Long-term fluctuations in river flow conditions linked to the Interdecadal Pacific Oscillation

By Waiology 17/11/2011

By Ross Woods

Within the Waterscape research programme, we’re doing some case studies on the potential effects of climate change on water resources, in water-limited parts of New Zealand, such as Canterbury and Hawkes Bay. In effect, we’re asking how these water resources might look in the future. In another post, I’ll look into the results of a couple of the climate change studies we’ve done that relate to water resources in Canterbury, as part of wider team efforts.

Before thinking about climate change, it’s a good idea to understand the variations in water resources that are happening already. The weather is different from one year to the next, and one decade to the next, so water resources also vary on these various time scales. Understanding those variations will let us put any potential future changes in the context of the changes that water managers already deal with.

Here I’ll look specifically at decadal-scale variability in river flows. This is a peek at some new work that isn’t published yet — I’ll be presenting a fuller story at the Hydrological Society’s Symposium in Wellington in December. Understanding decadal variability in streamflow can be critical for the design of water infrastructure for hydropower, irrigation and water supply. Without this understanding, it is difficult to use river flow data from the past as a guide to the future.

IPO? The Interdecadal Pacific Oscillation (IPO) is one of the causes of decadal-scale variability in New Zealand streamflow statistics, for some parts of New Zealand. The IPO is a cyclical change in the Pacific ocean-atmosphere system. A characteristic circulation pattern predominates for a 20-30 year period, and then the system changes to having a different characteristic circulation pattern. These patterns are known as phases of the IPO; the IPO was in a negative phase from 1945-77 and in a positive phase from 1978-99. During the positive phase, El Nino events and westerly winds are more frequent than usual, and rainfall in the west and south of the South Island is higher than usual. The opposite applies during the negative phase.

Does the IPO affect river flows? McKerchar and Henderson (2003) (PDF) showed that there was a significant difference between streamflow statistics for the periods 1945-77 and 1978-99, especially in the west and south of the South Island. For example, the mean flow in the Clutha River at Balclutha over 1978-99 was 14% larger than during 1945-77.

So why look at it again? In 2000, the IPO changed back to the negative phase (i.e., like 1945-78), though the magnitude of the recent IPO is closer to zero (neutral). So now I’m asking, in the ten years since 2000, have the river flow statistics changed back to the values they had for the negative phase (1945-77)?

Analysis. I looked at long river flow records for 35 sites across New Zealand. I calculated annual values of the mean flow, maximum flow, and 7-day low flow. The figure below shows an example of the results, for floods, mean flows and low flows in the Buller River at Te Kuha, on the West Coast of the South Island. The mean annual flood over the period 2000-09 was 17% lower than the corresponding value for 1978-99. Mean flows and mean annual low flows were also lower in 2000-09 than in 1978-99, by about 10%.

Results for 15 other rivers in the South Island also showed this general result: that all of these three flow statistics were lower in 2000-09 than they had been in 1978-99. Not all the differences were statistically significant, but in almost every case the flows for 2000-09 were lower than for 1978-99. Results for North Island rivers were more mixed, and didn’t display a consistent pattern of change. I’ll need to investigate more to understand the reasons behind both results.

Results. Flow data from 2000 onwards in the South Island support the notion that flows in those rivers are lower during the negative phase of the IPO. The data suggest that the post-2000 reduction in flow has been of the order 10%. We don’t know how long the IPO will remain negative for, but previous IPO phases have lasted 20-30 years, so the current negative phase may last another 10-20 years. Similarly, we don’t know whether the observed correlation between flow and IPO will continue.

Implications. If we look at this from a water resource manager’s point of view, it’s probably a good idea to treat South Island flow data from the period 1978-99 as being slightly higher than the long-term average. So when making forecasts of future water resources for planning purposes, it’s sensible to make allowance for the possibility that flows for the next 10-20 years could be lower than the long-term average. Towards the end of that time horizon (i.e. 2020-2030), we must begin to consider the potential effects of climate change, which is expected to increase precipitation in the South and West of the South Island. But that’s another story …

0 Responses to “Long-term fluctuations in river flow conditions linked to the Interdecadal Pacific Oscillation”