Stream channels change based on the environment around them, whether that be from topography, climate, or anthropogenic causes. Intense storms and flooding can induce changes in a channel, which is what we'll be looking at today. The intense flooding in 1927 influenced the Winooski River in Vermont, but how? We'll specifically discuss the recurrence interval and the geomorphic effects of this flood.

In order to study the effects of the flooding, an area on the Winooski River was selected and analyzed for change. The three aerial images above show a portion of the Winooski River in Burlington, VT and Winooski, VT overlooking the Intervale. Since 1927, there have been immense changes.

The 1927 image was taken shortly after the flood, and it is immediately clear that the channel is significantly wider than it is in 2004 and 2018. The trainbridge built across the river was destroyed, indicating that the water was likely higher than it was in that image. At the top left of the image, the water appears to move into an area of lower elevation that is occupied by vegetation today. Compared to today, there is little vegetation, suggesting faster erosion of the banks.

The channel in 2004 is narrower than the channel in 1927.The shape is around the same, except the river does not branch off to the left as it did after the 1927 flood. However, it is notable that there is a small vegetated island in the middle of the channel. An older web app suggests that the island was present before the 1927 flood but it may have been underwater following the flood. However, it did not appear to be vegetated in 1927 because the trees would have been higher than the water level. This suggests that the island may have formed due to objects falling into the river, which diverted the river around it and deposited sediment in the area as the water velocity decreased. It would explain the rather rounded shape around the island. Other than the island and the height of the water, it is difficult to tell whether the shape of the channel has changed significantly since 1927. The general shape appears to be the same, but the flooding makes it more challenging to see the more detailed changes.

In 2018, the channel does not appear to be drastically different than the 2004 channel, even after the significant 2011 flooding. The channel may be slightly lower, which can be inferred from the island in the middle of the channel. There is more sand towards the back which is not visible in the 2004 image, leading us to believe that the channel height has decreased at this point.

While these changes may not be drastic, rivers take years to form and develop, and seeing even small changes resulting from a flood is an indication that the flood had significant impacts on the Winooski River drainage basin.

Using data from the USGS, a table was constructed using the year, discharge, gage height, and streamflow from a section of the Winooski River in Essex, VT. These values were used to calculate the rank, recurrence interval, and probability of the largest flooding events in each year from 1928 to 2019. The mean, standard deviation, and percent standard deviation for the annual maximum flood and the annual average streamflow.

The flood statistics for the Winooski River at Essex can be seen in the table below. Comparing the percent standard deviations, it is important to note that the percent standard deviation of the annual maximum, 43.2%, is close to double that of the annual average, at 23.4%. This suggests that the annual maximum streamflow varies more than the annual average, which makes sense. Looking at the table of stream data below, the discharge has a much smaller range than the annual maximum, so it's reasonable that it has less variability. With extreme events such as the floods in 1927, 1936, and 2011, the annual maximum increases significantly, and even a few events of a large magnitude can cause large changes in the percent standard deviation. Most of the rainfall events in the area are not as extreme as the 1927 flood and would see a similar annual average each year. Looking at the recurrence intervals in the table, we see low values around 1 or 2 for a majority of them. Since these events occur frequently, the annual average discharge will not vary as significantly as the annual maximum.