Skip to main content
ARS Home » Plains Area » Lubbock, Texas » Cropping Systems Research Laboratory » Wind Erosion and Water Conservation Research » Research » Publications at this Location » Publication #327648
Title: A taxonomy-based approach to shed light on the babel of mathematical analogies for rice simulation

Author
item CONFALONIERI, ROBERTO - University Of Milan
item BREGAGLIO, SIMONE - University Of Milan
item MYRIAM, ADAM - Cirad, France
item RUGET, FRANCOISE - Inland Northwest Research Alliance, Inra
item LI, TAO - International Rice Research Institute
item HASEGAWA, TOSHIHIRO - National Institute For Agro-Environmental Sciences
item YIN, XINYOU - Wageningen University
item ZHU, YAN - Nanjing Agricultural University
item BOOTE, KENNETH - University Of Florida
item BUIS, SAMUEL - Inland Northwest Research Alliance, Inra
item FUMOTO, TAMON - National Institute For Agro-Environmental Sciences
item GAYDON, DONALD - Commonwealth Scientific And Industrial Research Organisation (CSIRO)
item LAFARGE, TANGUY - Cirad, France
item MARCAIDA, MANUEL - International Rice Research Institute
item NAKAGAWA, HIROSHI - National Agriculture And Food Research Organization (NARO), Agricultrual Research Center
item RUANE, ALEX - Nasa Goddard Institute For Space Studies
item SINGH, BALWINDER - Indian Agricultural Research Institute
item SINGH, UPENDRA - International Fertilizer Development Center (IFDC)
item TANG, LIANG - Nanjing Agricultural University
item TAO, FULU - Chinese Academy Of Sciences
item FUGICE, JOB - International Fertilizer Development Center (IFDC)
item HIROE, YOSHIDA - National Agricultural Research Organization - Japan (NARO)
item ZHANG, ZHAO - Beijing Normal University
item WILSON, LLOYD - Texas A&M Agrilife
item Baker, Jeffrey
item YANG, YUBIN - Texas A&M Agrilife
item MASUTOMI, YUJI - Ibaraki University
item WALLACH, DANIEL - Inland Northwest Research Alliance, Inra
item ACUTIS, MARCO - University Of Milan
item BOUMAN, BAS - International Rice Research Institute

Submitted to: Environmental Modelling & Software
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/8/2016
Publication Date: 9/16/2016
Citation: Confalonieri, R., Bregaglio, S., Myriam, A., Ruget, F., Li, T., Hasegawa, T., Yin, X., Zhu, Y., Boote, K., Buis, S., Fumoto, T., Gaydon, D., Lafarge, T., Marcaida, M., Nakagawa, H., Ruane, A.C., Singh, B., Singh, U., Tang, L., Tao, F., Fugice, J., Hiroe, Y., Zhang, Z., Wilson, L.T., Baker, J.T., Yang, Y., Masutomi, Y., Wallach, D., Acutis, M., Bouman, B. 2016. A taxonomy-based approach to shed light on the babel of mathematical analogies for rice simulation. Environmental Modelling & Software. 85: 332-341.

Interpretive Summary: Carbon dioxide in the Earth’s atmosphere is increasing mainly because humans burn fossil fuels for energy. Along with this rise in CO2 are projections on rising air temperatures. Globally, rice accounts for the majority of human food consumption. Rice is also known to be highly sensitive to both atmospheric CO2 and air temperature although different rice varieties can show wide differences in their response to CO2 and temperature. Because the types of precise environmental control in experimental system required to fully elucidate Rice’s potential response to these climate changes are often prohibitive due to cost constraints, some scientists resort to computer simulation models to arrive at answers concerning the World’s Food Security. This paper utilizes advanced statistical techniques to classify 13 different rice crop simulation models from around the world into 5 general clusters. They hypothesize that user subjectivity during model calibration often distorts simulations when attempting to fairly compare model performance.

Technical Abstract: For most biophysical domains, different models are available and the extent to which their structures differ with respect to differences in outputs was never quantified. We use a taxonomy-based approach to address the question with thirteen rice models. Classification keys and binary attributes for each key were identified,and models were classified into five clusters using a binary similarity measure and the unweighted pair-group method with arithmetic mean. PCA was performed on model output at four sites. Results indicate that (i)differences in structure often resulted in similar predictions and (ii)similar structures can lead to large differences in model outputs. A key hypothesis is that user subjectivity during calibration may hide expected relationships between model structure and behaviour. This explanation, if confirmed, highlights the need for shared protocols to reduce the degrees of freedom during calibrations, and to limit, in turn, the risk that user subjectivity influences model performance.