Quantification of climate change impacts on catchment water balance

Abstract

Water balance modelling at the catchment scale is valuable for regional planning, design and management of water resources systems as well as for risk assessment and disaster management. This is due to the fact that water-related disasters such as floods and droughts have become more frequent and destructive as a result of climate change which has been observed worldwide over past decades. As part of climate change, changes in precipitation and evapotranspiration are likely to have dramatic impacts on catchment water resources. This results in altering the catchment streamflow and runoff volume. Literature indicates that in practice current water balance models consider the effect of precipitation change rather than evapotranspiration, which is usually assumed to be time-invariant. On the one hand, precipitation estimation in a changing climate is considerably driven by extreme events, particularly by their magnitude and occurrence. The frequency analysis of extreme occurrence based on partial duration series (PDS) may outperform that based on annual maximum series (AMS), however this has not yet been examined with future data. On the other hand, the impacts of climate change on evapotranspiration at the catchment scale have not yet been finalized in terms of methods and data sources used to estimate evapotranspiration. This research thesis develops guidelines on future precipitation projections based on extreme events using frequency analysis of partial duration series. This is tested for cases of point precipitation at individual stations and areal precipitation over the North Island of New Zealand. The testing period ranges between 1945 and 2010. Statistically downscaled daily precipitation from CGCM3.1/T47 and GCM HadCM3 models with spatial resolution of 3.750 x 3.750 and 2.50 x 2.50 respectively, were compared to that directly obtained from RCM HadCM3 at 0.050 x 0.050 spatial resolution for the 1961-2090 period. Moreover, the variation of evapotranspiration across the Waikato catchment and its three forest and grass sub-catchments is accordingly examined for the first time using the integration of the FAO- 56 method coupled with very high spatial resolution RCM HadCM3 data. As a result, the combined effects of changing precipitation and evapotranspiration on the future runoff and volume of the three selected sub-catchment are projected. The results of this research indicate that, in general, precipitation and evapotranspiration have been changing from present to the predicted future and this change has a dramatic impact on catchment runoff. Thus the prediction is that daily precipitation will increase by 1.17% and 2.095% for the 2011-40 and 2041-70 periods, respectively, and the predicted annual precipitation will increase by about 0.89% until the end of 21st century. Likewise, the increase in daily and annual evapotranspiration is about 0.4% to 1.2% per 30-years. Water losses due to evapotranspiration are higher from grassed surface than that from forested surface. As a consequence, mean annual runoff is projected to decrease by 27.8% or to increase by 7.3% per 30-years from 2001 to 2090 depending on the sub-catchment, at the highest rate. In addition to above findings, this research also provides a guideline on the projection of future precipitation based on extreme events using frequency analysis of partial duration series. This could be useful for studying the variability of precipitation with details on how to cope with its complexity. This PhD research attempts to develop a deeper understanding of hydrological response subject to changing climate, as well as adding some case studies of a quantitative nature unlike most of the previous studies.