Low level jets (LLJs) describe conditions in which the wind speed reaches a local maximum with respect to altitude near the surface, and have been observed intermittently in the US Mid-Atlantic offshore environment. LLJs pose unique operating conditions for future wind turbines operating in the region, presenting negative shear and locally strong veer, but they are not typically considered in existing turbine standards. This work builds upon recent research that explains the formation and evolution of U.S. Mid-Atlantic LLJs through a simple analytical governing equation. We generate several LLJ inflow conditions with varying jet characteristics based on this analytical model and create monotonically-sheared (MS) analogues with constant veer in order to assess the impacts of the LLJ on turbine performance and loading. Using aeroelastic simulations with these inflow conditions on the IEA 15MW reference turbine, we find that the LLJ leads to a greater range of tower top pitching and yawing moments, which could contribute to larger accumulated structural fatigue in components compared to monotonically-sheared inflow. These preliminary results demonstrate a path toward a unified set of test cases for low-level wind maxima that can inform International Electrotechnical Commission standards related to offshore wind turbine design.

Accurately representing the formation of precipitation due to coalescence of cloud droplets remains a major challenge for atmospheric models. This study introduces and compares three data-driven approaches for modeling the time evolution of DSDs within a reduced-order latent space: a polynomial-based Sparse Identification of Nonlinear Dynamics (SINDy) framework, numerical integration of a neural-network-predicted time derivative, and direct prediction of state at the next timestep via neural network. All three methods are coupled and simultaneously trained with an autoencoder that discovers the latent space variables. We train with high-fidelity data from large eddy simulations equipped with the superdroplet method, which captures the stochastic nature of the coalescence process. Our results show that while all three approaches can successfully emulate the nonlinear dynamics of coalescence, the simplest, a polynomial-based SINDy representation, generalizes most robustly to out-of-sample cloud regimes, achieving strong performance and physical consistency. We further quantify model uncertainty using conformal prediction intervals, thereby providing rigorous estimates of predictive confidence across the modeling pipeline. These findings highlight that simple dynamical representations, when coupled to the correct prognostic variables, are a promising approach to advancing data-driven microphysics parameterizations toward operational atmospheric modeling.