%These are the initialization steps for the Lorenz_96 model

%apply perturbation to truth and create ensembles
truth(20)=truth(20)+initial_model_perturbation;
for j=1:ensemblesize
  ensemble(20,j)=ensemble(20,j)+initial_model_perturbation+randn*initial_ensemble_perturbation;
end

%create a matrix of indices
index=[1:modelresolution];
indices=[index+1;index-2;index-1];
indices=mod(indices,modelresolution);
indices(find(indices==0))=modelresolution;

%setup lorenz_96 model
deriv=inline('(x1-x2).*x3-x4+forcing');