Underwater wireless optical communication (UWOC) using Multiple-Input Multiple-Output (MIMO) technology has emerged as a promising solution for achieving high data rates in turbid water environments. However, non-line-of-sight (NLOS) scattering has a significant impact on system performance. This paper proposes an enhanced stochastic channel model for NLOS UWOC MIMO systems, incorporating adaptive weighting functions in the Weighted Gamma Function Polynomial (WGFP) model. Monte Carlo simulations (MCS) are utilized to derive the channel impulse response (CIR) and bit error rate (BER) under various conditions. The results demonstrate that the novel model achieves superior accuracy compared to existing WGFP models and provides a more robust estimation of scattering losses in high-turbidity environments.