You are viewing an old version of this page. View the current version.

Compare with Current View Page History

Version 1 Next »

ecCharts

ecCharts is a web-based presentation system to show a very wide range of ECMWF IFS model output worldwide in chart form, and in a very flexible way under the control of the user.  This allows ECMWF's medium-range forecasts to be viewed in far greater detail than was previously possible on the web, with panning and zooming being key functionalities that are not provided by the standard web charts.  ecCharts allows several forecast fields to be superimposed for ease of inter-comparison and as an aid to understanding the forecast structure of the atmosphere, and these charts may be animated throughout the forecast period.   HRES deterministic data and ENS probabilistic data may be selected together allowing assessment of the likelihood of an event or the threat of severe weather in a locality.  Several derived fields (e.g.vorticity, divergence, CAPE, CAPE-shear, etc) are available, also simulated IR and WV images derived from HRES.   A probe tool is provided to allow display of values at a selected point at the chart time.  Results from the latest and previous forecast runs are also readily available.  Time sequences of variables at a selected point can also be displayed in graph form.  

ecCharts documentation gives a clear and comprehensive description of options and practical use of ecCharts.

Illustration of the use of ecCharts

HRES weighting as an aid to deciding on weather hazards.

ecCharts allow exploration of the effects of HRES as a member of the ensemble.  HRES is likely to capture and forecast smaller scale features better than ENS in the short term due to its unperturbed analysis, higher resolution, finer orography, and convective precipitation is likely to be better described.  Later HRES loses the advantage over ENS owing to increasing potential for amplification of spurious features from small instabilities in the flow pattern.  HRES weighting when used as part of the ensemble can in principal compensate for this variation in forecast confidence, but the suggested weighting values may be varied at the discretion of the forecaster.  Fig10.2.1 and Fig10.2.2 illustrate the point where the weighting has been allowed to be higher than one might ordinarily use for Day4 (an HRES weighting of 8) in recognition of the energetic embedded instability, the dynamic forcing associated with an upper trough and the forced uplift over the Alpine region.

Fig10.2.1: ecChart for 00Z 12 August 2017 T+96 forecast from HRES and ENS runs data time 00UTC 8 August 2017.  On the chart the pin is near Trenčín in western Slovakia.  No weighting has been given to the HRES forecast and the probability of rainfall >7mm over last 6hr at Trenčín is 27%.  The meteogram for Trenčín shows the HRES had a major peak in precipitation at this time and location, possibly due to better modelling of embedded convection within the rain area.


Fig10.2.2: ecChart for 00Z 12 August 2017 T+96 forecast from HRES and ENS runs data time 00UTC 8 August 2017 as above, but with a weighting equivalent to 10 ENS members to recognise the likely better detail of forecast convective precipitation.  With this weighting the probability of rainfall >7mm over last 6hr at Trenčín is now greater at 44%.  In other cases this type of approach might cause a warning threshold to be exceeded.

Comparing fields for consistency

Ideally, simulated water vapour images can be compared with water vapour imagery to allow an assessment of any departure of the analysed and forecast fields from reality.  The structure of the IFS model atmosphere may also be explored by comparing the water vapour imagery with the potential vorticity fields.  Currently satellite imagery is not available on ecCharts and comparison of observed and simulated imagery cannot be done directly.

The simulated water vapour image is derived from a radiative transfer model applied to the IFS atmosphere.  As the distribution of moisture is governed by ascent, areas that are coloured light grey by convention (=low temperature emissions) commonly tally with regions of current or past ascent, and regions coloured dark grey commonly tally with current or past descent.  The potential vorticity (PV=2) pattern is essentially governed by the dynamics of the upper flow and variations induce ascent or descent at around the level of the tropopause (conventionally taken as indicated by 2 PV units; higher values represent stratospheric air as shown here by shades of purple).  In the vicinity of developing depressions it would be expected that descending air (here shown by stratospheric air at 300hPa) would match the confinement of moisture to lower levels. Comparing the detail of these patterns with satellite imagery can provide a powerful method of identifying discrepancies between the modelled atmosphere and reality.


Fig10.2.3: ecChart for 00Z 8 August 2017 T+00 from HRES run data time 00UTC 8 August 2017 showing the simulated water vapour image and 300hPa potential vorticity.  These can be compared with WV satellite imagery to highlight discrepancies between the modelled atmosphere and reality.   Stratospheric air (PV=2 taken as near the tropopause, greater PV values (shown purple)) imply stratospheric air has descended to 300hPa.  Descent is also implied by the darker areas of the simulated water vapour image.  Comparing WVimagery with simulated imagery and PV charts can provide a powerful method of identifying discrepancies between the modelled atmosphere and reality.

Considerations when using ecCharts

Although comprehensive and detailed investigations and presentations are possible with ecCharts, a few points should be considered when interpreting the data:

  • The values of isopleths of several parameters (e.g. CAPE, probabilities) are not marked on the charts but values are shown by colouring as shown in the legends window.  Use of the probe tool can be helpful.
  • Some of the isopleth intervals on the pre-defined chart styles are non-uniform (e.g. some probabilities have isopleths at 5%, 30%, 60%, 70%, 80%, 90%; others are different).  Use of the probe tool can be helpful.
  • There is no way to directly compare charts for a given time from different forecast runs.
  • There is no way to compare IFS model analyses and short-term forecasts directly with observational data such as imagery.
  • Some fields are only available at low temporal resolution (e.g. CAPE EFI is for 24h periods);  if these are selected then other parameters will not be displayed at their intervening steps (e.g. 3h precipitation would only be displayed for 21-24UTC each day if included with CAPE EFI).

 


  • No labels