I've just watched a fascinating episode of Horizon which examine the data collected since the Chernobyl disaster. It's a fine example of why you should collect data points for as much of your model as possible. One of the most interesting things the program discussed was the wildlife that is living in the area now. Although radiologically very active the animals show remarkably little sign of cell damage. Of course given the natural variation of background radiation on the planet its not surprising life has evolved to be able to cope with low dosage radiation.