HYDROLOGY OF THE BLACK SWAMP WETLANDS ON THE CACHE RIVER, ARKANSAS

The hydrologic characteristics of the Cache River wetlands between Pat terson and Cotton Plant, Arkansas, were investigated. The Cache River is an underfit stream with wetlands predominantly located in abandoned channels and backswamps. Much of the Cache River upstream of the stud y area has undergone extensive channelization to allow agricultural de velopment in the basin. Hydrologic measurements included streamflow ga ges at the upstream and downstream limits of the study area, water-lev el recorders inside the study area, a nest of deep and shallow ground- water wells that monitored variations in the underlying aquifer, a met eorological recording station that collected precipitation, air temper ature, and solar radiation data inside the study area, and regional pr ecipitation data. Analysis of the wetland's water budget showed that t he system is dominated by river discharges and that the magnitudes of other water-budget components are less than the error associated with well-maintained streamflow gages (5-10%). The system is characterized by floods occurring from late fall to late winter and again in mid-to- late spring. Peak flood discharges are approximately 185 m(3)/s for a 2-year event and 270 m(3)/s for a 5-year event. Peak discharges betwee n the upstream and downstream gages are reduced by 10-20% with greater attenuation occurring when the system is initially drier. Peak discha rge at the downstream gage lags the peak at the upstream gage by 4-8 d ays depending on antecedent conditions. The majority of overbank flood ing is produced by backwater from several constrictions in the downstr eam reach of the study area, rather than from the forward movement of the flood wave. Flood-peak attenuation between the upstream and downst ream gages is due mainly to floodplain storage, with flow resistance c ontributing minimally. Finally, the relationship between the hydroperi ods at different water-surface elevation gages was examined to determi ne if a long-term record could be used to estimate long-term hydroperi ods at interior gage locations (or perhaps at interior computer model locations) with shorter-term information.