WOFOST模型对江淮地区水稻生长发育模拟的适应性评价
Adaptability Assessment of WOFOST Model for Simulating Rice Growth and Development in the Jianghuai Region
通讯作者:
收稿日期: 2023-11-7 修回日期: 2024-05-8
基金资助: |
|
Received: 2023-11-7 Revised: 2024-05-8
作者简介 About authors
伍露,主要从事农业资源高效利用研究,E-mail:
通过收集2006-2012年江淮地区水稻生长发育和气象资料,对WOFOST模型进行调参验证,确定了关键发育阶段内所需积温、比叶面积、分配系数和枯萎速率等作物参数,并评价了模型在江淮地区的适应性。结果表明,WOFOST模型可以较好地模拟江淮地区水稻生长发育动态变化过程,开花期、成熟期、叶面积指数(LAI)、叶干物质量、茎干物质量、穗干物质量、地上干物质量和产量观测值与模拟值的RMSE分别为1.73~4.66 d、1.94~4.42 d、0.39~2.51、(0.43~0.86)×103 kg/hm2、(0.86~1.52)×103 kg/hm2、(0.52~1.21)×103 kg/hm2、(1.38~1.96)×103 kg/hm2和(0.45~1.33)×103 kg/hm2;NRMSE分别为0.75%~1.96%、0.74%~1.49%、8.74%~43.40%、14.94%~30.55%、18.16%~28.84%、9.44%~22.81%、11.33%~15.89%和5.49%~13.43%,其中合肥水稻发育期,叶、茎、穗和地上部干物质量,镇江水稻LAI和荆州水稻产量模拟效果最优。水稻LAI、叶和茎干物质量随移栽天数增加呈现先上升后下降的变化趋势,而穗和地上部干物质量随移栽天数增加呈现逐渐上升的趋势。
关键词:
This study collected rice growth and meteorological data from 2006 to 2012 in the Jianghuai region and validated the WOFOST model by adjusting plant genetic parameters. The following crop parameters for determining development stages and rice growth, such as accumulated temperature, specific leaf area, distribution coefficient and leaf senescence rate, were calibrated, and the adaptability of the model in the Jianghuai region was evaluated. The results showed that the WOFOST model can effectively simulate the dynamic changes in rice growth and development in the Jianghuai region. The RMSE of the measured and simulated values for flowering period, maturity period, leaf area index (LAI), leaf dry matter weight, stem dry matter weight, panicle dry matter weight, aboveground dry matter weight, and yield were 1.73-4.66 d, 1.94-4.42 d, 0.39-2.51, (0.43-0.86)× 103 kg/ha, (0.86-1.52)×103 kg/ha, (0.52-1.21)×103 kg/ha, (1.38-1.96)×103 kg/ha, and (0.45-1.33)×103 kg/ha, respectively and the NRMSE were 0.75%-1.96%, 0.74%-1.49%, 8.74%-43.40%, 14.94%-30.55%, 18.16%- 28.84%, 9.44%-22.81%, 11.33%-15.89%, and 5.49%-13.43%, respectively. Among different study stations, rice development stage, dry matter weight of leaves, stems, panicles, and aboveground in Hefei, LAI in Zhenjiang, and yield in Jingzhou presented the most accurate simulation results. The LAI, leaf and stem dry matter weight of rice first increased and then decreased with the increase of transplanting days, while the dry matter weight of panicles and aboveground parts showed a gradually increasing trend with the increase of transplanting days.
Keywords:
本文引用格式
伍露, 张皓, 杨霏云, 郭尔静, 斯林林, 曹凯, 程陈.
Wu Lu, Zhang Hao, Yang Feiyun, Guo Erjing, Si Linlin, Cao Kai, Cheng Chen.
作物生产的目标是提高光温资源利用效率和提升作物品种生产潜力,而光温的季节性变化是影响作物生产力[1]和品质形成过程[2]的主要环境因素,何强等[3]通过12种人工控制光温组合处理、短日遮光处理以及分期播种试验,探究了不同水稻品种的育性光温特性,并提出了光温双因子互作量化的光温效应连动假设。但田间试验往往存在时空限制性,而作物模型将光、温、水及土壤等条件作为环境的驱动变量,精细模拟了作物生理生态过程以弥补大田研究的不足[1,4]。WOFOST模型是荷兰瓦赫宁根大学和世界粮食研究中心共同开发研制的通用性作物生长模拟模型[5],以日为步长动态定量模拟小麦[6-7]、玉米[8-9]及水稻[10-11]等多种大田作物生长发育过程。目前,国内外利用WOFOST模型在土壤状况[12]、作物品种[13]和耕作制度[14]等对产量的影响评估方面已得到了广泛应用,提高了试验结果的外推性。黄健熙等[6]在冬小麦主产区分区的基础上,以2012-2015年气象数据驱动WOFOST模型,重点优化WOFOST模型中与品种相关的积温参数,并在站点尺度进行冬小麦物候期、叶面积指数(LAI)和单产动态模拟和精度分析。Roberto等[11]利用1989-2004年6个地点不同灌溉条件下水稻作物数据,比较分析了WOFOST模型对生物量分配、小穗不育性、水肥管理和土壤水文等方面的模型参数敏感性。目前,应用WOFOST模型对我国同一地区内单一水稻品种生长发育过程的适应性研究较多,但忽略了各省份不同水稻品种的适应性[4,15]。此外,在全球气候变暖导致极端天气频发的背景下,水稻生长季后期遭受高温损伤愈发严重[16],这就要求在应用作物模型评估气候变化对水稻生长发育和产量的影响之前,应先通过调参验证来检验模型的准确性和普适性获得适用于该地区不同水稻品种的模型参数,继而评估模型在该地区的适应性。
本研究以江淮地区多省份为研究对象,通过收集分析2006-2012年水稻生长发育和气象数据,对WOFOST模型进行调参验证,确定了关键发育阶段内所需积温、比叶面积、分配系数和枯萎速率等作物遗传参数,并评价了模型在江淮地区的适应性。研究结果为利用WOFOST模型评估气候变暖对江淮地区水稻生产力的影响提供理论基础和数据支撑。
1 材料与方法
1.1 试验材料
以江淮地区为研究区域,选取安徽合肥,湖北荆州、武汉,江苏兴化、徐州和镇江作为调查点,收集了2006-2012年各站点的逐日气象资料(数据来源:中国气象局
1.2 WOFOST模型水稻生长发育模块的遗传参数
WOFOST模型是一个动态的解释性模型,该模型的模拟基础是作物生理生态过程,主要包括同化作用、呼吸作用、蒸腾作用和干物质的分配等[15,17]。将各站点水稻田间观测资料分为调参组和验证组,且2组数据相互独立。本研究根据当地种植模式,设置模型初始发育期(developmental stage of initial,DVSI)和初始总干物质量(total dry weight of initial,TDWI)模型初始参数(表1),其中出苗期DVSI为0,开花期DVSI为1,成熟期DVSI为2,并采用试错法调试WOFOST模型水稻生长发育参数,包括控制水稻发育期的出苗期—开花期和开花期—成熟期所需积温模型关键参数(TSUM1和TSUM2)(表1)和控制水稻叶面积指数(leaf area index,LAI)、地上部生物量和产量的比叶面积(specific leaf area,SLA)、叶分配系数(fraction of leaf,FL)、茎分配系数(fraction of stem,FS)、穗分配系数(fraction of storage organ,FO)和枯萎速率(reduction factor of senescence,RFSE)模型关键参数(表2)。
表1 江淮地区WOFOST模型水稻发育期模块的遗传参数
Table 1
站点 Station | 初始 发育期 DVSI | 初始总干 物质量 TDWI (kg/hm2) | 出苗期— 开花期 所需积温 TSUM1 (℃·d) | 开花期— 成熟期 所需积温 TSUM2 (℃·d) |
---|---|---|---|---|
合肥Hefei | 0.30 | 350 | 1482 | 435 |
荆州Jingzhou | 0.23 | 330 | 1345 | 405 |
武汉Wuhan | 0.28 | 110 | 1350 | 440 |
兴华Xinghua | 0.00 | 310 | 1210 | 455 |
徐州Xuzhou | 0.35 | 180 | 1610 | 505 |
镇江Zhenjiang | 0.33 | 130 | 1490 | 610 |
表2 江淮地区WOFOST模型水稻生长模块的遗传参数
Table 2
站点 Station | 发育阶段 Developmental stage | 比叶面积 SLA (hm2/kg) | 发育阶段 Developmental stage | 叶分配系数 FL (kg/kg) | 茎分配系数 FS (kg/kg) | 穗分配系数 FO (kg/kg) | 发育阶段 Developmental stage | 枯萎速率 RFSE | ||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
合肥 Hefei | 0.10 | 0.0027 | 0.00 | 0.40 | 0.60 | 0.00 | 0.00 | 0.45 | ||||||||
0.70 | 0.0029 | 0.50 | 0.45 | 0.55 | 0.00 | 2.00 | 0.50 | |||||||||
0.98 | 0.0020 | 0.71 | 0.58 | 0.42 | 0.60 | |||||||||||
1.30 | 0.0029 | 0.95 | 0.10 | 0.30 | 0.80 | |||||||||||
1.40 | 0.0010 | 1.00 | 0.00 | 0.20 | 1.00 | |||||||||||
2.00 | 0.0010 | 2.00 | 0.00 | 0.00 | 1.00 | |||||||||||
荆州 Jingzhou | 0.10 | 0.0017 | 0.00 | 0.40 | 0.60 | 0.00 | 0.00 | 0.30 | ||||||||
0.58 | 0.0019 | 0.58 | 0.52 | 0.48 | 0.00 | 2.00 | 0.40 | |||||||||
0.86 | 0.0020 | 0.80 | 0.38 | 0.62 | 0.00 | |||||||||||
1.10 | 0.0015 | 0.85 | 0.00 | 0.00 | 1.00 | |||||||||||
1.40 | 0.0015 | 0.90 | 0.00 | 0.00 | 1.00 | |||||||||||
2.00 | 0.0010 | 1.00 | 0.00 | 0.00 | 1.00 | |||||||||||
2.00 | 0.00 | 0.00 | 1.00 | |||||||||||||
武汉 Wuhan | 0.10 | 0.0025 | 0.00 | 0.55 | 0.46 | 0.00 | 0.00 | 0.50 | ||||||||
0.66 | 0.0030 | 0.25 | 0.58 | 0.42 | 0.00 | 2.00 | 0.60 | |||||||||
0.90 | 0.0027 | 0.50 | 0.45 | 0.55 | 0.00 | |||||||||||
1.30 | 0.0018 | 0.72 | 0.51 | 0.49 | 0.00 | |||||||||||
1.40 | 0.0010 | 0.90 | 0.05 | 0.28 | 0.67 | |||||||||||
2.00 | 0.0010 | 1.00 | 0.00 | 0.01 | 0.99 | |||||||||||
2.00 | 0.00 | 0.00 | 1.00 | |||||||||||||
兴华 Xinghua | 0.00 | 0.0020 | 0.00 | 0.40 | 0.60 | 0.00 | 0.00 | 0.60 | ||||||||
0.60 | 0.0026 | 0.50 | 0.55 | 0.45 | 0.00 | 2.00 | 0.60 | |||||||||
0.90 | 0.0032 | 0.70 | 0.35 | 0.65 | 0.00 | |||||||||||
1.00 | 0.0022 | 0.90 | 0.30 | 0.15 | 0.55 | |||||||||||
2.00 | 0.0022 | 1.00 | 0.00 | 0.15 | 0.85 | |||||||||||
1.10 | 0.00 | 0.00 | 1.00 | |||||||||||||
2.00 | 0.00 | 0.00 | 1.00 | |||||||||||||
徐州 Xuzhou | 0.10 | 0.0017 | 0.00 | 0.40 | 0.60 | 0.00 | 0.00 | 0.50 | ||||||||
0.70 | 0.0019 | 0.50 | 0.45 | 0.55 | 0.00 | 2.00 | 0.40 | |||||||||
0.95 | 0.0024 | 0.85 | 0.38 | 0.62 | 0.00 | |||||||||||
1.10 | 0.0015 | 0.90 | 0.10 | 0.55 | 0.35 | |||||||||||
1.40 | 0.0015 | 1.00 | 0.00 | 0.20 | 0.80 | |||||||||||
2.00 | 0.0010 | 2.00 | 0.00 | 0.00 | 1.00 | |||||||||||
镇江 Zhenjiang | 0.10 | 0.0017 | 0.00 | 0.40 | 0.60 | 0.00 | 0.00 | 0.30 | ||||||||
0.58 | 0.0019 | 0.58 | 0.52 | 0.48 | 0.00 | 2.00 | 0.40 | |||||||||
0.86 | 0.0020 | 0.80 | 0.38 | 0.62 | 0.00 | |||||||||||
1.10 | 0.0015 | 0.85 | 0.00 | 0.40 | 0.60 | |||||||||||
1.40 | 0.0015 | 0.90 | 0.00 | 0.00 | 1.00 | |||||||||||
2.00 | 0.0010 | 1.00 | 0.00 | 0.00 | 1.00 | |||||||||||
2.00 | 0.00 | 0.00 | 1.00 |
式中,Xobs为观测值,Xsim为模拟值,
2 结果与分析
2.1 发育期验证
图1
图1
基于WOFOST模型的江淮地区水稻关键发育期验证
Fig.1
Validation of rice key development period in the Jianghuai region based on the WOFOST model
由图1a可知,WOFOST模型模拟江淮地区各站点水稻开花期的观测值与模拟值之间MAE为1.50~4.50 d,RMSE为1.73~4.66 d,NRMSE为0.75%~1.96%,其中荆州水稻开花期最早(221.50 d),兴华水稻开花期最晚(249.75 d),由NRMSE可知,各站点水稻开花期模拟精度大小依次是合肥、荆州、兴华、武汉、徐州、镇江。由图1b可知,WOFOST模型模拟江淮地区各站点水稻成熟期的观测值与模拟值之间MAE为1.75~4.00 d,RMSE为1.94~4.42 d,NRMSE为0.74%~1.49%,其中荆州水稻成熟期最早(249.75 d),镇江水稻成熟期最晚(299.25 d)。由NRMSE可知,各站点水稻成熟期模拟精度大小依次是合肥、武汉、徐州、荆州、镇江、兴华。综上所述,说明WOFOST模型可以很好地模拟江淮地区水稻发育期,其中合肥水稻发育期模拟效果最优。
2.2 生长指标验证
由图2可知,WOFOST模型对江淮地区水稻LAI观测值与模拟值之间的MAE为1.30,RMSE为1.78,NRMSE为31.88%,D值为0.86;对江淮地区水稻器官干物质量观测值与模拟值之间的MAE为(0.72~1.35)×103 kg/hm2,RMSE为(0.92~ 1.78)×103 kg/hm2,NRMSE为13.67%~22.92%,D值为0.85~0.98;对江淮地区水稻产量观测值与模拟值之间的MAE为0.74×103 kg/hm2,RMSE为0.88×103 kg/hm2,NRMSE为9.15%,D值为0.81。
图2
图2
基于WOFOST模型的江淮地区水稻生长过程验证图
Fig.2
Validation of rice growth process in the Jianghuai region based on the WOFOST model
图3
图3
基于WOFOST模型的江淮地区水稻生长指标动态模拟过程
Fig.3
Dynamic simulation process of rice growth indicators in the Jianghuai region based on the WOFOST model
由图2f可知,WOFOST模型对江淮地区各站点水稻产量观测值与模拟值的MAE为(0.43~1.25)×103 kg/hm2,RMSE为(0.45~1.33)×103 kg/hm2,NRMSE为5.49%~13.43%,其中合肥水稻产量最小(8.16×103 kg/hm2),兴华水稻产量最大(10.43×103 kg/hm2)。由NRMSE可知,各站点水稻产量模拟精度大小依次是荆州、合肥、武汉、徐州、兴华、镇江。
综上所述,说明WOFOST模型可以较好地模拟江淮地区水稻生长指标动态变化过程,并具有较高的模拟精度,其中镇江水稻LAI,合肥水稻叶、茎、穗和地上部干物质量,荆州水稻产量模拟效果最优。
3 讨论
4 结论
WOFOST模型可以较好地模拟江淮地区水稻生长发育动态变化过程,开花期、成熟期、LAI、叶干物质量、茎干物质量、穗干物质量、地上干物质量和产量观测值与模拟值的RMSE分别为1.73~4.66 d、1.94~4.42 d、0.39~2.51、(0.43~0.86)× 103 kg/hm2、(0.86~1.52)×103 kg/hm2、(0.52~1.21)× 103 kg/hm2、(1.38~1.96)×103 kg/hm2和(0.45~1.33)×103 kg/hm2;NRMSE分别为0.75%~1.96%、0.74%~1.49%、8.74%~43.40%、14.94%~30.55%、18.16%~28.84%、9.44%~22.81%、11.33%~15.89%和5.49%~13.43%,其中合肥水稻发育期、叶、茎、穗和地上部干物质量,镇江水稻LAI和荆州水稻产量模拟效果最优。水稻LAI、叶和茎干物质量随移栽天数增加呈现先上升后下降的变化趋势,而穗和地上部干物质量随移栽天数增加呈现逐渐上升的趋势。
参考文献
Evapotranspiration estimation using a modified crop coefficient model in a rotated rice-winter wheat system
Sensitivity analysis using Morris:Just screening or an effective ranking method?
Comparison of five wheat models simulating phenology under different sowing dates and varieties
Advancing crop modelling capabilities through cultivar-specific parameters sets for the Italian rice germplasm
Multi-metric evaluation of the models WARM, CropSyst, and WOFOST for rice
Comparison of CERES, WOFOST and SWAP models in simulating soil water content during growing season under different soil conditions
Forecasting future crop suitability with microclimate data
Simulation of wheat (Triticum aestivum L.) yield using WOFOST model under different management levels
A growth model based on standardized growing degree days for hydroponic fresh cut tulip in solar greenhouses
Manure and mineral fertilizer effects on crop yield and soil carbon sequestration: a meta‐ analysis and modeling across China
Optimization of water and fertilizer coupling system based on rice grain quality
Diversity analysis of antagonists from rice-associated bacteria and their application in biocontrol of rice diseases
Integrated Bayesian Multi-model approach to quantify input, parameter and conceptual model structure uncertainty in groundwater modeling
/
〈 |
|
〉 |
