Extremes, climate modes and reanalysis-based approaches to agricultural resilience

Research objective

Investigate new longer-term weather data from the 20th Century Reanalysis Project (20CRP) as a basis for estimating extreme weather conditions that affect many agricultural operations in Australia.

Project duration

2008 – 2011

About the 20CRP

Long-term weather observations are often of variable quality or non-existent for many Australian locations.

This hinders evaluation of weather and climate risk management using techniques such as seasonal forecasting and insurance pricing.

The 20CRP is reconstructing a 6-hourly, 4-dimensional global atmospheric database spanning 1871 to 2010 to place current atmospheric circulation patterns into a historical perspective.

In 2009, the 20CRP database for 1871–2008 became available.

The database is based on a collated history of surface pressure, ship-based sea surface temperature and sea-ice measurements.

These measurements are inputs to a modern weather forecast model at fine time resolution (better than 6 hours) and reasonable spatial scales (2 degrees latitude and longitude) running on a supercomputer.

The model is rerun 56 times for each time step.

For each run, the model produces values of surface and upper-level meteorological parameters.

Research conducted

Analysis using extreme value statistical methods can yield return periods (e.g. the 1-in-100-year event) and the variability of extremes over time and space.

Extreme events are more common when regional or hemispheric climate patterns are of a certain type.

We investigated dependencies of extremes on climate states such as ENSO, and applied the new understanding to innovative insurance products based on weather and climate indices.

We then investigated multiple climate indices as a means of forecasting extremes.

Using the 20CRP data for the Australian continent, we categorised extreme weather conditions such as intense rain, heatwaves, hail and drought.

We used long-term time series of the mode of climate indices to test the dependency of occurrence of extremes on regional characteristics and climate modes.

We evaluated the potential to use the relationships developed as a basis for index insurance (as an alternative to indemnity insurance) for hail, cattle heat stress and low wheat yield, providing a framework for future applications to agricultural insurance.

Outcomes

Hail and heatwave risks show similar patterns and trends, reflecting their common basis in severe convection.

Monthly totals of hail days and heatwave intensity were found to depend on local soil moisture and four climate modes.

This finding provides the possibility of a reasonable and transparent basis for risk management and potentially the introduction of new index insurance and seasonal forecasting schemes.

Practical application is limited by poor representation of upper-atmosphere conditions in the 20CRP database.

Implications

The 20CRP database continues to be developed and is likely to be useful on a world-wide basis, at least for the period 1900–2008.

Improvements to the 20CRP inputs and the use of a finer spatial resolution are required for reliable treatment of localised events.

This project demonstrates the value of the datasets that provide a framework for development for applications to agricultural insurance.

Products such as index insurance based on climate-mode indices have the potential to make risk-pooling easier for insurers and reinsurers.

Further research on the use of the 20CRP data for agricultural insurance and seasonal forecasting is needed as development of the database continues.

Publications

Best, PR, Stone, RC and Allan, R 2009. ‘Insuring agricultural resilience using climate modes and reanalysis’. 9th International Conference on Southern Hemisphere Meteorology and Oceanography, Melbourne, February 8–13.

Research contact

Dr Peter Best

University of Southern Queensland

cindualpk@bigpond.com

Phone: 07 3844 1777

You may also like...

Your email will not be published. Name and Email fields are required

seven − five =