Abstract
In this study, the authors explore simple concepts of persistence in
streamflow forecasting based on the real-time streamflow observations.
The authors use 15-minute streamflow observations from the year 2002 to
2018 at 140 U.S. Geological Survey (USGS) streamflow gauges monitoring
the streams and rivers over the State of Iowa. The spatial scale of the
basins ranges from about 7 km2 to 37,000 km2. Motivated by the need for
evaluating the skill of real-time streamflow forecasting systems, the
authors perform quantitative skill assessment of different persistence
schemes across spatial scales and lead-times. They show that temporal
persistence forecasts skill has strong dependence on basin size and
weaker, but non-negligible, dependence on geometric properties of the
river network of the basin. The authors show that anomaly persistence
forecasting can serve as a good reference for the evaluation of
real-time streamflow forecasts at scales of order 100 km2. Building on
results from this temporal persistence, they extend the streamflow
persistence to space through flow-connected river network. It simply
assumes that streamflow at a station in space will persist to another
station which is flow-connected, and refer to it as pure spatial
persistence forecasts (PSPF). The authors show that skill of PSPF
derived streamflow forecasts is strongly dependent on basin area-ratio
and lead-times, and weakly related to the downstream flow distance
between stations. They show that the skill depicted in terms of
Kling-Gupta efficiency (KGE) > 0.5 can be achieved for
basin area ratio > 0.5 and lead-time up to three days.
Adding complexities of hydrologic routing and rainfall QPF to the PSPF
further improves the skill. The authors discuss the implications of
their findings for improvements of rainfall-runoff models as well as
data assimilation schemes.