Improving estimation of in-season crop water use and health of wheat genotypes on sodic soils using spatial interpolation techniques and multi-component metrics
Article
Article Title | Improving estimation of in-season crop water use and health of wheat genotypes on sodic soils using spatial interpolation techniques and multi-component metrics |
---|---|
ERA Journal ID | 5245 |
Article Category | Article |
Authors | Choudhury, Malini Roy (Author), Mellor, Vincent (Author), Das, Sumanta (Author), Christopher, Jack (Author), Apan, Armando (Author), Menzies, Neal W. (Author), Chapman, Scott (Author) and Dang, Yash P. (Author) |
Journal Title | Agricultural Water Management |
Journal Citation | 255 |
Article Number | 107007 |
Year | 2021 |
Publisher | Elsevier |
Place of Publication | Netherlands |
ISSN | 0378-3774 |
1873-2283 | |
Digital Object Identifier (DOI) | https://doi.org/10.1016/j.agwat.2021.107007 |
Web Address (URL) | https://www.sciencedirect.com/science/article/pii/S0378377421002729 |
Abstract | Crop water use can be a useful indicator of in-season crop conditions including health and greenness on sodic soils. However, in-field manual measurements of crop water use can be tedious, time and labour-intensive and may also exhibit large spatial variability, and inefficient for surveying large areas. Here we propose a novel approach to estimate in-field crop water use of wheat using spatial interpolation techniques at critical crop development stages (tillering and close to flowering) on a moderately sodic and a highly sodic site in southern Queensland, Australia. The six spatial interpolation techniques; ordinary kriging, empirical Bayes kriging, inverse distance weighting, spline, local polynomial interpolation, and radial basis function were employed on ground-measured crop water use data and compared to accurately estimate the spatial distribution of crop water use. Initially, the comparison was made using root mean square error, mean absolute error, and coefficient of determination values using cross-validation. For a more clear comparison and to account for complexities of site-specific crop water use distribution, the standardized, multi-component model performance efficiency metrics, i.e. Kling-Gupta efficiency and spatial efficiency were adopted and tested over cross-validation techniques. Results showed that Kling-Gupta efficiency was slightly better at selecting the optimum interpolation models for crop water use estimation at a field scale than the spatial efficiency metric. Spline and local polynomial interpolation were relatively better estimators of crop water use for both the sites at tillering, while the radial basis function and spline were superior close to flowering. The estimated crop water use was positively and closely associated with seasonal GreenSeeker® normalized difference vegetation index data (R2 = 0.49 and 0.39 at tillering; and R2 = 0.71 and 0.62 at close to flowering, for the moderately sodic and highly sodic sites, respectively). The research improves our understanding of selecting appropriate spatial interpolation methods for in-season crop water use estimation that could help detect changes in crop health due to in-field and seasonal variations of crop water use in a sodic soil environment. |
Keywords | cross-validation; Kling-Gupta efficiency; spatial efficiency; normalized difference vegetation index; crop development stage; seasonal variations |
ANZSRC Field of Research 2020 | 490507. Spatial statistics |
401399. Geomatic engineering not elsewhere classified | |
300206. Agricultural spatial analysis and modelling | |
Public Notes | Files associated with this item cannot be displayed due to copyright restrictions. |
Byline Affiliations | University of Queensland |
School of Civil Engineering and Surveying | |
Institution of Origin | University of Southern Queensland |
https://research.usq.edu.au/item/q6779/improving-estimation-of-in-season-crop-water-use-and-health-of-wheat-genotypes-on-sodic-soils-using-spatial-interpolation-techniques-and-multi-component-metrics
138
total views7
total downloads9
views this month0
downloads this month