Field observations → Spectral indicators → PCA-weighted FHI → Validation → 2016–2025 forest-health dynamics Reliable assessment of forest health requires approaches that can integrate multidimensional field observations with spatially continuous environmental monitoring. However, standardized field assessments are inherently spatially limited, whereas satellite-based indicators may provide incomplete representations of forest condition when used without field validation. This study developed and evaluated an integrated framework for spatially explicit forest health assessment by combining standardized field observations with multi-temporal multispectral satellite imagery in the lowland Hyrcanian forests of northern Iran. Field surveys followed the USDA Forest Health Monitoring (FHM) protocol using a systematic 1 km × 1 km cluster sampling design comprising 42 sampling clusters and 168 plots. Four complementary dimensions of forest condition—Forest Diversity (FD), Crown and Stem Condition (CSC), Deadwood (DW), and Natural Regeneration (NR)—were integrated to derive a field-based Forest Health Score (FHS). Eight satellite-derived indicators were directionally standardized and objectively weighted using Principal Component Analysis (PCA) to construct a composite Forest Health Index (FHI). Landsat-8, Sentinel-2, and a fused Landsat-8/Sentinel-2 dataset were evaluated against field-derived FHS to identify the satellite approach providing the strongest representation of observed forest condition. The first principal component explained 72.4% of the total variance, with NDVI and MSAVI2 contributing most strongly to the composite index. Sentinel-2 showed the strongest agreement with field observations, explaining 63% of the variation in FHS (R² = 0.63), and outperformed both Landsat-8 and the fused dataset. Field observations revealed substantial spatial differences among the four forest districts, with Gisom-1 exhibiting the highest forest health and Tolaroud-3 showing the strongest evidence of ecological degradation. Multi-temporal Sentinel-2 assessment during 2016–2025 indicated a slight overall decline in forest health accompanied by pronounced spatial heterogeneity, including localized areas of improvement and degradation. The results demonstrate that field-validated satellite indicators can provide a scalable approach for translating multidimensional ecological observations into spatially continuous and temporally consistent information on forest condition. The proposed framework can support the identification of forest-health hotspots, prioritization of field inspections and management interventions, and long-term environmental monitoring in heterogeneous temperate forests.
@Roghayeh Jahdi
提示词
Field observations → Spectral indicators → PCA-weighted FHI → Validation → 2016–2025 forest-health dynamics Reliable assessment of forest health requires approaches that can integrate multidimensional field observations with spatially continuous environmental monitoring. However, standardized field assessments are inherently spatially limited, whereas satellite-based indicators may provide incomplete representations of forest condition when used without field validation. This study developed and evaluated an integrated framework for spatially explicit forest health assessment by combining standardized field observations with multi-temporal multispectral satellite imagery in the lowland Hyrcanian forests of northern Iran. Field surveys followed the USDA Forest Health Monitoring (FHM) protocol using a systematic 1 km × 1 km cluster sampling design comprising 42 sampling clusters and 168 plots. Four complementary dimensions of forest condition—Forest Diversity (FD), Crown and Stem Condition (CSC), Deadwood (DW), and Natural Regeneration (NR)—were integrated to derive a field-based Forest Health Score (FHS). Eight satellite-derived indicators were directionally standardized and objectively weighted using Principal Component Analysis (PCA) to construct a composite Forest Health Index (FHI). Landsat-8, Sentinel-2, and a fused Landsat-8/Sentinel-2 dataset were evaluated against field-derived FHS to identify the satellite approach providing the strongest representation of observed forest condition. The first principal component explained 72.4% of the total variance, with NDVI and MSAVI2 contributing most strongly to the composite index. Sentinel-2 showed the strongest agreement with field observations, explaining 63% of the variation in FHS (R² = 0.63), and outperformed both Landsat-8 and the fused dataset. Field observations revealed substantial spatial differences among the four forest districts, with Gisom-1 exhibiting the highest forest health and Tolaroud-3 showing the strongest evidence of ecological degradation. Multi-temporal Sentinel-2 assessment during 2016–2025 indicated a slight overall decline in forest health accompanied by pronounced spatial heterogeneity, including localized areas of improvement and degradation. The results demonstrate that field-validated satellite indicators can provide a scalable approach for translating multidimensional ecological observations into spatially continuous and temporally consistent information on forest condition. The proposed framework can support the identification of forest-health hotspots, prioritization of field inspections and management interventions, and long-term environmental monitoring in heterogeneous temperate forests.
模型
NANO-BANANA
类型
图片
点赞数
0
宽高比
1/1