Improving Climate Data for NSW

NSW DPI has been working with the Australian National University (ANU) since 2019 on ways to improve the quality and accuracy of climate data for NSW and Australia. This has large potential benefit to farmers and industry as we are working toward having a quality data set available for every kilometre across NSW.

Improving the quality of information is important. In our recent stakeholder survey farmers identified concern about data quality and trust as key barriers to using climate information in their decisions.

Interpolation science

DPI is working with world leading researchers at the ANU in the field of ‘interpolation science’.  In simple terms interpolation means ‘filling in gaps’ where we don’t have climate stations physically located.

The principal researcher is Professor Michael Hutchinson at ANU’s Fenner School. This work centres on the development and application of a powerful algorithm called ANUSPLIN.

ANUSPLIN uses factors like elevation and stable long-term averages to improve the interpolation, and it can yield accurate results even in places where the number of observation stations is relatively low.  Part of the process is stringent quality control and error detection in the underlying station network.

ANUSPLIN is used extensively around the world to turn station-based climate data into gridded climate data or maps.  You see applications of ANUSPLIN in other countries for example:

• New Zealand’s Virtual Climate Station Network (National Institute of Water and Atmospheric Research) which provides a climate station every 5kms across the country
• Providing daily and monthly climate data for Canada (Regional, national and international climate modelling; Natural Resources Canada)

In Australia ANUSPLIN underpins products like:

• The SILO Patched Point Data Set (Queensland Government)
• The Australian Water Availability Program (CSIRO)

ANUClimate- improved data for Australia

Over the last decade ANU has been undertaking research to work out how to optimally use ANUSPLIN on Australia’s climate network.  The outcome of this research is a nationally significant data set called ANUClimate. A description of the science behind ANUClimate and the improvements made is in progress.

ANUClimate is a suite of products including high quality monthly averages, monthly time series as well as daily time series for set periods, (e.g 1975-2020).  These ‘stable products’ are all free and openly available through the Australian National Computing Infrastructure.

Over the last few years DPI have worked with ANU to implement the ANUClimate process in near real time.  New climate data arrives every day, but this is for four days prior to the current date. This allows time for the data transfer from the field and quality control. This has allowed DPI to upgrade its near real time services, such as drought monitoring using the NSW Combined Drought Indicator .

If you want to obtain or use the near real time service, please get in touch.

More enhancements for accurate local farmer data

The improvements haven’t stopped there.  Over the coming years DPI and ANU are undertaking trials to further enhance the accuracy of climate data in NSW. This includes:

• Incorporation of climate data from an expanding network of private weather stations supported by other NSW Government programs like Farms of the Future program see how additional data improves the accuracy of rainfall and temperature interpolation.
• Small amounts of new high quality climate station data being used to test how much this improves the interpolation of rainfall.
• Satellite derived precipitation fields being used to improve daily rainfall interpolation.