Numerical weather prediction (NWP) models are being continually extended to higher altitudes. Most operational NWP models now reach the lower mesosphere. The increase in altitude is motivated by the goal of improving extended-range weather forecasts and improving the understanding of the middle atmosphere. To meet this challenge, NRL has developed a high-altitude version of the Navy’s Operational Global Atmospheric Prediction System (NOGAPS), designated NOGAPS-ALPHA.

An NWP system includes a data assimilation (DA) component in addition to the forecast model. The DA algorithm combines a relatively accurate “first guess” forecast with the available meteorological observations to produce an analysis that is the “best estimate” of the true atmospheric state. The analysis provides the initial state for forecasts and is an important product for many applications. NRL has developed a state-of-the-art four-dimensional variational (4D-Var) assimilation algorithm called NAVDAS-AR. It is an extension of the operational NRL Atmospheric Variational Data Assimilation System (NAVDAS), a 3D-Var algorithm. 4D-Var is expected to improve assimilation in the stratosphere and mesosphere relative to 3D-Var through a better representation of the dynamical changes during an assimilation time window and by providing a more dynamically balanced initial state to prevent forecast degradation from spurious wave generation.

An opportunity exists for innovative research in techniques for stratospheric and mesospheric assimilation. This includes techniques that make optimal use of current atmospheric measurements and improvements to the 4D-Var algorithm. It also includes research in the area of ensemble-based assimilation techniques and assimilation of trace gases such as ozone, water vapor, and N₂O.