Volumetric measurement of river bank erosion from sequential historical aerial photography

2017 
Abstract Understanding of the relative contribution of bank erosion to sediment budgets in New Zealand is limited. Few measurements of bank erosion rates exist, and this is a major limitation to the development of a locally calibrated model of bank erosion. The New Zealand sediment budget model, SedNetNZ, predicts bank erosion based on preliminary data, and this study aims to underpin the development of an improved model for bank erosion. Photogrammetric techniques and LiDAR were used to collect data on bank erosion rates for five different river reaches, ranging from 3 to 14 km in length, in the Kaipara Catchment, Northland, New Zealand. Changing river channel planform between the 1950s and 2015 was assessed using four to five well-spaced dates of historical aerial photographs. Changes in planform were combined with bank height, to calculate erosion and accretion volumes which were compared with SedNetNZ modelled estimates. Erosion and accretion is relatively evenly balanced in the study sites. The largest difference in terms of relative proportions of erosion and accretion are found along the Tangowahine River (13.4 km reach length), where 492,000 m 3 of sediment eroded between 1956 and 2015 compared to 364,000 m 3 of accretion. Lateral migration rates (erosion) for the five river reaches range between 0.14 m yr − 1 and 0.21 m yr − 1 and are comparable with those measured by previous assessments in New Zealand. The migration rates in channel widths per year for the three larger rivers (stream order 5–6) range between 0.4% and 0.8% of channel width per year. In contrast, the smaller streams (stream order 3–4) are retreating more rapidly, with width-averaged rates of 1.7% and 3.0%. Current SedNetNZ modelling tends to underestimate the bank height and greatly overestimates the migration rate.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    49
    References
    9
    Citations
    NaN
    KQI
    []