UAV-Based Estimation of Carbon Exports from Heterogeneous Soil Landscapes--A Case Study from the CarboZALF Experimental Area.

Abstract

The advantages of remote sensing using Unmanned Aerial Vehicles (UAVs) are a high spatial resolution of images, temporal flexibility and narrow-band spectral data from different wavelengths domains. This enables the detection of spatio-temporal dynamics of environmental variables, like plant-related carbon dynamics in agricultural landscapes. In this paper, we quantify spatial patterns of fresh phytomass and related carbon (C) export using imagery captured by a 12-band multispectral camera mounted on the fixed wing UAV Carolo P360. The study was performed in 2014 at the experimental area CarboZALF-D in NE Germany. From radiometrically corrected and calibrated images of lucerne (Medicago sativa), the performance of four commonly used vegetation indices (VIs) was tested using band combinations of six near-infrared bands. The highest correlation between ground-based measurements of fresh phytomass of lucerne and VIs was obtained for the Enhanced Vegetation Index (EVI) using near-infrared band b899. The resulting map was transformed into dry phytomass and finally upscaled to total C export by harvest. The observed spatial variability at field- and plot-scale could be attributed to small-scale soil heterogeneity in part.

Plant phenotyping relevance

UAVマルチスペクトル画像と植生指数を用いてルーサンの生体 phytomass を推定し、地上測定との相関で手法性能を比較・検証しているため、植物形質取得が中心的である。

Code and data availability

The supplied blocks describe UAV multispectral imagery, ground-based phytomass/carbon measurements, and image-processing workflows, but contain no public dataset deposit, no author code release, and no supplement with data. The only URLs present are the CC-BY license links, which are not paper-specific assets.

No evidence-backed public reproduction asset is currently recorded.