Topographic Shadow Correction with Optical Remote Sensing: A Case Study from Uttarakhand and Nagaland

Introduction: Remote sensing has emerged as a critical tool for environmental management, particularly in regions like the Indian Himalayan Region (IHR), where diverse ecosystems and rugged terrains pose unique challenges. Among these challenges, the accurate correction of topographic shadows in high-altitude terrains is essential for reliable terrain analysis, forest cover mapping, and land use classification. Traditionally, Synthetic Aperture Radar (SAR) and Light Detection and Ranging (LiDAR) data have been utilized to address this issue due to their ability to penetrate cloud cover and capture detailed terrain information. However, the high costs associated with SAR and LiDAR limit their practicality for large-scale or resource-constrained projects. To overcome this, our team in The Energy and Resources Institute (TERI), New Delhi developed a cost-effective solution using the Short-Wave Infrared (SWIR) and Near Infrared (NIR) bands of optical remote sensing data, specifically incorporating the Normalized Difference Moisture Index (NDMI) to enhance topographic shadow correction.

Problem Statement: Topographic shadows present a significant challenge in optical remote sensing, leading to inaccuracies in applications such as land cover mapping, change detection, and vegetation monitoring. Traditional correction methods that rely on SAR or LiDAR data are costly and resource-intensive, especially in remote and environmentally fragile regions like the IHR. Therefore, a cost-effective alternative is urgently needed — one that leverages existing optical remote sensing data to effectively mitigate the impact of topographic shadows.

Solution: Our approach focuses on using the NDMI for topographic shadow correction in optical remote sensing data. The NDMI is particularly effective in minimizing the effects of topographic shadows, a common issue in the high-altitude terrain of the IHR. This index evaluates vegetation water content and mesophyll structural information, making it a robust indicator for assessing vegetation health, leaf internal structure, and dry matter content. Unlike the Normalized Difference Vegetation Index (NDVI), which has limitations in mountainous regions, the NDMI integrates information from both the NIR and SWIR bands to provide precise data for vegetation assessment in rugged terrains.

The formula for NDMI is:

NDMI=(NIR+SWIR)/(NIR-SWIR)

Higher values of the NDMI indicate dense canopy cover, while lower values suggest sparse vegetation. This approach enables precise mapping of forest density, fractional vegetation cover (FVC), and land cover in selected Van Panchayats of Uttarakhand and Community Conserved Areas (CCAs) of Nagaland. By utilizing a machine learning-based Linear Spectral Unmixing algorithm, our team has generated FVC maps that effectively minimize the effects of topographic shadows.

Key Features:

• Cost-Effectiveness: By eliminating the need for expensive SAR or LiDAR data, our method significantly reduces the financial burden of topographic shadow correction in high-altitude regions like Uttarakhand and Nagaland.
• Enhanced Vegetation Health Assessment: The SWIR-based NDMI is sensitive to changes in vegetation water content and structure, making it a powerful tool for monitoring subtle changes in vegetation health, stress, and biomass — especially in areas vulnerable to environmental stressors.
• Improved Discrimination of Vegetation Types: The SWIR bands enhance the ability to differentiate between various vegetation types and land cover, a critical advantage in the diverse ecosystems of high-altitude regions.
• Reduced Specular Reflection: SWIR wavelengths are less prone to specular reflection from water bodies and wet surfaces, reducing potential errors in land cover classification and change detection.
• Improved Cloud Penetration: SWIR’s ability to penetrate through thin clouds and mist allows for more reliable imaging in areas prone to frequent cloud cover, such as high-altitude regions.
• Better Detection of Understory Vegetation: SWIR wavelengths can penetrate thin vegetation canopies to detect understory vegetation and ground features, which is particularly beneficial in high-altitude forests with dense canopies.

Case Study: To validate our approach, we conducted case studies in the Van Panchayats of Gairsain, Gagas, and Mukteshwar ranges in Uttarakhand, as well as in the CCAs of Nagaland. These areas are characterized by significant topographic shadows that challenge land cover mapping, forest cover change analysis, and carbon stock assessment. By applying our NDMI-based methodology to freely available multispectral satellite imagery, we successfully corrected topographic shadows and generated highly accurate fractional vegetation cover maps, land cover maps, and forest biomass distribution maps. Ground truth data and comparative analysis with SAR and LiDAR-based methods confirmed the high accuracy and reliability of our approach.

Conventional NDVI Based Approach

SWIR Based Moisture Indices based Approach

Conclusion: In conclusion, our approach to topographic shadow correction using optical remote sensing data represents a transformative breakthrough for forest cover change analysis & forest biomass mapping in high-altitude terrain of Uttarakhand & Nagaland. By offering a cost-effective, precise, and scalable solution, we have redefined the way remote sensing is conducted in this challenging terrain, enabling informed decision-making and sustainable management of natural resources. This work underscores the power of innovation and collaboration in overcoming complex environmental challenges and driving positive change for the benefit of communities in IHR and beyond.