Page History

...

The figures below (Fig8.10.1 2 and Fig8.10.23)  illustrate the significant differences that can occur between data shown by instantaneous type of precipitation charts and the important detail that can be concealed.   Histograms show probabilities of each type of precipitation measured as the proportion among the ensemble members of each type of precipitation forecast at the selected location.  The probabilities of all forecast precipitation types are shown, including potentially hazardous precipitation, even at low probabilities.

Users should always inspect the histogram of precipitation types.

Fig8.1.10-1A2A: HRES forecast type of precipitation chart and ensemble most probable type of precipitation. HRES DT 12UTC 10 Mar 2024, VT T+72 12UTC 13 Mar 2024.  At the arrowed location the HRES forecast shows freezing rain (red) but ensemble shows most probable precipitation is snow (50-70%).

Note: HRES and Ensemble Control Forecast (ex-HRES) are scientifically, structurally and computationally identical.  With effect from Cy49r1, Ensemble Control Forecast (ex-HRES) output is equivalent to HRES output shown in the diagrams.   At the time of the diagrams, HRES had resolution of 9km and ensemble members had a resolution of 18km.

Fig8.1.10-1B2B: Vertical profile at the arrowed location in Fig8.1.10-12.  DT 12UTC 10 Mar 2024, VT T+72 12UTC 13 Mar 2024.  The HRES forecast (solid line) shows warm layer above the surface typical of freezing rain.  The ensemble shows a broad range of temperatures but no clear indication of the structure of the warm layer.  

Note: HRES and Ensemble Control Forecast (ex-HRES) are scientifically, structurally and computationally identical.  With effect from Cy49r1, Ensemble Control Forecast (ex-HRES) output is equivalent to HRES output shown in the diagrams.   At the time of the diagrams, HRES had resolution of 9km and ensemble members had a resolution of 18km.

Fig8.1.10-1C2C: Type of precipitation histogram at the arrowed location in Fig8.1.10-12.  DT 12UTC 10 Mar 2024, VT T+72 12UTC 13 Mar 2024.  The histogram shows the main probability is for snow.  Note that types of precipitation diagrams separate intensities into heavy, moderate and light and it might be useful to sum the forecast probabilities of all intensities of the precipitation type.

Detail of precipitation types: 49% snow (purple, 0.2-1.0mm/hr), 13% light snow (light purple,0.04-0.2mm/hr), 14% wet snow (blue, 0.2-1.0mm/hr), 7%  light wet snow (light blue, 0.04-0.2mm/hr), 2% rain (green, 0.2-1.0mm/hr).  A small chance of freezing rain (red, 0.2-1.0mm/hr) is shown during the preceding 6 hours.  The probability of snow of any intensity is about 76%.

Fig8.1.10-2A3A: HRES forecast type of precipitation chart and ensemble most probable type of precipitation.  HRES DT 12UTC 10 Mar 2024, VT T+72 12UTC 13 Mar 2024.  At the arrowed location the HRES forecast shows ice pellets (orange) but ensemble shows most probable precipitation is uncertain (30-50%), but see also Fig8.1.10-2C3C.  

Note: HRES and Ensemble Control Forecast (ex-HRES) are scientifically, structurally and computationally identical.  With effect from Cy49r1, Ensemble Control Forecast (ex-HRES) output is equivalent to HRES output shown in the diagrams.   At the time of the diagrams, HRES had resolution of 9km and ensemble members had a resolution of 18km.

Fig8.1.10-2B3B: Vertical profile at the arrowed location in Fig8.1.10-23.  DT 12UTC 10 Mar 2024, VT T+72 12UTC 13 Mar 2024.  The HRES forecast (solid line) shows temperature fringes the 0ºC isotherm layer above the surface typical of ice pellets.  The ensemble shows a broad range of temperatures but no clear indication of the structure of the warmer layer.  

Note: HRES and Ensemble Control Forecast (ex-HRES) are scientifically, structurally and computationally identical.  With effect from Cy49r1, Ensemble Control Forecast (ex-HRES) output is equivalent to HRES output shown in the diagrams.   At the time of the diagrams, HRES had resolution of 9km and ensemble members had a resolution of 18km.

Fig8.1.10-2C3C:  Type of precipitation histogram at the arrowed location in Fig8.1.10-23.  DT 12UTC 10 Mar 2024, VT T+72 12UTC 13 Mar 2024.  The histogram shows a wide range of precipitation types, none individually greater than 40% probability so the most probable type of precipitation is coloured grey on the chart (Fig8.1.10-2A3A). There is, however, an indication of freezing rain which is a significant hazard which is not evident from the chart (Fig8.1.10-2A3A).  Note that types of precipitation diagrams separate intensities into heavy, moderate and light and it might be useful to sum the forecast probabilities of all intensities of the precipitation type.

...

• At shorter lead-times the ensemble solutions will usually be fairly similar.  The probability of the types of precipitation are likely to be high.  Charts normally show a lot of detail.
• As lead-time increases the spread of ensemble solutions will be greater.  The development, timing and location of precipitation events will become less certain.  Charts will show larger areas of grey (<50% probability of precipitation type).

Fig8.1.10-74: Multiplots of ensemble probability of type of precipitation forecasts all verifying at 12UTC 21 February 2018 from a series of ensemble forecast runs at 24hr intervals.  As lead-time increases the more hazardous, less common types of precipitation are less prominent or do not appear, and greys (total probability <50%) are more prominent.  

...

Nevertheless, IFS is able to predict freezing rain several days in advance despite the finely balanced vertical thermodynamics structure required.

Fig8.1.10-85: Sequence of precipitation types DT:00UTC 04 March 2023.   VTs:00UTC 04 Mar (T+00) to 12UTC 06 Mar (T+60).  Precipitation type shown by colours.  

...

The following plots illustrate how the handling of precipitation type by ensemble forecasts in a marginal rain-snow situation can make physical sense; and specifically how user-relevant subtle changes are predicted with some precision.

Fig8.1.10-126: Illustration of the probabilities of different instantaneous precipitation types.  The flow is mainly easterly over England and illustrates the impact of different length sea tracks over the relatively warm sea, and the impact of topography.  Consider an approximate low level trajectory from Belgium to Wales (dashed arrow).  Colour is used wherever the probability of some precipitation falling is 50% (from the ensemble).  The colour itself illustrates the most likely type, whilst the darkness of the shading indicates how likely that type is.  Dry snow over Belgium, wet snow and sleet over the relatively warm southern North Sea, wet snow over colder southeastern England, dry snow further away from the sea (most likely as the air re-cools partly via evaporation), then over Welsh mountains the probability of dry snow is very high (because the high probability of temperatures being below zero and of high probability of precipitation falling due to orographically forced ascent).

Fig8.1.10-137: Typical histograms of probabilities of instantaneous precipitation type for locations within the airstream.  Colours in Fig8.1.10-12 7 represent the most likely (i.e. produced by the greatest percentage of ensemble members) but don't show other types, even if they are marginally less likely.  The use of histograms gives a better overview of precipitation types and enables a more confident forecast to be made by the user.  Considering 12UTC 31 Jan 2019: In Belgium there is high confidence of dry snow with very small probability of any alternative precipitation type.  

...