Nonlinear Regression Models and Applications

Fernando Miguez, Sotirios Archontoulis, Hamze Dokoohaki

Research output: Chapter in Book/Report/Conference proceedingChapter


Nonlinear statistical models continue to be relevant because they have the advantages of simplicity, flexibility, and parsimony when used in the analysis and description of agricultural data. Often nonlinear models can be easier to interpret than other alternatives. These models have been applied to data ranging from leaf‐level to landscape‐level processes, such as photosynthesis, biomass accumulation, soil CO2 efflux, and temperature responses. Our objective is to provide a multistep framework that helps with the application of these models with specific considerations in each step. Even though objectives and applications may vary, common priorities should be set when fitting nonlinear models: obtaining acceptable parameter estimates, a good model fit, and meeting standard assumptions of statistical models. The following steps are considered: (i) choose candidate models, (ii) set starting values, (iii) fit models, (iv) check convergence and parameter estimates, (v) find the “best” model among competing models, (vi) check model assumptions (residual analysis), and (vii) calculate statistical indexes and confidence intervals. In particular, we emphasize the first step (choose candidate models) by providing an extensive library of nonlinear functions (77 equations with the associated parameter interpretations) and application examples in agriculture. We hope that this contribution will clarify some of the difficulties and confusion with using nonlinear models.
Original languageEnglish (US)
Title of host publicationApplied Statistics in Agricultural, Biological, and Environmental Sciences
EditorsBarry Glaz, Kathleen M Yeater
ISBN (Electronic)9780891183600
ISBN (Print)9780891183594
StatePublished - Aug 23 2018
Externally publishedYes

Publication series

NameASA, CSSA, and SSSA Books
ISSN (Electronic)2691-2341

Fingerprint Dive into the research topics of 'Nonlinear Regression Models and Applications'. Together they form a unique fingerprint.

Cite this