Back page

DATA ASSIMILATION

Workpackage number :

3

 

Start date or starting event:

+0

 

Partner code: Person-months

P1: 0

P2: 0

P3: 0

P4: 20

P5: 10

P6: 0

 

P7: 25

P8: 0

P9: 0

     

Objectives

The development of techniques for assimilation of radar data, among other remote sensing data, will be addressed to improve the quality of the description of the atmospheric state, to be used as initial conditions in Numerical Weather Prediction (NWP) methods. The impact of radar data on forecast quality will be studied through Observing System Experiments (OSE). Bearing in mind the constraints of the different numerical methods used in the proposal (i.e. nowcasting and very short-range forecasting over sub-national scale and short range forecasting over national scale), a spread of assimilation techniques will be investigated, ranging from laborious 3- and 4-dimensional variational assimilation (3D-Var and 4D-Var) to simpler interpolation techniques and nudging applied in continuous assimilation. Doppler radar data in super-observation form, prepared by workpackage 2, will be used as input to the variational data assimilation.

 

Methodology/work description

This is a central workpackage for the proposal. It is composed by a number of activities related to assimilation of radar and satellite data with three different techniques.

Variational Assimilation Partners 4 and 5 Implementation of an observation operator for radial winds in 3D-VAR and 4D-VAR for NWP purposes. Projection of the model state on the radar radial wind field. Coding of tangent-linear and adjoint observation operators. In the first phase, emphasis will be on using low elevation radar angles to minimise problems with projection of vertical velocities on radial wind vectors. Algorithms for quality control and ambiguity removal will also be developed and tested. Observation error statistics will be developed by comparison with other available observations (radiosonde data and aircraft wind data) and with model data. The impact of radial wind vectors on NWP forecast quality will be investigated by case studies (severe weather situations) and by Observing System Experiments, i.e. data assimilation and forecast experiments over extended periods with and without use of radar radial wind vectors.

Continuous Assimilation Partner 7 Implementation of 4-dimensional continuous assimilation, based on nudging (Newtonian relaxation) for radar, satellite as well as surface and radiosonde data.

 

WorkPackage Deliverables :

  1. Software modules for 4-dimensional assimilation; Delivery date: 24.
  2. Impact studies of radar radial winds on limited area NWP and an assessment of suitability of radar radial wind measurements for use in operational NWP; Delivery date:30.

 

Contribution to Project Milestones:

M1

M2 Overview for software modules to be applied for assimilation of radar and satellite data. Descriptions for ingestion of satellite and radar data. Description of observation operators, including tangent-linear and adjoint versions, for radar radial winds in 3D-Var. Results from first tests using simulated data in 3D-Var

M3 Implementation of the observation operators for radar radial winds in 4D-Var

M4 Implementation of continuous assimilation based on nudging

M5 Assessment of impact of radar wind in assimilation

M6 Dissemination of results

Back page