The future of ecological research: Quantum chaos and the Metadata Effect

Issue: 
Network News Fall 2012, Vol. 25 No. 3

Ecological Research – 2050

It is a hot summer’s evening in 2050. Laura, a postdoctoral researcher, is leaving the Arizona State University campus in Tempe, Arizona. She climbs into her electric car and commands it to take her home. While the car pulls out onto the street Laura connects to her apartment computer. She explains her latest hypothesis that an ecological trend she thinks she has identified probably began in the early part of the century and possibly hit a tipping point around 2025. Her computer works with her to refine the details, making a couple of suggestions, until together they re-articulate Laura’s research question. Laura asks the computer to look for ecological datasets that she might use to test her hypothesis.As Laura’s car drives on to the freeway a few moments later the computer confirms that it has identified 26,122 long term datasets that cover the period she is interested in. The computer also confirms that while 94 percent of the datasets from 2025-2050 have sufficient standard data and metadata to facilitate synthesis, only 60 percent of those from 2015-2025 and 30 percent from 2000-2015 meet these criteria.

As her car approaches 150 kilometers per hour, Laura’s computer reaches out simultaneously to the 4,761 data creators whose datasets are not fit for purpose.  Seven minutes later, just as Laura’s car drives itself into her neighborhood, she gets some additional feedback. Apparently, 30 percent of the data creators contacted were able to initiate a connection with their own data sources and provide additional metadata for the relevant datasets. Fifty percent of the data creators explained to the computer that ‘if I did not provide the necessary metadata to my information manager the first six times he asked, why would I provide it now?’ The rest could not be reached because they were dead.

Frustrated, Laura decides to park the car herself. Switching to manual mode, she proceeds to knock over the trash cans and then run over a stray cat because, of course, nobody drives their own car anymore, do they?

If only all the data and metadata had been available in a form suitable to complete Laura’s research she would not have felt inclined to drive her own car. And if she had not chosen to drive her own car that stray cat would (will in future?) still be alive. With this sad outcome in mind, perhaps we should all ask ourselves a searching question: “Do I want to be the reason that Laura runs over that cute little ginger tabby cat in 2050, or should I perhaps go back and make sure that metadata file is complete and, oh, while I’m at it I’ll just do one last quality check on those data?”

Ecological Research – 2100

It is the year 2100. Laura, professor emeritus, steps out on to the balcony of her rainforest apartment. She is reflecting on a successful career in academia. Her role as an info-social-urban-conservation-ecologist has taken her all over the world, finally retiring to the foothills of the Andes. Laura looks out over the Atacama mangroves towards the Pacific Ocean, wondering how long it will be before she has to move farther up the valley. She is mulling over a new hypothesis that an ecological trend she is currently observing began around 2025 and possibly hit a tipping point around 2050. She shares this thought with her neural net, which works with her to refine the details, making a couple of suggestions, until together they re-articulate Laura’s research question. Laura thinks the neural net should look for ecological datasets that she might use to test her hypothesis, and it does.

A few seconds later the neural net confirms that it has identified 762,340 long term datasets that cover the period that interests her. It also confirms that 99.99 percent of the datasets from 2075-2100 have sufficient standard data and metadata to facilitate synthesis, and only 98.1 percent from 2050-2075 and 88.4 percent from 2025-2050 meet these criteria.

While Laura pours another glass of wine her neural net reaches out simultaneously to the 423,440 data creators whose datasets are less than perfect.  Four minutes later, just as Laura is finishing her second glass of wine, she gets a summary response. Apparently, 78 percent of the data creators contacted were able to provide the additional information needed to make use of the relevant datasets. Two percent of the data creators explained to Laura’s neural net that “if I did not provide the necessary metadata to my information manager the first ten times he asked, why would I provide it now?”  Finally, the remaining 20 percent could not be reached because they were dead.

The good news was that Laura had enough to proceed, and relaxing over a third glass of wine, her neural net went to work. The results of her synthesis of these quality data were quite startling. In the period between 2025 and 2050, severe disease vectors combined with global climate change had caused a dramatic reduction in the worldwide Felis catus population. Extensive monitoring via intelligent city networks had discovered that a relatively small proportion of the population showed signs of being resistant to both the virulent diseases and temperature extremes. The sensor network eventually tracked down what they thought was the Holy Grail in the United States: the one animal that demonstrated complete resistance to these factors in Tempe, Arizona. Unfortunately, before conservation-geneticists could capture the animal it was run over by a car in an apartment parking lot. When they eventually recovered the decaying carcass from inside a dented trash can, the extreme heat of the central Arizona urban heat island had put paid to any hope of cloning the animal.

The death of “Super Tabby”, as he came to be known, was followed by a collapse of the cat population, which in turn was followed by an explosion of Rattus rattus and Rattus norvegicus numbers. Crop failures followed, along with the collapse of many small mammal populations. Ecosystem imbalances had continued to this day, including the one Laura was focused on at that moment. As the results of her research sunk in and Laura realized the part she had played in the major ecological trend of the second half of the 21st century, her wine lay forgotten on the table. Finally, Laura arose from her seat and threw herself off the balcony.

If only all the data and metadata had been available in a form suitable to complete Laura’s original research she would not have felt inclined to drive her own car. And if she had not chosen to drive her own car Super Tabby would (will in future?) have survived long enough to be cloned. And the rest, as they say, will be history. With this sad outcome in mind, perhaps we should all ask ourselves a searching question: “Do I want to be the reason that Laura throws herself from her apartment balcony in 2100, or should I perhaps go back and make sure that metadata file is complete and, oh, while I’m at it I’ll just do one last quality check on those data?”

By Philip Tarrant (CAP)