Lee JM. 2020.
Landslides triggered by the “Tasman Tempest” rainfall event, March 2017, in southeast Auckland, New Zealand. Lower Hutt (NZ): GNS Science. 20 p. (GNS Science report; 2020/29). doi:10.21420/XWY9-2V45.
Abstract
The Tasman Tempest storm event was generated by a tropical cyclone originating in the Tasman sea, bringing extreme rainfall to the North Island of New Zealand on the 7th to 12th of March 2017. The storm consisted of three high intensity rainfall pulses with the first occurring over the 7th to 8th, the second over 10th to 11th and the final pulse on the 12th of March. The storm caused thousands of landslides, flooding and power outages creating major disruption for people in southeast Auckland, with Hunua, Clevedon, Kawakawa Bay and Waiheke Island impacted the worst. The heaviest rainfall occurred south and east of Auckland over a 24-hour period from the 7th to 8th of March when 266 mm was recorded over 24 hours; an annual return interval that exceeded a 1 in 100-year event (HIRDS). The extensive landslides in the Hunua ranges south of Auckland compromised the water supply collection facilities in this area when silt from landslide deposits entered the reservoirs and overwhelmed the filtration capacities of the system. Landslides initiated by the storm were mapped using Google Earth imagery and high resolution aerial photographs. Landslides were mapped as points across the 882 km2 study area and in greater detail as polygons within a smaller study window (source areas, debris trails and debris deposits) using the Kaikoura landslide inventory mapping method (Massey et al. 2018). Gauge corrected rain radar data supplied by MetService was used to quantify rainfall for the storm and enable rainfall estimates to be made for individual landslides. Landslides were most commonly triggered across slopes which had a slope angle of 20–40°. Landslides were triggered predominantly on slopes that were underlain by greywacke, a reflection of the dominant rock type in the study area. Land cover consisted of native forest, forest plantations, pasture, scrub and other (settlements). The highest frequency of landslides was observed on pasture as expected, however, the highest landslide density was recorded within forest plantations (as mapped in 2012/2013). Significant areas of forest have been logged since LCDB v4.1 mapping was completed in 2012/3 which is not reflected in the ‘Forest’ category. It is probable the landslides occurred in recently logged areas. Maximum rainfall for the 24-hour period was between 260–280 mm, although this rainfall was located over the Hauraki Gulf. The maximum rainfall for a mapped landslide was in the range 200–220 mm. The highest density of landslides broadly coincided with the most rainfall of 180–200 mm. An exception was a small area near the coast (Waitawa Regional Park) where higher rainfall of 200–220 mm did not result in a higher density of landslides. The reason for this has not been investigated but may be due to coastal slopes being more resistant to high intensity rainfall events because they are exposed to coastal processes. (auth)
