Putting the “Ah hah!” in Data
In a little-used 3TIER conference room — the one where old desk chairs go to live out their twilight years — sit two stacks of cardboard boxes containing more boxes: slick, nearly featureless black boxes. One pile stores new CPUs awaiting introduction to the computing cluster.
The other pile stores hard drives shipped to us from the Space Science and Engineering Center at the University of Wisconsin-Madison. Although 3TIER is located in the most wired building in the Pacific Northwest, we calculated that it’s more bandwidth-savvy to ship empty hard drives to Madison and receive full hard drives back. The data acquisition team unpacks them one by one and joins their contents to a massive collection: 12 terabytes of satellite pictures of the planet. These images are the beginnings of the REmapping the World global solar map.
Ken Westrick once said, “If you keep data in a database, it’s worthless.” We want data to be worth your while. After all, data doesn’t become valuable information until it’s able to inform.
The software development team is one of the 3TIER groups that help add value to raw data. They write the computer programs that ingest data, model data into resource assessments and forecasts, and bank the refined data so that the web team can turn it into usable products – be it an interactive FirstLook map that’s free to the world or an individualized forecast for a particular client. (More on the web team in a future post.)
NOAA GOES satellite image courtesy of SSEC/Univ. Wisconsin-Madison.
It’s an interesting puzzle for our computer and atmospheric scientists: How to turn actual visible-band pictures of the planet into data about solar radiance on the ground? We developed an algorithm that considers the amount of cloudiness in the satellite pictures, simulates the position and angle of the sun, and thereby models the amount of sunlight that reaches the ground.
You might imagine that solar power plant developers wish only to know where and when the sun shines, but in fact they are also concerned with temperature and winds, which can affect the performance of their panels. But we’ve got that data, tons of it, from our numerical weather prediction models, so we’ll be able to provide clients who want a complete assessment with information about temperature, humidity, and wind speed in addition to solar radiance.
Developing a new solar map is just one of the many things that our software team does to turn data into information. Their to-do list is growing fast. If you’re a software engineer who thinks it would be cool to work in the private sector for a good cause with an academic-sized serving of CPUs, we’re hiring (and not just software engineers).
Tags: Assessment, forecasting, REmapping the World, visible satellite pictures
Author: Leah Kauffman
