TY - JOUR
T1 - Which multispectral indices robustly measure canopy nitrogen across seasons
T2 - Lessons from an irrigated pasture crop
AU - Patel, Manish Kumar
AU - Ryu, Dongryeol
AU - Western, Andrew W.
AU - Suter, Helen
AU - Young, Iain M.
N1 - Funding Information:
This work was supported by the Australian Government Department of Agriculture, Water and the Environment as part of its Rural R&D for Profit program (grant No. RRDP1715); the Australia-China Joint Research Centre for Healthy Soils for Sustainable Food Production and Environmental Quality (grant No. ACSRF48165); and Dairy Australia. We are thankful to Dr. Alexis Pang, Mr. Graeme Ward and Dr. Oxana Belyaeva who helped in maintaining the crop condition and organising the ground data collection. The authors would also like to acknowledge the timely services provided by TrACEES platform in chemical analysis at the University of Melbourne.
Funding Information:
This work was supported by the Australian Government Department of Agriculture, Water and the Environment as part of its Rural R&D for Profit program (grant No. RRDP1715); the Australia-China Joint Research Centre for Healthy Soils for Sustainable Food Production and Environmental Quality (grant No. ACSRF48165); and Dairy Australia. We are thankful to Dr. Alexis Pang, Mr. Graeme Ward and Dr. Oxana Belyaeva who helped in maintaining the crop condition and organising the ground data collection. The authors would also like to acknowledge the timely services provided by TrACEES platform in chemical analysis at the University of Melbourne.
Publisher Copyright:
© 2021 Elsevier B.V.
PY - 2021/3
Y1 - 2021/3
N2 - In precision farming, accurate estimation of canopy nitrogen concentration (CNC) is valuable for effective crop growth monitoring and nitrogen (N) fertiliser management. To date, many canopy multispectral indices have been proposed as indicators for CNC; however, many of these indices have also shown sensitivity to biomass and their performance drops at high biomass levels. Dependence on growth stage, season, or other environmental conditions limits their efficacy as generalized CNC indices. The objectives of this study were to assess the robustness of popular CNC indices across a wide range of biomass levels and fertiliser application levels; and for two contrasting seasons – winter and summer. To achieve this, we analysed the efficacy of seven canopy nitrogen indices, including canopy chlorophyll content index (CCCI), together with eleven other commonly used spectral indices. We used canopy level solar-induced hyperspectral reflectance data acquired using a hand-held optical spectroradiometer across four growth stages in winter (May-June 2018) and four in summer (January-February 2019) from an experimental field of irrigated perennial ryegrass with variable N application in Victoria, Australia. The field contained 40 plots, each with one of eight different N treatments. Almost all the indices exhibited similar correlation to CNC (%) when applied to individual stages (days) in both winter and summer; however, relationships between CNC and individual indices varied significantly between stages. We obtained similar results for canopy biomass. When the data across the entire range of growth stages and seasons were combined, the correlations between most canopy nitrogen indices and CNC became weak (R2 < 0.25, 0.9% ≤ RMSE ≤ 1.0%). PRI exhibited the highest correlation with CNC (R2 = 0.58, RMSE = 0.7%) for the combined data set. Even so, PRI's association with CNC and canopy biomass changed with the season. Most indices responded to both CNC and biomass simultaneously, and this confounds the estimation of CNC due to strong but growth stage-specific relationships between CNC and canopy biomass. This study shows that it is important to consider a wide range of conditions when evaluating multispectral CNC indices.
AB - In precision farming, accurate estimation of canopy nitrogen concentration (CNC) is valuable for effective crop growth monitoring and nitrogen (N) fertiliser management. To date, many canopy multispectral indices have been proposed as indicators for CNC; however, many of these indices have also shown sensitivity to biomass and their performance drops at high biomass levels. Dependence on growth stage, season, or other environmental conditions limits their efficacy as generalized CNC indices. The objectives of this study were to assess the robustness of popular CNC indices across a wide range of biomass levels and fertiliser application levels; and for two contrasting seasons – winter and summer. To achieve this, we analysed the efficacy of seven canopy nitrogen indices, including canopy chlorophyll content index (CCCI), together with eleven other commonly used spectral indices. We used canopy level solar-induced hyperspectral reflectance data acquired using a hand-held optical spectroradiometer across four growth stages in winter (May-June 2018) and four in summer (January-February 2019) from an experimental field of irrigated perennial ryegrass with variable N application in Victoria, Australia. The field contained 40 plots, each with one of eight different N treatments. Almost all the indices exhibited similar correlation to CNC (%) when applied to individual stages (days) in both winter and summer; however, relationships between CNC and individual indices varied significantly between stages. We obtained similar results for canopy biomass. When the data across the entire range of growth stages and seasons were combined, the correlations between most canopy nitrogen indices and CNC became weak (R2 < 0.25, 0.9% ≤ RMSE ≤ 1.0%). PRI exhibited the highest correlation with CNC (R2 = 0.58, RMSE = 0.7%) for the combined data set. Even so, PRI's association with CNC and canopy biomass changed with the season. Most indices responded to both CNC and biomass simultaneously, and this confounds the estimation of CNC due to strong but growth stage-specific relationships between CNC and canopy biomass. This study shows that it is important to consider a wide range of conditions when evaluating multispectral CNC indices.
KW - Canopy biomass
KW - Canopy nitrogen concentration
KW - N uptake
KW - PRI
KW - Vegetation index
UR - http://www.scopus.com/inward/record.url?scp=85100658597&partnerID=8YFLogxK
U2 - 10.1016/j.compag.2021.106000
DO - 10.1016/j.compag.2021.106000
M3 - Article
AN - SCOPUS:85100658597
SN - 0168-1699
VL - 182
JO - Computers and Electronics in Agriculture
JF - Computers and Electronics in Agriculture
M1 - 106000
ER -
