Ground-Based Remote-Sensing of Key Properties

Over the past two decades, the sophistication, and number of ground-based networks that have zenith-pointing millimeter radars, lidars, and radiometers as their cornerstones have increased significantly. Their measurements capture the structure of clouds, and precipitation, as well as the thermodynamical and dynamical state of the clear boundary layer. The vertically comprehensive and temporally continuous measurements that ground-based networks provide have filled a gap that was left by spaceborne and aircraft platforms. Toward improving our understanding of fast physics and its representation in numerical models, sophisticated retrieval algorithms have been developed to provide estimates of geophysical quantities. In parallel, forward simulators have been used to convert model output to measurements for apples-to-apples model-experiment comparisons, to quantify retrieval uncertainty, and to optimize experimental design. The multiparametric observations captured by ground-based observatories have supported numerical modeling in many ways, such as parameterization development, forcing datasets, and benchmarking. That said, the scale gap between high-resolution observations and large-scale models particularly complicates their intercomparison. Although methods exist to help bridge this gap, issues related to the representativeness of point observations across scales remain.While scanning sensors and spatially distributed instrument networks may help bridge this gap, further improvements in observing systems and model computational capabilities are desirable.