Data assimilation is the connection between models and observations. Data assimilation in numerical weather prediction (NWP) combines observations (which contain observation errors) with models that describe the time evolution of the atmospheric state (which are generally incomplete, resulting in forecast errors). Assimilation improves the model state based on the information contained in the observations. It provides the most probable state of the atmosphere, given the uncertainties in the observations and model forecast. This so-called "analysis" is used as forecast initial state.
The operational High Resolution Limited Area Model (HIRLAM
) at KNMI employs a 3D-Var assimilation scheme. Radiosonde, surface pressure observations and upper air temperature and wind information from aircraft observations are used to determine the initial state of the atmosphere. The horizontal resolution of HIRLAM is 11 km. Continuing advances in computer technology enable the representation of the atmosphere at finer grid scales. HIRLAM's successor HARMONIE
will be operated at 2.5 km resolution. A high density observation network is required to analyze the atmospheric state at the smallest scales that can be represented by the model.
Our activities focus on the use of additional (high-resolution) observations, including:
- scatterometer ocean surface winds
- Mode-S aircraft upper air wind and temperature observations
- Doppler radar wind observations
and a better use of the observations information content by taking into account the weather situation dependency of model errors.