After gathering information on river stages and various data distribution systems, the forecaster uses a rainfall-runoff model to assist in plotting the flow and stage for a given forecast point.

Information, such as rain gauge amounts, Doppler radar rainfall amounts, air temperature and sunshine amounts, all come together with a river forecast model and the forecaster's experience to create a river forecast.

River Stages
River stage information is gathered through gauges placed along a river bank and/or on the side of a bridge.

A gauge house is a well dug along the riverbank with an accompanying instrument package that records the water level.

Mounted to the side of a bridge, a wire weight gauge measures the river level by subtracting the distance from the bridge and the water from the distance from the bridge to the riverbed.

The third, and simplest, method is a ruler attached to the river bank or a bridge piling and the river level is read by noting the markings on the gauge.

Rainfall Amounts
Rainfall amounts are gathered through rain gauges located at observing sites, such as airports, remote and automatic sites, and volunteer observers.

Doppler radar is also used, basing the rainfall amount on the intensity of the radar echoes. It calculates the speed and direction of the precipitation areas, allowing forecasters to alert areas ahead of the rain. The information is used in conjunction with rain gauge data for a rainfall estimation.

Other Data Sources
There are various networks by which a river forecaster receives vital information. Through volunteer National Weather Service cooperative observers comes maximum and minimum temperatures, rainfall and snowfall, total snow depth and types of weather that occurred over the past twenty-four hours.

The GOES satellite collects cloud image, river stages, and precipitation reports among other data.

Information from various precipitation gauges is funneled through the Centralized Automatic Data Acquisition System (CADAS) and after processing, the data is disseminated to users. And some areas, such as Oklahoma, have a mesonetwork of hydro-meteorological reporting sites from which the forecaster can receive hourly data.

Computer Models
River forecast models estimate the amount of runoff a rain event will generate. It also computes how the water will move downstream from one forecast point to another and predicts the water flow at a given forecast point through the forecast period.
There are six different rainfall-runoff models the forecaster can use. The model also plots the flow and stage for a given forecast point.