An appraisal of the use of Ground Penetrating Radar as a tool to monitor the vulnerability of stopbanks in the Bay of Plenty region, New Zealand

Abstract

Following the 1908 Drainage Act, major channel realignment transformed low-lying swamplands along the Rangitaiki Plains in the Bay of Plenty region of New Zealand into four channels that now make up the lower courses of the Rangitāiki, Whakatāne, Kaituna, and Waioka-Otara Rivers. Subsequent concerns for flood protection prompted stopbank construction to support the development of agricultural lands and various towns including Whakatāne. Major disturbance events such as earthquake activity and large storms (floods) have impacted upon stopbank stability in this region. In 2017, a 44 year old concrete flood wall along the margins of the Rangitāiki River at Edgecumbe was breached on the back end of Cyclone Debbie, damaging local infrastructure. Since then, the Bay of Plenty Regional Council have critically examined their approach to stopbank design and construction in efforts to minimize prospective stopbank failures in response to rainfall and storm surge events up to a magnitude of a 100 year return period. This builds upon previous experiences following a stopbank breach in 2004. In this complex landscape, stopbanks are prone to many failure mechanism from engineering design, slope instability, to the influence of geomorphology, geology and seismology. This thesis presents an initial exploration of the use of a passive (non-invasive) technique, Ground Penetrating Radar (GPR), as a tool to supplement existing adaptive management approaches to assess stopbank design and stability. While the overall design of stopbanks has not changed in recent decades, small adjustments to design elements such as pore pressure relief wells have greatly influenced stopbank stability. Alongside recurrent soil sampling, passive techniques such as GPR can support subsurface analyses of prospective failure mechanisms. Three GPR surveys along stopbanks were completed in this thesis. The survey at College Road showed how the subsurface boundary between highly and lowly engineered soils may be susceptible to stopbank failure. The second survey on Rangitāiki Floodway identified the same failure mechanism along both the right and left bank of the stopbank, despite the right bank being a raised embankment stopbank. The third survey on the Whakatāne River surveyed pre-existing failures, with GPR detecting additional failure anomalies unknown prior to the field investigation. Results of RocSlide analysis showed the stopbank slope along the Rangitāiki River is sensitive to pore pressure and seismic activity under extreme conditions modelled for a 100 year return period. Based on these findings, and experiences gained in the conduct of this research, it is suggested that passive techniques can support proactive appraisal of stopbank stability. This could be incorporated within regulations. Importantly, enhanced record keeping on construction history and associated reports of monitoring programmes would support regional-scale analysis of stopbank stability.