Search

The objective of this study was to develop an effective method for predicting ecotype-specific cultivar parameters for the large-scale application of wheat crop simulation models. The Markov Chain Monte Carlo (MCMC) technique was explored for parameter calibration based on an existing model. We estimated the posterior probability distribution of ecotype-specific cultivar parameters at four ecosites (Huai'an, Zhengzhou, Weifang, and Shijiazhuang in China) by using the historical phenological stages of wheat and daily weather data from 1980 to 1995. 1000 sets of cultivar parameters which were randomly sampled from the posterior probability distribution at each ecosite from 1996 to 2005 were used to evaluate the MCMC-based method. Optimal 50 sets of parameters were chosen from the 1000 sets of parameters at each ecosite to represent the ecotype-specific cultivar parameters of the Jiangsu, Henan, Shandong and Hebei provinces to evaluate the MCMC-based method for the year 2005 at the regional scale. The results showed that the coefficients of determination (R 2 ) between the observed and estimated phenological stages ranged from 0.61 to 0.72, with a root mean square error (RMSE) of less than 3.6 days and a root mean square deviation ( RMSD ¯ ) of less than 3.7 days at the site scale. All of the RMSE and RMSD ¯ values for the three phenological stages obtained using the posterior probability distribution at the four ecosites were significantly lower than those based on the prior probability distribution. At the regional scale, R 2 between the observed and estimated phenological stages was greater than 0.86, with RMSE less than 3.4 days and RMSD ¯ less than 4.0 days in most of the grids for year 2005. The estimated phenological stages agreed well with the observations, suggesting that the MCMC technique has high reliability of for estimating multiple parameter combinations in a wheat phenology model. The combination of the present MCMC technique and a phenology model could be used for estimating the ecotype-specific cultivar parameters for the main wheat growing regions of China, which can be used to predict progress of wheat development at the regional scale. [ABSTRACT FROM AUTHOR]