The experience of using NDVI and color vegetation indexes of UAVs to identify phytocenotic parameters of steppe vegetation

Abstract

The spatial and temporal variability in the development of steppe fires is largely due to the peculiarities of seasonal, interannual and long-term vegetation conditions. At the same time, there is no objective understanding of which phytocenotic parameters of steppe vegetation are recorded by spectral vegetation indices (SVI) based on Earth remote sensing materials. In this regard, the purpose of the study was to identify these correspondences, as well as to test aerial photography materials as an additional source of detailed information about the spatial and temporal structure of vegetation cover. In the area located in the foothill steppes of the Southern Urals, field studies were conducted in 2023–2024, during which vegetation descriptions and mowing were carried out monthly during the growing season (April–October), aerial photography using a UAV with a standard RGB-sensor. Vegetation is represented by the Stipa zalesskii–Festuca valesiaca–Stipa capillata community and its pasture-degraded variant Ceratocarpus arenarius–Potentilla bifurca. The objective advantage of aerial photographs and color vegetation index (CVI) NDI, VARI, ExG, GLI, ExGR and ExR distribution schemes based on them is their high level of detail, which makes it possible to assess the features of spatial differentiation of vegetation cover, the direction of seasonal and long-term changes. It was revealed that the values of NDVI Sentinel-1–2 and CVI were most often highly correlated with the projective cover of green vegetation. A close relationship was observed with other phytocenotic indicators (total projected coverage, total phytomass reserves and green vegetation) in the hydrothermal conditions favorable for vegetation in 2024 due to the abundance of green vegetation. The results obtained suggest that the main factors of discrepancies between spectral and color indexes and the actual parameters of steppe communities are: a) the overlap of green vegetation with dead phytomass; b) the small share of diverse grasses in the structure of steppe communities; c) seasonal differences in the spectral response of dominant plants and in the aspects they create; d) morphological features of plant species, the volumetric structure of vegetation cover created by them. The revealed phenological reactions of communities to the features of hydrothermal conditions in 2023 and 2024 indicate a high degree of interannual and seasonal variability in their fire status.

Plant phenotyping relevance

UAV RGB画像、色植生指数、衛星NDVIを用いて植生被覆・植物量などの植物群落特性を推定し、現地測定との対応関係を検証しており、植物形質取得法が研究の中心です。

Code and data availability

The article describes UAV RGB imagery, field phytomass mowings, and CVI/NDVI analyses for two steppe sites, but contains no data availability statement, no public repository deposit, no author code URL, and no supplement reference. No paper-specific public asset is identifiable.

No evidence-backed public reproduction asset is currently recorded.