Case Study: Water Resources

Case study description:

The hypothetical "City A" is in western Victoria. It currently has adequate water resource security, however if there is an ongoing drop in runoff of 20% or more, major measures would be needed such as managing demand (e.g. restrictions) or managing supply (e.g. extra infrastructure such as reservoirs, or additional supply sources such as groundwater extraction and desalination). Non-climatic factors such as demographics and industry trends are also important. After considering all aspects, there are three broad policy settings for long-term management that the City sees as practical: "Business as usual", "Monitor and scope" and "Active management and phase-in". These are described in more detail in the table below.

Back of the envelope:

The catchments affecting City A (in southeast Australia) are known to have quite a large “inflation factor” between rainfall and runoff. For this example we assume the round number of 2 (e.g. an ongoing reduction of 10% in rainfall may lead to a reduction of 20% in runoff). A general projection for southern Australia is reduced annual rainfall. For western Victoria, projections (see the Cluster Report for Southern Slopes and/or Summary Data Explorer results) show a reduction of mean annual rainfall by 2090 under a RCP8.5 of -27% to +4% (median -9%), putting the possibility of a 10% reduction in rainfall as possible but not certain. Under a lower emission scenario (RCP4.5) the range projected by models is -15% to +3% (-7%), so reaching the threshold is possible there too.

Policy setting

Detail

Business as usual

No change to current water management

Monitor and scope

Set up monitoring procedures and scope the various management interventions that may be needed in the future

Active management and phase-in

Develop phased water management guidelines, from gentle to strict. Plan and develop finance for greater water infrastructure.

Since it is not obvious or unanimous that the 20% threshold will/will not be exceeded in all cases, more detailed analysis is suggested. Also, the seasonality, timing and periodicity of rainfall is important to runoff (e.g. 100 mm of rainfall in one day leads to more runoff than 5 mm for 20 days in a row), so projected change to the character of rainfall may influence the runoff projection markedly. Along with rainfall, the other aspect of runoff is evaporation and this may also be important.

All these factors indicate that more sophisticated analysis of runoff using specialised models is warranted, and this requires a more complex process of producing relevant datasets.

Risks inlcude:

  1. Being unprepared for potential change
  2. Financial costs and loss of good-will from introducing measures before they are needed

Set time-frame and risk:

The City primarily manages water on time-frames of up to a year, but also has long-term planning horizons of five years, 10 years, 50 years and 100 years. Climate projections are used to inform the 50 and 100 year time-frame planning. For these time-frames, managers have decided to manage using a ‘most likely climate scenario’ focus, but applying the precautionary principle where and when possible. This means they will mainly consider the "maximum consensus" projections scenario (see Introduction to Climate Futures ) but will also consider a "worst-case" climate scenario. Therefore, they want to consider the range of projections for an intermediate emissions future (RCP4.5) and a high emissions future (RCP8.5), but take particular note of the driest projection for high emissions.

Key context for choosing datasets

  1. Sub-setting global climate models and regional climate projections is important to assess – consider the low probability, high impact outcomes as well as the consensus view
  2. A low tolerance for bias – runoff models are calibrated to local observations
  3. All factors of the signal are potentially important (mean, variability etc.)
  4. Downscaling potentially adds value for this question – work is on a regional scale, we are most interested in rainfall, region is affected by topography and there are other regional drivers
  5. Mean scaling is probably inadequate, and complex scaling or bias correction seems warranted (but alternative, simpler approaches should be also assessed)

Choice of datasets:

Select a representative set of global climate models and downscaling (using the Climate Futures approach), use either complex scaling or bias-correction of internally-consistent series of rainfall and potential evapotranspiration to produce the input datasets for appropriate rainfall-runoff models.

Page updated: 1st August 2016