Q&A: CSIRO on big data to support the SDGs

CANBERRA, Australia — At the 2017 Crawford Fund Annual Conference, held in Australia’s Parliament House on August 8, speaker Dr. Mario Herrero was a conference highlight. As chief research scientist for the agriculture and food section of the Commonwealth Scientific and Industrial Research Organisation, Herrero is a leader in developing research methods and models supporting knowledge of farming impacts on the environment and global health.

In April, Herrero and his colleagues released new research analyzing associations between farm size, agricultural diversity and nutrient production, identifying smallholder farmers as an important element in increasing food diversity and nutrition options. The research is already benefitting development knowledge and is demonstrating the capabilities of big and open data to assist food security.

Speaking with Devex, Herrero discussed his work, and how CSIRO is helping with data to support the Sustainable Development Goals. The interview here has been edited for length and clarity.

Your research on the geography of nutrient production — identifying the importance of smallholder farmers in creating more option — was just released in April but has already been re-used in valuable ways. Can you discuss this?

When the data and research was released, I received a call the following week from the Global Burden of Disease project saying this data was what they had been looking for and asking if they could get access to the whole datacube. I told them to go for it.

They are using our data in their models to explain reductions in malnutrition and they have asked us specifically about our ability to produce time series data to monitor progress. This will allow them to drill down into specific places within specific countries, to monitor what is happening with nutrition and health.

There is also a good paper in Nature coming out that has used our research to demonstrate progress on reducing stunting. It’s showing incredible gains.

We are very excited at how our data is being used and we are happy to share it. To get access to the data you just contact us. It is also in the CSIRO portal, which is easy to access and use. That’s how it has to be, as we have received funding for this from organizations including the Bill & Melinda Gates Foundation and taxpayers. We need to give the data back and share it openly.

Monitoring and reporting on the SDGs requires continuous updating of data, not just a snapshot in time. Will you be providing updates, enabling the monitoring of changes and progress?

That is exactly the next step. That is where we want to go.

The methods have been refined so much, and there is such good satellite data that we can now start building a time series just to see how farms are changing. Are the evolving? Are they becoming more sustainable? We can see in real time if we are producing more food, and this is where we want to take our data and capabilities.

As part of our work, we are constantly releasing new datasets. It will take six months of work to provide an update for this data, but to update a dataset within less than a year of difference is a huge achievement. In the past, even five years ago, that would have been ridiculous. We wouldn’t have thought to do it.

The quality of data now is not just allowing us to produce and update quickly, but monitor data at a level that will impact the SDGs. Analyzing health in Nigeria is too big. We need to look at health within a town in Nigeria. Unless we get to that spatial resolution of town level, we cannot make real impact or gains. And to assist government to better target health programs, for example, data needs to enable them to target the regions most in need of support.

This is a great example of how data can be used to monitor the progress of the SDGs. Is CSIRO involved with the committees to identify data needs in supporting the SDGs?

I am involved in the committees directly on the agricultural side. We have to pick data associated with food security, gender, environmental protection and nutrition. But there are so many goals and so many indicators that we have to go step by step.

At the moment, it is about picking the low-hanging fruit. But our work is certainly exciting in demonstrating how we can monitor progress.

The data on farm sizes, production and nutrients is already being used in other areas, including health, to provide understanding and context. How can it assist other areas of development, such as understanding the impact of climate and environmental changes?

If we can collect good information about yield, and we supplement this with management information including fertilizers and other things farmers are doing, then we can calculate gas emissions.

We are making this data available to communities, and it has already been used to analyze the emission targets we can achieve from changing agriculture processes.

Spatially, we will also be able to look at water use. From information we have on yield and agrometeorological data, we can calculate water use efficiencies and see if farmers are improving their management practices.

The whole point is to link it to other information. We can go from the environment, to socioeconomic conditions, to health and nutrition. We need to connect sectors and it is in this process that the outcome will be a win-win for everyone — including the SDGs. Looking at agriculture by itself — forget about it! Don’t do it. But when all the sectors are interacting and connected, this is where we can create opportunities through the data.

You mentioned during the conference that there are “black arts” of analysis and modelling that need to go into creating these global datasets. How can local organizations working in developing countries assist in collecting, assessing and improving data?

There is nothing scientific about putting data together — it is a black art.

But local groups and NGOs have a very important role to play as part of our data rescue exercises. We now have, for example, 60,000 household surveys for Africa. This was just from knocking on the doors of project managers to ask for data from their surveys.

We have people clean this data, put it in databanks and, thanks to all of this assistance, we now have enormous coverage for Africa. It is incredible. And by consolidating small studies into a larger dataset, we are helping others to easily find, access and use data immediately rather than having to go out and collect this information again. If we have collected what you need we will give it to you.

As part of our process, we send maps back to the community and locals to review and provide us with feedback on our modelling — positive and negative. And they may offer better data for us to use. This is scientific transparency which helps in the process.

I don’t hold grudges if someone tells me the data is rubbish. As long as they give me better data.

Read more international development news online, and subscribe to The Development Newswire to receive the latest from the world’s leading donors and decision-makers — emailed to you free every business day.