---

作 者 ：薛昌颖;杨晓光;B A M Bouman;冯利平;Gon van Laar;王化琪

期 刊 ：作物学报 2005年 31卷 12期 页码：1567-1571

关键词：ORYZA2000；模拟模型；旱稻；

Keywords:ORYZA2000, Simulation model, Aerobic rice,

摘 要 ：利用北京昌平2年旱稻田间试验结果，对ORYZA2000模型模拟旱稻生长发育的适应性做了初步研究。根据2003年旱稻田间试验结果，对模型进行调试，获得了旱稻的基本作物参数，包括旱稻不同生育阶段的发育速率、干物质分配系数、比叶面积、最大根深等。利用2002年的数据对模型模拟的生物量、叶面积和产量等结果进行了检验。结果表明，ORYZA200能够比较准确地模拟旱稻的生物量、叶面积动态变化过程及最终产量，尤其是在模拟穗生物量方面具有较高的准确性。地上部总生物量、绿叶生物量、茎生物量、穗生物量、叶面积指数和产量的相对均方根误差NRMSE值分别为45%、35%、57%、37%、35%和23%。

Abstract:The ORYZA2000 model for aerobic rice was tested and verified via field experiments conducted at the Changping Experiment Station (40˚02’ N, 116˚10’E; elevation of 43 m) of the China Agricultural University in Beijing, in 2002 and 2003. From aerobic field results in 2003, the developmental rates of aerobic rice in different stages, partitioning factors of dry matter, specific leaf areas and maxinmum root depth were obtained. The biomass, LAI and yield were verified using the experimental data in 2002. The slope(β), intercept(α),and determination coefficient(R²) of the linear regression between simulated and measured values as well as the student‘s t-test assuming unequal variance and the normalized root mean square error(NRMSE) were used for the evaluation of ORYZ2000 model‘s effective.The results showed that ORYZA2000 model was acceptable in simulations of biomasses and LAI. In general, the simulations for total biomass and panicle biomass were more accurate than that for green leave biomass and stem biomass. NRMSE of total biomass, green leave biomass, stem biomass and panicle biomass were 45%,35%,57%and 37%, respectively(Table 1). The simulations of yield were lower than the messured ones with NRMSE of 23% and LAI‘s simulations were with NRMSE of 35%(Table2).

---