New method for runoff estimation under different soil management practices

Abstract

AbstractSoil erosion has been widely measured using different approaches based on models, direct runoff and sediment collections. However, most of the methods are, poorly applied due to the cost, the accuracy and their tedious nature. This study aimed to develop and test a new method for runoff characterization, which may be more applicable and adaptable to different situations of soil and crop management. An experiment was carried out on runoff plots under different cropping systems (sole maize, sole soybean and maize intercropped with soybean) and soil amendments (NPK, Biochar, NPK + Biochar and Control) in the Semi-deciduous forest zone of Ghana. The study was a two-factor experiment (split-plot) in which cropping systems constituted the main plot whereas soil management the subplot. To assess the quality of the method, different statistical parameters were used: p-values, coefficient of determination (R2), Nash-Sutcliffe efficiency (NSE), root mean square (RMSE) and, root square ratio (RSR). The NPK + Biochar under each cropping system reduced surface runoff than all other treatments. At p < 0.001, R2 ranged from 0.88 to 0.94 which showed good accuracy of the method developed. The dispersion between the predicted and observed values was low with RMSE varying from 1.68 to 2.66 mm which was less than 10 % of the general mean of the runoff. Moreover, the low variability between parameters was confirmed by the low values of RSR ranging from 0.38 to 0.46 (with 0.00 ≤ RSR ≤ 0.50 for perfect prediction). NSE values varied from 0.79 to 0.86 (≥0.75 being the threshold for excellent prediction). Though the sensitivity analysis showed that the method under high amount of runoff (especially on bare plots) was poorly adapted, the dimensions of runoff plots could be based on runoff coefficient of the region by analyzing the possible limits of an individual rainfall amount of the site. The findings provide alternative approach for monitoring soil degradation.