Modeling data assimilation

Kalnay. E.. 2003. Atmospheric Modeling, Data Assimilation and Predictability, Cambridge University Press. New York. 

How can weather forecast duration and reliability be improved by new space-based observations, data assimilation, and modeling ... 

Bioinformatics will play a vital role in overcoming this data integration and modeling challenge, because databases, visualization software, and analysis software must be built to enable data assimilation and to make the results accessible and useful for answering biological questions (p. 254). 

Fukumori I (2001) Data assimilation by models. In Satellite Altimetry and Earth Sciences. Academic Press, San Diego, p 237... 

Atmospheric Modelling and Data Assimilation Laboratory, Centre for Research in Earth and Space Science, York LFniversity, Toronto, Canada e-mail jwk wxprime.com... 

The current version of Enviro-HIRLAM has not previously been evaluated against ETEX-1 measurements. The ETX domain (Eig. 5.1) was used with at time-step fixed at 10 min, and initial and boundary conditions were post-processed from the European Centre for Medium-Range Weather Eorecasts operational model, IFS (Integrated Forecast System). No surface or upper air data assimilation was employed and the model was integrated 80 h into the future. The start time was on 23 October 1994 at 12 00 UTC, 4 h before the start of the release. Output was interpolated to measurement stations in order to compare to the observations and produce statistical measures. 

Coupling Global Atmospheric Chemistry Transport Models to ECMWF Integrated Forecasts System for Forecast and Data Assimilation Within GEMS... 

In IFS tracer data assimilation mode, the IFS tracer forecast mode is applied in the outer loops of ECMWF data assimilation system, i.e. the calculation of the trajectories runs of the complete model of the 4D VAR (Mahfouf and Rabier 2000) The iimer loops used in the minimisation step with the tangent linear and adjoint model are currently run uncoupled, i.e. without the application of the source and sink tendencies from the CTM. 

In the second mode, a consistent treatment of the emission injection and vertical transport would be achieved. In particular, the adjoint formulation of diffusion and convection in data assimilation would be consistent with the forward model. However, dislocation of the chemistry tendencies is more likely than in case 1 because the IPS concentration fields tend to differ more from the CTM fields. 

Data assimilation (DA) is important for daily chemistry forecasts, but it also is a challenge. At the moment there are much fewer stations with chemistry observations than meteorological observations, and the observations from these stations may not always be available in real time. Also, the gradients in chemical species often are very sharp, which the current observation network and data assimilation schemes are not capable of representing. And last but not least, it is difficult to determine the assimilation increments with so many possible species and so few detailed observations. This is why ECMWF has limited the chemistry in their model and data assimilation to five species (in addition to the cost in computer power). 

Data assimilation schemes for stand-alone atmospheric chemistry models are being developed by several groups in close collaboration with HIRLAM. Some of these schemes even have formulations close to the HIRLAM variational data assimilation. A coordination of these efforts for atmospheric chemistry data assimilation has a more long-term goal to extend the HIRLAM reference data assimilation to atmospheric chemistry. 

UAP models in operational UAQIFSs, as a rule, still use simple in-situ meteorological measurements which are fed into meteorological pre-processors (see Figure 9.1). Lacking an adequate description of physical phenomena and the complex data assimilation and parameterisations of numerical weather prediction (NWP) models, these pre-processors do not achieve the potential of NWP models in providing all the meteorological fields needed by modern UAP models to improve the UAQ forecasts. 

We have determined that nudging the models with the WTM surface observations alone appeared to have very limited impact on the mesoscale weather and dispersion forecasts. It is basically three-dimensional (two horizontal dimensions plus time) data assimilation over a small area. The impulses caused by the data insertion dissipated rapidly within 3 h after the turnoff of data nudging regardless of the length of relaxation time scale for model adjustment and nudging period. However, impulses caused by FDDA... 

An advanced mesoscale model may perform well in some cases but not in others, and the model responses to data assimilation are likely to vary with weather scenarios. There would be a potential benefit of using several model winds separately to run HPAC. A composite result of the HPAC runs would give a more complete depiction of the potential surface-based transport of hazardous agents. Also, the latest version of HPAC allows us to display dispersion output within a geographic information system (GIS) environment. We should explore all possible GIS applications in dispersion modeling and damage assessment. 

Data assimilation Inserting data into the weather prediction model to npdate the model state dnring the time integration step. 

Four-dimensional data assimilation (FDDA) A technique by which high-frequency observations are used to update the model state during the time integration step, improving the fidelity of meteorological model predictions. 

The ability to include data from weather forecasting models (assuming that a suitable meteorological data assimilation capability is attached to D2-Puff)... 

---