In groundwater numerical simulations, the interactions between surface and groundwater have received great attention due to difficulties related to their validation and calibration due to the dynamic exchange occurring at the soil–water interface. The interaction is complex at small scales. However, at larger scales, the interaction is even more complicated, and has never been fully addressed. A clear understanding of the coupling strategies between the surface and groundwater is essential in order to develop numerical models for successful simulations. In the present review, two of the most commonly used coupling strategies in detailed domain models—namely, fully-coupled and loosely-coupled techniques—are reviewed and compared. The advantages and limitations of each modelling scheme are discussed. This review highlights the strategies to be considered in the development of groundwater flow models that are representative of real-world conditions between surface and groundwater interactions at regional scales.

In the design practices of many engineering applications, gross information about the flow field may suffice to provide magnitudes of the parameters that are essential to complete the design with reasonable accuracy. If such design parameters can be estimated following simpler steps, it may be possible to abandon the need to conduct expensive numerical and/or experimental works to produce them. In this work, we are interested in providing a generalized power law that depicts the velocity profile for fully developed turbulent flows. This law incorporates two fitting parameters m and n that represent the exponents of (1) a nondimensional length scale and (2) an overall exponent, respectively. These two parameters may be determined by fitting the experimental and/or computational data. In this work, fitting benchmark experimental and computational fluid dynamics (CFD) data found in the literature reveals that the parameter m changes over a relatively smaller range (between 1 and 2), while the parameter n changes over a wider range (between 1 and 12 for the range of Reynolds number considered). These two parameters (m and n) are, generally, not universal, and they depend on the Reynolds number (Re). A correlation was also developed to correlate n and Re in the turbulent flow region. In order to preserve the continuity of the derivative of the velocity profile at the centerline, a value of m equals 2 over the whole range of Re is recommended. Apart from the near wall area, the new law fits the velocity profile reasonably well. This generalized law abides to a number of favorable stipulations for the velocity profile, namely the continuity of derivatives and reduction to the laminar flow velocity profile for lower values of Re.