Advancing fire regime assessment in the Indian Himalayan forests using GEDI LiDAR

Abstract

Wildfires are the leading cause of forest degradation and reduce the carbon sequestration capacity of forest ecosystems. As climate change intensifies fire regimes, accurately mapping burn severity and biomass loss remains difficult, particularly in complex mountainous terrain. Traditional remote sensing approaches based on optical data often fail to capture sub-canopy and structural variations, while field-based methods are constrained by scale, access, and logistical complexity. To address these limitations, this study utilises NASA’s Global Ecosystem Dynamics Investigation (GEDI) LiDAR data to capture three-dimensional forest structure and monitor changes in aboveground biomass. Focusing on the Indian Himalayan Region (IHR), we utilised GEDI structural metrics for fire severity classifications derived from NASA’s Fire Information for Resource Management System (FIRMS) data. Spatial fire density mapping revealed that Evergreen Needleleaf (ENL), Evergreen Broadleaf (EBL), and Deciduous Broadleaf (DBL) plant functional types (PFTs) are particularly vulnerable to recurring fires. The GEDI Level 4 A AGBD product, when tested in the IHR, showed a tendency to underestimate biomass relative to regional field studies, indicating the need for localised calibration in rugged terrain. We found that the pre-monsoon fire season resulted in a biomass loss of 14.33 Mg/ha for the ENL PFT, while the EBL PFT showed a biomass gain of 3.15 Mg/ha. A Random Forest model utilising GEDI structural inputs for fire severity classification achieved an overall accuracy of 85 %. We found that vegetation structural complexity, rather than canopy height, is a more important predictor of high fire severity. We also evaluated the Waveform Structural Complexity Index (WSCI), a novel GEDI metric that effectively characterised fire severity across the under-storey vegetation. These findings underscore the importance of incorporating 3D vegetation structure into fire impact assessments, moving beyond the limitations of conventional two-dimensional optical datasets.

Plant phenotyping relevance

GEDI LiDARによる森林構造・バイオマス損失・火災重症度の推定を中心に、既存AGBD製品の地域較正と新規指標WSCIの評価を行っており、植物状態の取得・推定手法が中核である。

Code and data availability

公開状態または取得可能な本文経路を確認できませんでした。

No evidence-backed public reproduction asset is currently recorded.