North-Central Section–40th Annual Meeting (20–21 April 2006)

Paper No. 25
Presentation Time: 8:00 AM-12:00 PM

A STELLA MODEL FOR A VERY SMALL HEADWATER HYDROLOGICAL SYSTEM AT NEW CONCORD, OHIO


ORESS, Matthew P.1, LAW, Eric2 and VAN HORN, Stephen2, (1)Geology, Muskingum College, 163 Stormont St New Concord, New Concord, OH 43762, (2)Geology, Muskingum College, 163 Stormont St, New Concord, OH 43762, moress@muskingum.edu

A hydrological system which includes the flows of surface water and groundwater over a small drainage area over the campus of Muskingum College at New Concord, Ohio is studied based on a STELLA model. The drainage area is slightly larger than one square mile and it includes two first-order creeks. One is significantly larger than the other. Both creeks drained immediately into a reservoir which is the campus lake. The total topographic relief is about 100 ft (from 950 ft, which is the regional high elevation). Factors considered in the model include surface water input, surface runoff, soil water, evapotranspiration, groundwater infiltration and groundwater discharge. Lithologic character of the shallow ground is also considered. It includes two thick layers of nearly horizontal silty shale sandwiched one median bed of sandstone. The flow of the large creek passes over a section of flat bedrock and the whole large creek drainage passes through a pipe into the reservoir. This situation allows an easy and good estimation on the amount of base flow, groundwater discharge and surface runoff. A level logger is used to measure the change of water level in the lake. And the out flow of the lake through a draining well can also be accurately estimated. The advantages of this model include conditions that allow relatively accurate measurement on the rainfall; the base flow and surface runoff. The drainage area is high enough in elevation so that the source of out flowing groundwater is likely to be all local. At this initial stage of study, the parameters on each hydrological property are adjusted so that the model would give results that have the closest match to the measured data.