Thiruvallur District is situated in the northern part of Tamil Nadu, which encompasses a diverse landscape with a mix of urban , peri-urban and rural land uses. Spanning an area of 3,426 Sq.Kms approximately, which is strategically located adjacent to Chennai, making it crucial for development and expansion. Bordered by the Bay of Bengal, it experiences a tropical climate with monsoon rains influencing its agricultural productivity and hydrological dynamics. With rapid urbanization and industrialization it drives demographic shifts and land use changes. Urban centers like Ambattur, Avadi and Thiruvallur Towns are focal points of economic and infrastructural development, residential expansion and industrial development. Concurrently, the rural hinterlands of Thiruvallur District support intensive agriculture, majorly paddy fields, orchards and vegetable cultivation.

The remote sensing data for the years 2010 and 2023 consist of multi-spectral imagery captured by high-resolution satellites, including bands in the visible, near-infrared, and shortwave infrared regions. These images are selected to ensure a comprehensive spectral coverage, enabling a detailed analysis of land cover changes 2.

Prior to analysis, the imagery undergoes thorough preprocessing. This includes radiometric and atmospheric corrections to standardize the data across both time periods. Additionally, geometric corrections are applied to rectify any distortions caused by terrain variations or sensor artifacts. The aim is to enhance the accuracy and reliability of subsequent analyses.

The unsupervised classification in ERDAS Imagine employs clustering algorithms to group pixels based on their spectral characteristics without prior training. The resulting clusters represent distinct land cover classes. The choice of an appropriate number of classes is determined through iterative refinement, considering spectral signatures and expert knowledge.

The classified images from 2010 and 2023 are then compared to identify changes in land cover. This involves overlaying the two datasets and generating a change mask highlighting areas where alterations have occurred. Post-classification comparison techniques are employed to distinguish between stable and changing land cover classes, enabling a detailed understanding of the spatial dynamics of change.

Ground truth data is collected through field surveys and high-resolution imagery, providing reference points for accuracy assessment. Sample points are randomly selected across the study area, and the classified images are compared with these ground truth points to calculate accuracy metrics such as overall accuracy, producer's accuracy, and user's accuracy. This process ensures the reliability of the change detection results.

The unsupervised classification revealed several distinct land cover classes, including built-up areas, agricultural land, vegetation, water bodies, and the HVF area. Change detection analysis highlighted significant increases in built-up areas, indicating rapid urbanization. The HVF area exhibited changes, potentially associated with anthropogenic activities and natural processes. The accuracy assessment results indicated a high level of reliability in the change detection outcomes. Overall accuracy metrics demonstrate the effectiveness of the unsupervised classification approach in capturing and quantifying land cover changes. These results provide a robust foundation for the interpretation and implications of the observed changes .

In conclusion, the analysis of land cover changes in Thiruvallur District from 2010 to 2023 underscores the dynamic nature of the landscape, especially in the context of urban expansion and ecological alterations. The increase in built-up areas, particularly in the HVF zone, raises concerns about potential environmental impacts. The findings emphasize the need for sustainable urban planning and conservation efforts to mitigate adverse effects on ecosystems.

The accuracy assessment lends confidence to the reliability of the change detection results, affirming the suitability of the methodology employed in this study. The outcomes contribute valuable information for policymakers, urban planners, and environmentalists, guiding future decision-making processes aimed at fostering sustainable development while preserving the ecological balance in Thiruvallur District. Continuous monitoring and periodic reassessment are recommended to track evolving land cover patterns and inform adaptive management strategies 4.