Regional variation of recession flow power‐law exponent

Abstract

Recession flows of a basin provide valuable information about its storage–discharge relationship as during recession periods discharge occurs due to depletion of storage. Storage–discharge analysis is generally performed by plotting −dQ/dt against Q, where Q is discharge at time t. For most real world catchments, −dQ/dt versus Q show a power‐law relationship of the type: −dQ/dt = kQα. Because the coefficient k varies across recession events significantly, the exponent α needs to be computed separately for individual recession events. The median α can then be considered as the representative α for the basin. The question that arises here is what are the basin characteristics that influence the value of α? Studies based on a small number of basins (up to 50 basins) reveal that α has good relationship with several basin characteristics. However, whether such a relationship is universal remains an important question, because a universal relationship would allow prediction of the value of α for any ungauged basin. To test this hypothesis, here, we study data collected from a relatively large number of basins (358 basins) in USA and examine the influence of 35 different physio‐climatic characteristics on α. We divide the basins into 2 groups based on their longitudes and test the relationship between α and basin characteristics separately for the two groups. The results indicate that α is not identically influenced by different basin characteristics for the two datasets. This may suggest that the power‐law exponent α of a region is determined by the way local physio‐climatic forces have shaped the landscape.

Estimation of streamflow recession parameters: New insights from an analytic streamflow distribution model

“…The RAMs discussed in this paper have been developed in the past to obtain insights in the water storage‐release behaviour of catchments and remain a key tool to understand low flows and water storage at catchment scale (e.g., Floriancic et al, ; Staudinger et al, ). Recent studies try to understand how the storage‐release relationship is related to geomorphological and other physiographic catchment features (Biswal & Marani, ; Mutzner et al, ; Patnaik, Biswal, Kumar, & Sivakumar, ), with the ultimate goal to predict catchment‐scale recession behaviour from observable catchment characteristics. But the use of traditional RAMs for hydrological process analysis and comparative hydrology is inherently limited by the nontrivial interactions between methodological choices and resulting recession descriptors (estimated parameter values).…”

An Empirical Reevaluation of Streamflow Recession Analysis at the Continental Scale

“…We assessed each for its predictive power in describing both Rmed b values and the likelihood of convexity using a linear model, a logarithmic model, and an exponential model. Confirming Patnaik et al's (2018) results from a smaller data set, b values were poorly predicted by each individual catchment attribute. The single best predictor of b was the log of the fraction of the watershed comprised of soil group HGB (a deep permeable soil; r 2 = 0.147), with other properties of soil, precipitation (PPT), and topography being the next best predictors with r 2 values generally below 0.06.…”

Spatial and Temporal Patterns in Baseflow Recession in the Continental United States

“…Promisingly, the typical nonlinearity of individual events strongly predicts the nonlinearity of data point clouds (as well as multiple baseflow indices), indicating that regional-scale physical mechanisms likely underlie recession nonlinearity (Tashie et al, 2020). However, these physical mechanisms have proven difficult to identify (Patnaik et al, 2018).…”