Research Topics » Flashy Streams

By the Casperkill Assessment Group, Vassar College
Published: February 18, 2010

Flooding is a natural process that occurs in all streams, whether the watershed is in a forested or an urban landscape. However, flooding occurs more frequently and more rapidly in heavily developed watersheds.


In healthy watersheds, most rainwater will infiltrate into the soil, where it flows slowly toward the stream as groundwater or is taken up by vegetation. A small amount of water flows across the ground surface as runoff. In more urbanized areas, pavement and other impervious surfaces prevent the water from infiltrating into the soil, trapping a much larger volume of water above the surface. This stormwater runoff is usually transported as quickly as possible to the nearest body of water by a stormdrain system. While about 10% of precipitation becomes runoff in vegetated areas, up to 55% of precipitation in urban areas becomes stormwater runoff.

As a result, urbanization increases the total amount of runoff and the speed with which it reaches streams. Urban streams often have higher flow rates and more rapid rises and falls in water level – this is considered “flashy” behavior. Larger volumes of water lead to a greater frequency of flooding, and the increased velocity of water gives the stream greater erosive power. Both flooding and erosion can be major issues for landowners near streams, and it is crucial that we understand the causes behind these problems before we try to solve them.


For the same size storm and the same size watershed, the stream's peak flow (Q) is higher and occurs earlier after urbanization (Q after) than before urbanization (Q before). This higher and more rapid peak flow would be considered "flashy."
(Image from the Federal Interagency Stream Restoration Working Group, 2001)

Back to top


In July 2007, researchers from the Vassar College Environmental Research Institute installed a stream gauge in a residential area of the Casperkill Creek (in the Town of Poughkeepsie) to monitor flow rates. The flow of water is typically measured in cubic feet per second (cfs).

The data shows that flow in the Casperkill is highest during the spring snowmelt period and lower during summer months. Over the course of a year, flows varied between 2 and 90 cfs. The stream overflows its banks at the site after reaching approximately 56 cfs, a flow that was exceeded 15 times in the period from July 10, 2007 to Oct. 30, 2008.


In undisturbed watersheds, streams fill to their banks only once every 1‐2 years, so the fact that the Casperkill so frequently overflows suggests that the channel is not in equilibrium with the amount of water it is trying to convey (Charlton 2008). Potential causes include:

  • a change in climate toward wetter conditions
  • an increase in the amount of impervious surface in the watershed that speeds runoff into the stream channel
  • a localized constriction in the channel that backs up water upstream
  • loss of wetland and floodplain water storage
  • or some combination of these factors.

The fact that the Casperkill overflows its banks during even relatively small storms suggests that impervious surface and loss of storm water storage capacity through wetland infilling is the likely cause.

When streams are in disequilibrium, they typically respond with increased erosion as the stream deepens and widens its channel to accommodate higher flows (Riley 1998). Erosion is a problem for some property owners along the Casperkill and other local streams, who have witnessed channel migration that threatens loss of land.



Two different areas along the Casperkill where property is rapidly eroding into the stream.

Erosion is also a problem for aquatic ecosystems, which suffer when sediments are washed into streams, smothering filter feeders and changing the grain size of materials on the channel bottom. Stream-bottom organisms use the nooks and crannies between gravel and small rocks (called cobbles) as habitat; if they fill in with sediment, this habitat is destroyed. In addition, some fish species require a particular grain size of sediment to shelter their eggs. As erosion proceeds, their spawning grounds may disappear.


The more deeply bured cobbles are, the less habitat they provide for aquatic organisms (image from Behar & Cheo 2004).

Research suggests that the Casperkill may suffer somewhat from sedimentation issues. At some sites, cobbles were deeply buried in finer grained materials. In addition, several sites could not be sampled because cobbles could not be found. It is unclear whether these sections of the stream simply lacked these small rocks, or if cobbles were 100% buried.

Back to top


Flooding and erosion issues are often addressed through various “hard stabilization” engineering practices. These include dredging the channel to deepen or widen it so that it will convey more water and armoring channel banks with rip‐rap (small boulders), gabions (cobbles contained within wire baskets), or concrete. While temporarily effective, all of these practices ultimately fail and require repeated costly interventions in the channel (Riley 1998). Furthermore, they can also shift problems up‐ and down‐stream of the “stabilized” reach, causing problems for adjacent landowners.

raingardenBetter solutions to flooding and erosion problems come with changes in land management practices. Proper protection of wetlands and avoiding development on floodplains are the first steps towards preventing the causes of stream disequilibrium. Local governments can work collaboratively with property owners to restore wetlands or floodplains that were developed to a more natural state. Increasing vegetated areas and using rain gardens to allow stormwater to infiltrate into soils can dramatically decrease runoff going directly into stream channels (Buttle 1994; Dietz & Clausen 2005). There are many options for implementing “Better Site Design” principles that allow storm water to infiltrate on site by reducing the amount of impervious cover (Hood et al. 2007). The picture to the right shows a rain garden used to manage stormwater at Vassar College.  (For more information about rain gardens, click here.)

To reduce the impact of impervious surfaces on the Casperkill, the Town could:

  • require residential streets be built to the minimum required width based on traffic volume
  • require new development to have shorter streets while maximizing the number of homes along each street
  • limit the number of residential cul‐de‐sacs, provide incentives for shared parking spaces
  • minimize parking space dimensions in parking lots
  • mandate on‐site stormwater management, including use of rain gardens in parking lots where runoff can slowly infiltrate into the ground.

Additional steps can be taken to deal with local erosion problems. The New York State Department of Environmental Conservation’s (NYSDEC) Hudson River Estuary Program promotes the replanting of Hudson Valley stream corridors through “Trees for Tribs.” The “Trees for Tribs” program offers free native trees, shrubs and planting materials as well as technical assistance to project partners, including town governments, watershed groups, and private institutions. The non‐profit group Trout Unlimited has also provided shrubs for riparian buffer restoration. Planting native plant species, such as willows, have a proven superiority over hard stabilization techniques (Riley 1998). Homeowners (through watershed or non‐profit groups) and town governments alike are encouraged to take advantage of the “Trees for Tribs” initiative to stabilize the banks of the Casperkill and other streams in the area.

Back to top


To study flow in the Casperkill Creek, a Hobo water pressure sensor was installed in July of 2007. The sensor measures water pressure variations related to changes in stream surface elevation, and is calibrated to provide flow data based on discharge measurements from a propeller flow meter (Charlton 2008).

Flow data from the Hobo was compared with precipitation data from a weather monitoring station at the Vassar Farm and Ecological Preserve, located less than half a mile away.

To test whether increasing availability of hard substrate habitat in the stream (gravel, cobbles, or boulders) influenced benthic macroinvertebrates, we conducted a simplified Wolman pebble count (Wolman 1954) on stream bottom sediments by walking back and forth across the stream and recording substrate types (silt/clay, sand, gravel, cobbles, or boulders) at each foot fall (Behar & Cheo 2004). Transects of the stream moved from upstream to downstream within the riffle zone until a minimum of 50 measurements were made, from which we calculated relative abundance of the different substrates. No sites had exposed bedrock in the substrate. For the analysis presented in this report we used percentage cobbles to represent substrate because of correlations among size classes.

For more information, contact This e-mail address is being protected from spambots. You need JavaScript enabled to view it .

Back to top


  • Behar, S. & Cheo, M., 2004, Hudson Basin River Watch Guidance Document, River Network, 138 p.
  • Buttle, J.M., 1994, “Hydrological response to reforestation in the Ganaraska River basin, southern Ontario,” Candadian Geographer, v. 38, p. 240-253.
  • Center for Watershed Protection, 1998, Better Site Design: A Handbook for Changing Development Rules in Your Community, Center for Watershed Protection.
  • Charlton, R., 2008, Fundamentals of Fluvial Geomorphology, New York: Routledge, 234 p.
  • Dietz, M.E., & Clausen, J.C., 2005, “A field evaluation of rain garden flow and pollution treatment,” Water, Air, and Soil Pollution, v. 167 p. 123-138.
  • Hood, M.J., Clausen, J.C., & Warner, G.S., 2007, “Comparison of stormwater lag times for low impact and traditional residential development,” Journal of the American Water Resources Association, v. 43, p. 1036-1046.
  • Riley, A., 1998, Restoring Streams in Cities: A Guide for Planners, Policymakers, and Citizens,” Washington D.C., Island Press, 423 p.
  • Wolman, M.G., 1954, “A method of sampling coarse river-bed material,” Transations of the American Geophysical Union, v. 35, p. 951-956.

Adapted from:

Vassar College Environmental Research Institute. “Health of the Casperkill.” February 2009.

Back to top



Back to Research Topics



  • Urban Wilderness Festival

    August 8th

    Date: Saturday, August 8th, 2015
    Time: 11:00am - 4:00pm
    Location: Vassar Farm and Ecological Preserve

    On Saturday, August 8th from 11am to 4pm, join us for a free Urban Wilderness Festival celebrating the beauty and diversity of the Vassar Farm and Ecological Preserve. Come on one of our many nature walks, play around with crafts, get your face painted, and go on a scavenger hunt! Enjoy food by local vendors, browse our tabling booths, and listen to some great music. It will be a blast! Parking will be in the main parking lot by the Poughkeepsie Farm Project.

    For more information on the this event, see the Facebook event page

  • Mitzvah Day 2015

    May 3rd

    Date: Sunday, May 3rd, 2015
    Location: Vassar College, Poughkeepsie, Room TBA

    Join the Jewish Family Services of Dutchess County and the Casperkill Watershed Association for Mitzvah Day, an annual day of community service. This year the CWA will be removing invasive vines from the vegetation surrounding the Casperkill Creek. Invasive species, especially vines, can crowd out beneficial native species and negatively impact the stream. Clippers, waders, boots, gloves, and garbage bags will be provided.

    Join the Jewish Family Services of Dutchess County and the Casperkill Watershed Association for Mitzvah Day, an annual day of community service. This year the CWA will be removing invasive vines from the vegetation surrounding the Casperkill Creek. Invasive species, especially vines, can crowd out beneficial native species and negatively impact the stream. Clippers, waders, boots, gloves, and garbage bags will be provided.