Comparison of wheat simulation models under climate change. I. Model calibration and sensitivity analyses

Abstract

A comparison of the performance of 5 wheat models was carried out for 2 sites in Europe with considerably different agroclimatic conditions: Rothamsted, UK, and Seville, Spain. The models were calibrated against field data sets from both sites. For Rothamsted the measured time courses of crop growth, evapo-transpiration and nitrogen uptake were reproduced reasonably well by the different models, except for leaf area index. For Seville, the experimental data set was insufficient for such a detailed comparison and mainly simulated results were compared. The sensitivity of the model results to stepwise changes In individual weather variables was then determined. In the different model runs a temperature rise generally resulted in lower yields, an increase in precipitation and atmospheric CO2 concentration resulted in higher ylelds, and increased variability of weather variables often resulted in lower yields with increased yleld varlab~lity.

A framework for testing the ability of models to project climate change and its impacts

“…Models are often, but not always, calibrated. Experience has shown that even models that through calibration have been forced to achieve similar performance in a calibration period may exhibit quite different projections for future climates (Wolf et al 1996;Velázquez et al 2012).…”

“…It is good practice in many earth science disciplines to conduct rigorous testing of models before they are applied for predictions (Jørgensen 1995;Refsgaard and Knudsen 1996;Wolf et al 1996). Such tests may be denoted validation tests, history matching or similar depending on the terminology used.…”

Impacts of Heat Stress on Wheat: A Critical Review

“…During seed-filling stage the oil and protein concentrations were inversely related in response to heat stress [94]. Pipolo et al [95] concluded similar quadratic response for oil and protein concentrations of soybean seeds cultured in vitro. A positive correlation was observed between protein and oil concentrations which was a function of rate of dry matter accumulation of soybean seeds.…”

“…Increasing temperatures from optimum 25°C to 28°C, the oil content increased however, above 28°C the oil content declined [94]. In case of soybean seed protein content remained constant at temperatures between 16°C and 25°C but increased at temperatures above 25°C [94,95]. During seed-filling stage the oil and protein concentrations were inversely related in response to heat stress [94].…”

“…A positive correlation was observed between protein and oil concentrations which was a function of rate of dry matter accumulation of soybean seeds. Thus, temperature oriented changes in overall seed-growth rate influences the seed size may, which in turn are likely to be dependent on carbon (C) and N supply to the seeds [95]. Machado and Paulsen [96] in response to heat and drought stress and found that water relations adjusted to heat stress when the soil was maintained at field capacity, whereas heat stress interacted strongly with drought and exacerbated its effects when water was withheld.…”