Comparing gully length and average gully width relationships on two geological sediments of varying vulnerability to erosion in the north central Nigeria

Abstract

Gully erosion is an environmental problem that confronts the international community. This study was undertaken to investigate the relationships between gully length and average gully width on two geological sediments of varying resistance to erosion. Thirty-seven and five gully samples formed respectively on two homogenous geological sediments (the Ajalli Sandstones (AS) and the Upper Coal Measures (UCM)) underlying the Idah-Ankpa Plateau of the North Central Nigeria were measured and their lengths (L) and average widths (W) were subjected to correlation analyses and the sample bivariate regression to examine and compare their relationships on the two formations. Results show that on the highly erodible AS, L is positively, poorly, and not significantly correlated with W (R = 0.201, R² = 0.040), whereas the correlation is stronger, significant, but negative at the 0.05 level on the UCM (R = -0.930, R² = 0.865). The regression equations show that L is a poor predictor of W on the AS (W₁ = 6.387 + 0.03 L₁, R² = 0.040), and a very good predictor of the parameter (W) on the UCM (W₂=18.135 – 0.033 L₂, R² = 0.865), which appears as a negative value. And the valid range of gully lengths for the prediction of the average width on the AS can be ranked as 0 = L < ∞ (infinity), whereas the ranking is of the form 0 ≤ L ≤ 549.55 m on the UCM. The study suggests that the correlation of W with L becomes stronger, negative and significant as the resistance to erosion of the geological sediments increases. It also suggests that the predictability of W using L as a tool improves and transforms to a negative value as the resistance to erosion of the formation increases.

Keywords: Bivariate regression; correlation; geological sediments; gully erosion; soil erosion; vulnerability