centuries-old Math Theorem Delivers Exact Water Depths for Natural River Channels

A centuries-old theorem from the era of Joseph-Louis Lagrange is now solving one of hydraulics’ most stubborn computational problems: determining how deep water will flow in a natural river or channel section. In a study published in Water Resources Management, Ahmed A. Lamri of the University of Houari Boumediene in Algiers and Said M. Easa of Toronto Metropolitan University have derived explicit analytical formulas that compute the normal depth of natural channels using the Darcy-Weisbach friction formula, eliminating the trial-and-error and graphical procedures that engineers have relied on for generations. The work is striking not only for its elegance but for its precision: the new solutions achieve maximum relative errors of just 0.02 percent in rough-flow regimes and 0.06 percent in smooth-flow regimes, where existing regression-based approximations can drift by 2.4 percent and 5 percent respectively. For a field where normal depth underpins flood control, irrigation design, and reservoir operations, a hundredfold improvement in worst-case accuracy is a genuinely consequential advance.

Normal depth is the depth at which water flows steadily down a channel under uniform conditions, where gravity’s driving force along the bed slope is exactly balanced by the resisting friction of the channel surface. It is the anchor quantity of open-channel hydraulics, entering virtually every design and assessment calculation for canals, rivers, drains, and culverts. For simple prismatic shapes such as rectangular or circular sections, the governing equations can sometimes be rearranged or inverted with known special functions. Natural channel sections, however, whose boundaries follow irregular cosine-like profiles shaped by sediment dynamics and stable-channel theory, resist closed-form inversion. The flow area, wetted perimeter, and top width all change nonlinearly with depth, and when these are coupled with the Darcy-Weisbach formula and its Reynolds-dependent friction factor, the resulting equation for depth is transcendental. The depth is buried inside both the geometry and the friction term, and no amount of elementary algebra extracts it.

Traditionally, engineers have responded in two ways. The first is iterative: guess a depth, compute the discharge it implies, adjust the guess, and repeat until convergence. This works, but it is slow, can fail to converge for poor initial guesses, and is awkward inside optimization loops or real-time control systems where thousands of evaluations may be needed. The second approach is regression: fit an approximate power-law expression to numerical solutions and publish it as an explicit formula. Regression formulas are fast, but they inherit the errors of their fitting, and for natural sections in smooth-flow conditions, where the friction factor varies strongly with the Reynolds number, those errors have historically exceeded 5 percent, large enough to matter when sizing spillways or setting freeboard margins.

Lamri and Easa took a different route, turning to the Lagrange inversion theorem, a result from classical mathematical analysis that provides the coefficients of the power series expansion of an inverse function. If a variable can be expressed as a function of a parameter, the theorem allows the parameter to be expanded as a convergent series in powers of the variable, even when the inverse function has no elementary closed form. The authors reformulated the normal-depth problem for natural sections so that the dimensionless normal depth appears as the inverse of a tractable functional relationship involving the flow geometry and flow regime parameters. Applying the Lagrange inversion theorem then yields explicit, rapidly converging power-series expressions in which the depth is computed directly from the discharge, bed slope, roughness height, kinematic viscosity, and section-shape constants. Crucially, the series coefficients involve well-behaved functions, including complete elliptic integrals of the second kind and gamma functions, all of which are available in standard mathematical software and even in spreadsheet environments.

The paper develops the solutions through two complementary pathways. The first applies the Lagrange inversion theorem separately to the rough-flow and smooth-flow regimes, producing dedicated power-series solutions for each. The second addresses the transition region between them, where the friction factor interpolates between fully rough and smooth behavior and where a single series becomes less uniformly accurate across the parameter range. For this zone, the authors use curve-fitting refinement, constructing a compact two-term explicit solution calibrated to the transition behavior. The result is a three-piece analytical toolkit that together spans the full range of flow conditions encountered in natural channel practice, with each piece retaining the fully explicit character that makes direct, iteration-free computation possible.

The accuracy gains are documented through systematic error analysis against the exact numerical solution of the governing equations. In rough flow, the proposed series delivers a maximum relative error of 0.02 percent, two orders of magnitude below the 2.4 percent of the best existing regression method. In smooth flow, where legacy approximations are weakest, the new series holds the maximum error to 0.06 percent against 5 percent for the incumbent approach, roughly an 83-fold reduction. In the transition region, the two-term curve-fitted solution achieves a maximum relative error of 0.16 percent compared with 3.25 percent for the existing method, a twentyfold improvement using a formula simple enough for hand calculation. The authors also performed a rigorous convergence analysis showing that truncating the series after only two or three terms introduces negligible error across the entire range of applicability, meaning practitioners do not need to evaluate long expansions to reach design-grade accuracy.

Because every engineering input carries uncertainty, from surveyed bed slopes to estimated roughness heights, the authors subjected their formulas to Monte Carlo uncertainty analysis, propagating typical parameter uncertainties through thousands of randomized trials. The proposed method remained robust, with mean errors below 0.03 percent and 95th-percentile errors below 0.09 percent, confirming that the analytical solutions do not amplify parameter noise and can be trusted within standard engineering tolerances. This statistical validation matters for practical adoption, since a formula that is exact at the design point but erratic under perturbation would be of limited value in field applications where inputs are inherently approximate.

The practical implications extend well beyond textbook convenience. The fully explicit form of the solutions enables instantaneous computation, making them well suited to real-time decision support, where flood forecasting and reservoir operation models must evaluate channel hydraulics continuously as conditions change. They are equally suited to automated design optimization, in which algorithms explore thousands of candidate channel geometries and need a fast, reliable normal-depth evaluator at each step. As a companion contribution, the paper also derives an exact closed-form solution for the critical depth of natural stable sections, obtained from the minimal specific energy condition, giving designers a matched pair of explicit tools for the two most important reference depths in open-channel flow. The datasets and models supporting the study are available from the corresponding author upon request.

The work continues a research line in which Lamri and Easa have previously applied special-function methods, including Lambert W-function solutions and Lagrange-theorem treatments of the Colebrook equation, to problems that hydraulics engineers had long accepted as solvable only by iteration. By importing a theorem from eighteenth-century analysis into modern water resources computation, the new study demonstrates that some of engineering’s persistent numerical bottlenecks yield not to faster computers alone, but to sharper mathematics. For the designers of irrigation networks, flood defenses, and reservoir systems, water depths in natural channels can now be calculated directly, precisely, and in a single step.

Subject of Research: Explicit analytical solution of normal depth in natural open channels using the Lagrange inversion theorem.

Article Title: Explicit Lagrange-Theorem Solution for Normal Depth of Natural Channels

Article References: Lamri, A. A., & Easa, S. M. (2026). Explicit Lagrange-Theorem Solution for Normal Depth of Natural Channels. Water Resources Management, 40(11), Article 524. https://doi.org/10.1007/s11269-026-04888-6

Image Credits: AI Generated

DOI: 10.1007/s11269-026-04888-6

Keywords: natural channels, normal depth, Darcy-Weisbach, Lagrange inversion theorem, open-channel hydraulics, explicit solution, flow resistance, flood control, irrigation design, reservoir operations, Explicit, Lagrange-Theorem