Saturday, May 23, 2009

Filling in data gaps with mathematics

Science Daily: Studies of climate evolution and the ecology of past-times are often hampered by lost information – lost variables needed to complete the picture have been long thought untraceable but scientists have created a formula which will fill in the gaps of our knowledge and will help predict the future.

A novel method of reconstructing missing data will shed new light on how and why our climate moved us on from ice ages to warmer periods as researchers will be able to calculate lost information and put together a more complete picture.

…By developing a novel Hamiltonian approach to the problem, using a mathematical algorithm, assuming the dynamics of each system has unknown parameters and that the data are distorted by random fluctuations, the researchers from California and Lancaster were able to successfully recreate measurements in a study on a vole-mustelid community.

Many small mammalian species have cyclic population dynamics, periodically oscillating between large and small communities, a behavioral phenomenon which has puzzled ecologists for decades. Reconstructed data on such predator-prey dynamics could now give new insight into why some species suddenly decline.

Climate evolution is subject to similar cyclical variations, which could be uncovered by applying the method to measuring the distribution of isotopes in sediments taken from the ocean floor, potentially giving further insight into the reasons behind climate change….

A herd of caribou

No comments: