Meteorological Drought Projections for New Zealand Using CMIP5 Data
Droughts are one of the most damaging natural hazards, and anthropogenic climate change will continue to impact drought sensitive sectors such as primary production, industrial and consumptive water users. Drought monitoring and early warnings are essential for the development of mitigating strategies. The overall aim of this thesis is to develop a methodology to project droughts and its severity in the future through a multi-scenario and multi-model approach using the latest Coupled Model Intercomaprison Project Phase5 (CMIP5) models. All sixteen regions of New Zealand are included in the analysis. To achieve the above objective, the analysis was initially carried out to select the most applicable meteorological drought index – Standardised Precipitation Index (SPI) for New Zealand. Temporal changes in historic rainfall variability and the trend of SPI were investigated using non-parametric trend techniques to detect wet and dry periods across the regions of New Zealand. The first part of the analysis was carried out to determine annual rainfall trends using Mann-Kendall (MK) and Sen’s slope tests for the sixteen regions with long historical records (109 years) of the data set. For SPI trend analysis, it was observed that, results obtained showing significant trends; direction of SPI trends were similar to annual precipitation (downward and upward trends). In addition, the rate of occurrence of drought events were examined in the temporal trends. The fact that all regions showed positive slopes indicated that the intervals between events were becoming longer and the frequency of events was temporally decreasing. From the SPI trends, it was also observed that some of the regions over New Zealand will face more dry periods leading to increased drought occurrence. Information similar to this would be very important to develop suitable strategies to mitigate the impacts of future droughts. This main objective of this thesis is to assess the drought projections for the regions of New Zealand using General Circulation Models (GCMs) under two emission scenarios – Representative Concentration Pathways (RCP4.5) and RCP8.5 for three future periods (2010-2039, 2040-2069, 2070-2199). Drought severity and spatial extent are analysed for 12-month (SPI12) events. A novel concept centric on improving the GCM data was successfully derived for the regions using an innovative bias correction algorithm. This algorithm removes errors from climate models in comparison with historical observations. The quantile mapping bias correction applied to the GCMs improved the rainfall projections thus reliable SPI values for the drought projections were generated. Drought projections vary substantially depending on the GCM, emission scenario, region, season and definition of drought. Overall, climate change enhances drought conditions across the study region, with marked increases projected for the northern islands under both emission scenarios; reductions in moderate droughts are projected for the regions in the South Island. The interannual variability of precipitation tends to enhance drought conditions caused by mean precipitation changes, or to moderate or reverse their reductions. Greater agreement in the direction of change tends to occur in the northern island regions. Projection ranges tend to increase with time and magnitude of warming. The implications of the large uncertainties include that decision-making should be based on multi-scenario and multi-model results, and with consideration of drought definition. Many parts of New Zealand have experienced their worst droughts on record over the last decade. With the threat of climate change potentially further exacerbating droughts in the years ahead; a clear understanding of the impact of droughts is vital. The information on the probability of occurrence and the anticipated severity of droughts will be helpful for water resource managers, infrastructure planners and government policy-makers with future infrastructure planning and with the design and building of more resilient communities.