Writing
River Discharge and Water Level Are Not the Same Measurement
People living near a river usually want to know how high the water may rise. A model may instead provide discharge. Flow and water level are related, but one cannot be read directly from the other without local river geometry and conditions.
People living near a river usually want to know how high the water may rise. A model may instead provide discharge. Flow and water level are related, but one cannot be read directly from the other without local river geometry and conditions.
Discharge and stage even use different units. If a dashboard labels both vaguely as river condition, the distinction disappears. The axis and label should state which quantity is actually being shown.
When a modeled river grid point is displayed beside a gauge observation, the locations matter too. Two points that look nearby on a map may not represent the same part of the river system. Without local validation, a model value should not be presented as a guaranteed water level in front of someone's home.
I want river data products to show the distance between a number and the decision people may make from it. A change in discharge can be useful warning information, but it is not a complete local risk measurement. The unknown relationships need to stay visible.
Source: official reference.