Daily flow time series patching or extension: a spatial interpolation approach based on flow duration curves

Abstract

An algorithm is described that was initially developed as a simple method for patching and extending observed time series of daily streamflow. It is based on the use of 1-day flow duration curves for each month of the year and on the assumption that flows occurring simultaneously at sites in reasonably close proximity to each other correspond to similar percentage points on their respective duration curves. The algorithm has been incorporated into a model that allows flows at a destination site to be estimated from flows occurring at several source sites. The model has been applied to six groups of catchments within southern Africa and the resulting streamflow simulations compare favourably with those obtained using a semi-distributed, physically-based, daily time step rainfall-runoff model. The current limitations of the approach and its future potential value are discussed.

Developing best practice for infilling daily river flow data

“…The catchment area scaling technique essentially seems too simple to capture the influences affecting the target and even very closely related stations seldom exhibit a linear relationship with catchment area (Hughes and Smakhtin, 1996).…”

“…These may also necessitate the use of localised data and highlight the need to maintain comprehensive user guidance information alongside hydrometric records (Dixon, 2010). Other means of localising datasets are to consider wet and dry epochs separately (Hughes and Smakhtin, 1996) or to group flows according to the month or season that they correspond to, which has been demonstrated to offer significant improvement (Raman et al, 1995). Ongoing work by the authors is exploring such approaches, by applying the same infilling techniques to both full and localised datasets.…”

“…This may allow easier identification of instances when particular techniques surpass others at certain flow magnitudes and could also isolate favourable technique combinations. A number of studies has previously advocated that a single technique is unlikely to be optimal for all occasions of missing data (for example: Hughes and Smakhtin, 1996;Gyau-Boakye and Schultz, 1994).…”

“…What constitutes a good donor is arguably a research topic in its own right, but common considerations include proximity and similarity (in terms of responsiveness and catchment physiography) to the target (Rees, 2008). Where available, multiple donors can enhance the likelihood of capturing the many influences affecting a target, but a single donor could be adequate if located very close to the target or on a major upstream tributary (Hughes and Smakhtin, 1996).…”

Estimation of Missing Streamflow Records in the Euphrates Basin using Flow Duration Curves and Regression Models

“…Ayrıca bir havzadaki mevcut Akarsu Gözlem İstasyonu (AGİ) verilerindeki eksikliklerin giderilebilmesi için debi süreklilik çizgilerinden de (DSÇ) faydalanılabilir (ör. Hughes and Smakhtin, 1996;Mohamoud, 2008). Kullanılan yöntemlerden bağımsız olarak öncelikle istasyonlar arasında yüksek bir korelasyonun varlığına bakılmalıdır (Burgess and Webster, 1980).…”