Design and optimisation of a large-area process-based model for annual crops
Design and optimisation of a large-area process-based model for annual crops
复制标题
DOI：
10.1016/j.agrformet.2004.01.002
复制
发表时间：
2004-07-20
影响因子：
6.2
通讯作者：
Grimes, DIF
中科院分区：
文献类型：
--
作者：
Challinor, AJ;Wheeler, TR;Grimes, DIF
提出了一种新的基于过程的作物模型--一年生作物通用大面积模型（GLAM）。该模式的设计是为了在与全球和区域气候模式相当的空间尺度上运作。它旨在模拟气候对作物产量的影响。描述了模型参数确定和优化的过程，并对花生（即花生;花生）的预测进行了演示。1966年至1989年期间印度的产量。最佳参数（如消光系数，蒸腾效率，收获指数的变化率）是稳定的空间和时间，提供的产量技术趋势的估计是基于整个24年的时间。该模型有两个位置特定的参数，种植日期，和产量差距参数。后者在空间上变化，并通过校准来确定。当使用不同的输入数据时，最佳值略有不同。该模型使用2.5度x2.5度网格上的历史数据集进行测试，以模拟产量。三个网站进行了详细检查，从古吉拉特邦在西部，安得拉邦向南，和北方邦在北部的网格单元。观察到的和建模的产量之间的协议是可变的，相关系数分别为0.74，0.42和0。技能是最高的气候信号是最大的，相关性相当或大于季节平均降雨量的相关性。将来自所有35个细胞的产率汇总以模拟全印度产率。实测产量与模拟产量的相关系数为0.76，平均产量的均方根误差为8.4%。该模式可以很容易地扩展到任何一年生作物的气候变率（或变化）对作物产量的影响在大面积的调查。(C)2004 Elsevier B. V.保留所有权利。
The formulation of a new process-based crop model, the general large-area model (GLAM) for annual crops is presented. The model has been designed to operate on spatial scales commensurate with those of global and regional climate models. It aims to simulate the impact of climate on crop yield. Procedures for model parameter determination and optimisation are described, and demonstrated for the prediction of groundnut (i.e. peanut; Arachis hypogaea L.) yields across India for the period 1966-1989. Optimal parameters (e.g. extinction coefficient, transpiration efficiency, rate of change of harvest index) were stable over space and time, provided the estimate of the yield technology trend was based on the full 24-year period. The model has two location-specific parameters, the planting date, and the yield gap parameter. The latter varies spatially and is determined by calibration. The optimal value varies slightly when different input data are used. The model was tested using a historical data set on a 2.5degrees x 2.5degrees grid to simulate yields. Three sites are examined in detail-grid cells from Gujarat in the west, Andhra Pradesh towards the south, and Uttar Pradesh in the north. Agreement between observed and modelled yield was variable, with correlation coefficients of 0.74, 0.42 and 0, respectively. Skill was highest where the climate signal was greatest, and correlations were comparable to or greater than correlations with seasonal mean rainfall. Yields from all 35 cells were aggregated to simulate all-India yield. The correlation coefficient between observed and simulated yields was 0.76, and the root mean square error was 8.4% of the mean yield. The model can be easily extended to any annual crop for the investigation of the impacts of climate variability (or change) on crop yield over large areas. (C) 2004 Elsevier B.V. All rights reserved.