1: \begin{abstract}

2: 

3: We develop information-theoretic measures of spatial structure and

4: pattern in more than one dimension. As is well known, the entropy

5: density of a two-dimensional configuration can be efficiently and

6: accurately estimated via a converging sequence of conditional

7: entropies. We show that the manner in which these conditional

8: entropies converge to their asymptotic value serves as a measure of

9: global correlation and structure for spatial systems in any

10: dimension. We compare and contrast entropy-convergence with

11: mutual-information and structure-factor techniques for quantifying and

12: detecting spatial structure.   

13: 

14: % Insert PACS numbers on next line

15: PACS:

16: %02.50.Ey  % Stochastic processes 

17: %02.50.Ga % Markov processes 

18: 05.20.-y % Classical statistical mechanics

19: 05.45.-a  % Nonlinear dynamics and nonlinear dynamical systems

20: %05.45.Tp  % Time series analysis

21: 65.40.Gr % Thermodynamics of solids: Entropy and other thermodynamical

22:           % quantities  

23: 89.70.+c % Information science 

24: 89.75.Kd  % Complex Systems: Patterns 

25: 

26: Santa Fe Institute Working Paper 02-12-06X

27: ~~~~~~~arxiv.org/abs/cond-mat/0212XXX

28: 

29: \end{abstract}

30: