From field scale eco-hydrological process understanding to landscape scale water fluxes

从水文的角度来看，由于生物、气候、地形和物理景观特征，许多陆地表面模型（LSM）在表征与水相关的反馈机制方面存在重大缺陷。这导致对水文过程的描述过于简化，影响了这些模型在模拟水通量方面的可信度。在过去，由于缺乏计算能力和对水文过程的理解，这些简化往往是必要的妥协。在RESPECT研究小组的第一阶段，我们已经开始改进新开发的LSM“HUMBOL-TD”中的水文程序。这是基于对山地雨林土壤-植物-大气连续体中水文过程和与水有关的反馈机制的调查。我们讨论了土壤中不同的流动机制，特别是优先流动过程，这经常被lsm忽略。基于我们在第一阶段的发现和模型改进，并考虑到RESPECT研究小组在第二阶段的总体目标，我们确定了该后续项目的四个工作包（WP）中要解决的三个目标。对于我们的第一个目标，我们将我们的实地实验研究扩展到山地干燥森林和雨林（WP1），并使用我们创新的土壤水采样系统（WP2）沿着吸力压力梯度更详细地研究水的稳定同位素的混合过程。这些数据将在WP2中用于验证新开发的HUMBOL-TD同位素模块，并改善LSM的整体水文模型性能。作为我们第二个目标的一部分，我们将平衡模型性能和运行时间，因为我们希望在WP3中以景观规模运行HUMBOL-TD。我们将确定性能良好的水文模型结构，这些结构与我们从WP2中获得的现场尺度过程理解相匹配。为此，我们将根据系统模型分解的概念实现半自动搜索算法，并为每个子集水区分配最佳表现的模型结构。除了我们的现场实验数据外，我们还将使用从我们自己的长期监测站的排放数据中获得的水文特征以及国家机构的进一步观测来进行模型校准和验证。最后，在我们的第三个目标中，我们将研究多标准假设检验（WP4）。我们首先采用严格的多标准模型评估洪堡- td的水、碳和能量通量，使用我们的数据，以及其他子项目的进一步数据。在共同的努力下，我们将在气候和土地利用变化的背景下测试RESPECT的中心假设。

From a hydrological perspective, many Land Surface Models (LSM) have major deficiencies in the representation of water-related feedback mechanisms due to biotic, climatic, topographic and physical landscape features. This results in an oversimplified description of hydrological processes, affecting the credibility of these models in simulating water fluxes. In the past, those simplifications where often the necessary compromise due to the lack of computational power and hydrological process understanding.In the first phase of the research group RESPECT, we have started to improve the hydrological routines within the newly developed LSM “HUMBOL-TD”. This was based on investigations of hydrological processes and water related feedback mechanisms within the soil-plant-atmosphere continuum of the mountain rainforest. We addressed different flow mechanisms in the soil and particularly preferential flow processes, which are frequently ignored by LSMs.Building on our findings and model improvements made in phase I and in light of the overall objectives of the research group RESPECT in phase II, we have identified three objectives tackled in four work packages (WP) for this follow-up project. For our first objective, we extend our field experimental studies to mountain dry and rain forests (WP1) and study the mixing processes of stable isotopes of water in more detail along a suction pressure gradient using i.e. our innovative soil water sampling system (WP2). These data will be used in WP2 for the validation of the newly developed isotope module of HUMBOL-TD and to improve the overall hydrological model performance of the LSM. As part of our second objective, we will balance model performance and runtime in view of our ambitions to run HUMBOL-TD on the landscape scale in WP3. We will identify well-performing hydrological model structures that match our derived field scale process understanding form WP2. For this, we will implement a semi-automatic search algorithm following the concept of systematic model breakdown and assign the best performing model structure to each sub-catchment. In addition to our field experimental data, we will use hydrological signatures derived from discharge data from our own long-term monitoring stations and further observations from national agencies for model calibration and validation. Finally, in our third objective, will work on multi-criteria hypothesis testing (WP4). We start with a rigid multi-criteria model evaluation of the water, carbon and energy fluxes in HUMBOL-TD using our, but also further data from other sub-projects. In a joint effort, we will then test the central hypotheses of RESPECT under the light of climate and land use changes.