Spatial Hydrographs of River Flow and their Analysis for Peak Event
Detection in the Context of Satellite Sampling
Abstract
The study of river dynamics has long relied on the analysis of
traditional in situ hydrographs. This graphical representation of
temporal variability at a given location is so ubiquitous that the mere
term “hydrograph” is widely recognized as a time series. While such a
“temporal hydrograph” is well suited for in situ data analysis, it
fails to represent hydrologic variability across space at a given time;
a perspective that characterizes satellite-based hydrologic
observations. Here we argue that the concept of “spatial hydrograph”
should be the focus of its own dedicated scrutiny. We build “space
series” of river discharge and present their analysis in the context of
peak flow event detection. We propose the use of peak event spatial
coverage, referred to as “length”, as an analog to event duration. Our
analysis is performed in the Mississippi basin using a dense in situ
network. We reveal that peak flow events range in length from around 75
to 1,800 km with a median (mean) value of 330 (520) km along the basin’s
largest rivers. Our analysis also suggests that spatial sampling needs
to be a factor of 4 (2) finer in resolution than peak flow lengths to
detect 81±13% (70±20%) of events and to estimate their length within
84±3% (67±12%) median accuracy. We evaluate the connection between
temporal and spatial scales of peak flows and show that events with
longer durations also affect larger extents. We finally discuss the
implications for the design of satellite missions concerned with
capturing floods across space.