Multi-objective optimization of spatial sampling

Lark, R.M.. 2016 Multi-objective optimization of spatial sampling. Spatial Statistics, 18 (B). 412-430. https://doi.org/10.1016/j.spasta.2016.09.001

Before downloading, please read NORA policies.

Preview

Text (Open Access Paper)
1-s2.0-S2211675316300562-main.pdf - Published Version
Available under License Creative Commons Attribution 4.0.
Download (646kB) | Preview

Official URL: http://dx.doi.org/10.1016/j.spasta.2016.09.001

Abstract/Summary

The optimization of spatial sampling by simulated annealing has been demonstrated and applied for a range of objective functions. In practice more than one objective function may be important for sampling, and there may be complex trade-offs between them. In this paper it is shown how a multi-objective optimization algorithm can be applied to a spatial sampling problem. This generates a set of solutions which is non-dominated (no one solution does better than any other on all objective functions). These solutions represent different feasible trade-offs between the objective functions, and a subset might be practically acceptable. The algorithm is applied to a hypothetical example of sampling for a regional mean with the variance of the mean and the total distance travelled between sample points as the two objective functions. The solutions represent a transition of sample arrays from a loose grid to a tight loop. The potential to develop this approach and apply it to other spatial sampling problems is discussed.

Item Type:	Publication - Article
Digital Object Identifier (DOI):	https://doi.org/10.1016/j.spasta.2016.09.001
ISSN:	22116753
Date made live:	07 Dec 2016 09:42 +0 (UTC)
URI:	https://nora.nerc.ac.uk/id/eprint/515432

Item Type:

Publication - Article

Digital Object Identifier (DOI):

https://doi.org/10.1016/j.spasta.2016.09.001

ISSN:

22116753

Date made live:

07 Dec 2016 09:42 +0 (UTC)

URI:

https://nora.nerc.ac.uk/id/eprint/515432