| 摘要: |
| 为探究不同积温计算方法对作物发育期模拟效果的影响情况，以北京密云冬小麦1989-2011年返青-拔节、拔节-抽穗和抽穗-成熟期为例，统计了基于3种线性假设的京冬8号和不区分品种的冬小麦活动积温及有效积温，对比分析了6种积温（GDD）模型和发育时间（DD）模型统计变量的变异系数（CV）及模拟偏差（SD），结果表明：GDD模型对阶段积温模拟的稳定性较DD模型对发育日数的模拟效果更佳；区分品种的变量稳定性相对不分品种的稳定性更大且模拟效果更好；对比模型有效率（ME）、均方根误差（RMSE）及标准化均方根误差（NRMSE）得出结论：6种GDD模型相对DD模型对发育期的模拟效果均达到优秀；分品种的模拟效果较不分品种的模拟更优；同一线性假设中，活动积温比有效积温模拟效果更好；3种生长假设间，线性假设B对发育期模拟效果优于A和C。 |
| 关键词: 积温 生长假设 变异系数 模型有效率 模拟效果 |
| DOI：10.11841/j.issn.1007-4333.2016.10.15 |
| 分类号: |
| 基金项目:北京市科委专项（Z411100003614052）；“十二五”国家科技支撑计划课题（2012BAD20B02）；国家自然科学基金项目（41475104） |
|
| Comparison of the simulation effect of different growing degree-days calculating methods on crop development processes |
|
WU Yu-jie1,2, YE Cai-hua3, JIANG Hui-fei2, YAN Meng-ling2, NI Zi-xuan2
|
|
1.Institute of Meteorological Sciences of Jilin Province, Changchun 130062, China;2.College of Resource and Environment, China Agricultural University, Beijing 100193, China;3.Beijing Climate Center, Beijing 100089, China
|
| Abstract: |
| To explore the impact of methods of calculating growing degree-day on simulation effect of crop development processes,the data from 1989 to 2011 of the 3 developmental phases of Miyun winter wheat including reviving-jointing,jointing-heading and heading-maturity,were analyzed.The active and effective growing degree-days (GDD) were calculated by adapting six calculating methods based on three linear (Linear A, B and C) growth assumptions on single cultivar Jingdong 8 or all cultivars.The stability of coefficient of variation (CV) and simulation deviation (SD) of development days (DD) model and GDD models were analyzed.Results showed that:Simulation effects of GDD models for different developmental phases were excellent compared with DD model;Simulation effect of single cultivar was better than that of all cultivars;The model efficiency (ME),Root-mean-square error (RMSE) and Normalized RMSE (NRMSE) of DD model and GDD models were compared.Results showed that:Simulation effects of 6 GDD models are all excellent compared with DD model; Simulation effect of single cultivar was better than that of all cultivars;Simulation effect of active GDD was better than that of effective GDD for same linear growth assumptions;Simulation effect of developmental process was significantly improved by linear B than that of linear A and C among three linear growth assumptions. |
| Key words: growing degree-days growth and developmental assumption coefficient of variation model efficiency simulation effect |
