|
|
|
|
The Evolution of Flood Risk and Erosion Damage in the Shaker Lakes/ Doan Brook Urban Watershed Primary Researchers:
Abstract The Shaker Lakes/Doan Brook watershed of Cleveland, Ohio is an important environmental and cultural resource that suffers from typical urban watershed problems; increased flow and decreased environmental quality. Ongoing research projects are evaluating how the development of historic hydraulic structures have lead to the current flood risks and structural damage
The Shaker Lakes/Doan Brook watershed is a 11.7 mi2 (30.3 km2) system that begins in the "heights" east of Cleveland (the Cities of Shaker Heights, Cleveland Heights and University Heights) atop the Portage Escarpment and drops nearly 550 feet (168 m) as it flows West and North through University Circle (the cultural and educational center of Cleveland), Wade Park, Rockefeller Park and Gordon Park on its way to Lake Erie (see Fig. 1). The system contains four impounded lakes constructed by the Shakers in the early 19th century, and 8.1 miles (13.0 km) of stream channel that varies from "urban wilderness" (approximately 1 mile (1.6 km) sheltered by deep gorges in the Bedford and Cleveland shale) to sections channelized with elegant stone revetments. There is also a 1.3 mile (2.1 km) reach culverted under University Circle. Lower Shaker Lake was constructed in 1826 to power a saw mill and is believed to be the oldest impoundment in the state of Ohio. However, by the turn of the century, the "heights" were urbanizing and by 1940 had become some of the most glorious urban residential real estate in the country. Land use in the watershed is predominantly residential (85%) and is nearly saturated. Only 12 % of the area is undeveloped, and much of this is park land adjacent to the stream. As one might expect, man´s influences have resulted in substantial problems within the watershed. The lakes suffer from eutrophication and extensive sediment accumulation. There is evidence of dredging activity dating back as far as 1893. Water quality has also suffered and has continued to be a problem even as combined sewer overflows were eliminated. Furthermore, stormwater flows yield dangerous conditions in the channelized downstream reaches, and the possibility of plugging the culverted section yields serious flood risks to University Circle. There is also evidence of recent erosion damage implying that water quantities may be increasing, and that the whole aging infrastructure requires repair. Historic Development of the Lower Reaches of the Doan Brook The Doan Brook is named for Nathaniel Doane who built a log cabin on what is now Euclid Avenue in 1798, and is believed to be the first non-native American to see the brook. We have no record of what the local native Americans called the brook. The Doan Brook began its transition into a public resource in 1872 when Jeptha H. Wade began the development of 63.5 acres (25.7 ha.) of land north of Euclid Avenue as a public park. He presented this to the City of Cleveland in 1882 on the condition that walks, drives and a pond be developed. During this same period William J. Gordon developed a 122 acre (49.4 ha.) estate at the mouth of the Doan Brook which was donated to the City in 1893 after his death. Because the Wade and Gordon donations were on the North and South ends of a very distinctive reach of the Doan Brook, by 1890 the Cleveland Park Commissioners were being urged to acquire the lands between and to preserve the "wild romantic valley. . . as nature has formed and adorned it". Thus, when in 1893 the State of Ohio passed a law allowing park commissioners to raise funds and acquire park lands, the commissioners pursued this goal. Between 1894 and 1985 they acquired over 200 parcels (200 acres (80.9 ha.)) of land in the valley. It appears that the park commissioners hoped to acquire the whole Doan Brook stream valley. They acquired substantial portions of the gorge area (Ambler park) from the Ambler Family in 1894 and a large parcel of land (Shaker Heights Park) containing both Horseshoe Lake and Lower Shaker Lake from the Shaker Heights Land Company owned by John D. Rockefeller in 1896. Although these areas were not in the lower reaches of the brook, engaging Rockefeller in the concept of a continuous park led to its completion. On July 22, 1896, ClevelandŐs Centennial Day Celebration, John D. Rockefeller presented the park board with lands to connect Ambler park to Shaker Lakes Park and 254 acres (102.8 ha.) of land to be designated Rockefeller Park connecting Wade and Gordon Parks. Therefore, as of 1896, the whole lower reach of the Doan Brook became one continuous park. By 1900 the entire area was connected by drives overlooking the park (High Level Drive now known as East Blvd.), paralleling the Brook (Low Level Drive now known as Martin Luther King Jr. Blvd.) and numerous bridges crossing the brook and valley. It has been reported that on a single Sunday in 1896 more than 43,000 people used High Level Drive in carriages, on bicycles and on foot. However, this is not just a history lesson. Consider the hydraulic implications of this activity. In 1900, beautiful stone arch culverts were completed so Wade Park and Superior Avenues could cross the park. Between 1915 and 1939, ethnic groups built "Cultural Gardens" as World War I commerations. As more features were constructed in the park, flooding was perceived to be a greater problem. Between 1902 and 1926 this led to channelization of the entire brook, and to inevitable confrontations between the constructed works of man and the hydraulics of nature´s storms. Hydraulic Design of Doan Brook Structures We know when hydraulic works were constructed in Doan Brook and, because so many remain (an amazing fact given how they were designed) we know their physical dimensions. It is not as easy to uncover the "hydraulics" basis for these designs. From correspondence between project engineers we know that empirical formula were used to determine maximum design flows, but with few exceptions, we do not know the formulae or their predicted results. One exception is mention of use of "TalbotŐs Formulae" (an early version of the Rational Method) based on a rainfall intensity (I) of 0.79 inches per hour, a drainage area of 7,000 acres and a runoff coefficient of 0.18. A time of concentration of 120 minutes was also indicated, but it is not clear how that was applied. If one multiplies these values in the conventional way (i.e. ignoring negligible unit conversions) the result is 995 cfs.
Based on this (7000 acres (10.9 mi2, 28.2 km2) is a reasonable approximation of the basin´s size), and post-hoc evaluations of the structures themselves, it seems apparent that the engineers based their designs on a maximum flow of about 1000 cfs (28.3 m3/sec). By todayŐs calculations, this is not a "conservative" value. Given that the hydraulic structures were undersized, one would expect Doan Brook to be a constant source of flooding. In this regard, the Brook has disappointed no one. Therefore, it is surprising that the majority of the original channel structures still exist. One might assume that they should have been washed away years ago. This did not occur. They have been damaged and repaired repeatedly, but they retain what appears to be their original hydraulic capacities. The salvation of the system seems to be the (unintended ?) hydraulic capacity of Low Level Drive. When flooding occurs, Martin Luther King Jr. Blvd. acts as a floodway and carries flow that cannot be contained by Doan Brook. This allows the elegant but undersized hydraulic structures to survive. Major Floods on the Doan Brook The average dry weather flow of the Doan Brook is modest. The discharge of Lower Shaker Lake is generally 10-20 cfs. This increases as one moves down stream, but the greatest increase is probably due to groundwater discharge into storm sewers within the "sewershed". Under dry weather conditions, flows remain below 10% of existing channelŐs capacities. Flows are much different during wet weather conditions. The channelization was completed by 1926. By 1932 an additional $800,000 had to be spent to repair damage and to construct new walls. In 1955 another major round of renovation was required. Prior to 1959 it is not clear whether this damage accumulated gradually or was the result of a few major storms. This changed when on June 1, 1959 a rainstorm dropped 3 in. (7.6 cm) of rain in one hour on the drainage basin. The result was a flood flowing 10 feet (3 m) deep through University Circle. A similar storm dropped 1.5 in. (3.8 cm) of rain on June 4, 1962 yielded flood depths of 4 ft.(1.2 m). The worst event recorded in the basin occurred on Aug. 25, 1975 when a sudden storm dropped 6.0 in. (15.2 cm) of rain yielding flood depths of 11 ft. (3.3 m) in University Circle and causing more than $1,000,000 in damage even though capacity of the Doan Brook Culvert had been increased and Buildings had been equipped with flood gates. This was followed by a similar storm one week later that produced similar effects. Estimates of Flood Events Repeated flood damage has led to surprisingly few flood predictions. In 1964 the Stanley report predicted flood magnitudes for the Doan Brook at University Circle for current, and for "pre-development" (c. 1890) conditions. A similar Havens and Emerson report in 1968 predicted significantly lower values based on routing attenuation. Vidra (1976) also includes flood estimates for Doan Brook (see Table 1).
Table 1 - Previous Flood Predictions for the Doan Brook at University Circle
In the current research project, flood predictions were also developed for pre-development and current conditions. Estimates for pre-1890 conditions were developed using the method of Koltun and Roberts (1990) for flood discharges from rural, unregulated streams in Ohio. Calculations were based on a pre-development watershed size of 9.38 mi2 estimated from topography alone. Flood estimates were also developed using the method of Sherwood (1986) for urban watersheds. The "sewershed" was estimated to have expanded to 19.51 mi2 (50.5 km2) based on our analysis of the local storm sewer systems. These predictions are summarized in Table 2.
Table 2 - Additional Flood Predictions (cfs) for Doan Brook at University Circle
The observed record of flooding and the values of Tables 1 and 2 make it clear that a design flow of 1,000 cfs is inadequate to handle even low probability storms. Summary and Conclusions The historical development of hydraulic structures in Doan Brook was probably based on undersized flows. The subsequent urbanization of the drainage basin yielded increasing flood magnitudes and constant structural damage. Repairs of the hydraulic structures have resulted in a channel with undersized and mismatched flow capacity that can not contain the Brook. Given the nature of the drainage basin, it is unlikely that watershed management practices will resolve this quandary. The solution will require a bold redesign of the Doan Brook channel. References:
Note: This text is an abbreviated version of one from a set of papers to be presented at a session on Small Urban Watersheds at Water Resources and the Urban Environment in Chicago, June 10, 1998.
CWRU Department of Civil Engineering Communication
|
