What does one mm of rain mean

This site also uses cookies

Precipitation falls on the earth's surface in the form of rain, snow, ice, hail or dew.

All rainfall is in Millimeters (mm) water column specified.

We give our high-resolution precipitation forecasts in pointMeteograms, Pictocast, rainSPOT and in the maps.
Other representations and scales can be set up on request.


meteoblue predicts the following types of precipitation:

  • Rain (liquid).
  • Snow (crystals).
  • Ice (frozen water) freezing rain.
  • Dew (condensation surfaces) for AGRO Services.

The calculation of other types of precipitation, such as hail (solid ice), can be offered as a special service.

Convective and stratiform precipitation

Precipitation comes in three different forms:

  • Convective Precipitation falls as showers with rapidly changing intensity. Since convective clouds have a certain horizontal extent, they fall over specific areas and usually for a short time.
  • Orographic Precipitation falls when air masses are pushed against land elevations, such as mountains, and then lifted by the wind.
  • Stratiforms Precipitation is caused by fronts and falls over larger areas and for longer periods of time.

meteoblue shows the amounts of precipitation as totals. In the meteograms, precipitation is divided into convective (both convective and orographic) into light blue (showers) and the sum (by adding stratiform precipitation) into dark blue. Convective precipitation is likely to be more erratic and unevenly distributed than stratiform precipitation, so its current values ​​and probabilities are more likely to vary compared to their averages than stratiform precipitation.

Snowfalls are displayed with "***" and freezing rain with "!". Hail is only displayed for special services. If precipitation falls as snow, the height of the snow cover can be displayed separately - otherwise, multiply the amount of water (in mm) by a factor of 10 to get the height of the snow cover.

When temperatures are above 3 ° C, rainfall is usually rainfall. This limit of 3 ° C is indicative. In order for there to be snowfall, the atmosphere must be below zero in the range of 0.5-5 km above the ground. Temperatures on the ground are almost always higher than above, and sometimes above 0 ° C when snowflakes are falling. If the temperatures are too high, the snowflakes will melt before they hit the ground

In some cases (e.g. warm fronts) warm moist air can flow over cold air. The precipitation is then originally rain, which turns into snow on the way to the ground when it reaches the cold air layer. In some cases (with a very cold layer of air above the ground) the precipitation can freeze, creating "freezing rain" made entirely of liquid droplets. The raindrops are supercooled as they fall through a layer of supercooled air several hundred meters thick above the ground and then freeze when they hit a surface such as the ground, trees, electrical lines, airplanes and cars. The resulting ice is called clear ice (or "lightning ice") and can be several centimeters thick. The METAR code for freezing rain is FZRA.

Nierderschlag - 5 days meteogram


Precipitation is in quantity for a particular one Period measured, for example in millimeters per hour. It is illustrated in millimeters (mm) water columns. They apply to the previous period of the specified time. One millimeter of rain corresponds to 1 liter of water per square meter of surface, or approximately 10 millimeters of snow.

The measurement is valid for the period before the specified time. For example: three hours of precipitation at 12:00 UTC shows the total amount in millimeters (= l / m²) between 9:00 UTC and 12:00 UTC.

Precipitation measurement is the attempt to represent a spatially unevenly distributed process with the help of sampling methods (weather stations) on a large scale using raster images (radar or satellites).

Unless otherwise stated, meteoblue compares the precipitation forecast with measurements based on the WMO standards.


Precipitation can be measured using 3 methods:

  1. Local weather stations: with rain gauges ("Pluviometer", "Rain gauges").
  2. Remote sensing: with the help of radar, the distribution of precipitation in the atmosphere can be calculated.
  3. Local indirect observations: with the help of containers, subsidence, surface runoff or river water levels.

Units of measurement are millimeter, Centimeters or inches; 1 millimeter corresponds to 1 liter per square meter.
Snow precipitation is measured in centimeters or inches.
meteoblue uses metric units to show precipitation for the specified time period (day or hour). Solid precipitation (snow) is shown in millimeters of water equivalent. Solid precipitation (snow) is in the mm water equivalent (WÄ) or as snow height in cm for the specified period.


In order to be able to classify precipitation chronologically, one can imagine a time-quantity diagram: The y-axis (abscissa) indicates the precipitation quantity in mm water columns and the x-axis (ordinate) indicates the time.

There are roughly four different types of precipitation in terms of amount and time:

  • heavy, short precipitation: it rains a large amount in a short time (e.g. thunderstorms, downpours)
  • Heavy, long-lasting precipitation: it rains a large amount over a longer period of time (e.g. continuous rain, monsoon rain)
  • weak, short precipitation: it rains a smaller amount in a short time (e.g. short drizzle)
  • weak, long-lasting precipitation: it rains a small amount over a longer period of time (e.g. long drizzle)

In reality, however, it is difficult to distinguish between these different variants, since the precipitation is usually a combination or hybrid of the individual variants.

In addition, it must be noted that the weather is not based on the date line, so precipitation events can be postponed by a few hours and then slide to the following day, which changes the pictograms of the respective days enormously.

Since precipitation events can occur sporadically or temporally (stochastically), we give a precipitation probability that is calculated from 20 ensemble runs. The probability is the frequency with which precipitation occurs in the 20 different forecast calculations.

The probability of precipitation refers to precipitation events with more than 0.1 liter per hour or more than 1 liter per day. The duration of the precipitation is irrelevant, so short showers are also recorded. It does not provide any information about the amount, duration, form (rain, snow, freezing rain, etc.) or the intensity of the precipitation. Drizzle can occur with a very high probability, with a rainfall of less than 1mm per event. A thunderstorm, on the other hand, can arrive at the forecast location with a very low probability, but if it does occur it will rain down more than 10mm per event.

A precipitation probability of 50% means that it will rain with a probability of 50%, i.e. in 1 out of 2 cases, so the forecast is relatively uncertain. A probability of 25%, for example, would be more unambiguous; this would mean that rain would occur in 1 out of 4 cases. With the help of the accuracy one can classify, check and evaluate the predictions better. On the home page, the accuracy is shown with the help of bars.

The amount of precipitation is calculated with a higher resolution, this can lead to inconsistencies between the amount of precipitation and the probability of precipitation.

The probability of precipitation is usually calculated for an area (grid cell) of 50x50km, also for days 1-7. For the first 7 days, the results of the 12km calculation are proportionally included and for the first 3 days the 3km calculation. Using the 3km calculation for every day would be too time-consuming and expensive.

3 and 14 day forecast

It must be noted that the probability of precipitation for the first 3 days is much more accurate than for days 4-7 or for days 7-14. This is due to the 3x3km grid used there, in contrast to the 12x12km grid that is used for days 4-7 or the 50x50km grid for 14 days. Effectively, in mountain regions, during days 4-7, excessive amounts of precipitation are predicted. This is a special effect of our European high-resolution model: the differences between mountain and valley cannot be included, the model leads mountains and valleys into a grid cell. With the 3x3km grid, however, mountains and valleys are easy to distinguish. So far we have not yet found a "simple" solution to correct this effect, because artificially lowering the precipitation forecast amount would also reduce the precipitation forecast on the mountains (which are less populated, but also important). That is why we decided not to artificially influence this effect, but to make the interpretation easier with this explanation.

Contradictions between the amount of precipitation and the probability of precipitation

It is possible that the amount of precipitation and the probability of precipitation contradict each other. For example, the predicted amount of precipitation can be 0mm, while a high probability of precipitation is indicated. Or, in reality, it rains, although there is only a very low probability of precipitation. These scenarios can be partially explained. It must always be borne in mind that the probability of precipitation relates to a larger, surrounding area, while the amount of precipitation is calculated precisely for the desired location. In the event of a thunderstorm in the area, the probability of precipitation for the selected location can be relatively high, but the predicted amount of precipitation can be 0mm.

Short thunderstorms or showers are relatively difficult to predict, as they usually occur very quickly and it is difficult to assess exactly where they will pass. This also applies to precipitation that is delayed or shifted (due to the terrain). These are mostly narrow shower bands.

Another cause of contradictions can be the rounding off of small quantities. This occurs, for example, with high fog, where the resulting precipitation is often less than 0.4mm. Then this is rounded down to 0mm. In reality, however, these small amounts can still fall to the ground.

The probability of precipitation and the sun can also be predicted at the same time. This can be due to the fact that the precipitation is a brief thunderstorm, in which one may not yet know for sure exactly where it will pass, what amount of precipitation it will rain or when it will occur exactly.


The rainSPOT offers a regional overview of the precipitation. It shows the precipitation for the area around the selected location for the time intervals before the specified time. The probability of precipitation relates to the entire area of ​​the rainSPOT, whereas the amount of precipitation is specified precisely for the selected location. It can happen that no precipitation is predicted at the selected location (predicted amount of precipitation = 0mm), but the probability of precipitation is high. This is due to the fact that precipitation is to be expected in the immediate vicinity. This can also be seen from the coloring of the rainSPOT.


The representation of the weather pictograms always applies to exactly one day, i.e. for 24 hours. If precipitation is forecast for one day, the meteogram shows a rain cloud. The appearance of the rain symbols is independent of the probability of precipitation. An indication of the predicted amount is the number of raindrops under the cloud. The different cloud representations are calculated from the degree of coverage and the different cloud layers.

If precipitation is predicted at the date line (at night), it is possible that the pictogram assumes sun for this day because the nighttime precipitation is less relevant and because it could "migrate" into the next day.

The date line can also make earlier forecasts or pictograms look wrong: For example, if rain is forecast, the pictograms have a precipitation symbol. If this precipitation shifts backwards from, for example, 11 p.m. by a few hours, then it does not fall until the next day. This small shift causes an enormous change in the daily pictograms: on the first day the precipitation symbol disappears, while on the following day precipitation is displayed, provided that there is no other precipitation on the respective days. This huge change may at first glance indicate a false forecast, but it has only shifted time by just a few hours while all other factors have remained the same. But it is still the "same" precipitation.


Precipitation forecasts and actual amounts of precipitation are among the most important in meteorology. By means of short-term verification, the deviation of weather forecasts from real weather can be shown.
However, it is very difficult to reliably get the actual amount of rain. Incorrect forecasts would be worse than none at all, which is why Meteoblue has so far not carried out such a short-term verification of precipitation for the following reasons:

  1. Precipitation amounts are only measured by a fraction of the weather stations.
  2. The type of measurement for precipitation amounts differs greatly: some stations measure hourly (e.g. at Meteo-Switzerland), some every 3 hours, some twice a day (morning and evening), some only once a day. Hourly comparisons are therefore practically impossible. Making a uniform daily amount from the different time intervals does not work reliably either.
  3. Getting reliable measurements of precipitation amounts is even more difficult: errors occasionally occur at many stations. Conclusion: measurements without tests are worse than predictions.
  4. Failures can hardly be noticed with amounts of precipitation: zero can either be no precipitation or a failure of the weather station.
  5. We offer our services worldwide: Reliable measurements in a network of stations (such as in Switzerland or Germany) still do not allow us to provide reliable information on the amount of precipitation, e.g. in Italy or Austria.

A verification of the amount of precipitation therefore requires a reliable measuring network, a certain area coverage and a thorough quality check, which is time-consuming and therefore not able to be updated automatically up to now.
However, we have the topic on the list and will be introducing such comparisons on a monthly or annual basis in the future. As a first step, our history + service will be supplemented with satellite data this year, where you can then make such comparisons across the board and download the data for comparisons with your own measurement data.