Combining spatiotemporal climatological data, vegetation indices and solar-induced chlorophyll fluorescence to observe heat stress and water deficit in UK rapeseed

Abstract

Climate change introduces unprecedented abiotic challenges for agro-food sectors, making monitoring their interactions an imperative task. For UK rapeseed, little is understood about the spatiotemporal distributions of temperature stress and water deficit and their impacts on rapeseed yields. Moreover, whether these stresses are observable in remote sensing imagery has not been attempted at a large scale. This study identifies the spatiotemporal dynamics of heat stress, cold stress and water deficit for UK rapeseed cropping, and utilises spaceborne dataset including 5 km solar-induced fluorescence (SIF) based data (GOSIF, the apparent fluorescence yield SIF_yield and the physiological fluorescence yield Ф_F) and 500 m Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation indices (VIs) and gross primary productivity (GPP) and 10 m Sentinel-2 (S2) VIs to track the stresses and assess crop productivity.
Results demonstrate that from 1961 to 2020 heat stress was increasing for flowering and the whole of the reproductive period, with southeastern UK experiencing higher levels of stress. Cold stress was decreasing and negatively correlated to latitude. Particularly, during flowering heat stress was increasing at a faster rate than cold stress was decreasing, causing an increasing potential production loss with differences between UK regions (p < 0.05). Additionally, during flowering for 1981-2020, both water deficit (p < 0.05) and heat stress were increasing slowly across UK regions, with heat stress occurring over a larger area. The stresses had overall negative relationships with rapeseed yields during 2016-2022, with water deficit indices having greater explanatory power and UK regional relationships being stronger than at national scale. Remote sensing analyses revealed that the studied remotely sensed data excluding S2_EVI_10, GPP and GOSIF had higher sensitivity to rapeseed’s stresses during annual flowering at the UK level during 2016-2022. At the 8-day scale for both national and regional levels, fluorescence yield generally exhibited negative relationships with identified acute stresses and had faster responses particularly within flowering than VIs, GPP and GOSIF. Water deficit indices possessed stronger relationships with remote sensing data than heat stress metrics particularly for the flowering period. Despite regional differences, MODIS VIs particularly MODIS_NDVI_500 (R² ~ 0.60) exhibited less variable and stronger regression relationships for rapeseed yields of UK regions including those where rapeseed has the largest cultivation areas, indicating greater abilities to explain yield variability and value for yield estimation. Combining the findings of stress detection and yield prediction, these results highlight the competitiveness and applicability of fluorescence yield (Ф_F and SIF_yield) for large spatial scale acute stress assessment and MODIS VIs for production estimation for crops including rapeseed. However, at the 8-day scale apart from the low levels of stresses observed for UK rapeseed fields, the mismatch of spatial resolutions of stress and remote sensing data might affect the assessment of their relationships. This work highlights the ongoing impacts of climatic warming on crop production and the utility of remote sensing for monitoring this.

Date of Award	2025
Original language	English
Awarding Institution	University of Dundee
Supervisor	Alexandra Morel (Supervisor) & Mark Cutler (Supervisor)