Ideas for Teaching
Graphing Data
Page 1 | Page 2 | Page 3 | Page 4
Radar Sampling
The storm appears to have a steady height in the first two hours because it is traveling along a nearly tangent line to the radar during that time. Around 21 UTC the storm height from radar starts jumping around wildly, falling as much as 6 km over a short period of time. Could that possibly be real? Satellite estimated heights are steady during that time.
The following graph repeats the thick, colored radar-derived heights, and adds in thin colored lines the heights at which the radar dish is sampling.
Now we can see that during the first few hours we really don't know the radar height. The 6.0 degree elevation from the radar is sampling the storm and the 9.8 degree elevation isn't. But there are 6 km distance between them! Do we really know the storm height between 19 and 21 UTC? Probably not. As the storm moves away from a tangent path to the radar between 20 and 21 UTC the height rises with the radar sampling line.
Once the storm is close to, and moving toward, another radar we start to see the erratic behavior in radar-estimated height. This, too, is simply from sampling, for the most part. For example, look at the heights between about 21:30 and 22 UTC. The radar labeled KTLX does not sample heights between about 12 and 17 km, so the storm height appears to fall with the radar sampling until the next higher elevation angle of the radar receives echo from the storm.
What can you do to visualize data in your classroom? Sometimes your inspiration will come from knowing there is a problem and why the problem exists. Be careful, though, not to draw conclusions from coincidence. Have you ever seen the graphic showing how an inverted solar radiation curve matches the New York Stock Exchange prices for the year 1929? Those variables - solar radiation and stock prices - have no reason to be related, but plotting them together could appear to show a correlation.
Last updated:
March 26, 2002
Created by: DSZ