Summary of OHRFC Case
This quick review of the hydrologic forecasts of stage height from the OHRFC is intended to provide guidance regarding techniques used to verify forecasts. The technique used here had several important steps.
- Look at basins within each response time group to make sure that basin aggregation makes sense.
- Evaluate measures of bias and error and how those vary with forecast scenarios, lead time, and basin response time.
- Examine correlation between observations and forecasts and how that varies with lead time.
- Evaluate forecast skill, in particular, the skill of forecasts with respect to QPF input and lead time.
- Assess whether QPF may be a significant source of forecast error.
The study appears to show that including QPF has a greater positive influence on the stage forecasts than the modifications to the forecasts that forecasters perform within the hydrologic model. Since QPF has the greatest positive impact on stage forecasts, that could mean that QPF errors have significant negative impact on hydrologic forecasts. However, more study would have to be done looking at, among other things, the impact of QPF for different QPF amounts.
Forecast quality as measured by error, bias, and correlation decreases as lead time increases. This decrease in forecast quality occurs more quickly with fast-response basins than with slow-response basins. However, the forecast quality degrades more slowly for forecasts that contain QPF input. In other words, when QPF input is used, the forecast show positive skill as lead time increases compared to forecasts with no QPF input.
As stated above, this study did not attempt to look at the impact of certain QPF amounts. In addition, we cannot conclude that the results would be the same in other regions with different precipitation regimes. But the point here was to demonstrate how verification can be structured to answer questions about hydrologic forecast performance and error sources.