Conversion of canopy intercepted radiation to photosynthate: review of modelling approaches for regional scales.
Conversion of canopy intercepted radiation to photosynthate: review of modelling approaches for regional scales.
复制标题
将冠层截获的辐射转化为光合产物：区域尺度建模方法的回顾。
作者：
B. Medlyn;D. Barrett;J. Landsberg;P. Sands;R. Clement
大多数陆地碳平衡模型的一个基本组成部分是估算由冠层截获的辐射驱动的植物冠层光合吸收。在这篇文章中，我们回顾了用于模拟辐射转化为光致发光的方法。由于这一过程在叶尺度上得到了很好的理解，建模问题基本上是一个扩大到树冠、区域或全球尺度的问题。因此，我们的审查集中在缩放的问题，包括模型识别，参数化和验证大规模。四种不同的方法通常被用来模拟大规模的光合产物生产：最大生产力模型、资源利用效率模型、大叶模型和遮阳模型。代表这些方法的模型进行了讨论和模型预测的总初级生产力估计来自涡度协方差数据测量以上的锡特卡云杉林。遮阳模型被认为在所有的时间尺度上表现最好。然而，其他模型也有显著的优点，包括实施简单，以及能够将该模型与植被辐射截获的遥感信息联合收割机结合起来。我们的结论是，所有四种方法都可以成功地用于模拟光合吸收，在给定的情况下，最好的方法将取决于模型的目标和数据的可用性。
A fundamental component of most models of terrestrial carbon balance is an estimate of plant canopy photosynthetic uptake driven by radiation interception by the canopy. In this article, we review approaches used to model the conversion of radiation into photosynthate. As this process is well understood at the leaf-scale, the modelling problem is essentially one of up-scaling, to canopy, regional or global scale. Our review therefore focuses on issues of scaling, including model identification, parameterisation and validation at large scales. Four different approaches are commonly taken to modelling photosynthate production at large scales: the maximum productivity, resource-use efficiency, big-leaf, and sun-shade models. Models representing each of these approaches are discussed and model predictions compared with estimates of gross primary productivity derived from eddy covariance data measured above a Sitka spruce forest. The sun-shade model was found to perform best at all time scales considered. However, other models had significant advantages including simplicity of implementation and the ability to combine the model with remotely-sensed information on vegetation radiation interception. We conclude that all four approaches can be successfully used to model photosynthetic uptake and that the best approach in a given situation will depend on model objectives and data availability.