In view of the problem of limited usage scenarios as well as inaccuracy of state duration probability density function (SDPDF) in existing land mobile satellite (LMS) channel model, we propose a narrowband LMS channel model which is capable of describing much more operational scenarios accurately and effectively. The state sequence is achieved by using Semi-Markov chain which approximates the SDPDF as a lognormal distribution. Within each state the carrier-to-noise ratio (C／N) time series is generated and defined as a Loo distribution by combining with propagation parameters. Doppler frequency shift is applied in slow fading components to increase parameter database and extend usage scenarios. The state probabilities, the first-order statistical characterization of C／N time series, and the required margins of 90％ signal availability are investigated and compared with measurement results under different scenarios conditions. Simulation results show that the proposed model is effective for the representation of shadowing conditions on much more scenarios. In addition, it has a good applicability for different elevations and driving directions in different environments, and can be used in system performance analysis.