliable scenario forecasting of urban expansion and fragmentation.
Constructed a multi-modal multi-kernel convolutional long short-term memory (MM-MK-ConvLSTM) model integrating city-level contextual encoding with attention-based multi-kernel encoder–decoder layers.
Trained the network on the Seoul Metropolitan Area, benchmarked predictions against ConvLSTM and DS-ConvLSTM architectures in Yongin City, and simulated 12 urban growth policy scenarios through 2040.
The MM-MK-ConvLSTM outperformed ConvLSTM and DS-ConvLSTM baselines across all validation metrics for late 2010s predictions while achieving faster training convergence and fewer computational operations.
The architecture captured fine-grained fragmented development patterns that single-scale baseline models failed to detect.
Scenario forecasts to 2040 showed that strict slope regulations inadvertently displace fragmented sprawl onto flat farmlands, whereas pairing high-density targets with development-scale constraints minimizes both slope loss and fragmentation.

Cite This Study

Nam et al. (2026) studied this question.

synapsesocial.com/papers/6a9fd82058e84d0ff5b47565 https://doi.org/10.1080/15481603.2026.2719955

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

View Full Paper

Ask AI

Bookmark