Field-level analysis of phenological cycles and dynamics of sunflower (Helianthus annuus L.) and oil seed rape (Brassica napus L.) flowering within various regions of Hungary

Phenological observations are expensive and demanding in terms of manpower to monitor the vegetation stages. Therefore, satellite products have opened new possibilities for easier and more widespread data collection. Nowadays biomass estimations extensively rely on these tools, given their extensive spatial and temporal coverage, which are defined by indicators such as vegetation indices, which describe the biomass growth, canopy structure, vegetation health and even water management etc. However, the detection of flowering stages through remote sensing is less explored, with fewer established methods available.

This study investigates temporal phenological changes during the blooming period of the most widely cultivated oilseed crops in Hungary in 2021, specifically the sunflower (Helianthus annuus L.) and the winter-cultivated oilseed rape (Brassica napus L.). The objective is to characterize the blooming phase and dynamics of these two crop species utilizing various vegetation indexes and satellite-derived products. The investigation is conducted across seven distinct regions, using honey bees as bioindicators of the fields.

Methodologies outlined in prior scientific literature, focusing on the analysis of anthesis timing and duration in major nectar-producing crops utilizing Sentinel-1 SAR and Sentinel-2 optical products, serve as the basis for this research. Within each radius study areas, the parcel-averaged and smoothed daily time series were acquired. The estimation of the blooming phases was achieved through parcel smoothing methods using daily non-parametric local regression (loess) approach which showed better performance compared to the Savitzky-Golay (SG) algorithm. In our flowering detection analysis, we also examined the differences in the ascending and descending orbits and their combined results. In the case of oil seed rape, the NDVI index reached its maximum after flowering, while for sunflower it varied. Additionally, we investigated the outcomes of all polarization and method combinations within each crop type.

Our study enables the comprehension of temporal flowering patterns in bee pasture crops through the integration of SAR and optical measurements. Additionally, it supports the utilization of beehive scales to provide field-based reference data for estimating anthesis.

The research was funded by the National Multidisciplinary Laboratory for Climate Change, RRF-2.3.1-21-2022-00014 project. Project No. 993788 has been implemented with the support provided by the Ministry of Culture and Innovation of Hungary from the National Research, Development and Innovation Fund, financed under the KDP-2020 funding scheme.