The design and assessment of air interfaces for modern Land Mobile satellite systems between 1 and 4 GHz require the use of reference generative propagation channel models. Among possible generative modeling approaches, the empirical stochastic method has been widely used in the past two decades for its simplicity, reduced number of input parameters and efficiency in terms of computation time. A model based on 3-state Markov chain approach for SISO links in particular was used in various standardization groups whereas available input parameters generally were affected by measurements biases and uncertainties. Recently, an evolution of such model with more flexibility, the so-called enhanced 2-state SISO model, has been developed. New experimental datasets collected at 2 and 4 GHz in different environments and for various elevation angles have been used to slightly improve this initial enhanced 2-state model and to fully parameterize it. A dual polarization MIMO extension has also been developed from these new experimental results. This paper aims at presenting the upgraded enhanced 2-state channel model and at providing testing results.Index Terms-SISO, MIMO, Land Mobile Satellite, generative chanel model.