Approximations made to the Reynolds-averaged equations of turbulence to allow solutions for flow and turbulence variables. The Reynolds-averaged equations contain statistical correlations such as the variance or covariance between dependent variables such as velocity or temperature. The equations that forecast lower-order correlations often contain unknowns of higher statistical order, a difficulty known as the closure problem. When the higher-order terms are approximated as empirical functions of lower-order terms and of known independent variables, the resulting approximate equations can then be solved. These approximations, known as closure assumptions, must satisfy parameterization rules.