Automated quantification of contouring as support practice for improved soil erosion estimation considering ridges

Using soil conservation practices such as contouring are able to reduce soil loss on arable land parcels. In the empirical model of the Universal Soil Loss Equation (USLE), these measures are taken into account by the P-factor for support practice management. In the context of application, there is usually a lack of sufficient data or suitable methodology to accurately determine the P-factor within a plot-specific analysis. In this study, we demonstrate the effort and benefit of deriving an individual P-factor for each land parcel within a typical application scale. For this purpose, we apply the Fast Line Detector algorithm to open remote sensing data of Google Earth from May 2016 in German low mountain range. The algorithm detects lines from tire tracks and seed rows, which allows to determine an individual main cultivation direction for each land parcel. The success rate was 94.9 % for 2495 land parcels with 26 different crops. The results show a major time advantage for the automated method when considering a large number of parcels. Subsequently, we used the detailed information obtained to calculate the P-factor under regional German conditions using the German standard DIN 19708 and, secondly, an approach based on revised USLE. It is apparent that the current German standard cannot be applied with the necessary level of detail for 78.1 % of all land parcels in this low mountain range study due to unsuitable equations and validity ranges for slope steepness and length and a non-consideration of ridges and off-grade contouring and therefore needs to be revised to avoid being restricted in its application.