Precipitation distribution in the Lake Pukaki Catchment, New Zealand
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
Mountain precipitation, as a major component of global ecology and culture, requires diverse observation-based distribution studies to improve process characterisation and so enhance environmental management and understanding. Analysis of data from an array of precipitation gauges within the nationally important, and internationally extreme, mountainous Lake Pukaki catchment in New Zealand has been undertaken in an effort to provide such a study, while also improving local hydrological understanding. An objective observation based undercatch-corrected 1971-2000 average annual precipitation distribution has been prepared for the mountainous Lake Pukaki catchment, New Zealand. Precipitation records from 58 gauges at 51 sites, augmented with 10 new gauges, were used in preparation of the distribution. The assessed undercatch correction of 17 % across the catchment indicates that mountain hydrological investigations in New Zealand that use precipitation data and yet do not consider undercatch will be in considerable error. The average annual distribution confirms the existence of high precipitation magnitudes and horizontal gradients in the catchment in comparison with other mountain regions around the world. The high magnitude is unusual when its position in the lee of the principal orographic divide is considered indicating rare precipitation distribution processes occur in the region. Consideration of river flows, glacial change and evaporation led to a confirmation of the gauge derived average catchment precipitation. Precipitation to wind direction relationships identified the predominant westerly wind to be the primary precipitation generating direction with large magnitude events biased towards the northerly direction. All directions from the eastern side of the mountain divide had the lowest frequency and daily precipitation magnitude. Derivation of wind-classed precipitation distributions identified a distinctive south east to north west precipitation gradient for all wind directions, most severe for the north west direction and least severe for the easterly direction. Precipitation extent was greatest for the northerly direction and least for the south south westerly. The wind-classed distributions enable the estimation of daily precipitation likelihood and magnitude at any location in the catchment based on knowledge of the synoptic wind flow direction and precipitation at just one reference site. Improved river flow and lake inflow estimates resulted from the use of wind classed daily precipitation estimates validating the quality of the wind classed distributions. From 1939 to 2000 there has been no statistically significant trend in precipitation magnitudes, frequencies, or extremes in the catchment. At Aoraki/Mt Cook village, in the upper catchment, there have been significant increases in magnitude, frequency and extremes associated with the phase change of the Interdecadal Pacific Oscillation (IPO) in 1978. This change can be explained by the increase in strength of westerly winds for the different IPO phases but not by a change in frequency of different wind directions. In the lower catchment the IPO relationship is of an opposite sense to that observed in the upper catchment, indicating that the areas operate under two different climate regimes with different precipitation controls. The significant relationship to the IPO phase indicates that it is more important than climate warming in terms of future precipitation distribution in the Lake Pukaki catchment, and by extension the Southern Alps. The distributions prepared provide a valuable tool for operational and academic hydrological applications in the region. In addition, they provide a valuable characterisation of the precipitation in a Southern Hemisphere mid-latitude lee to predominant westerlies glacierized mountain catchment. From this standpoint they highlight the contrast to Northern Hemisphere mountain precipitation distributions commonly used in model validation studies, thereby providing an extension of locations with which to refine orographic precipitation process understanding.