Lorenz '96 differential model, for LibBi.
Matlab Shell Perl
Switch branches/tags
Nothing to show
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
data
oct
LICENSE
Lorenz96.bi
Lorenz96Deterministic.bi
Lorenz96Stochastic.bi
MANIFEST
META.yml
README.md
VERSION.md
bifurc.conf
config.conf
filter.conf
init.sh
parse_time.pl
posterior.conf
posterior_sir.conf
prediction.conf
prior.conf
qsub_run.sh
qsub_run_bifurc.sh
qsub_time.sh
run.sh
run_bifurc.sh
time.sh

README.md

Lorenz '96 model

Synopsis

./run.sh

This samples from the prior and posterior distributions, and performs a posterior prediction. The oct/ directory contains a few functions for plotting these results (GNU Octave and OctBi required). In particular, after ./run.sh, the plot_and_print function will produce SVG figures in the figs/ directory.

./time.sh

This can be used to reproduce the timing results in Murray (2013).

A synthetic data set is provided, but a new one may be generated with init.sh (GNU Octave and OctBi required). A number of qsub_*.sh scripts are also provided that may assist with setting up the package to run on a cluster.

Description

The original, deterministic Lorenz '96 model (Lorenz 2006) is given by

$$\frac{dx_{n}}{dt}=x_{n-1}(x_{n+1}-x_{n-2})-x_{n}+F,$$

where $\mathbf{x}$ is the state vector, of length 8 in this package, with subscripts indexing its components in a circular fashion. $F$ is a forcing parameter. This form of the model is given in the Lorenz96Deterministic.bi file.

A stochastic extension of the model adds an additional $\sigma$ parameter and rewrites the above ordinary differential equation as a stochastic differential equation:

$$dx_{n}=\left(x_{n-1}(x_{n+1}-x_{n-2})-x_{n}+F\right), dt+\sigma, dW_{n}.$$

This form is specified in Lorenz96.bi and used for inference in LibBi.

The interest in the Lorenz '96 model is that its dimensionality can be scaled arbitrarily, and that, according to this number of dimensions and $F$, the deterministic model exhibits varying behaviours from convergence, to periodicity, to chaos. The stochastic model exhibits similar behaviours.

The model is one of the examples given in the LibBi introductory paper (Murray 2013). The package may be used to reproduce the results in that paper.

References

E. N. Lorenz. Chapter 3: Predictability -- a problem partly solved. In Predictability of Weather and Climate, Cambridge University Press, 2006, 40-58.

L. M. Murray. Bayesian state-space modelling on high-performance hardware using LibBi. 2013.