EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Spatially balanced subsampling improves stand-level leaf area index estimation in mature Norway spruce stands

Jakub Černý, Jaroslav Čepl, Jan Cukor, Zdeněk Vacek, Jan Stejskal, Holger Lange

Frontiers in Forests and Global Change · 01.07.2026 · 9

DOI: 10.3389/ffgc.2026.1864662 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Frontiers in Forests and Global Change, online erschienen am 01.07.2026. Verfasst von Jakub Černý und 5 weiteren Autor:innen. Die Arbeit nennt 66 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Introduction

Leaf area index (LAI) estimation is sensitive to sensor field of view (FOV), within-plot spatial heterogeneity, and sampling layout. Because LAI influences canopy radiation transmission, microclimate and vegetation–atmosphere exchange, robust field estimation is important for biometeorological and ecosystem research.

Methods

We evaluated these effects in mature Norway spruce [ Picea abies (L.) H. Karst] stands in the Czech Republic across 15 sites at elevations of 407–1,019 m a.s.l., using a combined gap-fraction approach based on LAI-2200 PCA measurements and digital hemispherical photography. At each site, a measured 9 × 9 grid of 81 below-canopy measurement points with 2-m spacing was used as an operational within-plot benchmark for mean optically derived LAI and spatial structure. Monte Carlo subsampling was then used to compare how reduced layouts, including random, row-wise, column-wise, block-wise, and spatially balanced block–row–column layouts, reproduced the full-grid benchmark mean.

Results

Narrower FOVs produced higher stand-level optically derived LAI estimates and greater within-plot variability. The full grids also showed directional spatial structure, with stronger autocorrelation along the north–south column direction than along the west–east row direction, although this contrast weakened under the narrowest FOV. Reduced layouts with more even spatial coverage outperformed random sampling. The spatially balanced block–row–column layout performed most consistently, whereas the row-wise layout provided little improvement. Approximately 25–36 well-distributed below-canopy measurements were sufficient to keep reduced-layout LAI deviations within 0.15 m 2 m −2 of the full-grid benchmark, whereas 7–9 measurements were enough to remain within 5% of this benchmark in most stands.

Discussion

Spatially balanced sampling can therefore improve the robustness and efficiency of stand-level optically derived LAI estimation, with relevance to forest biometeorology, ecosystem monitoring, and the validation of satellite-derived LAI products.

Zitieren

APA

Černý, J., Čepl, J., Cukor, J., Vacek, Z., Stejskal, J., & Lange, H. (2026). Spatially balanced subsampling improves stand-level leaf area index estimation in mature Norway spruce stands. Frontiers in Forests and Global Change, 9. https://doi.org/10.3389/ffgc.2026.1864662

Vancouver

Černý J, Čepl J, Cukor J, Vacek Z, Stejskal J, Lange H. Spatially balanced subsampling improves stand-level leaf area index estimation in mature Norway spruce stands. Frontiers in Forests and Global Change. 2026;9. doi:10.3389/ffgc.2026.1864662

BibTeX
@article{ern2026,
  author = {Černý, Jakub and Čepl, Jaroslav and Cukor, Jan and Vacek, Zdeněk and Stejskal, Jan and Lange, Holger},
  title = {Spatially balanced subsampling improves stand-level leaf area index estimation in mature Norway spruce stands},
  journal = {Frontiers in Forests and Global Change},
  year = {2026},
  volume = {9},
  doi = {10.3389/ffgc.2026.1864662}
}
Kostenlos starten →