From local CRNS network observations to meso-scale soil moisture distributions

在Cosmic Sense的第一阶段，我们重点研究了几个研究点（Fendt, w<s:1> stebach, Marquardt）在高田尺度（0.1 - 1 km²）上的土壤湿度（SM）变化。CRNS位置的选择是为了在最大化足迹重叠和空间覆盖之间取得平衡。因此，可以通过假设空间上一定程度的连续性来识别足迹内部和足迹之间的水平SM梯度，这是由大的传感器足迹与高网络密度相结合来证明的。然而，对于较大间距的CRNS传感器，这种假设可能不成立。智能覆盖模块（Smart Coverage module， SC）的主要任务是弥合分布稀疏的传感器之间的差距。在继续项目第一阶段MC模块的实地规模工作的同时，我们现在将研究扩展到更大的规模。一个关键的问题是：我们如何使用CRNS技术来表示SM在中尺度（~ 10 - 100 km²）上的时空分布，而详尽的CRNS覆盖是不可实现的？为此，我们将运行小型CRNS集群和单个CRNS探针的“智能”组合，作为一个网络，其探针数量与第一阶段的联合野外战役相似，但具有优化的密度和位置。我们还将考虑一系列代理和支持测量，这些测量可以告诉我们在未监测的地点的SM，例如：(1)已知影响SM的静态景观属性（土壤质地，地貌，土地利用）；(2)遥感产品表征地表SM动态和植被特性；(3)整合影响根区气象强迫的水文模型；以及(4)利用地面和机载平台在离散时间点（RA）沿着横断面量化SM变化的流动观测。智能覆盖是指识别与基于CRNS的SM的关系尚未得到充分约束和理解的位置属性，因此具有此类属性的新位置可以通过额外的CRNS测量来覆盖。智能覆盖的另一个方面是，我们将使用识别的关系在固定CRNS网络中找到集水区尺度的最佳位置，从而使集水区尺度SM的总体估计不确定性最小化。这些想法将与其他研究模块一起在联合野外战役（JFC）中进行探索，也将在具有合成SM和代理数据以及模拟CRNS观测的“虚拟”JFC中进行探索。此外，我们将探索将新发展整合到CRNS网络的运行中，如高灵敏度探头、定向屏蔽和本地传入校正。我们还将继续缩减规模。总体而言，该研究模块将结合技术改进、方法发展、来自其他模块的中尺度数据和扩展的解释能力，通过在密集和稀疏网络设置中使用CRNS获得集水区尺度SM。

In the first phase of Cosmic Sense, we focused on soil moisture (SM) variability at the upper field scale (0.1 - 1 km²) at several study sites (Fendt, Wüstebach, Marquardt). CRNS locations were selected to achieve a balance between maximizing footprint overlap and spatial coverage. Thus, horizontal SM gradients could be identified within and between footprints by assuming some degree of continuity in space, justified by the large sensor footprint in combination with a high network density. For larger spacings of the CRNS sensors, this assumption may, however, not be valid.Bridging this gap between sparsely placed sensors is the main subject of the Smart Coverage module (SC). While continuing the field-scale work of the MC module of the first project phase, we now expand research to larger scales. A key question is: How can we use the CRNS technique to represent spatio-temporal distribution of SM on the meso-scale (~ 10 - 100 km²) where an exhaustive CRNS coverage is not feasible? To that end, we will operate a "smart" combination of small CRNS clusters and individual CRNS probes as a network with a similar number of probes as in the Joint Field Campaigns of the first phase, but with optimized density and locations. We will also consider a range of proxies and supporting measurements that inform us about SM at unmonitored locations, e. g. (1) static landscape attributes known to affect SM (soil texture, geomorphology, land use); (2) remote sensing products (RS) to represent surface SM dynamics and vegetation properties; (3) hydrological models to integrate the meteorological forcing that affects the root zone (RZ); as well as (4) roving observations using ground-based and airborne platforms to quantify SM variability along transects at discrete points in time (RA).Smart coverage refers to the identification of location properties for which the relationship to CRNS-based SM is not yet sufficiently constrained and understood, so that new locations with such properties can be covered by additional CRNS measurements. Another aspect of smart coverage is that we will use the identified relationships to find optimal locations within a stationary CRNS network at the catchment scale such that the overall estimation uncertainty of catchment-scale SM is minimized. These ideas will be explored, together with other research modules, in a Joint Field Campaign (JFC), but also in a "virtual” JFC with synthetic SM and proxy data and simulated CRNS observations.Furthermore, we will explore the integration of new developments into the operation of CRNS networks, such as high sensitivity probes, directional shielding, and local incoming correction. We will also continue our downscaling efforts. Overall, this research module will combine technical improvements, methodological developments, meso-scale data from other modules and extended interpretation capabilities to obtain catchment scale SM from the employment of CRNS in both dense and sparse network settings.