Predicting the annual erosion rates on a small stream by the BANCS model

Abstract

The erosion of streambanks causes soil loss and degrades the stream habitat. To optimize the prevention of bank erosion, we first need to determine the most vulnerable places on banks. This can be done by the BANCS model. However, data are still missing on its accuracy in small streams. We measured the real annual erosion rates on 18 experimental sections established on the Lomnická stream. Using the Near Bank Stress (NBS) and Bank Erosion Hazard Index (BEHI) we developed the erosion prediction curves and evaluated the relationship between these two indices and the real annual erosion rates. We found a strong relationship between BEHI and real annual erosion rates, with R<sup>2</sup> = 0.72. The relationship between the NBS index and real annual erosion rates was also strong, with R<sup>2</sup> = 0.53. Then we constructed erosion prediction curves for very high and extreme BEHI and for moderate and high BEHI. Despite the strong correlation between BEHI and annual erosion rates, the prediction curves had no real relationship with real annual erosion rates, with R<sup>2</sup>= 0.004 and 0.15, respectively.