By Peter Williams
CTO, IBM Big Green Innovations
During the past year, we’ve seen extreme weather conditions, from crippling drought in many parts of the United States and Europe to floods in Italy, Thailand, China and more. According to the Environmental Protection Agency, climate change may increase the probability of some ordinary weather events reaching extreme levels or of some extreme events becoming more extreme – so in essence, we can expect a continued rise in extreme weather condition and events.
Even without climate change, floods are not rare; in fact, they are the most common natural disaster in the United States. Although we typically have some advance warning of their arrival, thanks to satellite forecasts, there is always the possibility (and likelihood) that a flash flood will behave in unpredictable ways, causing untold damage. To add insult to injury, dry, desert lands are often the hardest hit by floods, in areas where water is the most precious.
Clearly, we can’t fight the weather. Floods and droughts are a fact of life. We can, however, better predict how they affect us and protect ourselves from harm. Most flood modeling systems look at the main stems of large rivers. These forecasts provide valuable information, but often times the real action is in the thousands of small river branches and the tributary networks where flooding actually starts.
The computing power required to forecast the behavior of tens of thousands of river branches was simply not available in past decades. Today, using new flood prediction technology, we can predict the behavior of millions of river branches simultaneously. IBM and the University of Texas are using this modeling technology to predict the behavior of the 230 mile-long Guadalupe River and more than 9,000 miles of tributary networks in Texas. In only one hour, the system can generate models for up to 100 hours of river behavior.
Armed with this sort of knowledge before floods strike, cities and counties can draw up better response or preventive plans. And that kind of information can also be linked to other data regarding the structural condition of levees, so that targeted evacuation orders could be issued if necessary.
The value of technology in water management isn’t limited to cities — there are clear applications for data analytics in agriculture as well.
Although agriculture is an inherently risky business due to unpredictable weather patterns, farmers can minimize the risk with data analytics. In fact, many are already doing so. Looking at historical weather, crop yield and irrigation data, they can design their activities to take advantage of natural weather patterns and have a better understanding of the exact amount of water each type of crop needs. Sun World, a California fruit grower, has used data analytics to reduce fuel and water consumption, as well as labor costs. By analyzing financial data and increasing the use of drip irrigation systems, the company cut water use by 9% over four years’ time. A redesign of grape trellises – which didn’t require workers to bend over to pick the fruit — increased productivity and reduced physical injuries and the number of pickers needed at harvest time. This kind of “precision agriculture” can save water (and thus energy), improve crop yield (more crops with less water), and reduce runoff.
As one Pennsylvania farmer told the Altoona Mirror, “[Farmers] are big risk takers and weather’s a big risk we just have to deal with every year.” But with the help of data analytics and flood modeling systems, weather doesn’t have to be a force to contend with or a challenge to overcome — it can be an energy source harnessed for good.