Water level - measuring the stage on a gauge

Methods guide · June 20, 2026

There is a certain rhythm in the work of people who watch over a river or lake: the daily, almost reflexive glance at the staff gauge and noting the number. Seemingly a trifle - yet in fact one of the oldest and most important hydrological measurements. It is from such daily readings, kept for years by observers, naturalists and water communities, that the knowledge of how water lives is built: when it rises, when it falls, how it responds to rain.

This measurement has something soothing and deeply sensible about it. It needs no expensive equipment - a fixed gauge and consistency are enough. And because the water level responds to the weather, the readings come into their own only when set against rainfall and the catchment situation. And this is exactly where the simple habit of noting the water level turns into a real understanding of the river.

What we measure

Water level (stage) is the height of the water surface relative to a fixed reference point (the gauge zero), usually given in centimetres. On its own it says “how much water there is now”; combined with a discharge measurement, it lets you build the so-called rating curve and estimate the discharge from the level alone.

What exactly is the “gauge datum”

It is a conventionally adopted reference level, levelled into the national geodetic network and expressed as an elevation in metres above sea level - but the water-level reading itself is given relative to that datum, not to sea level. The datum is deliberately set below the lowest expected stage, so that bed erosion does not push readings into negative values.

An important consequence: water level is NOT river depth. It is the height of the water surface above a conventional point - two different staff gauges on the same river may have completely different datums, so stages from two gauges must never be compared directly, only each against its own history.

Datums are also shifted as bed erosion progresses. Poland offers a worked example: on 1 November 2025 the national hydrological service lowered the datum elevation at the Warszawa-Bulwary and Gusin gauges by 100 cm, because hydrological drought and bed erosion threatened negative readings - the same water surface previously read as 6 cm reads as 106 cm afterwards. That is not a change in how much water the river carries, only a rescaling - but if you compare a series from before and after such a change, you must know that it happened. Whichever country you work in, check the gauge’s datum history before you interpret a long series.

How to measure it - step by step

1. Use a fixed staff gauge set in one, unchanging place and referenced to an established zero.
2. Read the level at a fixed time (classically in the morning), looking at the gauge perpendicularly (avoiding parallax error).
3. Record the value with the date, time and any notes (ice, drifting branches, backwater).
4. Where possible, a sensor (pressure/ultrasonic) records the level continuously - the gauge then remains as a check.

The same procedure covers structures outside the river network: a gauge board at a forest dam or a pond is the whole instrumentation of small-scale retention monitoring.

Read the water level together with the weather

The water level is the catchment’s response to what fell from the sky and what flowed in from upstream. That is why it gains its full meaning when set against:

• rainfall (rain, snowmelt),
• the level and discharge upstream in the river network.

Such data is made publicly available by national hydrological services (e.g. IMGW-PIB in Poland). Comparing your own reading with rainfall immediately explains why the water is rising or falling - and lets you catch unusual situations (the level is rising without rain? something is happening upstream).

The size of the jump alone does not say what caused it - only comparing it with rainfall and with stages in the river network upstream lets you tell these four situations apart.

The most common mistakes

• A shifted or damaged gauge - it changes the reference point, and the series loses continuity.
• Parallax error - reading at an angle over- or underestimates the result.
• A variable time - with rapid level changes, the hour matters.
• Ice and jams - note the conditions, as they can falsify the reading.

How often

Classically daily at a fixed time, and during floods more often. Regularity and a fixed point are everything here - it is from a continuous series that a picture of the water regime emerges.

Sources and further reading

Data on rainfall, levels and discharges of rivers in Poland is run and shared by IMGW-PIB. The topic is covered in hydrology textbooks and in materials on the measurement-and-observation network.

📊 See it live: Reef & Coast Watch - a public dashboard from six NOAA tide gauges (Hawaii, Guam, Florida, US Virgin Islands, Puerto Rico): water temperature and water level fetched automatically, with a station map. No login.

How to make it easier

The water level says the most as a series and in combination with the weather - and manually splicing your own readings with rainfall data is tedious work that is easy to put off “for later”.

LimnoLog does this for you - and here connectors come in:

• you record the gauge reading right in the field (also offline), in a fixed station and session;
• a connector automatically pulls data from public APIs (e.g. rainfall and the hydrological situation from IMGW-PIB) and saves it alongside your measurements - on the same chart and timeline;
• so you see the water level and rainfall together: a rise after rain becomes obvious, and anomalies stand out;
• you can share the results via a public link or export them.

And, as always - the same system covers, alongside the level, discharge, temperature, turbidity or any other indicator. Your measurements and the weather context in one place.

Early users get free access for now and keep their features for good. If you watch over your water - take a look and try it.