欢迎访问《生态学杂志》官方网站,今天是 分享到:

生态学杂志

• 方法与技术 • 上一篇    下一篇

基于锦州春玉米田间试验的WOFOST模型参数的确定及性能评价

蔡福1,米娜1,纪瑞鹏1,明惠青2,冯锐1,张淑杰1,张慧3,赵先丽1,张玉书1*   

  1. (1中国气象局沈阳大气环境研究所, 沈阳 110166;2辽宁省气象服务中心, 沈阳 110166;3锦州市生态与农业气象中心, 辽宁锦州 121000)
  • 出版日期:2019-04-10 发布日期:2019-04-10

Determination of crop parameters for WOFOST model and its performance evaluation based on field experiment of spring maize in Jinzhou, Liaoning.

CAI Fu1, MI Na1, JI Rui-peng1, MING Hui-qing2, FENG Rui1, ZHANG Shu-jie1, ZHANG Hui3, ZHAO Xian-li1, ZHANG Yu-shu1*   

  1. (1Institute of Atmospheric Environment, China Meteorological Administration, Shenyang 110166, China; 2Liaoning Province Meteorological Service Center, Shenyang 110166, China; 3Jinzhou Ecological and Agricultural Meteorology Center, Jinzhou 121001, Liaoning, China).
  • Online:2019-04-10 Published:2019-04-10

摘要: 利用2011—2015年在锦州地区开展的玉米分期播种试验资料,确定了玉米初始总干物量(TDWI)、比叶面积(SLA)、出苗时叶面积指数(LAIEM)及发育参数,并采用试错法确定出其他生理参数。利用相关系数、相对误差以及均方根误差验证模型对发育期、叶面积指数(LAI)和不同器官生物量的模拟性能来评价所确定参数的适应性。结果表明:80%花期的平均模拟误差为1.25 d,而75%成熟期的平均模拟误差为3.47 d,LAI和总地上生物量(TAGP)模拟值对实测值的解释能力分别达到0.8412和0.8945;模型模拟性能存在较明显的年际和播期差异,在干旱年较差,2012年(土壤水分适宜)和2015年(干旱)LAI和TAGP的均方根误差分别为0.38 m2·m-2、9.40 kg·hm-2和0.44 m2·m-2、22.65 kg·hm-2;模型性能随播期的偏离而下降,4月30日(适播期)与5月20日播期LAI和TAGP的均方根误差分别为0.30 m2·m-2、15.19 kg·hm-2和0.43 m2·m-2、25.66 kg·hm-2。本研究将为东北春玉米作物模型参数的确定提供重要参考。

关键词: PM2.5, 大气颗粒物, 激光粒度仪, 毛白杨, 洗脱称量粒度分析法

Abstract: Correct parameters are the key prerequisite for the prediction of crop models. In this study, a series of important crop parameters including initial total crop dry weight (TDWI), specific leaf area (SLA), leaf area index at emergency (LAIEM), developing parameters and other physiological parameters were measured and determinated using the observation data and trialanderror method based on the experiment of different sowing dates for maize in Jinzhou during 2011-2015. The adaptability of those parameters in WOFOST model was evaluated by testing the simulation accuracy of growth periods, leaf area index (LAI) and the biomass of different organs of maize with the methods of correlation coefficient, relative error (RE) and root mean square error (RMSE). The results showed that mean simulation errors for antheses and mature stages with 80% and 75% of the total samples were 1.25 and 3.47 days, respectively. The explanatory abilities of simulation to observation for LAI and total aboveground production (TAGP) were 0.8412 and 0.8945, respectively. Besides, there were obvious interannual and sowing date differences for the model performances in simulating abovementioned variables and their simulation accuracy were lower in dry year. Specifically, the RMSEs of LAI and TAGP were 0.38 m2·m-2 and 9.40 kg·hm-2 in 2012 respectively when soil water content was under the suitable condition during the whole growing season, and were 0.44 m2·m-2 and 22.65 kg·hm-2 in 2015 when maize suffered longterm drought. Furthermore, the model performed worse when the sowing date was deviated from the normal range. The RMSEs of LAI and TAGP were 0.30 m2·m-2 and 15.19 kg·hm-2 for the optimum sowing date on 30 April and 0.43 m2·m-2 and 25.66 kg·hm-2 for the sowing date on 20 May. Our results provide an important reference for determining the crop model parameters of spring maize in northeast China.

Key words: PM2.5, atmospheric particulate matter, laser particle size analyzer, Populus tomentosa, elution-weighingparticle sizeanalysis (EWPA).