Stochastic modeling of interannual variation of hydrologic variables

David Dralle, Nathaniel Karst, Marc Mueller, Giulia Vico, Sally E. Thompson

Research output: Contribution to journalArticlepeer-review

9 Citations (Scopus)

Abstract

Quantifying the interannual variability of hydrologic variables (such as annual flow volumes, and solute or sediment loads) is a central challenge in hydrologic modeling. Annual or seasonal hydrologic variables are themselves the integral of instantaneous variations and can be well approximated as an aggregate sum of the daily variable. Process-based, probabilistic techniques are available to describe the stochastic structure of daily flow, yet estimating interannual variations in the corresponding aggregated variable requires consideration of the autocorrelation structure of the flow time series. Here we present a method based on a probabilistic streamflow description to obtain the interannual variability of flow-derived variables. The results provide insight into the mechanistic genesis of interannual variability of hydrologic processes. Such clarification can assist in the characterization of ecosystem risk and uncertainty in water resources management. We demonstrate two applications, one quantifying seasonal flow variability and the other quantifying net suspended sediment export.

Plain Language Summary We present a method to predict interannual/interseasonal streamflow variation. Predicting such variation is a long-standing target of hydrological modeling, yet remains challenging to achieve using existing methods, which are generally derived from purely meteorological data. Here we develop an alternative approach that accounts for the random occurrence of rainfall and the process by which rainfall is transformed (via infiltration into the soil and eventual drainage through a shallow groundwater system) into stream discharge. By summing over these variations, the method addresses a major source of interannual variability, arising from the stochasticity of precipitation. The proposed technique can also be generalized to predict interannual variability of physical processes that depend on flow (e.g., suspended sediment export), meaning that the method has broad applicability across the geosciences.

Original languageEnglish
Pages (from-to)7285-7294
Number of pages10
JournalGeophysical Research Letters
Volume44
Issue number14
DOIs
Publication statusPublished - 28 Jul 2017
Externally publishedYes

Fingerprint

Dive into the research topics of 'Stochastic modeling of interannual variation of hydrologic variables'. Together they form a unique fingerprint.

Cite this