Scholarlink

<p>Climate change is shifting the structure and composition of vegetation across the northern forest-tundra ecotone. Multi-spectral satellite images are frequently used to model vegetation dynamics at landscape to continental scales. However, validating these models over large areas remains challenging. As a result, analyses conducted at fine spatial scales are needed to interpret changes derived from medium-resolution satellite imagery. In this research, we used high-resolution air photographs (<i>n</i> = 27) captured in the 1930’s and 1970’s paired with contemporary satellite images from the 2020’s to map forest and shrub expansion across four transects crossing the northern treeline in Northwest Territories, Canada. We used these data to compare the likelihood of forest and shrub expansion in relation to distance to the forest limit, climate and fire history. We found that shrub and forest expansion was greater in close proximity to, or above the forest limit at three of the four transects. Forest advance was most limited at our northernmost transect, which was drier, contained white spruce (<i>Picea glauca</i>) woodlands, and had steeper topographic gradients. Vegetation dynamics at the southern sites were driven by large fires and recovery within closed black spruce (<i>Picea mariana</i>) forests. To better understand how well multi-spectral satellite indices can capture these dynamics, we then compared photo-interpreted change to a classification of vegetation change using Landsat time-series from 1985 to 2021, providing a direct link between two widely used change detection methods with different spatial scales and characteristics. Agreement between datasets varied by vegetation type and transect. Correlation was higher for increases in forest cover vs shrub cover and when using the 1970’s historical photos vs the 1930’s photos. These results suggest that vegetation change occurring between the 1930’s and 1970’s decreased correlation with the satellite record beginning in 1985. These findings provide a multi-site comparison of long-term vegetation dynamics across the Canadian forest-tundra ecotone and highlight the use of historic aerial photographs to corroborate medium-spatial-resolution satellite time-series.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Long-Term Vegetation Changes Across the Canadian Forest-Tundra Ecotone Using Historic Air Photos and Satellite Time-Series

  • 期刊论文

  • 发表日期：

    2026年4月30日

查看全文
  • H. Travers-Smith,
  • N. C. Coops,
  • M. A. Wulder,
  • T. C. Lantz,
  • D. D. Ignace

摘要

Climate change is shifting the structure and composition of vegetation across the northern forest-tundra ecotone. Multi-spectral satellite images are frequently used to model vegetation dynamics at landscape to continental scales. However, validating these models over large areas remains challenging. As a result, analyses conducted at fine spatial scales are needed to interpret changes derived from medium-resolution satellite imagery. In this research, we used high-resolution air photographs (n = 27) captured in the 1930’s and 1970’s paired with contemporary satellite images from the 2020’s to map forest and shrub expansion across four transects crossing the northern treeline in Northwest Territories, Canada. We used these data to compare the likelihood of forest and shrub expansion in relation to distance to the forest limit, climate and fire history. We found that shrub and forest expansion was greater in close proximity to, or above the forest limit at three of the four transects. Forest advance was most limited at our northernmost transect, which was drier, contained white spruce (Picea glauca) woodlands, and had steeper topographic gradients. Vegetation dynamics at the southern sites were driven by large fires and recovery within closed black spruce (Picea mariana) forests. To better understand how well multi-spectral satellite indices can capture these dynamics, we then compared photo-interpreted change to a classification of vegetation change using Landsat time-series from 1985 to 2021, providing a direct link between two widely used change detection methods with different spatial scales and characteristics. Agreement between datasets varied by vegetation type and transect. Correlation was higher for increases in forest cover vs shrub cover and when using the 1970’s historical photos vs the 1930’s photos. These results suggest that vegetation change occurring between the 1930’s and 1970’s decreased correlation with the satellite record beginning in 1985. These findings provide a multi-site comparison of long-term vegetation dynamics across the Canadian forest-tundra ecotone and highlight the use of historic aerial photographs to corroborate medium-spatial-resolution satellite time-series.