Drought decision points or triggers are the threshold values of an appropriate indicator of drought that can potentially determine when management actions should begin and end. Ideally they should specify the value, the time period, the spatial scale, the drought level, and whether conditions are improving or worsening.

There are a number of options to determine decision points that define areas affected by drought, and support decisions by governments, industries and producers. The most salient, credible and legitimate appear to be those that use thresholds based on set values, percentiles or moving windows for these indicators: • Rainfall for meteorological drought • Water storage or soil water content for hydrological drought • Production for agricultural drought • Measured farm financial for socio-economic performance

Rainfall is a core component of many definitions of drought, is adaptable to agro-ecological regions, cost-efficient, and easy to communicate and administer. However, rainfall alone does not translate well to estimates of agricultural production. Consideration of other meteorological indicators such as temperature, wind-run and relative humidity should be made in the development of a drought index.

Agricultural and hydrological indicators closely link to the impacts on agricultural and environmental outputs. They try to integrate several climatic drivers with management factors to indicate what should be a ‘normal' output.

Socio-economic indicators closely link to the desired outcome from managing in a stressed environment. However, there are many other drivers that can modify this final effect or outcome.

Decision points for drought need to also consider regional differences in climate, soil, land use and current weather and future climate conditions. Climate change is likely to complicate the use of many drought decision points and decision making frameworks because the potential impacts of future temperature and precipitation patterns are likely to differ from region to region.

These changed risks will have implications for the threshold levels of drought. Essentially the thresholds may need to be ‘renormalised' to accommodate these changes. Since the amount of climate change is largely in the future any method for accommodating climate change will likely need to use some data from climate models.