Articles
Detection and simulation of water stress in processing tomato
Article number
1351_7
Pages
39 – 46
Language
English
Abstract
Water stress inducing open field experiments were conducted for 4 years at two different locations in Hungary (in 2017-2019 at Szarvas and in 2021 at Gödöllő) with processing tomato.
Different water supply levels were set regarding 50%, 75% and 100% of crop evapotranspiration (ETc) and were compared to control with no regular irrigation.
Different measurements were performed to detect and monitor water stress in plants.
Thermal camera was used for measuring leaf surface temperature and then for the computation of crop water stress index (CWSI). Hyperspectral spectrometer was used for reflectance measurements and then vegetation indices were computed from the data.
Crop growth and water related stress were simulated in the AquaCrop model.
Simulation of biomass and fruit yield was fairly accurate at harvest (nRMSE=11.6), but significant errors were found during mid-season predictions in some cases.
Prediction errors for the whole season were increasing with the growing stress levels.
Good correlation could not be found between CWSI and modelled stress indicators, but trends can be identified in the case of severe stress treatment.
CWSI was able to detect different levels of water stress especially in 2017 and 2018. Differences were lower in 2019, because of the lower level of stress.
More data are needed regarding the use of reflectance indices but there are promising results in the case of mNDVI and CLgreen.
Different water supply levels were set regarding 50%, 75% and 100% of crop evapotranspiration (ETc) and were compared to control with no regular irrigation.
Different measurements were performed to detect and monitor water stress in plants.
Thermal camera was used for measuring leaf surface temperature and then for the computation of crop water stress index (CWSI). Hyperspectral spectrometer was used for reflectance measurements and then vegetation indices were computed from the data.
Crop growth and water related stress were simulated in the AquaCrop model.
Simulation of biomass and fruit yield was fairly accurate at harvest (nRMSE=11.6), but significant errors were found during mid-season predictions in some cases.
Prediction errors for the whole season were increasing with the growing stress levels.
Good correlation could not be found between CWSI and modelled stress indicators, but trends can be identified in the case of severe stress treatment.
CWSI was able to detect different levels of water stress especially in 2017 and 2018. Differences were lower in 2019, because of the lower level of stress.
More data are needed regarding the use of reflectance indices but there are promising results in the case of mNDVI and CLgreen.
Publication
Authors
S. Takács, Z. Pék, T. Bíró, P. Szuvandzsiev, G. Palotás, E. Czinkoczki, L. Helyes
Keywords
AquaCrop, CWSI, infrared, thermography, index, reflectance, vegetation indices
Online Articles (42)