Profiling optimal-Estimates for RaIn-CLoud Efficiency Study (PERICLES)

降水被一致认为是全球水和能量循环的中心变量之一，主要是因为它对人类、农业和地球上的生命的水的可用性具有直接意义，而且还因为它通过相关的潜热释放对能量收支和大气环流产生影响。在数值天气预报和大气环流模式中，降水过程对天气现象和气候演变具有决定性的控制和预测作用。尽管水对地球上的所有生物和整个地球系统都很重要，但人们对云和降水的生命周期却知之甚少；像暖雨的快速形成这样一个看似简单的过程仍然令人困惑，而且离有一个社区共识的解释或模型还很远。支撑云向雨过程演变的微物理过程的复杂性是这一领域取得重大飞跃的主要障碍，迫切需要将建模和观测相结合。虽然大涡模拟和云解析模式的时空尺度现在都适合研究云的生命周期，但遥感观测（唯一实际可能观察这种现象的方法）一直受到不适定反演问题带来的不确定性的影响。例如，根据定义，雷达反射率信号强烈依赖于波束体积中散射体（例如雨滴）的水滴大小分布，因此其解释与负责水滴大小分布形成及其演变的微物理过程有关。一个独特的部署（将于2010年底完成），在所有ARM设施中使用辐射模式的多波长扫描雷达，将提供前所未有的独立观测，这将缩小检索过程中的不确定性，并提供云演变的所有阶段的详细观测，从开始到上升气流和下降气流的发展，到水流星在时间和空间上的演变，再到凝结水划分为降水和外流顶。我们建议通过开发一种能够以一致的物理方式集成不同传感器的最佳估计方法来利用这一即将到来的机会。我们将结合主动（雷达反射率）和被动（亮度温度）测量，因为两者在整个垂直延伸过程中产生不同类型的云微物理信息：云和天气雷达允许距离分辨云结构，而被动微波信号包含关于视距综合水/冰含量的信息。我们提出的技术将测量（及其误差特征）与先验信息（以及关于其代表性的知识）结合到一个最佳估计框架中，以提供大气状态和不确定性估计。为了最佳地利用遥感观测的信息内容，对大气状态的首次猜测通过正演模式进行迭代-将大气状态与测量相连接-直到测量和先验信息与检索到的大气状态最匹配的点。检索的最终结果由观测到的大气柱的云和降水含水量剖面表示，这将在计划于2011年在俄克拉荷马南部大平原站点进行的MC3E活动期间通过独立的方法进行广泛验证。这个尖端的产品将有助于开发，评估，并最终改善数值模式中的云降水过程的参数化。作为测试平台，将与MC3E同时进行针对不同微物理封装评估的详细云解析模型研究。

Precipitation is unanimously recognized as one of the central variables of the global water and energy cycle, mainly because of its direct significance for the availability of water for human beings, agriculture and life on Earth in general, but also because of its impact on the energy budget and the atmospheric circulation through the associated latent heat release. Precipitation processes play a decisive role in controlling and thus predicting both weather phenomena and climate evolution in numerical weather prediction and general circulation models. Despite the importance of water to all creatures on Earth and to the Earth system as a whole, the life cycle of clouds and precipitation is not well understood; a seemingly simple process like the rapid formation of warm rain is still puzzling, and remains far from having a community-consensus explanation or model. The complexity of the microphysical processes underpinning the cloud evolution into the rain process represents the major obstacle for a considerable leap forward in this field and urgently calls for an effort towards combining modeling and observations. While the temporal and spatial scales of both Large-Eddy Simulation and Cloud Resolving Models are now suitable for studying cloud lifecycles, remote sensing observations (the only practically possible to look at such phenomena) have always suffered by the uncertainties deriving from ill-posed inversion problems. For instance the radar reflectivity signal is by definition strongly dependent on the drop size distribution of the scatterers, e.g., raindrops, in the beam volume and its interpretation is therefore related to the microphysical processes responsible for the formation of drop size distributions and their evolution. A unique deployment (to be completed by end of 2010) of multi-wavelength scanning radar with radiometric mode at all ARM facilities will provide unprecedented independent observations which should narrow down the uncertainties in the retrieval process and provide detailed observations of all phases of cloud evolution, from initiation, to development of updrafts and downdrafts, to hydrometeor evolution in time and space, to partitioning of condensate into precipitation and outflow anvils. We propose to take advantage of this upcoming opportunity by developing an optimal estimation approach capable of integrating different sensors in a consistent physical way. We will combine active (radar reflectivity) and passive (brightness temperatures) measurements because both yield different kinds of cloud microphysics information throughout the vertical extension: cloud and weather radars allow to range-resolve cloud structure, whereas passive microwave signals contain information about along-sight integrated water/ice contents. Our proposed technique combines measurements (and their error characteristics) with a priori information (and knowledge about its representativeness) into an optimal estimation framework to provide the atmospheric state together with uncertainty estimates. In order to optimally exploit the information content of remote sensing observations a first guess of the atmospheric state is iterated through the forward model - connecting atmospheric state with the measurement - up to a point where measurements and a priori information best match the retrieved atmospheric state. The ultimate product of the retrieval is represented by profiles of cloud and precipitation water content for the observed atmospheric columns, which will be extensively validated by independent methodologies during the MC3E campaign, planned for 2011 at the Oklahoma Southern Great Plain site. This cutting-edge product will help in developing, evaluating, and ultimately improving parameterization of cloud-precipitation processes in numerical models. As a test bed, a detailed cloud resolving model study oriented at the evaluation of different microphysical packages will be conducted in coincidence with MC3E.