Incorporation of Wheat Canopy Temperatures into Agroecosystem Models by Using a Meta-model☆
Incorporation of Wheat Canopy Temperatures into Agroecosystem Models by Using a Meta-model☆
复制标题
使用元模型将小麦冠层温度纳入农业生态系统模型â
DOI：
10.1016/j.proenv.2015.07.230
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
Böttcher
中科院分区：
文献类型：
--
作者：
Ahrends;Neukam;Böttcher
冠层温度是预测热胁迫对粮食产量影响的重要变量。然而，已建立的农业生态系统模型通常使用高度简化的方法，例如基于气温，来估计冠层/冠层内温度对环境条件的响应。因此，会导致较大的模型不确定性。由于影响作物生长的现象（例如热应激）通常只能通过解决昼夜周期来考虑，因此需要使用每日时间步长来获取这些信息并将这些信息整合到现有的植物生长模型中。在我们的研究中，我们的目标是提出一种方法，如果这些数据不可用，则可以将冠层温度的昼夜变化的影响纳入作物生长模型中。使用红外辐射计对冬小麦植株进行观察，在基尔大学 Hohenschulen 实验农场的两个生长季节进行三种不同灌溉处理的田间试验中，对模拟每小时冠层温度的模型进行了验证。更具体地说，我们的目标是（I）从昼夜循环中识别出适合使用每日时间步长来表征和汇总模型中冠层温度响应信息的参数，II）使用模型输出来估计每日气象数据、植物特征和已识别参数之间的统计关系，并基于这些关系，（III）推导出元模型以在现有气象数据的基础上模拟这些参数。我们建议，将所得元模型纳入农业生态系统模型中，将通过改进对高温和干旱胁迫对作物产量影响的预测来进行更现实的情景分析。
The canopy temperature is an important variable for predicting the impact of heat stress on grain yield. However, established agroecosystem models often use strongly simplified approaches, e.g., based on the air temperature, to estimate the response of canopy/canopy-within temperature to ambient conditions. Consequently, large model uncertainties are caused. Since phenomena affecting crop growth, such as heat stress, can often only be considered by resolving the diurnal cycle, methods are needed to gain & integrate these information into existing plant growth models using daily time steps.In our study, we aim at presenting an approach to include the impact of the diurnal variability of canopy temperatures in crop growth models, if these data are not available. A model simulating hourly canopy temperatures is validated using IR radiometer observations of winter wheat plants obtained during a field trial for three different irrigation treatments during two growing seasons at the Hohenschulen experimental farm of the University of Kiel. More specifically we aim at (I) identifying parameters from the diurnal cycles that are suitable to characterize and aggregate information on the canopy temperature response in models using a daily time step, II) using the model output to estimate the statistical relations between daily meteorological data, plant characteristics and the identified parameters and, based on these relations, (III) deriving a meta-model to simulate these parameters on the basis of existing meteorological data. We suggest that incorporating the resulting meta-model in agroecosystem models will lead to more realistic scenario analysis by improved predictions of the impact of heat and drought stress on crop yield.