02703naa a2200253 a 450000100080000000500110000800800410001902200140006002400530007410000220012724501390014926000090028852019060029765000120220365000290221565000210224465000130226565000120227865300110229070000200230170000180232170000260233977300840236521497272023-02-14       2022    bl uuuu     u00u1 u    #d  a2352-93857 ahttps://doi.org/10.1016/j.rsase.2022.1008672DOI1 aLIMA, G. S. A. de  aCarbon estimation in an integrated crop-livestock system with imaging sensors aboard unmanned aerial platforms.h[electronic resource]  c2022  aBiomass and carbon estimates in integrated crop-livestock (iLP) systems remain scarce. Recently, sensors aboard an unmanned aerial vehicle (UAV) have been used for such estimation and can represent an efficient and low-cost method for acquiring accurate remote-sensing data. This study aimed to estimate the relationship between vegetation indices (VIs) and carbon stock in an upland rice field intercropped with Brachiaria, an exotic grass species in the savanna ecosystem of Brazil (Cerrado biome), using multispectral aerial imaging with a high cartographic scale. The research was conducted in an experimental iLP system located at the Brazilian Agricultural Research Corporation (Embrapa Arroz e Feijão), Goiás State. During the data collection stage, flights were conducted with a fixed-wing UAV (eBee Plus RTK) on the same day or close to collecting biomass samples in the field. In the processing step, a correlation was performed between the normalized difference vegetation index (NDVI), simple ratio (SR), soil-adjusted vegetation index (SAVI), and modified photochemical reflectance index (MPRI) VIs and the biomass measured in the field. The results indicated that the vegetation index (VI) that used the near-infrared (NIR) band showed a stronger correlation with biomass than the VI that used only the visible band information. All VIs were significantly correlated with the estimated field biomass (P &lt 0.05) using Pearson's correlation. The regression models were able to predict the biomass and carbon stock in Phase 1 (3.93 Mg ha-1, 1.66 Mg ha-1, R2 = 0.748) and Phase 2 (7.53 Mg ha-1, 3.19 Mg ha-1, R2 = 0.816). From the best model selected for each phase, maps with the spatial and temporal distribution of the biomass of the iLP system were plotted, demonstrating one of the advantages of employing remote sensing techniques based on drones and multispectral imaging sensors.  aBiomass  aUnmanned aerial vehicles  aVegetation index  aBiomassa  aCarbono  aDrones1 aFERREIRA, M. E.1 aMADARI, B. E.1 aCARVALHO, M. T. de M.  tRemote Sensing Applications: Society and Environmentgv. 28, 100867, Nov. 2022.