The National Weather Service has undergone a dramatic evolution in the field of hydrology since the Organic Act of 1890. Early forecasting to meet the responsibility of "gauging and reporting on rivers" was based on empirical rules and personal knowledge of river basins. By the late 1930s, empirical rules gave way to use of physical data such as rainfall and snow melt and the development of the hydrograph. During the mid 1940s River Forecast Centers (RFC) were formed to concentrate on the problem of river forecasting and to develop and refine hydrologic forecasts and procedures.

The latter half of the 20th century was characterized by infusion of science and technology to provide improved river forecasts. Rapidly changing technology and evolving forecast procedures posed considerable challenges to the RFC hydrologist. For the past two decades, the Missouri Basin River Forecast Center and North Central River Forecast Center of the National Weather Service have collected river forecast data for the Missouri and Mississippi Rivers, respectively, and calculated forecast verification statistics. This paper will evaluate forecast error over that time period with particular attention to improving or degrading trends, citing possible explanations of those trends.