Natural Buffer Placement and Downstream Flood Mitigation in Rural Hawkes Bay, New Zealand
Small scale field studies from around the world have shown that agricultural land management has a significant effect on the timing and magnitude of flood peaks. One land management technique called ‘soft’ engineering utilises strategically planted trees, wetlands, and other natural buffers to temporarily store flood water in upland catchments. This helps mitigate lowland flooding by delaying the release of water into the river system which dampens the peaky response and therefore reduces the pressure on urban areas downstream. With these issues in mind, this MSc thesis examines the landscape benefits arising from both existing and optimally located natural buffers within the Hawkes Bay region of New Zealand, quantifying their capacity to mitigate flooding under varying soil and climatic conditions through; a) Collating existing data and knowledge; b) Collecting further targeted data on buffer impacts; and c) Using this data to inform and apply a flood mitigation model to examine options for buffer placement and simulate flow response times under different land management scenarios. The ability of any model to make practical predictions is largely dependent on the quality of data input. This research established that the nationally available 25m Digital Elevation Models (DEMs) are not suitable for detailed hydrological modelling at the farm scale. A 10m DEM was the coarsest resolution considered appropriate. In addition, the nationally available soil information while generally appropriate benefited from moderate “ground truthing” to better represent the soils “true” hydraulic properties. Further targeted data relating to the influence of trees on soil infiltration and storage capacity was collected. Measurements of hydraulic conductivity found that soil under individual populous spp. trees and a Cupresses macrocarpa shelterbelt were 3.1 and 5.5 times as conductive respectively as soil under pasture at 10m from the trees. The soil was also less compacted near the trees when the livestock were excluded. This improved the structure and thus water storage capacity of soil. These results informed the buffer assumptions when simulating rainfall-runoff under the different land management scenarios. The modelling results suggest that the capacity of natural buffers to reduce quickflow is strongly influenced by soil antecedent conditions. Under very wet soil conditions the buffers had little extra capacity to store water when subjected to large rainfall events. In drier soil conditions large rainfall events were absorbed by the buffers with considerable reductions in quickflow. This suggests that buffers occupying a relatively small amount of land but sited in areas of high flow accumulation could prove very effective at mitigating intense rainfall, especially in drier summer months e.g. sub-tropical storms. Although the results from the modelling are speculative, the outcome is never the less encouraging. Results from both the model simulations and field measurements of hydraulic conductivity suggest that strategically placed ponds and small scale planting can be used to improve the infiltration and water storage capacity of extensive areas of grazed pasture. This will likely reduce runoff and erosion rates and thereby improve stream water quality and farm productivity at both the farm and wider catchment scale. Considering that flooding is the most frequent and costly natural hazard worldwide, natural buffers with their low maintenance costs and recognized ecosystem co-benefits could offer a cost effective and sustainable solution as part of future flood management planning.