Wind-tracking algorithms produce Atmospheric Motion Vectors (AMVs) by tracking water vapor across spatial-temporal fields. Thorough error characterization of wind-track algorithms, otherwise known as uncertainty quantification, is critical in properly assimilating their produced AMVs into forecast models. Uncertainty quantification has two key quantities of interest: accuracy— the systematic difference between a measurement and the true value, and precision— a measure of variability of the measurement. Traditional techniques for uncertainty quantification through machine learning have focused on characterizing accuracy but often struggle when estimating precision. By pairing a random forest algorithm with unsupervised parametric clustering (using a Gaussian Mixture Model), we propose a machine learning based method of building uncertainty models characterizing both accuracy and precision using limited experimental data. In particular, we develop a Gaussian Mixture Model to cluster the principle quantities of interest in our training dataset— water vapor, measured AMVs, and true wind speed— into discrete regimes each with a distinct precision and accuracy. Concurrently, we train a random forest to predict true wind speed given the outputs of a wind-tracking algorithm, which works to model some of the extreme error in the algorithm. Combining these, we build a model which can place a retrieved AMV into a distinct regime with a characterized accuracy and precision.
Even though vertical motion is resolved within convection-permitting models, recent studies have demonstrated significant departures in predicted storm updrafts and downdrafts when compared with Doppler observations of the same events. Several previous studies have attributed these departures to shortfalls in the representation of microphysical processes, in particular those pertaining to ice processes. Others have suggested that our inabilities to properly represent processes such as entrainment are responsible. Wrapped up in these issues are aspects such as the model grid resolution, as well as accuracy of models to correctly simulate the environmental conditions. Four primary terms comprise the vertical momentum equation: advection, pressure gradient forcing, thermodynamics and turbulence. Microphysical processes including their impacts on latent heating and their contributions to condensate loading strongly impact the thermodynamic term. The focus of this study is on the thermodynamic contributions to vertical motion, the shortfalls that arise when modeling this term, and the observations that might be made to improve the representation of those thermodynamical processes driving convective updrafts and downdrafts.
Recent technological advances have enabled the miniaturization of microwave instruments (radars and radiometers) so they can fit on very small satellites, with enough capability to measure atmospheric temperature, water vapor and clouds. The miniaturization makes these systems inexpensive enough to allow scientists to contemplate placing several examples in low-Earth orbit concurrently, to observe atmospheric dynamics in clouds and storms. To identify the most important weather and climate problems that can be addressed with these new observations, and to develop corresponding observation strategies using these "distributed" systems, specific analyses were conducted and used to justify "distributed" measurement requirements and quantify their expected performance. This presentation will describe the types of convoys, the expected observations, and their applications.