System description

So far, we used the name meteorogram although in many languages it is used interchangeably with meteogram - that is why in our newer services it appears in the latter form.

We are introducing information on storms on an experimental basis for the time being. The colours of the circle fillings indicate: • white: possibility of thunderstorms • yellow: weak storm up to 10 discharges per hour in one mesh (4x4 km) • orange: moderate thunderstorm up to 30 discharges per hour • red - severe storm above 30 discharges per hour How storms are determined is described in a separate section.

Storms are calculated by the UM model. First, model nodes are found where the GWP, or graupel water path exceeds a threshold value. The electrical charge is then parameterised at these nodes. If this charge exceeds a certain value or a multiple thereof, a discharge (or several discharges) is considered to have occurred. The remainder is passed on to the next calculation step. The discharge density presented is the quotient of the number of discharges per time period divided by the area belonging to the model node (16 km2). For example: a value of 1 on the storm map corresponds to 16 discharges for the last 3 hours in the model node. The representation of storms on the meteogram is explained in a separate question.

The longer the forecast, the lower its reliability. In principle, there is no technical obstacle to counting the forecast over such a horizon, but then its accuracy will not be significantly better than global long-term forecasts, which operate on a sparser computational grid. The results of the 14-day GFS model forecast are available on the site.

The cloudiness classification is according to the following table (indicative values): Very low - 1000 mb - from 0 to 111 m Low - from 1000 mb to 800 mb - from 111 m to 1949 m Medium - from 800 mb to 500 mb - from 19449 m to 5574 m High - from 500 m to 150 m - from 5574 m to 13608 m mb - pressure in millibars of standard atmosphere

Octant is the eighth part. It is used to determine cloudiness, e.g. cloudiness of 4/8 means half the sky is covered by clouds and 8/8 is full cloud.

At the weather station, when it snows, the condition (thickness) of the snow cover and the mass of water from the melted snow that has fallen into the rain gauge are measured. Thus, the increase in snow cover and the amount of precipitation in millimetres in terms of the thickness of the layer of water formed from this precipitation is measured. The numerical model determines the phase state of the precipitation (snow, snow with rain, rain), but does not give the thickness (increment) of the snow cover, which it could do when it is only snowing. The reason is that the density of freshly fallen snow varies widely: from 10 kg/m3 to 150, or even, up to 170 kg/m3, according to meteorological literature. The amount of precipitation, regardless of the phase state, is given by the model in kg/m2 and when it is only snowing, the user would like to know how many centimetres of this snow have fallen. This relationship can only be estimated.

No. New forecasts appear at times described in a separate answer. Computers constantly checking for new meteograms as a backup will be blocked because they put unnecessary strain on ICM servers and network links.

Some quantities shown in the meteogram are averages of values from neighbouring grid nodes. For the UM 4 km model, this is averaged over 7x7 model nodes. We also provide values from the midpoint and minima or maxima from the whole area. See the key of each meteogram for a detailed description.

This is the approximate horizontal distance between the points at which the parameters describing the weather condition (temperature, pressure, etc.) are determined. More technically, it is the horizontal distance between consecutive nodes of the computational grid.

This is because there has been a change from winter (CET) to summer time. After the time change, the difference between the official time in Poland (CEST) and UTC, at which meteorological measurements are taken and according to which calculations are made, is two hours instead of one. Therefore, model results will appear one hour later.

It depends on the model and the start time of the forecast: • UM 4 km: 00 and 12 o'clock forecast - 120 hours (5 days) • UM 4 km: 06 and 18 o'clock forecast - 60 hours (2.5 days) • UM 1.5 km: 00 and 12 o'clock forecast - 78 hours (~ 3.5 days) • WRF: 00, 06, 12 and 18 o'clock forecasts - 72 hours (3 days) • GFS (non-ICM): 00 and 12 o'clock forecasts - 14 days

In principle, that is how it should be. It happens, however, that the first few hours of the forecast take the "start-up" of some phenomena in the model, e.g. convection, so it is also worth looking at the results of the forecast calculated as the previous one. In addition, forecasts starting at different times may be based on larger or smaller initial data sets, which also sometimes influence the development of certain weather phenomena.

The UM model is calculated 4 times a day for the initial conditions at 00, 06, 12 and 18 o'clock (according to UTC time). Forecasts appear on the website approximately 4.5 hours after the time specified as the start of the forecast. WRF model updates take place 2 times per day for initial conditions at 00 and 12 UTC. The visualised calculation results are available on the website approximately 6.5 hours after the start of the forecast. The start of the forecast is at 03 and 15 UTC.