Research Article 2014 WRR Burdekin sediment budget

Water Resources Research 10.1002/2013WR014386 gauged subcatchment sites upstream of the BFD (Upper Burdekin, Cape, Belyando, and Suttor Rivers) were used to determine individual subcatchment discharge and sediment load contributions into the dam for each water year. The ungauged Lower Burdekin subcatchment annual contributions to end-of-river export were then calculated by subtracting the BFD overﬂow from the Burdekin River (Inkerman) site. Contribu- tions from the Bowen River (Myuna) site are represented within the Lower Burdekin contribution. Measure- ments of uncertainty for each of the sediment loads are represented as 80% conﬁdence intervals in parenthesis after each load. Annual dam trapping estimates calculated in Lewis et al. [2013] have also been included within the sediment load budgets. LRE measured uncertainties in annual dam trapping estimates are also reported in parentheses, as 80% conﬁdence intervals. Long-term mean annual discharge based on all available recorded ﬂow years at each gauge, and a 5 year mean sediment load for each site calculated over the study period are also provided. 3.4.1. Clay, Fine Silt, and Coarse Sediment Load Budget As sediment particle-size data were available over the ﬁrst four water years (2005–2009), an additional sedi- ment load budget was constructed using 4 year averaged sediment load contributions from each of the seven gauged sites including the proportions of clay, ﬁne silt, and coarse sediment. This 4 year averaged budget is not summative and does not represent a complete mass balance from subcatchment source to export. However, this 4 year period covers a range of rainfall and hydrological regimes, with particle-size class contributions from each site relatively similar from year to year, particularly the ratio of the clay/ﬁne silt component to the coarse sediment fraction (data not shown). Therefore, we contend these data are rep- resentative of longer term sediment particle-size trends within this catchment. Clay, ﬁne silt, and coarse sediment loads were calculated for each of these sites by the following process: (1) linear interpolation was used to calculate daily particle-size distribution for days lacking sample data pro- vided data existed prior to and following each interpolated day; (2) the daily suspended sediment load cal- culated by the LRE tool was multiplied by the corresponding particle-size distribution data for that day, and then each day was summed for each water year (2005/2006 to 2008/2009); (3) these size-fractioned load- ings were then scaled-up to represent each full water year using the total annual suspended sediment load for any ‘‘ﬂow/load’’ period outside of the sample collection dates and (4) the clay, ﬁne silt, and coarse sedi- ment load fractions were then calculated for each site, as a sediment load weighted mean of the four water years. The numbers of available particle-size samples for each site are displayed in Table 1. The Bowen River (Myuna) site had a number of days where multiple samples were collected during a 24 h period. In this case, particle-size distribution data for these samples were averaged for that day. 3.4.2. Minor Tributary Volunteer Network Sites Available TSS data for each of the ungauged minor tributary volunteer network sites were averaged over all discrete ﬂood events and water years where water samples were collected to determine a mean TSS con- centration per site. The number of wet seasons monitored for each site varied from each location depend- ing on the occurrence of ﬂood events in any given wet season (i.e., wetter versus drier years) and the availability of the landholder to collect samples during such events (see supporting information Table A1). Load-based mean annual concentrations (MAC) for each of the end-of-subcatchment gauged sites from 2005 to 2010 were also calculated for context with the ungauged minor tributary volunteer network sites located within these subcatchments. A Burdekin-wide mean TSS concentration was also calculated using compiled TSS data from all Burdekin minor tributary and gauged subcatchment locations. 3.5. Sources of Error Although the sampling techniques applied capture the clay and ﬁne silt sediment fractions of interest in this study, the collection of water samples at the surface may miss the sand fraction transported as sus- pended, bed and saltation load, resulting in underestimates of this fraction. In an attempt to quantify uncer- tainties in ﬁeld collection and laboratory analysis, experimental cross-section transect samples were collected at each gauged site (e.g., triplicate water samples collected at the left bank, center, and right bank of each stream channel). These data conﬁrmed that the surface of each river was laterally well-mixed in rela- tion to TSS concentrations, providing conﬁdence in the surface grab sampling approach and the laboratory methodology; on average, each individual set of triplicate TSS samples were within 10% (RSD). Further sta- tistical analysis of the data show the variation in TSS concentrations across the stream proﬁles was not sig- niﬁcant (see supporting information section). This variability in TSS concentration measured through our BAINBRIDGE ET AL. C 2014. American Geophysical Union. All Rights Reserved. V 9074

Research Article 2014 WRR Burdekin sediment budget | Page 8