Moist observations in HARMONIE 4DVAR and diagnosis of its performance (MSO project)

Brief description of the project:  Currently, many observations are not used or, at least, not optimally used by the KNMI high resolution model HARMONIE. In this proposal we want to remedy this by exploring an alternative data assimilation approach in conjunction with a very much increased use of humidity observations. We envisage this by extending the current three-dimensional variational (3DVAR) data assimilation system with a time dimension, which does justice to the fact that observations are spread in time. With this 4DVAR set-up, we want to focus on humidity observations from Global Navigation Satellite Systems (GNSS) and radar data. We expect that by better handling of observations in general and by massively increasing the use of humidity observations, we are much better positioned in providing improved HARMONIE forecasts in situations of, for example, strong convection and low visibility.

Infographic MESO4DVAR
Performance of HARMONIE forecasts for the daily cycle of cloud cover by using conventional and Mode_S EHS observations with 3DVAR (purple), 4DVAR (green). The observed daily cycle of cloud cover is given by the blue line.