Channel instability is a ubiquitous problem for streams and rivers in northeast Kansas, caused in part by widespread land cover change from deep-rooted prairie grasses and trees, to shallow-rooted commercial crops and rangeland (Verry, 2004; Thorne, 1991). Channel instability is especially prevalent in regions with highly erodible soils and no bedrock channel controls – the same conditions that tend to favor agriculture (Shields et al., 1998). The sandy soils of the Kansas River and its tributaries typically have low cohesiveness, making them especially vulnerable to rapid erosion. The effects of channel instability are far-reaching. For farmers of riparian lands, channel instability and erosion can result in the loss of thousands of acres of valuable arable land (Geyer, 2000). Changes in channel structure and sediment regimes can also negatively affect native fish, mussel, and macroinvertebrate populations.
To identify eroding banks, we used National Agriculture Imagery Program (NAIP) infrared imagery to run a Normalized Difference Vegetation Index (NDVI) analysis to quantify the health of riparian vegetation along the Kansas river corridor. To determine where vegetation was being lost along streambanks, we calculated NDVI for the years 2012 and 2006 using NAIP infrared imagery. The data was then masked so that only values indicating vegetation (>0.3) were being used in the calculations. The 2006 layer was subtracted from the 2012 layer, giving vegetation gain or loss over a 6-year period. The results present a very clear representation of eroding banks, as well as moving islands within the stream channel (vegetation loss on the upstream end, gain on the downstream end). These results were used to create an “eroding banks” layer, shown in the map below. This layer is also displayed on the combined river inventory map.