中国农业气象 ›› 2018, Vol. 39 ›› Issue (11): 725-738.doi: 10.3969/j.issn.1000-6362.2018.11.004

• 论文 • 上一篇    下一篇

基于两种方法建立辽宁大豆产量丰歉预报模型对比

王贺然,张慧,王莹,李晶,米娜,王若男,李琳琳,董巍,张琪,苏航   

  1. 1.中国气象局沈阳大气环境研究所,沈阳 110166;2.辽宁省气象科学研究所,沈阳 110166;3.锦州市生态与农业气象中心,锦州 121001;4.辽宁省气象装备保障中心,沈阳 110166;5.中国气象局气象干部培训学院辽宁分院,沈阳 110166;6.沈阳中心气象台,沈阳 110166
  • 出版日期:2018-11-20 发布日期:2018-11-13
  • 作者简介:王贺然(1985-),女,博士,工程师,主要从事农业气象业务和科研工作。E-mail:wangheran001@aliyun.com
  • 基金资助:

    辽宁省科技厅重点研发计划指导计划项目(2017210001);中国气象局沈阳大气环境研究所中央级公益性科研院所基本科研业务费重点项目(2016SYIAEZD1);辽宁省气象局博士启动金项目(D201604);国家自然科学基金青年项目(41505120)

A Comparative Study on Forecast Model for Soybean Yield by Using Different Statistic Methods in Liaoning Province

WANG He-ran, ZHANG Hui, WANG Ying, LI Jing, MI Na, WANG Ruo-nan, LI Lin-lin, DONG Wei, ZHANG Qi, SU Hang   

  1. 1.Institute of Atmospheric Environment, China meteorological Administration, Shenyang 110166, China; 2.Liaoning Institute of Meteorological Sciences, Shenyang 110166; 3.Jinzhou Ecology and Agriculture Meteorological Center, Jinzhou 121001; 4.Liaoning Meteorological Equipment Support Center, Shenyang 110166; 5.Liaoning Centre, China Meteorological Administration Training Centre, Shenyang 110166; 6.Shenyang Central Meteorological Observatory, Shenyang 110166
  • Online:2018-11-20 Published:2018-11-13

摘要:

利用辽宁省56个气象站1992?2016年逐日气象资料和5个代表农业气象站的大豆发育期资料,计算不同生育期关键气象因子和气候适宜度指数,分别建立基于关键气象因子和气候适宜度的辽宁省大豆逐候产量动态预报模型,并进行回代检验和预报检验。结果表明:基于关键气象因子的预报模型在6月16日、7月21日、7月26日、8月1日、8月26日和9月16日可以进行产量预报(P<0.05),基于气候适宜度的预报模型在8月16日-10月1日每候可进行1次产量预报(P<0.05);两种预报模型的平均回代检验准确率均高于83.0%;基于气候适宜度的预报模型回代检验准确率和预报检验准确率的变幅较小,稳定性更高;应用两种预报模型,辽宁省大豆产量趋势预报业务得分>0的年份约占60%。说明利用两种模型对辽宁省大豆产量进行动态预报均能满足业务服务需求;进行趋势预报时,可以优先考虑基于关键气象因子的预报模型,而在未出现重大气象灾害的正常年份,可以赋予基于气候适宜度的预报模型更多权重,以减少预报时次。

关键词: 大豆, 关键气象因子, 气候适宜度, 产量预报, 辽宁

Abstract:

Two different dynamic forecast methods for soybean yield based on key meteorological factors and climatic suitability were established in this study. The key meteorological factors and climatic suitable index were calculated by daily meteorological data from 56 meteorological stations, soybean growth stage records of 5 key agrometeorological stations and yield data in Liaoning province from 1992 to 2016. The results showed that the key meteorological factors-based forecast model was suitable to forecast soybean yield on June 16, July 21, July 26, August 1, August 26 and September 16 (P<0.05), and the climatic suitability-based forecast model could be used for every 5 or 6 days from August 16 to October 1 (P<0.05) in Liaoning province. The average accuracies of return test of both models were higher than 83.0%. Compared with the key meteorological factors-based model, the climatic suitability-based model was stable, which indicated by smaller variation of average verified accuracy and average forecast test accuracy. From 1997 to 2016, both models could be applied in soybean yield tendency forecast, according to scoring criteria, core above zero reach up to 60% by the two models. In conclusion, the yield forecast models based on key meteorological factors and climatic suitability could meet the basic needs for agrometeorological services in Liaoning province. For soybean yield tendency forecasting, the method based on key meteorological factors should be preferred. In the normal years without severe meteorological disaster, the method on climatic suitability could be given priority to yield quantitative forecasting to reduce forecast times.

Key words: Soybean, Key meteorological factors, Climatic suitability, Yield forecast, Liaoning