Glossary: level, discharge and water data

Height of the water surface at a given station. Depending on the network it is published as a local centimetre reading or as an absolute altitude in metres above sea level. The level is not the same as water depth.

Volume of water passing a measurement cross-section per second. 1 m³/s equals 1,000 litres per second. Discharge is usually derived from the level using a level–discharge relationship established through separate discharge measurements.

A fixed location where an operator records level, discharge, temperature or other quantities. Values apply to that place and time; they cannot simply be transferred to the whole water body.

A measurement comes from a sensor or a station calculation derived from it. A modelled value is produced by a scientific model. A forecast describes a possible future development. Seepegel labels the data type and source.

The central value of a sorted series; for an even number of values we use the average of the two central values. At least half the values are less than or equal to the median, and at least half are greater than or equal. Ties are possible. The median is less sensitive to individual extremes than the arithmetic mean.

Comparison with values for the same calendar day or window across several years. FOEN level comparisons often use 1991–2020; the stated period is what matters. Comparing the past 30 days is not a long-term seasonal reference.

A percentile is a threshold in a sorted distribution, such as its 95 per cent boundary. Percentile rank instead places a particular value within the series. In the 30-day station comparison, lower daily means count fully and equal ones count half. A constant series therefore sits at 50 per cent, not at the top. Neither percentile nor rank is an official hazard level.

A daily mean averages the values of a day. A short peak may be much higher. Seepegel's 30-day station comparison averages available readings per UTC day; missing readings are not filled in. Having a daily mean therefore does not prove that the entire day was measured without gaps.

For a measurement, the timestamp identifies the measurement time; for model and forecast data, the time for which the value is valid is essential. This differs from the time of the model run or page visit. During transmission problems an older value may remain visible, so the time reference is part of every interpretation.

Tributaries bring water into a lake, the outflow leads it out. Precipitation, evaporation, groundwater and possibly regulation add to this. The difference of the visible river stations is therefore only a partial balance, not a complete volume calculation.

An official assessment of a possible danger. It is based on water-body specific thresholds and further situational information. Colours or statistical deviations on a data platform do not replace this warning level.