Lack of spatial data infrastructure limits oceanographic research in Brazil

Researchers and students spend time on basic tasks because of a lack of structured primary databases, weakening the impact of their studies.

 

By Fábio de Castro

Agência FAPESP – The lack of an open-access, integrated spatial data infrastructure is hindering oceanographic research in Brazil, according to Professor Jarbas Bonetti of Universidade Federal de Santa Catarina’s (UFSC) Coastal Oceanographic Laboratory. He presented this analysis during the international workshop on Marine Data Management held at FAPESP’s headquarters on April 11-12. 

According to Bonetti, many scientists and graduate students use most of their research time generating basic information because of the lack of an open-access structured platform with reference data. This lack of infrastructure limits their studies. The country must invest in creating unified data repositories, he says.

“Many doctoral students spend the majority of their research time structuring a primary database. However, this effort is often unnecessary because the data exist but are dispersed, are not interoperable or are not in an open-access format. Accordingly, the student invests years of exhaustive effort but ultimately, during the interpretation phase, has little time or energy to contribute effectively to advance knowledge through more sophisticated analyses,”Bonetti comments in an interview with Agência FAPESP.

Bonetti analyzed the impact of the lack of structured databases using his own research experiences for a post-doctoral degree at the French Research Institute for Exploitation of the Sea in Brest, France from 2007 through 2008.

In the study, Bonetti cross-referenced data from several reference bases to develop a map applying the concept of “marine landscapes.” Later, he encountered serious difficulties in attempting to replicate the study in Brazil.

“The French institution had a structured database, and that’s why I found it very easy to obtain conclusive and useful results for managers. The majority of my efforts consisted of developing a conceptual framework and proposing a methodological alternative to integrate diverse spatial data,” he  says.

Within the concept of marine landscapes, according to Bonetti, one can start from a set of relatively generic data – such as depth, seabed temperature, substrate type, light penetration and current intensity – to understand how the biological community is organized given the environmental characteristics that support the establishment of diverse habitats.

“With this methodology, based on relatively simple data, one can obtain an initial diagnostic of the seabed structure. And this can optimize the selection of sensitive locations for more effective management or to conduct more verticalized research,” he explains.

Based on the basic data available on the region of the Iroise Marine Park, Bonetti built a model of marine landscapes in an area of Brittany, France. “When I finished my post-doctoral degree, I was excited about the results and had the opportunity to replicate the study in Brazil’s Arvoredo Marine Biological Reserve in Santa Catarina State, a very important area from the point of view of the Brazilian conservation units system,” he said.

However, Bonetti soon realized the difficulty of finding this type of data in an accessible format. The scientist had no other choice but to dedicate long hours to collecting reference data.

“I relied on hard labor to generate extremely basic information. I also relied on the collaboration of colleagues who shared raw data, based on contacts with former mentees who had information available from original tables. I had to learn many things that were secondary to the investigation and spend a lot of time that could have been invested in deeper analysis,” he explained.

After the experience in France, Bonetti became convinced that structuring a common, open-access primary database would allow researchers to dedicate less effort to the primary treatment of data, allowing them to invest more time in the analysis and the development of alternatives to working with spatial data, seeking to reestablish relations and identify relationships and how different variables behave jointly in space.

“There are data with a precise and specific character. I am referring to primary and fundamental data such as bathymetry – data relating to depth,” he affirms. There is no open-access bathymetric database available online, Bonetti says. The existing nautical charts, for example, are only available in a format similar to digital photos and not in a vector format, which would allow reprocessing.

“When we need these data, we have to upload these nautical charts to the computer, georeference them, create a mosaic and use a mouse click on each of the pixels that have greater depth. This turns a task that would take just days into a job that lasts weeks or months,” he affirms.

The topographical sheets on the scale most widely used for oceanographic research associated with the inner continental shelf, according to Bonetti, are based on aerial photographs from the 1960s. These images are difficult to fold, making it difficult to match the coastlines from one chart to the next. Altimetry data on coastal areas, fundamental for anyone working on the dynamics of beaches and the evaluation of coastal susceptibility to rising sea levels, are also scarce.

“Occasionally, some city halls have this data or some groups have conducted studies of this type, but all of this information is dispersed and difficult to access. Much of the effort to collect data is connected to individual research projects. These data end up being very restricted to the groups that produced them and access depends on personal contact,” he affirmed.

Another recurring problem, he says, is the lack of metadata – information that explains and contextualizes the data. “Without metadata, the data lose their reliability and cannot be used with other similar data,” he comments.

Bonetti suggests that to sidestep the problem, it is fundamental that the public authorities invest in research programs that have continuity and in which data provision services function regularly and efficiently. “Reference data, such as coastlines, altimetry, bathymetry, type of seabed and soil usage, must be available to any scientist at the click of a mouse,” he says.

According to Bonetti, it is crucial to raise awareness that data collected by researchers funded with public money should be made available to the public.

“The intellectual property rights for data must be guaranteed, by giving the researcher time to publish, and a standard must be established so that the data can later be compared to other data. But the main thing is to create a culture in which data financed by the state and public foundations are viewed as public property,” he adds.