Page History
...
Note: HRES and Ensemble Control Forecast are scientifically, structurally and computationally identical. With effect from Cy49r1, Ensemble Control Forecast output is equivalent to HRES output where shown in the diagrams. At the time of the diagrams, HRES had resolution of 9km and ensemble members had a resolution of 18km.
Structure of the lower atmosphere and boundary layer
Modelling the structure of the lowest, near surface, layers is very important for analysis and forecasting, especially for cold temperatures at 2m. For this reason there are many model levels in the boundary layer. Capturing the structure of the temperature and moisture can be difficult and often imprecise. The structure is particularly difficult to define where there are low level inversions or complex orography. Modelling wind shear and mixing processes also has an impact.
...
Users should assess how well the model has analysed the structure of the inversion and take into account the extent and persistence of the cloud area under the inversion during the past few days.
Fig9.1-1: A comparison of observed (orange) and model analysed/forecast (green) temperature and dewpoint structures. Errors are due to assimilation issues coupled with the difficulties handling the cloud physics. In this case the surface cool and moist layer was analysed to be slightly deeper than in reality. This retarded fog clearance and therefore delayed heating and overturning of the boundary layer through the morning. So by 12UTC the forecast inversion was too low compared with reality and it had also not captured the stratocumulus from the convective overturning within the boundary layer. Consequently the true radiation balance around midday was not captured.
Fig9.1-2: Examples of the difficulty of describing the boundary layer temperature and moisture structure. Dew-point is used here as a measure of moisture. Forecast values shown in red; observed values shown in black. DT:00UTC 19 Jan 2025; T+12 VT12UTC 19 Jan 2025. There was a large anticyclone over Europe at this time with extensive low cloud but with some breaks mainly around the edges. In this case the Ensemble control (Ex-HRES) handled the boundary layer evolution rather poorly.
- At Budapest the boundary layer was observed to be saturated with thick low cloud but the forecast had thinner cloud with a higher base. Forecast brighter conditions allowed the near surface temperature to rise a little. The error in 2m temperature was about 2°C (Temperatures: Observed -3°C, forecast -1°C).
- At Wien the boundary layer was observed to be mostly saturated with low cloud but the forecast had broken the cloud. Forecast morning insolation allowed the near surface temperature to rise a rather more. The error in 2m temperature was about 4°C (Temperatures: Observed -3°C, forecast 1°C).
- At Sofia was observed to have fog in the morning and to remain with low cloud all day despite the thin layer of low cloud shown on the vertical profile. The forecast had readily dispersed the cloud. and unhindered morning insolation allowed the near surface temperature to rise a great deal. The error in 2m temperature was about 12°C (Temperatures: Observed -4°C, forecast 8°C).
Potential causes of poor stratus and stratocumulus forecast
IFS forecasts of low Stratus or stratocumulus sheets depend upon correct and detailed modelling of the lowest layers of the model atmosphere. This depends upon valid data. In particular availability of:
...
Before accepting the the presence or absence of areas of stratus or low stratocumulus, users should assess how well the boundary has been modelled by IFS. Compare the analysed and observed structure of the boundary layer and assess critically the validity of the data used. Low level winds and turbulent effects can also play a part as turbulence from a breeze can clear stratus by mixing down drier air while stagnant air can aid persistence.
...
Fig9.1-3: An example of low stratus not captured by ensemble control (ex HRES). FigA shows an area of stratus running in across the White Sea. Kem' observation shows mist, but lies somewhat inland from the coast. FigB shows that the ensemble control (ex HIRES) missed the low cloud area. The vertical profile shows some reasons - 1: the radiosonde data is incomplete below 900hPa (unrealistic straight lines) and boundary layer detail is lost. However, the moisture up to about 940hPa is clear. 2: the model has not assimilated the radiosonde data and is far too dry. Also, model low level winds (not shown) are stronger than the almost calm conditions observed contributing to overturning of the lowest layers of the boundary layer.
Fig9.1-4: An example of low stratocumulus not captured by ensemble control (ex HRES). The cloud layer is thin - thin enough to show Kármán vortices. Ensemble vertical profiles reveal that none of the ensemble members managed to create low clouds although all are close to saturation (see vertical profile). Nevertheless, the forecast of cloud, thin or not, can have a large effect on insolation and radiative cooling.
...
Overview
Community Forums
Content Tools