摘要
为探索热区环境下冬播玉米物候发育的温度效应,建立具有较好解释性的冬玉米物候发育模拟模型。基于云南陇川、瑞丽、芒市三试点分品种和播期的3年试验资料,比较不同模拟方法对冬玉米物候期的预测效果。结果显示以0℃、6℃、8℃、10℃为下限温度的积温模型,预测精度均不高,根均方差（RMSE）为7~11天。因此本研究在WOFOST模型基础上,以日最高温、日最低温2个驱动变量取代日均温,为不同物候期分别设置基点温度,基于物候期分段函数、建立了新的冬玉米物候发育模拟模型。模型采用独立试验数据进行检验,对两供试品种物候期预测的RMSE分别为4.42天和4.19天。模型体现了日温变化对发育速率的影响,体现了不同物候期温度效应的差异,具有较好的机理性,模型预测精度较高。
The paper aims to quantify the temperature effect on phenological development of winter maize planted in tropic region, and build a new phenological model with better explanation. Field experiments were conducted with different maturity cultivars and sowing dates in three sites in Yunnan tropical region（Ruili, Longchuan, and Mangshi County） for three years, to compare the accuracy of different predicted methods. With 0℃, 6℃, 8℃ and 10℃ as lower bound, four accumulated temperature models did not predict accurately with RMSEs from 7 days to 11 days. A new maize development model was developed based on WOFOST, in which the daily maximum temperature and daily minimum temperature took place of the daily average temperature, and a piecewise function was set with particular base temperature for each phenophase. This new model was validated by independent experiment data. The RMSEs of this model was 4.42 d and 4.19 d for two cultivars respectively. This new model showed higher accuracy and was more explanatory. It could explain the effect of temperature variation within-day and the difference of different phenophases.
出处
《中国农学通报》
2015年第24期53-58,共6页
Chinese Agricultural Science Bulletin
基金
云南省自然科学基金"冬玉米发育与产量形成的温光生态及模拟研究"(2011FZ093)
关键词
冬玉米
物候期
模拟模型
积温
winter maize
phenological development
simulation model
accumulated temperature
