Extracting Land Use Change Patterns of Rural Town Settlements with Sequence Alignment Method

Senkai Xie, Wenjia Zhang, Yi Zhao, De TongLand 2022 · DOI

TL;DR

This paper introduces a novel application of the Sequence Alignment Method (SAM) to analyze land use change patterns in rural town settlements (RTSs) across Guangdong, China, from 1980 to 2015. By treating land use trajectories as "DNA sequences," the study identifies 10 distinct development patterns, revealing significant spatial polarization between the rapidly urbanizing Pearl River Delta (PRD) and less developed regions.

What Problem Does This Paper Address?

Understanding how rural areas evolve is critical for effective planning, especially in rapidly developing regions like China. While land use change is a well-studied phenomenon, most research focuses on macro-level or urban areas. There's a significant gap in understanding the specific, longitudinal land use change patterns of rural town settlements (RTSs) and the underlying factors driving these changes. Existing methods often rely on cross-sectional data or case studies, lacking the ability to capture the dynamic, sequential evolution of RTSs and to cluster them into meaningful development patterns. This makes it difficult to diagnose regional development issues and formulate targeted policies for rural revitalization.

Methods

The core innovation of this paper is the application of the Sequence Alignment Method (SAM), a technique traditionally used in bioinformatics for comparing DNA or protein sequences, to analyze land use trajectories of RTSs.

Here's a breakdown of their approach:

  • Data Preparation: Land use data from Landsat TM imagery (1980-2015) was manually interpreted and classified into five primary types: cultivated land, urban land, rural residential land, nonresidential construction land, and ecological land. The areal size of these land use types was used as input.
  • Trajectory Construction: For each of the 1158 RTSs, a "land use sequence" was created, representing the dominant land use type in each of the seven waves of data. This sequence was treated as a "DNA sequence" of the RTS's evolution.
  • Sequence Alignment Method (SAM):
    • Pairwise Alignment: The SAM algorithm, based on the Needleman-Wunsch algorithm, was used to calculate the similarity (or distance) between any two RTS land use sequences. This involves defining scoring rules for matches, mismatches, and gaps (representing missing data or transitions).
  • Multiple Sequence Alignment: To cluster all RTSs, a progressive alignment approach was employed. This involved:
  • Pairwise alignment of all sequences to create a distance matrix.
  • Constructing a guide tree using the UPGMA (unweighted pair-group method with arithmetic averages) clustering algorithm to reflect evolutionary relationships.
  • Sequentially aligning sequences based on the guide tree.
  • Clustering: The output of the SAM analysis was used to group RTSs with similar land use trajectories, resulting in 10 distinct "DNA types" or development patterns.

Datasets

The study utilizes a comprehensive dataset of Landsat TM imagery for Guangdong Province, China, spanning seven time points from 1980 to 2015. This data was manually interpreted and classified, providing a spatial resolution of 30 meters. The study focuses on 1158 Rural Town Settlements (RTSs) within Guangdong Province, using their official boundaries from 2008 as a consistent reference.

Key Insights and Ideas

  • Rural Evolution as "DNA": The novel framing of land use trajectories as "DNA sequences" is a powerful conceptual tool. It allows for the application of established bioinformatics techniques to understand complex spatio-temporal patterns in human-modified landscapes.
  • Spatial Polarization: A striking finding is the clear spatial divide in land use change patterns. The rapidly developing Pearl River Delta (PRD) is a hotbed for changing RTSs, primarily urbanizing and agricultural upgrading types, while stable land use patterns are more prevalent in the northern, eastern, and western parts of Guangdong. This highlights the uneven development across the province.
  • Dominance of Stable Patterns: Despite the focus on change, the majority of RTSs (88%) exhibit stable land use patterns over the 35-year period. This suggests that while some areas are undergoing rapid transformation, many rural settlements maintain their traditional land use structures.
  • Urbanization's Footprint: The study clearly shows that RTSs undergoing significant change are heavily concentrated in the PRD, driven by urbanization. This includes the conversion of agricultural and ecological land to construction land.
  • Driving Forces are Multifaceted: The research identifies location, landforms, industries, and policy factors as key drivers of these land use evolution patterns, underscoring the complex interplay of factors shaping rural landscapes.

Key Results

  • 10 Land Use Change Patterns: The SAM analysis successfully identified 10 distinct clusters of RTS land use trajectories, categorized into stable and changing patterns.
  • 73% of Changing RTSs in PRD: A significant majority (approximately 73%) of RTSs that underwent land use changes were located within the Pearl River Delta (PRD).
  • Stable RTSs Dominate: 88% of all RTSs in Guangdong exhibited stable land use patterns throughout the study period.
  • Four Stable Types: Four types of stable RTSs were identified: cultivated ecological, ecologically dominant, rural construction, and mixed construction. Notably, no stable "urban construction" RTSs were found, suggesting that constant urban land dominance from 1980 is rare.
  • Six Changing Types: Six types of changing RTSs were identified, illustrating transitions from traditional land use types to more developed ones, particularly towards mixed and urban construction.

Limitations and Future Work

  • SAM Limitations: The authors acknowledge that progressive SAM, while efficient, provides an approximate solution and may not find the absolute optimal alignment. They also note that evaluating the quality of multiple sequence alignment results can be challenging due to the lack of a perfect objective function.
  • Future Work Suggestions:
    • Network Analysis: Exploring the conversion of sequence alignment into network-related data and applying complex network analysis for clustering is suggested as a potential solution to overcome some SAM limitations.
  • Broader Spatial Elements: Future studies should incorporate other spatial elements beyond land use, such as road network structure, RTS spatial patterns, and settlement structure, to provide a more holistic understanding.
  • Coupling Driving Factors: Investigating the interplay between multiple driving factors and multiple spatial evolution sequences is crucial for developing models that couple driving forces with spatial patterns in rural areas.
  • Focus on Backward Regions: The study emphasizes the need to focus more on the development processes of backward regions outside the PRD, using longitudinal data and pattern mining to achieve coordinated regional development.

Summary updated

Abstract

Understanding land use change patterns of rural town settlements (RTSs) is crucial for rural and small-town planning; however, few studies have explored pattern mining approaches to RTS trajectory analysis. In this study, we adopted a novel method by building sequence alignment method (SAM) to detect representative trajectory clusters of land use change of 1158 RTSs in seven waves from 1980 to 2015 in Guangdong, China. The results suggest that there are 10 clusters of RTSs with varying trajectories of land use change, implying their differences in the development processes and underlying socioeconomic, demographical, and institutional factors. A spatial distribution map of RTSs shows that stable cultivated ecological and stable ecologically dominant RTSs are distributed in the northern, eastern, and western parts of Guangdong, whereas stable rural construction and stable mixed construction RTSs are mostly located around the provincial boundary. Notably, 73% of the RTSs that have undergone changes in land use types are located in the Pearl River Delta (PRD), including urbanized and agricultural upgraded RTSs. The analysis presented here summarizes the driving forces of the spatial evolution of RTSs, including the location, landforms, industries, and policy factors. This study provides dynamic policy implications to understand longitudinal and sequential spatial restructuring and regional coordinated development in the fast-growing PRD area.
BibTeX
@inproceedings{xie2022extracting,
  author = {Senkai Xie and Wenjia Zhang and Yi Zhao and De Tong},
  title = {{Extracting Land Use Change Patterns of Rural Town Settlements with Sequence Alignment Method}},
  booktitle = {Land},
  year = {2022},
  doi = {10.3390/land11020313},
}
Figure