四川单季稻区主栽水稻品种的生长模拟

利用四川单季稻区7个农业气象观测站5个主栽品种的田间观测数据，结合当地栽培管理措施、土壤条件及逐日气象资料对ORYZA2000模型进行参数调试，并确定四川单季稻区5个主栽品种的作物参数值；利用4～5a各主栽品种的观测数据对单季稻生育期、叶面积指数、生物量和产量等指标的模拟结果进行验证和评价。结果表明，合系39营养生长期发育速率最大，而生殖生长期发育速率最小，Ⅱ优838营养生长期发育速率最小，而D优63和汕优2生殖生长期发育速率最大；模型对5个单季稻主栽品种的生育期模拟效果较好，各品种开花期与成熟期的相对模拟误差均在1～2d，归一化均方根误差（NRMSE）均小于1%；各品种产量的NRMSE在5.26%～10.01%，叶面积指数的NRMSE为10.37%～19.19%，地上部总生物量、茎生物量、绿叶生物量及穗生物量的NRMSE分别为13.17%～18.69%、14.31%～20.41%、18.95%～24.74%和20.85%～25.39%。由此可见，ORYZA2000模型能够较为准确地模拟四川单季稻区5个主栽品种的发育及产量形成过程，适应能力较强，可以应用于四川单季稻生产。

Growth Simulation of Major Rice Varieties in Single Cropping Rice Area of Sichuan Province

To provide the reference for regional adaptation and application of ORYZA2000 in Sichuan province, the model was calibrated by the field data of five rice varieties at seven agrometeorological stations, and evaluated by the dataset from above stations with the five rice varieties. The daily meteorological data over the growing period of single cropping rice for the studied sites were used as driving variables in ORYZA2000 model. Management measures and soil data were used as the input of the model. By the comparison between measured and simulated values of development stages, leaf area index (LAI), biomass and yield, the simulation capacity and performance of ORYZA2000 model were evaluated in single cropping rice area. The results showed that nutrition stage growth rate of Hexi-39 was largest, but reproductive stage growth rate of Hexi-39 was smallest. The nutrition stage growth rate of Ⅱyou-838 was smallest, but reproductive stage growth rate of Dyou-63 and Xianyou-2 were largest. ORYZA2000 model could simulate the phenology of five rice varieties with 1-2 days’ difference for flowering and maturity stages. Furthermore, the normalized root mean square errors (NRMSE) of different development stages were less than 1%. While the NRMSE of other rice parameters were 5.26%-10.01% for yield, 10.37%-19.19% for leaf area index, 13.17%-18.69% for aboveground biomass, 14.31%-20.41% for stem biomass, 18.95%-24.74% for green leaves biomass, and 20.85%-25.39% for panicle biomass．These results showed that ORYZA2000 model could satisfactorily simulate the dynamical process of growth and yield of five major rice varieties. In general, we could conclude that ORYZA2000 model was adaptable and could be applied in scenarios analysis study in single cropping rice area of Sichuan province.